PAGENO="0001"
COMPETITIVE PROBLEMS IN THE
DRUG INDUSTRY
~, 7O ~2 ~ii~
HEARINGS
BEFORE THE
SUBCOMMITTEE ON MONOPOLY
OF THE
SELECT COMMITTEE ON SMALL BUSINESS
UNITED STATES SENATE
NINETIETH CONGRESS
FIRST AND SECOND SESSIONS
ON
PRESENT STATUS OF COMPETITION IN THE
PHARMACEUTICAL INDUSTRY
PART 5
DECEMBER 14, 19, 1967, JANUARY 18, 19, AND 25, 1068
Printed for the use of the Select Committee on Small Business
U.S. GOVERNMENT PRINTING OFFICE
81-280 WASHINGTON 1968
~
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price $2.
PAGENO="0002"
SELECT COMMITTEE ON SMALL BUSINESS
[Created pursuantto S. Res. 58, 81st Cong.J
(90th Cong., first and second sess.)
GEORGE A. SMATHERS, Florida, Chairman
JOHN SPARKMAN, Alabama JACOB K. JAVITS, New York
RUSSELL B. LONG, Louisiana HUGH SCOTT, Pennsylvania
WAYNE MORSE, Oregon NORRIS COTTON, New Hampshire
ALAN BIBLE, Nevada PETER H. DOMINICK, Colorado
JENNINGSRANDOLPH, West Virginia HOWARD H. BAKER, Ja., Tennessee
E. L. BARTLETT, Alaska MARK 0. HATFIELD, Oregon
HARRISON A. WILLIAMS, Ja., New Jersey
GAYLORD NELSON, Wisconsin
JOSEPH M. MONTOYA, New Mexico
FRED R. HARRIS, Oklahoma
WILLIAM T. MCINARNAY, Staff Director and General Counsel
JAMES H. GROSSMAN, Minority Counsel
MONOPOLY SUBCOMMITTEE
GAYLORD NELSON, Wisconsin, Chairman
JOHN SPARKMAN, Alabama HUGH SCOTT, Pennsylvania
RUSSELL B. LONG, Louisiana MARK 0. HATFIELD, Oregoa
WAYNE MORSE, Oregon JACOB K. JAVITS,* New York
GEORGE A. SMATHERS,* Florida
officio member.
BENJAMIN GORDON, Staff Economist
SUSAN H. HEWMAN, Research Assistant
II
PAGENO="0003"
CONTENTS
Statement of-
Binkley, Howard, vice president for administration and planning,
Pharmaceutical Manufacturers Association, Washington, D.C.;
accompanied John M. Firestone, Ph. D., professor of economics, the Page
City College of New York, New York, N.Y 1690
Comanor, William S., Ph. D., professor, Department of Economics,
Harvard University, 7146 Cambridge Street, Cambridge, Mass~.. - 2041
Conrad, Gordon R., senior staff associate, Arthur D. Little, Inc.,
25 Acorn Parkway, Cambridge, Mass 1667
Cootner, Paul H., Ph. D., professor of finance, Massachusetts Insti-
tute of Technology, 77 Massachusetts Avenue, Cambridge, Mass_ 1618
Cutler, Lloyd N., special counsel, Pharmaceutical Manufacturers
Association, Washington, D.C 1615
Firestone, John M., Ph. D., professor of econOmics, the City College of
New York, New York, N.Y.; accompanied by Howard Binkley,
vice president for administration and planning, Pharmaceutical
Manufacturers Association, Washington, D.C 1690
Kelly, William H., member, Bureau of Economics, Federal Trade
Commission, Washington, D.C.; accompanied Dr. Willard F.
Mueller, Chief Economist and Director, Bureau of Economics~_ - 1807
Markham, Jesse W., Ph. D., professor: of economics, Princeton Uni-,
versity, Princeton, N.J 1667
Mueller, Dr. Willard F., Chief Economist and Director, Bureau of
Economics, Federal Trade Commission, Washington, D.C.; ac-
companied by Dr. Russell C. Parker, Assistant Director; and
William H. Kelly, member, Bureau of Economics 1807
Parker, Dr. Russell C., Assistant Director, Bureau of Economics,
Federal Trade Commission, Washington, D.C.; accompanied
Dr. Willard F. Mueller, Chief Economist and Director, Bureau of
Economics 1807
Plotkin, Irving H., consultant in econometrics and operations research,
Arthur D. Little, Inc~, 25 Acorn Parkway, Cambridge, Mass 1638
Schifrin, Leonard G., - Ph. D., director, Department of Economics,
College of William and Mary, Williamsburg, Va 1863
Squibb, George S., former vice president, E. R. Squibb & Sons,
3-Cedar Island, Larchmont, N.Y 1555
Steele, Henry B., Ph. D., associate professor of economics, University
of Houston, 105-A Heyne Building, Houston, Tex 1901
Whitney, Simon N., Ph. D., professor of economics, New York
University, New York, N.Y 1725
APPENDIXES
Correspondence between Mr. Benjamin Gordon and Dr. P. M. Costello
re technological progress in the pharmaceutical industry 2113
Paper, "Risk and Corporate Rates of Return," by I. N. Fisher and
G. R. Hall, The Rand Corp., Santa Monica, Calif 2120
Additional Submissions by the Pharmaceutical Manufacturers
Association 2129
III
PAGENO="0004"
lv CONTENTS
SALIENT EXHIBITS
Report, "Risk and Return in American Industry-An Econometric Page
Analysis," by Gordon R. Conrad and Irving H. Plotkin 1746
Study, "Trends in Market Shares for Ethical Pharmaceutical Products,"
by Gordon R. Conrad 1783
Supplemental statement of Dr. Willard F. Mueller, Director, Bureau of
Economics, Federal Trade Commission 1843
Article, "The Ethical Drug Industry: The Case for Compulsory Patent
Licensing," by Dr. Leonard G. Schifrin 1890
Article, "Monopoly and Competition in the Ethical Drugs Market," by
Dr. Henry B. Steele, from Journal of Law and Economics, October
1962 1950
Article, "Patent Restrictions and Price Competition in the Ethical Drugs
Industry," by Dr. Henry B. Steele, from Journal of Industrial Eco-~
nomics, July 1964 1971
Article, "The Fortunes of Economic Reform Legislation: The Case of
the Drug Amendments Act of 1962," by Dr. Henry B. Steele, Rice
University 1998
Article, "Income Opportunities and Physician Location Trends in the
United States," by Dr. Henry B. Steele and Dr. Gaston V. Rimlinger,
from Western Economic Journal, spring 1965 2011
Article, "An Economic Interpretation of the Spatial Distribution of
Physicians in the United States," by Dr. Gaston V. Rimlinger and
Dr. Henry B. Steele, from Southern Economic Journal, July 1963 2025
Article, "Research and Competitive Product Differentiation in the Phar-
maceutical Industry in the United States," by Dr. William S. Comanor,
from Economica, November 1964 2069
Article, "Research and Technical Change in the Pharmaceutical Industry,"
by Dr. William S. Comanor, from Review of Economics and Statistics,
May 1965 2078
Article, "The Drug Industry and Medical Research," by Dr. William S.
Comanor, from Journal of Business of University of Chicago, January
1966 2086
Article, "Advertising Market Structure and Performance," by Dr. William
S. Comanor and Dr. Thomas A. Wilson, from Review of Economics and
Statistics, November 1967 2092
HEARING DATES*
December 14, 1967 1555
Afternoon 1579
December 19, 1967 1609
Afternoon 1661
January 18, 1968 1807
January 19, 1968 1863
January 25, 1968 2041
* The testimony for May 15, 16, 17, June 7 and 5, 1967, appears in pt. 1 of these hearings; the testimony for
June 27, 28, 29, July 24, and Aug. 8, 10, 1967, appears in pt. 2 of these hearings; the testimony for Sept. 13, 14,
29, and Oct. 13, 1967, appears in pt. 3 of these hearings; the testimony for Oct. 31, Nov. 9, 15, 16, and 28, 1967,
appears in pt. 4 01 these hearings.
PAGENO="0005"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. SQUIBB. Not at all, Sena
Senator N1~soN. All right, if
time to time, as questions occur
Mr. SQUIBB. I have a few mt
a little background which I maj
Senator NELSON. All right, ~
read into the record your biogra
Mr. SQUIBB. Yes, I will do th
Squibb. I live at 3 Cedar Islai
at the request of the committe
Harvard University, I have fi
[Hatfield.
don, staff economist; James H. Gross-
H. Hewman, research assistant; and
ive director, staff of Senator Nelson.
of the Monopoly Subcommittee of the
ee will come to order.
mnittee is very pleased to have the op-
a Mr. George S. Squibb. He was former
& Sons, and a leader in the drug in-
ne has been advocating reform within
egards pricing policies. Mr. Squibb, I
~timony very carefully. Your analysis
ik, opened the door to consideration of
lustry. It is reassuring to know that
udgment, on many quality drugs, and
you are willing to fight for reforms
The committee is most pleased to have
u may present your testimony in any
you do not have any objection to inter-
oceed to present your testimony.
or. As you see fit any time.
rou will proceed, Mr. Squibb, and from
e will interrupt you.
oductory remarks as to who I am and
read first.
r. Squibb, you may proceed. Will you
~hical sketch?
t now, Senator. My name is George S.
d in Larchmont, N.Y., and I am here
For 30 years, since graduating from
led a wide variety of positions with
THURSDAY,~ DECEMBER 14, 1967
U.S. SENATE,
MON0~OLY SUBcoMMrrrei OF THE
SELECT C~OMMITrEE ON SMALL BusINEss,
Washington, D.C.
The subcommittee met, purskiant to call, at 10:20 a.m., in room 1202,
New Senate Office Building, S~nator Gaylord P. Nelson (chairman of
the subcommittee) presiding.
Present: Senators Nelson an
Also present: Benjamin Go
man, minority counsel; Susa:
William B. Clierkasky, li~gisla
Senator NELSON. The hearin
Senate Small Business Commi
Today, the Monopoly Subc
portunty to hear testimony fro
vice president of E. R. Squib
dustry. Mr. Squibb for someti
the industry, particularly as
have read your 28 pages of t
of the drug industry has, I thi
major reforms within that h
prices can be reduced in your
that industry leaders such as
which will make this possible.
you here this morning, and y
fashion you see fit. I trust that:
ruption for questions as you p
STATEMENT OF GEORGE S. QUIBB, FORMER VICE PRESIDENT,
E. R. SQUIBB & ONS, LARCEMONT, N.Y.
1555
PAGENO="0006"
1556 COMPETITIVE P1 )BLEMS IN THE DRtTG INDUSTRY
TE. R. Squibb & Sons, and
that organization.
I am a great-grandson
Squibb, and represent the
pany, which was founde
Squibb I have served in
the plants and laboratori
in production planning
then served as headqua
points. I was elected secr
boards of directors of all
Chemical Corp. I was th
director of administratio
ment as director of trade
ant director and directo
marketing.
After a series of corp
vice president for legal a
the company.
During my career I h
management committees,
types of management fun
I have participated as
of professional and trade
I have been active in co~
major industry tradi~r4
man of the National Phar
I am a lawyer, member
study at the Food and Dr
business backward and fc
concerned with its future
all matters affecting pubi
us in the years ahead.
I have followed very
nesses who have a~pea~
studied it, and familiariz
least two major conclus
all those concerned with
health plans insofar as
the industry will not adn
help clear the air of misu
I would like to try to
marks based on long mdi
ful analyses of things pa
in the future.
Now, if I may go ahead
Senator NELSON. Yes; i
Mr. SQUIBB. The publ
prices are too high-tha
scribed medicines-and t
cal companies were not so
Because the Governme
am currently on a consultantship basis with
of Dr. Edward R. Squibb, the founder of
fourth generation of the family in that corn-
in 1858. During my 30 active years with
very part of the organization. I worked in
~. I was a salesman and detail man. I served
nd control. I managed a branch office and
ters manager for all Squibb distribution
tary of the company in 1949, and served the
~quibb Cos. until the merger with Mathieson
n made assistant to the president, and later
* In 1954 I started back in the sales depart-
and customer relations, progressed to assist-
of sales and finally to vice president for
rate management changes I was appointed
d public affairs and am now a consultant to
ye served on boards of directors, executive
patent and research committees, and in all
;ions at all levels.
peaker or panelist on programs of all kinds
groups in nearly every State in the country.
aitteewoçk for the PMA and NPC, the two
ips, and was elected to two terms as chair-
riaceutical Council.
of the New York Bar, with special graduate
ig Law Institute. I know the pharmaceutical
rward, most everybody in it, and am vitally
both personally and from the conviction that
c health must be of vital importance to all of
losely all the testimony of the various wit-
I before this subcornmitee. I have re,ad it,
d myself with it, and have come up with at
ons: First, there is a real problem facing
~he establishment of Government-supported
[rug procurement is concerned, and second,
it it, talks around it, and contributes little to
derstanding, suspicion, and distrust.
mprove on this performance with some re-
stry experience, careful study, and thought-
t with some specific suggestions for changes
Senator?
ease proceed.
c believes that pharmaceutical prescription
they should be paying less for their pre-
at they could be paying less if pharmaceuti-
greedy-or used less wasteful sales methods.
it, State and Federal, now is preparing to
PAGENO="0007"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1557
pay so mudh of the medical bill of the public-a bill the potentiaJ of
which incidentally still is grossly understated and misunderstood-
legislators have seized upon this public conviction of high prices on
drug products to attack the industry for its behavior and to use it
as the justification for proposals of new regulation and control of a
magnitude far beyond anything that has been felt to date.
Senator NELSON. At the end of the first sentence you say: "Or used
less wasteful sales methods." Would you elaborate on what you mean
by wasteful sales methods?
Mr. SQUIBB. I am talking here, of course, about the expenditures
which seem to be large, are actually large for the detail man and for
the advertising and promotional methods of one type or another that
the industry uses to further the sales of its products. These seem on
the surface, to the public and to the industry critics, to be unneces-
sarily large and unnecessarily indiscriminately used. I develop that
a little bit later specifically when I get to some of these areas of
expense.
Senator NELSON. All right.
Mr. SQUIBB. The problem that faces the pharmaceutical industry
today relative to public criticism of its pricing and profit structure
can be stated just that simply. What to do about it, if indeed anything,
is not so easy for the pharmaceutical manager to determine. All sorts
of complex social forces and economic facts must be understood, sorted
out, and balanced against each other in an effort to foresee the long-
range effects of this concept of pharmaceutical prices, a concept which
is now generally accepted outside of the industry itself.
The only answer that the pharmaceutical industry has given to
date is to "tell its story," to "improve its image," to "explain its con-
tributions to medical knowledge and the health of the people." These
efforts for the most part, have been ineffectual in changing the mind of
the public about the nature of the prices it pays for its drugs, and yet
the industry keeps at it with dogged persistence issuing new studies
in depth-new comparative analyses of profits country by country
and industry by industry, new interpretations of the value of its prod-
uct contributions, new explanations and justifications for its modus
operandi. To all of its critics, temperate or intemperate, expert or lay,
legislative or academic, it keeps on saying, in effect: "You just do not
understand us-there is the truth which only we can properly explain."
The industry has been primarily concerned, and on doubt properly
so, with its own internal problems induced for the most part by
reason of its phenomenal expansion and by the possibilities for even
further growth in widely divergent areas. It has tried to meet, some-
times successfully and sometimes not, the day-to-day problems arising
from a complex product mix in a complex marketing situation all
the while keeping a sharp eye on earnings and corporate health. It
has tried honestly and hard to produce good products, and has gen-
erally not been satisfied to rest on today's achievement, but to press
on to something better. It has been diligent within its own self-estab-
lished limits in assuming its responsibilities, and it is truly hurt by
those who suggest anything to the contrary.
However, diligence, devotion to ideals, and just plain hard work
do not necessarily imply intelligence and good judgment, and it is in
PAGENO="0008"
1558 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
this light that the pharmaceutical managers should look today at their
policies and their operational Procedures if they are to control their
own futures.
Mr. GORDON. What do you mean by "within its own self-establisI~ed
limits"?
Mr. SQUIBB. Well, I think the industry sets forth for itself the areas
in which it is properly concerned. It does not accept limits that are
proposed from outside, critics or from the public, even generally
as to what should be the proper limits of its responsibility. The in-
dustry decides what its responsibility is, aild has carried out that ic-
sponsbility very carefully and diligently. It is making its own internal
judgments as to what are its own duties in this concept.
Let us take a look, just as to how the industry prices its products.
Every product in the marketplace has a cost, a price and a value.
Costs of products in all industries are difficult to determine precisely,
not only because manufacturers are naturally reluctant to discuss
them, but also because there is a. real difference in opinion as to what
they actually are within a given company's financial structure. Meth-
ods of accounting, allocation of overhead and variable production
expenses, problems in separating out indirect distribution costs and
applying them fairly to diverse items, all cloud over any effort to fix
clearly just what the "cost" of a. product is. This is often an even
more irritating difficulty to tile manufacturer than to the outside in-
vestigator, and many a firm has undergone real financial distress
because it could not properly establish its product costs.
All this applies especially to the pharmaceutical industry. With
their multiproduct lines of greatly dissimilar products requiring all
sorts of different production leadtimes, storage and thipping vari-
ants, and sales promotion programs ranging from the simplest to the
most complex techmcally of any that can be imagined, Pharmaceutical
companies find product cost determination a real problem even with-
out the almost overwhelming difficulty of proper allocation of basic
and specific research and development costs which are a substantial
part of their annual operating expenses. However, it seems reasonably
certain that costs of major volume products are a much smaller per-
centage of trade prices than is usual in consumer goods industries.
This can be deduced from much published evidence of costs estimated
and real, and the facts that price ranges are enormous, and profits
better than average.
Senator HATFIELD. Mr. Squibb, could I interrupt? Tile sentence be-
fore last: "However, it seems reasonably certain that costs of major
volume produc.ts are a. much smaller percentage of trade prices than
is usual in consumer goods industries."
I presume here ot.her consumer goods industries. Could you give
us a. little further explanation of comparisons which you are drawing
here?
Mr. SQUIBB. Well, what I am trying to compare here are the factory
costs, the direct labor and tile actual out-of-pocket costs to create a
prodluct for the pharmaceutical business, compared with consumer
products, such as food products and pac.kaged goods products that are
distributed and bought in the general markets by the public. I think
that we have a large mass of evidence that has been deduced and put
PAGENO="0009"
COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY 1559
together from various directions, `that the actual costs, of pharmaceu-
tical products, are quite low compared to the price that is set. I do not
have tables, and `this evidence here.. This is in the background. Much
of this is in testimony already giveli `to the committee.
Senator NELsON. Just for clarification, are you referring to the kind
of testimony that appeared sometime ago in `the Kef'auver hearings,
and I think once or twice here, in which efforts were macic by the
committee or by some witnesses `to establish the cost of the basic pro-
duct and then acid reasonable markup? I think we had some testi-
mony `a while back `that `the cost of making 100 tablets of a particular
drug was $1.80 or something like that including profit, and the tablet
was selling for $16 or $18 a 100. Is that the kind of differential that
you are `talking about?
Mr. SQUIBB. That is `the cost I am:'talking `about, the basic cost of
`the package. I think we have seen `also laid out before the committee
at various times in the past weeks the other elements of expense that
go into the final pricing of the product, and because profits, research,
and `selling expenses are larger in this industry percentagewise than in
most other industries, it follows inevitably that the cost has to be
smaller. I think this is a matter of general `agreement by anyone who
examines the subj ect `objectively.
Mr. GORDON. Mr. Squibb, to corroborate your statement you may be
interested to know that according to the 1961 Annual Survey of
Manufacturers, U.S. Bureau of the Census, the wages as a percentage
of value added for all manufacturing is 33.5 percent. For chemical and
allied products it is 17 percent. For drugs it is 11.1 percent. That is
labor cost.
With respect to materials cost, it is also pretty low for the drug
industry; and with respect to investment data, fixed assets, as a per-
cent of sales for `all manufacturing is 28.5 percent, chemicals and
allied products 37.6 percent, and drugs 25.6 percent. So the statistics
of the Bureau of the Census certainly corroborate w'hat you have said.
Mr. SQUIBB. I think this is a basic fact which actually determines
much of what happens later on in setting the pricing structure for
any product the fact that the cost is comparatively low.
Generally speaking and within a given industry, the price of a
product has a predictable relationship to the cost, and in turn the
value is close to the price, such being the nature of our competitive
economic system. In the drug business, ho'wever, prices are unpredict-
able; often they seem to bear no normal relationship to cost. More
often than not there are several widely divergent prices for the same
product, the highest frequently in~ the largest market and the lowest
in the smallest market. Indeed it `is difficult to `determine just what
the price is on many important prescription drug products.
Senator NELSON. We have had considerable testimony, which I take
it you have read, showing the difference between the high prices
charged for drugs `to the retail pharmacist, `and the dramatically lower
price that the same companies bid for the same drugs to the Defense
Supply AgenQy or the Veterans' Administration or the city of New
York or major hospitals. These differences amount to a couple thou-
sand percent. There are cases where 100 tablets would sell on the retail
market for $17 or $18, while the same company bids $1.20 to New
PAGENO="0010"
1560 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
york for the same drug. What explanation do you have for the same
drug on the same day being sold to the Defense Supply Agency or
to large hospitals for a tenth or one-twentieth of the price that is
charged by the same company to a retail pharmacist?
Mr. SQUIBB. Well, you put your finger on the whole difficulty. There
seems to me to be no good explanation for this. These price differ-
entials have developed over the years out of a series of market con-
ditions and other pricing situations which I am going to expand on
a little bit, but there is no rational explanation that I think can be
offered for these very large price variants. This is precisely what has
led the industry into the problems that exist today. I believe that this
is the point that the industry should face up to, namely, the fact that
they cannot give a reasonably accurate explanation of these differ-
entials that is acceptable to outside economists or even the public
generally. If they would accept this fact and examine their pricing
policies more specifically and more realistically, more in view of the
way the outside world looks at them, I think a good part of their
difficulties would be over, because they would obviously cut out the
practice that you have just mentioned.
Senator NELSON. It seems to me in viewing the situation that there
might be some explanation in the fact that when drug companies offer
bids to big government agencies or to the city of New York they are
in a tough competitive situation. Bids are solicited and a number of
companies come in with an offer, and compete with each other for the
business of the particular big hospital or government agency, whereas
in the reta.il market the use of a brand name seems to eliminate coin-
petition. The physician does the prescribing, and if a trade name is
established among a certain percentage of the doctors for a drug that
they want to prescribe, does not that tend to eliminate genuine
competition.
Mr. SQUIBB. Well, I think there is a bigger factor than the fact that
the doctor is presold on what he wants by the trade name, that does
tend, as you say, to eliminate competition at the retail level. This is
the fact that the price that is paid for the prescription by the patient,
is determined basically by the price that the manufacturer has charged
the retailer for the merchandise. Now, this price is set and is very rigid.
There is no competitive activity on this price from retailer to retailer,
from large retailers to small retailers, in any way. This price is affected
as I point out later on, by Robinson-Patman concepts, and there is a
rigidity of pricing at the retail level. This has existed remarkably
firmly, and I think this more than anything else sets the top level for
the situation which you have just described.
Senator NELSON. But we do have in the testimony that I am sure
you have read, an instance of a drug on the market that is selling for
as mitch as $17 or $18 a hundred by a large firm down to as low as $2 a
hundred by another distinguished brand name company and for as
little as 78 cents by some lesser known companies. Yet, the company
that is charging $17 or $18 a hundred on the retail market and holding
a large share of the market does turn around on the same day and bid
one-tenth or one-fifteenth as much as to the New York City or the
Defense Supply Agency.
Mr. SQUIBB. That is right.
PAGENO="0011"
COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY 1561
Senator NELSON. Now, they would not sell anything to New York
City or the Defense Supply Agency, if they were charging $17 or $18
a hundred because they would be outbid every single time. So they
offer a price of one-tenth or one-fifteenth as much. On the other hand,
they go into the retail market and charge the $17 or $18 a hundred or
whatever it may be, depending upon the drug, several times as much,
and do manage to hold a share of that: market and make a substantial
profit. This is the part that is difficult to understand, unless it is simply
the fact `that the trade name has become well enough established with
the doctors who will prescribe their, drug regardless of what the
price is.
Mr. SQUIBB. That is exactly, of course, what has happened. There
is no price competition or very little price competition at the so-called
retail level. If you get your trade name established with your doctors
through your promotion activities and your advertising and the gen-
uine value of your product, the largest drug chain in the country or the
smallest drugstore or any buying organization of the retail pharma-
cist still has to pay the $16 because the price is established and pro-
tected for competitive purposes. There is not any price competition
that the manufacturer has to worry about at that level, and he is very
happy, I think, about this.
`Senator NELSON. Go ahead.
Mr. SQUIBB. Value, continuing at the top of page 4, value on the
other hand, is often far above any of the prices set for the product.
If health and indeed life itself are measured in terms of dollar value
against the cost of the medicines that contribute to them `and in many
cases guarantee them, drug products give probably the finest value of
anything that can be purchased today. Yet, in ~pite of these tremen-
dous values, the public belief persists that prices of drugs are too high
and costs inflated by unnecessary and ,wasteful operational procedures
by their manufacturers.
The reason for this seems to be in the unusual, to say the least,
pricing policies of the pharmaceutical industry. It must be understood,
however, that the industry operates in a complex-partly controlled
and partly open-market in which it deals on a direct basis with the
widest possible range of customers both in term's of their capacity for
product utilization, and in terms of their competitive and noncompeti-
tive natures.
At the retail level, that is the retail pharmacy, prices on direct sales
to pharmacies large and small-some are very large, others very
small-are affected by Robinson-Patman. Identical prices must be
charged to the largest chain and its smallest independent competitor
for the same drug package.
Senator NELSON. I am not an authority on the Robinson-Patman
Act. That act does not prohibit a comp'any from giving advertised
price reductions for quantity purchases, does it?
Mr. SQUIBB. As long as it gives those advertised quantity price re-
ductions to everybody on an equal basis, and that t'hey are really bona
fide quantity price reductions.
Senator NELSON. That they are related to the cost?
Mr. SQumi~. Related to the cost. They have to be related to the cost
and also related, as I point out here in this paragraph, to what the
PAGENO="0012"
1562 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
customer could normally buy. The price cannot be established for a
large quantity that only the biggest chain drugstore could reasonably
use. This is a trick wl~ich can be done, but which is not done in the drug
business. It just cannot be done by coming out with a 100,000-tablet
price which only Wralgreen, for example, could use. This would be a
sham, giving Wa.lgreen a price advantage which is improper because
the little fellow cannot either physically or financially buy 100,000
of these tablets. I think the pricing practice in the drug business in this
area is extremely careful, is extremely closely advised by legal coun-
sel, and is strictly within the spirit and the letter of this law which
protects the smaller merchant against the larger potential commercia.l
competition that lie has. I think this is a very important fact which has
affected prescription prices and is affecting today's prescription prices.
Sometimes I think it is not always completely appreciated. No matter
what you think about Robinson-Patman, you have to realize that this
is one of the controlling factors in keeping these Prices rigid and
keeping them at the level at which they were originally established.
There are iio pressures to bring them down, no normal competitive
pressures in the marketplace.
Senator NELSON. We have just talked about the situation where a
company will sell 100 tablets of the same drug to the retail pharmacist
for $17 and $1 to New York City. If you then allow a company to buy
100,000 tablets, and give it the price which you give to New York
City, then you would end up driving the people out of business who
are paying the $17 per 100.
Mr. SQUIBB. Out of business, that is the idea..
Senator NELSON. If you didn't have that price differential in the
first place then you would have quite a different problem.
Mr. SQUIBB. Yes, and the fact that the industry has permitted it
more and more and still seems to be following the practice as you
pointed out, of having this tremendously wide variance, continues to
keep them in trouble at their end. They do not seem to learn. They
keep developing these very wide variants and I think it is going to get
worse because the other area which we have not talked about yet, the
private hospital area, is developing more into competition with the
retail pharmacy area of operation, and in this latter area Robinson-
Patman at this moment does not apply. Hospitals are nonprofit and
are not judged to be in competition.
Senator HATFIELD. Mr. Squibb, let me clarify the record a bit by
asking this. As I understand it, you take issue with some of the in-
dustry witnesses we have had who have stated very specifically that
quantity buying does have a distinct impact upon the pricing dif-
ferentials. You say in your testimony that package size and quantity
discounts are minimal.
Mr. SQUIBB. I Say that they are minimal at the retail level; that is,
in sales to retail phar~nacies. They certainly are not minimal in over-
all sales to private institutions, Government and privately controlled
hospitals. I am talking here about the rigidity of the retail price list,
in which five-percent quantity discounts or ten-percent quantity chs-
counts are the maximum that are given.
Senator HATFIELD. As it would relate to the average consumer pur-
chasing through a corner drugstore, the quantity buying by that drug-
PAGENO="0013"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1563
store would have an effective price differential which might exist be-
tween that drugstore and some other one?
Mr. SQUIBB. Very little, because even if you go to buy your prescrip-
tion from the largest chain, that chain taking full advantage of what-
ever quantity discounts that are, can only buy perhaps, a or 10 percent
less at the most. That is all that any retttiler can get in advantage over
another retailer. There is very little at that level.
Senator HATFIELD. That is in spite of what other witnesse.s have
said?
Mr. SQUIBB. Well, I am sure tha.t thete are, if you examine the trade
list of manufacturers, quantity discounts to retail pharmacies, but
they are small. I am not talking about hospital lists. There are special
Prices there, but quantity discounts to retail pharmacies are minimal,
m my opinion. They go 5 to 10 percent, and they cannot be any more
than that or you are in trouble right away with the Robinson-Patman,
unless you can give this discoi.mt to everyone.
Senator HATFIELD. But, they are significant in terms of the large
institutions?
Mr. SQUIBB. Oh, very significant. I think we have covered the para-
graph at the end of page 4 and the top of page 5, which I was talking
to Senator Hatfield about.. Price cutting, discrimination, favored
customers, under-the-counter deals, hidden discounts are no factor in
sales of prescri.ption products direct to the retail pharmacy trade by
pharmaceutical manufacturers. It must be stressed that the retail
Pharmacist does most of the prescription business in the country, and
it is from this source, generally speaking, that the industry not only
gets most of its sales volume, but an overwhelming percentage of its
profit.
Pharmaceutical manufacturers also sell to wholesalers who, in turn,
fill orders from retail pharmacies. There are sharply competing whole-
salers in every city and State in the Nation, and the price to each of
them on a given pharmaceutical product from a given manufacturer
is identical. It has to be for obvious legal reasons, and there are no
exceptions to this rule. The wholesaler price is generally discounted
from 5 to 20 percent from the so-called trade-retailer__price. Price
competition among wholesalers in passing along to the retailer part
of tihis discount is a common procedure, but this also is controlled by
Robinson-Patman, and wholesalers must offer these discounts equally
to all his competing customers. Attempts have been made to tie in
services, order quantity and order timing factors, and credit privi-
leges with discount sharing, but generally speaking, ttie variance in
price produced by this wholesaler practice is not large and it remains
but a small factor in the determination of the ultimate price of the
product on the consumer's prescription. Occasionally and especially
in large metropolitan areas when competition among wholesalers for
retail accounts is bitter, products can `be purchased by the pharmacist
from the wholesaler more cheaply than he can buy them direct, which
fact in itself reflects the rigidity of the direct-selling manufacturers
list., and the firm observance of the general principle of Robinson-
Patman.
While prices of drug products remain generally rigid to the retail
and wholesale druggist, quite a different picture is shown on prices to
PAGENO="0014"
1564 COMPETITIVE PROBLEMS 1_N THE DRUG INDUSTRY
hospitals-non-profit or Government-operated. Here the widest varia-
tions exist and from this area of pricing, evidence is deduced that other
pricing is non-competitive, unreasonable and unfair. While many
pharmaceutical manufacturers maintain so-called hospital pricelists,
most hospitals, large and small, find ways to buy their requirements
at lesser figures particularly on important large volume products
which bear relatively high-price tickets to start with. It should be
understood at once that Robinson-Patman does not apply to non-
profit institutions who are, per se, not juged to be in competition with
each other. Therefore, while Walgreen Drug Co., the largest drug
chain in the country, pays $5 for a bottle of a brand of diuretic tablets
even if that purchase is repeated a 100 times in a week as compared
with the one-a-month $5 price for the item by an independent phar-
macy, a large hospital can and often does beat that price down to
$3.50 or less, and even the small hospital can expect to get at least
10 percent off the retail price if it has any buying acumen at all. This
is done by one or more of several procedures which hospitals are using
to ever-increasing effect as their size and importance grow under cur-
rent expansive trends of medical treatment supported by Government
and insurer funds. Request for competitive bids, formation of buying
associations, threat to manufacture or subdivide themselves, control
over brand specification within its own walls by therapeutic or formu-
lary committees, exposure of products to large numbers of physicians
who can be expected outside the hospital to use a product with which
they have become fan'iiliar within-such usage incidentally, at consid-
erably higher prices-all these things contribute to the hospital's
ability to obtain drug products at lower prices than the retailer or
wholesaler.
Senator HATFIELD. Mr. Squibb, could I interrupt here? WThat would
happen to prices if the Robinson-Patman Act were applied to hos-
pitals, in your opinion?
Mr. SQUIBB. The prices would firm up, take the same rigid nature
that they have in the retail trade. The manfuacturers would probably
welcome it in terms of removed competitive factors in pricing to hos-
pitals. Somewhere in the middle of the presei~t wide range you would
have a hospital price that would be arrived at after a little give and
take for a while. There would be a rigidity of prices and a lack of
competitive bidding, which I think would tend to raise the overall
average of drug prices very substantially.
Senator HATFIELD. It would raise the prices?
Mr. SQUIBB. It would raise the average.
Senator HATFIELD. WTould the companies be able to establish uni-
formity on prices? Would they be subjected at that point to-
Mr. SQUIBB. Oh, no, they do it themselves. They have a list. They
would set a hospital pricelist. The price would be $5 or whatever on
an item a.nd that would be the published hospital list price available
for all to see, including the Government regulatory agency. The price
would be $5 and there would be no hospital large or small of any type
which gets anything different than $5. It could not, if Robinson-
Patman were applied. This would be the price, just as $5 or whatever
the price may be to the retailer.
Now, they could have quantity discounts to hospitals, bift they
would be published quantity discounts. There would be a 25,000 price
PAGENO="0015"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1565
or 100,000 price on a tablet. This would be published in your list and
all hospitals would be allowed to buy at~this level.
Senator HATFIELD. You mean from all the drug companies within
say $5 for the specific product?
Mr. SQUIBB. No, they would establish a pricelist.
Senator HATFIELD. Individual price?
Mr. SQUIBB. Individual pricelist.
Senator HATFIELD. You would still have the competition, though?
Mr. SQUIBB. Oh, yes, you would have cross competition, because a
similar product from another company would be at approximately
the $5 level, but prices would tend to freeze at that level, that rigid
level, and do away with price competition on that product.
Senator HATFIELD. Well, are drug companies taking losses then?
Could I infer from your testimony that drug companies are taking
losses in selling to hospitals on the basis that they feel that they can
make up for those losses by the influence, through having their product
in the hospital, upon the physician who goes outside of the hospital
and engages in private practice?
Mr. SQUIBB. This answer to the question of taking losses in hospital
sales or Government sales at low prices is a very difficult one to estab-
lish. Obviously if they sold their entire line, their entire product mix or
even all of any one product at that price level, they might well take
losses. The accounting difficulty here in applying your costs to the
hospital part of your sales is fantastic. Within a given company you
would probably have to say that if, they sold their entire product
output at the hospital level, they would take a loss.
Senator HATFIELD. That they would ?
Mr. SQuIBB. They probably would frequently, because their whole
structure, selling structure and other structures, is predicated on their
ability to sell at a higher level to the retail pharmacy.
Senator HATFIELD. You have given me some generalities here, but
what happens in the specific case of Squibb?
Mr. SQUIBB. Squibb has a hospital pricelist.
Senator HATFIELD. Do you take any losses?
Mr. SQUIBB. You would be very hard put to it to say that we lost
at a given price to the hospital. I think the answer to your question
would be "No." But you would lose, if you put the burden of advertis-
ing and selling expenses at that leiel. You do not price out your
product return at each different area of sales. We would probably
sell the same product to the Government even cheaper than we sell
it to a hospital. There may be another lower price yet.
Senator HATFIELD. But, do you have any lines that if you sold
out at that price you would take a loss?
Mr. SQUIBB. Oh, yes; you have a overall cost figure for your product,
and you have an average price level which has to certainly cover your
basic cost. Plus your selling costs, research costs, other costs in other
brackets of your operating statement. Overall actually, when you
look generally at a product you sell, and I cover this later on, there
is an average selling price, a return to the company which is the sum
of all the sales at all the prices, including the retail price, the hospital
price level and the Government prices, and these factors vary pro-
portionately greatly from product to product in each category.
PAGENO="0016"
1566 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator HATFIELD. Can Squibb calculate a certain impact of having
their product in the hospital and affecting the attitude of physicians
who are using their product in the hospital and then what they hope
the physician will do outside of the hospital where there would be
a different price?
Mr. SQtTIBB. Yes; I think they can generally. We have seen this. We
know that if a certain product is used in a teaching hospital, where a
large proportion of the physicians of a community practice, that the
use of this product outside the hospital on private practice prescrip-
tions will be larger. We learn this through prescription analyses and
studies of the various market research prescriptio1i services. I think
this is a real factor in the success of a product.
Senator HATFIELD. This is pretty generally practiced among all the
pharmaceutical houses?
Mr. SQUIBB. Yes.
Senator HATFIELD. As well as Squibb?
Mr. SQUIBB. Yes; this is well known.
Senator HATFIELD. And then, actually you feel that part of that
price that you are giving to the hospital will be more than compen-
sated for by the influence it will have on the physicians outside of the
hospital practice?
Mr. SQUIBB. That is right, and certainly the younger physicians,
physicians in residence, and interns coming along who see in their
teaching hospital a certain product being used and are familiar with
it, who watch its reactions and its effect on patients for a number of
years, this is a product which they understand. They will naturally
pick it up later on in practice. This makes sense. We can prove it by
market research examination, and, therefore, this is a reason to justify
going to special lengths, pricewise, to get your product into this
hospital.
Senator HATFIELD. Do you feel that in the teaching hospital or in
academic pursuits of medical education there is an adequate academic
program in which the physicans, or the medical students and interns,
are being given appropriate education in order to make judgments on
pharmaceutical products for use?
Mr. SQUIBB. No; surely there is not. Now, I judge this just on per-
sonal opinion. I have not gone into the subject in specific detail but in
conversation with many hundreds of physicians and examination of
the testimony before the committee of the leading men that you have
had here, it is obvious that there is not adequate background in this
area. When I say there is not maybe there can never be, because of the
size of the problem and the demands of other areas.
Senator HATFIELD. So then, the pharmaceutical houses really sort of
depend upon this in some instances, or let us say, they perhaps exploit
it?
Mr. SQUIBB. They exploit it.
Senator HATFIELD. In such terms as getting their products into the
hospitals and teaching hospitals at a lesser cost, in order to influence
this intern or this young physician who has not been fully and ade-
quately prepared in this area through the academic programs, and so
it comes in primarily as a sales approach and a promotional approach,
rather than so much a matter of analyses or evaluations or
comparisons?
PAGENO="0017"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1567
Mr. SQUIBB. That is right. Actually the sales in this area are studied
and evaluated to establish whether this really does the thing that
we think it does. Prescription studies and market research studies
indicate that it does in fact.
Senator HATFIELD. But, you feel that once he is convinced that
Squibb is really a superior product he will then be using and prescrib-
ing Squibb products outside?
Mr. SQUIBB. That is right. Certainly the specific product that he
has used, prescribed and seen used in the hospital, and hopefully this
will spread even to others.
Senator HATFIELD. That is where you make your profit.
Mr. SQUIBB. That is where we make the bigger profit outside.
Mr. GoRDoN. Mr. Squibb, even then the products are not being sold
to the hospital at a loss, are they?
Mr. SQUIBB. No; I do not think so.
Mr. GORDON. They are making some money?
Mr. SQUIBB. Yes; they are not selling at a loss. The difficulty of ac-
counting here from company to company is a major problem. I do not
think products are being sold to hospitals at a loss.
Senator NELSON. I have the impression from what I have read and
from testimony before the committee, that the use of formularies by
hospitals and by cities in the purchase of drugs is a rapidly expanding
practice. Is that correct?
Mr. SQUIBB. Yes.
Senator NELSON. Is this a trend that has been relatively accelerated
in the past 10 years?
Mr. SQUIBB. Yes. I think the development of formularies by hospi-
tals has sort of been forced upon them by the problem of inventory,
by the problem of drug selection which has become so vast and so com-
plicated in the last 10 or 15 years with the emergence of so many new
potent drugs. There is a problem of selection. There is a problem of
inventory investment by the hospital which is under financial pres-
sure anyway. The problem of just carrying every type of product there
is just for an occasional use. There is a great deal of effort being spent
to bring some order out of this very difficult `situation by the develop-
ment of formularies. This has been since the war or even more recently
than that. I would say in the last 10 years formularies have come into
use in just about every hospital for this reason.
Senator NELSON. We have had testimony on formularies from a
number of pharmacologists, as well as representatives of some very fine
hospitals, such as Montefiore in New `York, Los Angeles County Gen-
eral Hospital, and a group health organization in the State of Wash-
ington. They have testified that they make their formulary available
for sale and use by doctors outside the hospital. In hospitals in smaller
communities, 25-bed, 50-bed, 75-bed hospitals, which may not have the
expertise to development of a good formulary of their own, is there
anything comparable available for use and if so, are these hospitals
using it.?
Mr. SQUIBB. Well, I think even the smallest hospitals are using for-
mularies. I think the American Association of Hospital Pharmacists
has developed a st.a.i~dard or a basic formulary which is suitable for hos-
pitals of any size, I think every hospital w-ould, as a matter of fact, have
S1-2S0-----GS----pt. 5-2
PAGENO="0018"
1568 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
a formulary. I think it is required now because of the rules of medicare
and reimbursement by the Government in that area, that the products
be on a formulary in order for it to be reimbursed by the Government.
So every hospital has a formulary.
I think that there is no problem now on the part of any hospital, in
developing a formulary, or borrowing somebody else's if they are so
small they have not the time or the know how to get one of their own
individually developed. I think most every hospital, whatever its na-
ture or size, a. 50 or 100 bed hospital, develops its own, on its own local
requirements, from. its own medical committee, pharmacists, doctors,
technicians working together. This is a very important development in
medicine right now, no question about it.
Senator NELSON. I do not even have a good guess as to what the ef-
fect will be, but it does seem to me that there is going to be a dramatic
effect in several ways on the prescribing of drugs and the price of drugs
when the time arrives that practically every hospital in America is
using a formu~ary, and when practically every hospital is taking com-
petitive bids, and when all the bids are requested on a generic basis.
Mr. SQUIBB. That is right.
Senator NELSON. I do not know what effect this is going to have, but
it seems to me, that the use of formularies will have some dramatic
effects in the field of prescribing of drugs by generic name, even though
the physician may identify the company. And it seems to me, when
formularies are widely available, and when the Government is more
deeply involved and concerned about costs because of medicare and
medicaid and when all the information is available to develop good
formularies, that some kind of dramatic changes are going to occur.
I do not know how to evaluate it, but I would like your comments on it.
Mr. SQUIBB. You have right in the rec.ord of this hearing the testi-
mnony of Dr. Williams, I believe his name is, from Grady Hospital of
Emory University in Atlanta, in which he describes for you his own
experience in developing a formulary there, and the extraordinary
results it has had on his drug bill; and not only on the drug bill,
the price paid for drugs themselves, but also on the way drugs were
prescribed in that particular hospital. I think this is an extraordinary
clear exposition of what happens when a large hospital, well run, with
high quality standards without any question, sets itself into the for-
rnulary problem. It is an extraordinarily clear example of how this
develops.
Senator NELSON. I would like your comments on this. I also meant
to add that I assume if the medical profession is to improve the way it
strives to, it will not be many years before almost every practicing
physician will necessarily be affiliated with a hospital. So when formu-
lanes are available, competitive bidding is the rule, and differentials
in prices are visible to the doctor who sees what the hospital pays and
then sees what his patients are paying in the retail market, there are
going to be some tremendous pressures developing not only within the
medical profession, but also within the hospitals, at the govermimental
level, and with the public itself for changes in pricing practices. Do you
think that is the case?
Mr. SQTJIBB. There is no question about it. This is exactly what the
pharmaceutical industry does not appear at this moment to fully ap-
PAGENO="0019"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1569
preciate what the effects will be of this development in this noncon-
trolled area, where there is price competition and where there will con-
tinue to be more and harder price competition as the years go on. This
will have a major effect on the phariñaceutical industry, its pricing
policies and the operational practices that it can afford to pay for and
that even that are advisable to pay for. This is the major changing
trend in the medicine business, to use the word generally, in the United
States. You have put your finger on it right there.
Senator NELSON. And, I take it from your testimony, that you feel
that the industry is not creatively undertaking to resolve this prob-
lem at this time?
Mr. SQUIBB. No, no they are not, because they are riding along with
their profits and their procedures geared to their major area of opera-
tions, which is at present the return from the retail pharmacist at a
rigid price level. This has given them in the past all they needed and
more than they needed, and although there are rumblings and, of
course, discussions, they do not seem to see and certainly they do not
seem to see in their publicly issued statements, the problem that I see
and perhaps others see in this area, the problem for them which is
developing.
Senator HATFIELD. Mr. Chairman. Speaking of these formularies,
Mr. Squibb, recognizing that these people who are going to be gov-
erning the establishment of formularies in hospitals around the coun-
try, are some of the same people if not exactly the same people, who
have been siThj ected to your industry's influence through previous con-
tacts, can you tell me what kind of objective criteria will be used to
e~t~bli~h these formularies? Are they not going to be actually estab-
lished by the same people whom we are talking about now? Are they
going to somehow; in some transformation, come into a new under-
standing of these competitive drugs? How are we going to know that
we have some sort of objective criteria upon which these formularies
are based, if we are dealing with primarily the same people we are
falking about today?
Mr. SQUIBB. Of course, you are and they will be making th~ir selec-
tions. For example, the Grady Hospital, as spelled out in your hearings
here by Dr. Williams, has to use a diuretic. They do not want to buy
large quantities of all 15 or 20 diuretic tablets that are on the market.
They have to have one. Their committee, which is made up of physi-
cians in private practice in Atlanta, and in hospital practice at Grady,
makes a selection, an arbitrary selection if you will, after due consid-
eration. How do they do it?
Senator HATFIELD. Arbitrary?
Mr. SQUIBB. Well, they have to decide on one out of 15. It is arbi-
trary. I say it is arbitrary because when it gets right down to it they
have to cross out 14 and leave one. Now, what goes on in their minds to
pick the one that they leave there is, of course, interesting. Certainly,
the decision is subject to the background they have had in experience
with the drug in its advertising and presentation to them. We find in
different formularies around the country in different states, different
diuretics, to just take one product, for example. Different products are
selected in different areas. In other areas this Grady Hospital choice
is not selected.
PAGENO="0020"
1570 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator HATFIELD. So, there is no real objective criteria that is un-
related to the individual physician's own background, influence, ex-
perience, exposures and pressures?
Mr. SQUIBB. No. I do not think you could ever have a physician ar-
rive at a position of prominence without having some kind of a back-
ground in which he has been subjected to varying degrees of what you
are talking about.
Senator HATFIELD. On this matter of prices which you said might
become one of the most important of all criteria, what is the differential
between the percentages of profit that the industry makes between
your three major groups to whom you distribute, the hospitals, the
retail trade, and the wholesale trade? You say here that the over-
whelming percentage of profit of the industry comes from the retail
trade. What is that overwhelming percentage in relation to these two
other maj or distribution groups?
Mr. SQUIBB. I am deducing that from background rather than from
any specific figure, because still the private prescription is the major
source of outgo of the industry's products. Most of these prescriptions
still go through the retail pharmacy, so they are at the high price levels.
Therefore, just from deduction most of the profit comes from there.
Senator HATFIELD. 70 percent, 60 percent?
Mr. SQUIBB. Oh, I would say 70 percent.
Senator HATFIELD. 70 percent, or 80 percent?
Mr. SQUIBB. It varies for certain products. There are some products
that are hospital products, of course, but I would say basically it is
75 percent.
Senator HATFIELD. How does it divide as between wholesale and
hospitals?
Mr. SQUIBB. Well, I am putting the wholesaler into the retailer
activity, because the wholesaler is merely a middleman with a corn-
mission in the same rigid market.
Senator HATFIELD. But, can you divide that for me between the
retailer and wholesaler?
Mr. SQUIBB. Well, it depends on the company. Some companies do
not have any wholesaler business at all. Others have nothing but
wholesale.
Senator HATFIELD. How about Squibb?
Mr. SQUIBB. Squibb has some wholesale business. Squibb splits, say,
around 30 percent wholesale business.
Senator HATFIELD. So, you could then take a ratio of that 30 percent
within the 70? And in the 30 percent of the total overall volume of
profit will be from hospitals.
Mr. SQUIBB. From hospitals, although some of the wholesale business,
if you have wholesale business, goes to hospitals, although most hos-
pitals prefer to buy direct.
Senator HATFIELD. Most do?
Mr. SQUIBB. Most do. They prefer it, if they can.
Senator HATFIELD. So, the overwhelming percentage of profit then
would be wholesaler and retailer combined and would be about 70
percent total profit?
Mr. SQuIBB. I would think so, or more.
Senator HATFIELD. Or more?
PAGENO="0021"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1571
Mr. SQUIBB. Or more.
Senator HATFIELD. Do you know specifically what the Squibb per-
centages are?
Mr. SQUIBB. No; I do not know. It is not really a matter of interest
in that category, although it should be. I think if they studied it
more specifically, particularly the growing effect of their hospital
business, they would see that the business is growing here, and that
they had better watch out and prepare ourselves more firmly pricewise
and procedural wise for this changing trend.
Senator HATFIELD. What is the trend of profits from drug sales to
hospitals today, up or down.
Mr. SQUIBB. Up, mostly going up because the volume is going up.
They have to be going up. The volume is going up there very sharply.
Senator HATFIELD. I am speaking now of profits from the sales to
hospitals from the drug companies.
Mr. SQUIBB. As the volume goes up the profits would go up.
Senator NELSON. As concerns the formulary, I would just like to
read into the record the comments of Dr. Williams from Grady Hos-
pital, when he testified before the committee about some of the very
kind of problems that Senator Hatfield has raised. He says:
It was not easy in the beginning. As you may imagine, the medical profession
is conservative. This-
lie is referring to the adoption of the formulary-
this was a radical departure of performance at Grady Hospital. There were com-
plaints that the committee was trying to dictate the type of medicine practiced
at the hospital, that Grady Hospital patients would be poisoned by cheap inferior
drugs, that the change from one color pill to another would upset the Grady
Hospital patients in an irremedial manner. That we should buy trade name
expensive items to support research done by the large drug companies, and that
the committee was attacking the American free enterprise system. We persisted
with the support from a large part of the faculty and the total support of the
administration of the hospital who were interested in cutting this enormous
drug bill. We have been fortunate. Estimated savings during the first year ran
as high as $150,000 on a `budget of $480,000 for drug purchases.
Now, on the question of how a formulary is developed, it is correct,
is it not, that in a good hospital with all the specialties represented
the therapeutics committee is composed of the urologist and the heart
specialist and the various internists and pharmacologists and pharma-
cists, and that based upon experience from medical literature, upon
their own experience, they come to a conclusion about what ought to be
included in the formulary. When they concluded from clinical evidence
that several drugs of the same compound appear to be of equivalent
clinical value, then the pressure, of course, is to select the one that
is the most clinical. Is not that about the way a formulary is developed?
Mr. SQUIBB. I think that would certainly enter into the picture.
`F he pharmacist is generally charged with the responsibility of bring-
ing in the price background of the different products, and at that
juncture, price would certainly be a determining factor, all other
things being equal.
Senator HATFIELD. But would you not agree, Mr. Squibb, that it
is not purely a clinical basis upon which they establish tha formulary?
It is, as the chairman has very clearly pointed out, based upon the
experience and the involvements of that physician-
PAGENO="0022"
1572 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. SQUIBB. Oh, yes.
Senator HATFIELD (continuing). Serving as a member of that
committee?
Mr. SQUIBB. That is right.
Senator HATFIELD. And his background having been influenced by
your industry?
Mr. SQUIBB. That is right, certainly. We even see the committee of
physicians in a given hospital take turns, rotate in this responsibility
over the years, and as a result you see the products on a formulary
changed just as the physicians change in this committee. This is
evidence of what you are talking about.
I think we have probably covered the pharmaceutical pricing back-
ground. I would like to raise or point out something here which I
think has not gotten into the testimony yet too clearly, and that is
on page 8, the top paragraph there. It is significant to note that the
industry's pricing policies and practices have not only brought on
sharp criticism from those in academic, press, and legislative circles
outside the pharmacy and medical professions, but also from those
within the professions. The drug press, pharmacy convention pro-
ceedings, records of legislative hearings, all are full of complaints
of discriminatory pricing on the part of manufacturers from retail
pharmacists, wholesalers, their trade associations; from puzzled
physicians, purchasing agents, and institutional personnel both pri-
vate and govermnent-employed who are faced with responsibilities
to maintain and dispense inventories of drugs. Friend and foe alike
of the industry point with all sorts of degrees of alarm, shame, clis-
gust, distrust, and perplexity at what is certainly a situation at best
hard to justify, and at worst completely unreasonable, chaotic as well
as shortsighted and stupid.
The criticism of drug pricing is even more bitter within the in-
clustry, within its customer framework, than it is in the full light of
testimony in hearings such as this. Industry is struggling with this,
but it does not seem to come to grips with it.
Now, let us see what can be done about pharmaceutical pricing, the
public conception of which threatens a whole industry, and which
couldi lead if not properly handled, to irretrievable steps of retribution
andi regulation which very wEll might have serious implications for the
whole pattern of medical care in the United States-and not neces-
sarily for its improvement. It is important that public confidence in
the fairness of drug prices be restored, and that in so doing the future
of an important industry's contribution to medical progress not be
limited or proscribed to absolutely no one's benefit.
First, how can the industry by itself revise its price structure and
restore some sanity into its marketing programs? It must be im-
mediately realized that the concern here is with prices, and this being
so, a concerted, cooperative approach by pharmaceutical companies is
impossible, antitrust and conspiracy concepts being what they are.
Actions must be taken audi solutions found, company by individual
company working within its own framework of costs, prices, and
profits-not to mention its own competitive standing in a bitterly
contested market.
In the first place and by far the most important of all, pharmaceu-
tical management must understandi the nature of the social climate
PAGENO="0023"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1573
relative to medical care in which it now lives and works. This is not the
climate of 10 years ago, or even of last year. It is a moving, expanding,
developing concept of prepaid or guaranteed protection from the fi-
nancial impact of illness as a basic personal requirement with the
added growing idea of a real, affirmative responsibility resting in the
public health authorities-State and Federal-for establishing and
maintaining conditions leading to optimum health for all. There is no
longer any question that it is the duty of government to protect the
public from itself as well as to provide for the means to insure sound
health in every physical and 1nental aspect of individual well-being.
The public wants the government, State and National, to pay for most
of this, and for what it does not actually pay for, the public expects its
governments to assume roles of stern regulators, overseers, and active
supervisors.
It is precisely because the public looks for and even demands these
services and responsibilities from its elected officials that all of those
privately engaged in medical care in any form must exercise extreme
caution to conform their activities-particularly financial-to what
are considered normal, fair, and reasonable-such judgments to be
made and construed within a very narrow range of acceptable conduct
and its results. Definitely the pharmaceutical industry must accept
special responsibility for its operations, and special burdens and limi-
tations not regularly assumed by other industries because of its place
in the scheme of medical care now an object of particular public con-
cern. It must adapt itself at the very least to the standards of normal
economic and industrial practices and profits, and be able to live with
them and at the same time continue to make the contributions and
perform `the services that it has pointed to so proudly in the past.
Probably it must go even further than this, and accept the idea that its
profits should be less than those of average comparable industries in
recognition of this special social responsibility which is being firmly
and irrevocably forced upon it by the general public.
Senator HATFIELD. Mr. Squibb, would you identify for me the aver-
age comparable industries? You are speaking here of the social respon-
sibility of the drug industry?
Mr. SQUIBB. Yes.
Senator HATFIELD. And, the relation of that to its profits, that they
should be less than those of average comparable industries?
Mr. SQUIBB. In `the testimony that has been given before the com-
mittee, the industry has been compared in its innovative fUnction with
industries such as office equipment and other electronic industries on
the one hand and with the textile industries and the food industries
on the other hand, which are rigid or noninnovative industries. What 1
am really saying here is that you have to take an average return on
investment by the representative industries that cut across a broad seg-
ment of public life, not by special industries that do not affect a mass
of people.
Senator HATFIELD. You mean like food?
Mr. SQUIBB. Like food, definitely food.
Senator HATFIELD. Clothing.
Mr. SQUIBB. And clothing, that is right, that type of an industry
that everybody or most everybody in the country contributes to by
some sort of purchase.
PAGENO="0024"
1574 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator HATFIELD. Tobacco?
Mr. SQUIBB. Tobacco, yes. I should think tobacco is now an essential,
that is right.
Senator HATFIELD. Then you would apply, if I understand you
correctly, a double standard in a sense, to the drug industry and to
these so-called comparable industries as it relates to their profits?
Mr. SQUIBB. Yes, a double standard for the pharmaceutical industry,
because of its nature, a special standard.
Senator HATFIELD. I see. And in relation to this whole matter of
picking out these so-called comparable industries, and, of course,
they would no longer be comparable, would-
Mr. SQUIBB. No. What I am saying when I talk about comparable,
I mean an industry whose products are widely used by the consuming
public as are pharmaceuticals.
Senator HATFIELD. This is because of the special and unique social
responsibilities I understand that you feel the drug industry has.
Mr. SQUIBB. Tha.t is exactly it.
Senator HATFIELD. And as it relates to this protecting of the public
against itself, you have led into this particular point, what about
the tobacco industry? As a former member of the National Cancer
Society board and as one who is familiar somewhat with these reports
made by our Surgeon General, do you not agree that it does threaten
the health and well-being of great numbers of American people?
Mr. SQUIBB. Well, never having smoked a. cigarette in my life and
not a participant in, or supporter of, the tobacco industry I just
know what I read in the papers on this. I think that tobacco, how-
ever, in spite of those statements, is a necessary fact of life to millions
of people, and as such is a very important consumer industry at the
moment. It may be hurting people and it may not be, I do not know,
but-
Senator HATFIELD. But there is a social responsibility, as you point
out here, on the part of the drug industry based upon its relation
to the health of the people.
Mr. SQUIBB. Excuse me, I see what you mean. Yes, if a product
made by the tobacco industry, or a product made by any industry,
even an automobile has its damaging functions. Any one who makes
an instrument that can damage people has some degree of social
responsibility, as has been recently expressed. But I think that the
drug business far transcends any of these, either tobacco, or any other,
because it is given at a time for purposes which are unique and
desperately involve life itself in some cases.
Senator HATFIELD. Even though we have statistical evidence a~bout
lun~ cancer in relation to cigarette smoking?
Mr. SQUIBB. Yes. I am not too well informed about that, but even
e'oing along with you, although I really do not know about the harm-
ful nature of tobacco, but if it is in fact harmful, then the industry
has a very heavy social responsibility.
Senator HATFIELD. How do we protect the public against itself
when we determine there is a sOcial responsibility by a particular
industry that is not being upheld, merely regulating profits?
~~:rr. SQUIBB. In a lot more than profits.
Senator HATFIELD. What would you suggest further?
PAGENO="0025"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1575
Mr. SQUIBB. Well, you have to regulate the whole performance
of an industry that is engaged in this type of activity; quality, its prac-
tices of distribution, its practices of promotion, presentation of its
products.
Senator NELSON. Theoretically, `as to quality, with a sufficient num-
ber of personnel to inspect, the present required standards are de-
signed to insure quality, although these standards are not always
complied with as we know from drug recalls and various other things.
But there is legally at least a quality standard established for this
industry. This is true maybe of some other industries, too.
Mr. SQUIBB. Oh, yes. Every industry has its quality standards.
Senator NELSON. I take it from what you were saying about profits
that you are simply setting some limits. You think the pricing prac-
tices as now exist ought to be reformed, and as you have stated, at
the very least, the standards for pricing should :be set to accord with
the standards of normal economic and industrial pricing practices
and profits.
Your other sentence is: "Probably it must go further."
Mr. SQUIBB. Yes.
Senator NELSON. I am assuming what you are saying is that the
industry at least must conform to noEmal practices and might even
have to go further than that in some cases.
Mr. SQUIBB. Perhaps go further than that, because it is so urgently
under the eye of people who are concerned with health care problems
today.
Senator NELSON. As to your comment about social responsibility, I
think it is interesting to read a statement from page 139 of the re-
port of the Committee on the Judiciary of the U.S. Senate, June 27,
1961:
In industry generally, the views expressed by Jefferson and Franklin with
respect to patent monopolies appeared to fade rapidly, but for a time they con-
tinued to prevail in the drug industry, reflecting a recognition of the peculiar
and unique relationship of this industry to the public health. In 1854 when
Dr. E. R. Squibb, founder of E. R. Squibb & Son, managed to distill for the first
time pure ether of uniform strength, he declined to take out patents. Instead
lie published his discovery in September of 1856 issue of the American Journal
of Pharmacy. The essential difference between most other countries and the
United States on this matter is that the views held by Dr. Squibb have continued
to prevail abroad but have long since been abandoned here.
But at least it was a matter of discussion and concern over 100 years
ago when your great grandfather founded your distinguished company.
Mr. SQUIBB. That is right, and I think it always will be. It always
will be from the very nature of the problem we are dealing with.
I would like to continue on page 11 and point out this area here
which I think adds a little bit to what we are talking about right now.
A pharmaceutical company management-now enjoying, let us say,
a 10-percent return on sales after taxes-sets its objective for the next
year at 12 percent-and then 15 and even 18 percent over a 5- or 10-year
plan. Often in the past this type of profit objective has been obtained
through new products and intensive marketing programs. Recently,
such growth has been more exceptional lit actuality, although it still
is quite possible. However, pharmaceutical firms still project and plan
for steady growth in profits, as does most industry. These pressures
PAGENO="0026"
1576 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
for better profits, quite normal and accepted elsewhere as good busi-
ness practice, take on peculiar overtones when sales of essential health
care products are involved. The consent of more and more profits from
the miseries of the sick, the aged, and the malnourished segment of the
population seems to run counter to the swelling trend toward State-
supported medicine. If the Government takes over the responsibility of
ameliorating or improving the health of a large part of the Nation,
this should not result in enriching beyond normal levels at the very
most those who participate in the job. Expenditures for medicine by
tax-supported Government agencies bring with them strong pressures,
both sociological and economic, for the avoidance of profiteering, waste,
and misappropriation. The latter two pitfalls are the responsibility of
those who administer the program-the first is the prime responsibility
of the supplier, in this instance the pharmaceutical manufacturer.
Until and unless he sees this clearly, the threat of Government-im-
posed profit regulations will always be imminent. Profit objectives
must be made subordinate to productivity objectives.
Senator NELSON. Some of the testimony we have had from the indus-
try, and other statements made by members of the industry and the
medical profession have asserted that in order `to attract capital for the
purposes of expansion and for research, these higher profit rates are
necessary because of the great risk and so forth. Do you have any
comment. about that?
Mr. SQuIBB. I commented on risk in my earlier paper. I do not feel
that the pharmaceutical industry itself, under the definition of the
word "risk," as I understand it, is a risky business. It is a profitable
business, one in which most of the large companies have been long en-
gaged a.nd have many years of profit history behind them. While it is
true that a given product in their structure may come and go and may
often be relatively short-lived, their financial risk, it seems to me, in my
experience, has been very little in the pharmaceutical business.
A well-managed pharmaceutical company is as good an investment1
as good a place to put your hard-earned dollars, as safe a spot as any I
know of. I do not `think `that there is a risk involved here in terms of
cvha't is commonly `thought of `as a risky speculative sort of a business.
The drug business is not.
Senator NELSON. I have one other question I intended to ask some
time back. Although it is not directly related to the context of your
present statement, I would like to ask it here.
The committee staff has gathered statistics on prices of `a number `of
drug products manufactured in this country and sold in foreign coun-
tries. We have found by checking through the State Department and
the embassies in a number `of European countries as well `as Latin
America. and others that it is quite common for `the price charged in
the marketplace here to be anywhere from 50 percent higher to twice
as high or even 10 times `as high as a price charged by `the same corn-
pany for the same product in Paris, London, Berne, `Switzerland,
Rome and &sewhere.
I have not been able `to get an explanation that `is satisfactory to me
as to why that clifferential-why the manufacturer of `a drug here,
ships it overseas, and charges a lower price `abroad. What justification
is `there for charging a lower price to people in foreign countries than
char~ed to our own nationals?
PAGENO="0027"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1577
Mr. SQUIBB. I think it is very simple. The justification for this
lower price you are referring to is simply the fact that that is the price,
lower or higher or whatever it may be with which the manufarturer
can get into that particular market. He could not sell the product at
the price that he is selling it here. He could not sell it in adequate
quantities. He could not make an entry in the market at this level.
So, I think that the pric.e that is established in any given country is
the price that the market there will bear, and return him some ade-
quate sales. That is the judgment decision that is made when a product
is marketed in any country. It is as simple as that. I do not think
there is any other explanation. There cannot be any other explanation.
Senator NELSON. Is there a.ny difference between the marketing and
the advertising of drugs, in foreign countries and here? For example,
if doctors are convinced in Switzerland that this very fine miracle drug
called by a certain brand name is going to help their patients, why
cannot the firm establish the same high price that brand names com-
mand in our own market even where the same drug is available for
far less?
Mr. SQUIBB. I think they probably could if they wanted to put the
promotion money, the selling and advertising effort in establishing a
trade name in Switzerland, to use your example. But the money is
not there, the time is not there, the people are not there to do that, and
the overall market, the Swiss market or the French market or what-
ever market you are talking about, the total market is probably not
large enough to justify a great deal of~ money and effort in this direc-
tion.
Senator NELSON. Is there any difference in the laws applicable to
the advertising of drugs in foreign countries? Or, is there any differ-
ence in the traditional practice of the medical profession in those
countries that might account for this differential?
Mr. SQUIBB. I think this is part of it. Promotion and advertising
in some countries is limited by law and, restricted as to distribution
of samples, distribution of free materials. The media, that is the
medical journal media that are available differ widely. The actual
procedures of medicine in terms of who buys the product and who
`ays for it differs from country to country, and, of course, basically
the size of the market itself differs. Some of the smaller countries
have less sophisticated developments in this area. But, basically the
price level is predetermined. If you decide to go into the Algerian
market, you look at the price you are going to have to charge for
your product in that market to get any sales at all, and that is the price
you set, irrespective of what your priëe may be here.
Senator NELSON. Is there any question in your mind as to whether
the firms make a profit in these foreign countries, where they charge
one-third, one-fourth, one-fifth a.s much as they do in the marketplace
here?
Mr. SQUIBB. No, because of w-hat we have already talked about, the
basic cost of the product is still low in comparison with the price. So,
it gives you a wide range of opportunity here.
Mr. GORDON. Mr. Squibb, aren't you really saying that prices are low
overseas because of competition, and that they are high here because of
the lack of competition?
ii\'Ir. SQUIBB. No.
PAGENO="0028"
1578 COMPETITIVE PROBLEMS IX THE DRUG INDUSTRY
Mr. GORDoN. You are not saying that?
Mr. SQUIBB. No, I am not saying that. It could be or could not be. It
might or might not be the case, but I am definitely not saying that they
are low because of competition. There may be competition in terms of
what they are thinking about in a given foreign country, and there is
certainly competition here. Competition is engendered at the high level
by trademark promotion and competing for the doctor's attention,
which we are doing here, and spending money to do it.
Mr. GORDON. I am talking about. price competition now.
Mr. SQUIBB. You can set your price high if you put on your strong
sales promotion, and produce large volume.
Mr. GORDON. But, that kills competition. Why do you have sales pro-
motion? In order to try to differentiate your product, to show that it
is a different product from anyone else's.
Mr. SQUIBB. That is right. I think we are talking about competition
in a different framework. You are right. What you are talking about,
our efforts here is to kill competition by taking all of the market your-
self. That is true.
Mr. GORDON. Now, coming back to the risk business, how many PMA
members have gone out of business in the last 10 or 15 years? Do you
know of any?
Mr. SQUDiB. I do not know of any.
Mr. GORDON. Can you tell us how many times during, say the last 10
years, have the profits of the large PMA members fallen below the
average for the. industry as a whole?
Mr. SQUIBB. Well, the studies I have seen in this connection have
always indicated that the pharmaceutical industry represented by
these top producers has been higher than average, and among the top
one or two industry groups. They have always been in the last 10 years
well above what might be considered average or normal return.
Mr. GoRDoN. Can you tell us how the drugindustry might be affected
by cyclical factors, say depressions?
Mr. SQUIBB. Well, certainly in the past, the drug industry has only
been slightly affected by this, which can be expected because purchases
for medicines are probably the last thing that you cut out when you are
hard up. Also in good times you do not go out, rush around and buy a
lot of unneeded medicines. But I think that there is going to be even
less such effects in the future as the Government steps in buying more
and more medicine and reimbursing more and more of the medical
bill. Psychological forces are going to have even a lesser effect. They
will have very little effect.
Mr. GORDON. Now, as to product obsolescence, are there many prod-
ucts which have been losses to the drug firms? What is the extent of
the losses?
Mr. SQUIBB. Well, there are many products which are on a down-
ward sales curve, and there are many products at a given moment on
any given sales list, I believe, which are losers in themselves and by
themselves. They have been winners in the past and now they are
losers.
Mr. GORDON. Why do they retain them?
Mr. SQUIBB. Well, I think that one of your witnesses touched on this
very well. He said sales departments find it hard to let these things go.
PAGENO="0029"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1579
I think that they do tend to hang on because there are a few physicians
who still want to use them and they want to keep, those physician's
goodwill. Eventually as these losers grow to a bigger part of the
pharmaceutical companies list they are cut out and disappear. Once in
a while there is a big washing out of losers.
Mr. GORDON. How great is this? What is its extent?
Mr. SQUIBB. Every year it is a continuing problem, but it cannot
be great in terms of percentage of the total, because the loss is not great
enough to damage your profit picture. Any company that lets its losers
get above the 5-percent sales level is in certainly bad management
trouble.
Senator NELSON. If there is no objection on the part of the witness,
we will take a leisurely lunch and come back in 30 minutes.
Mr. SQUIBB. All right.
Senator NELSON. We may have some voting this afternoon. I would
rather break now and start again in 30 minutes if you do not have any
objection.
Mr. SQUIBB. None at all. Thank you, Senator.
(Whereupon, at 11 :45 a.m., the hearing was recessed, to reconvene
at 12:15 p.m., this day.)
AFTERNOON SESSION
Senator NELSON. The hearing of the subcommittee will resume.
Mr. Squibb, I believe you were on about page 11, is that correct?
STATEMENT OF GEORG~E S. SQUIBB, FORMER VICE PRESIDENT,
E. R. SQUIBB & SONS-Resumed
Mr. SQUIBB. I was going to pick up at the top of page 12, Senator.
Senator NELSON. Go ahead.
Mr. SQUIBB. As another necessary step to pricing policy reform-
the price of the pharmaceutical product should become a function of its
cost-a function approximately the magnitude of those in major con-
sumer goods industries and varying only in terms of the sales. burden
imposed. Prices which are set at nine or 12 or 20 times factory cost only
serve as incentives for public attack on the various means and methods
that permit such action, patent protection, product monopoly, success-
ful sales pressures, or whatever. Exploitation of the value factor of
medicines used in life preserving and lifesaving situations, by setting
prices far above the cost is what must be deliberately and conscien-
tiously avoided, no matter what justification or economic temptation is
felt by the manufacturer. For example, if a new drug will largely
depopulate State mental institutions by permitting home care of the
patients, all of the extraordinary economic gain from this product
must be felt by the State, and none by the pharmaceutical company
that supplies it. Ordinary profits, yes, but windfalls, no. This may come
as a shocking idea to those who set the prices on major volume pharma-
ceutical products which are used for treatment of patients in tax-
supported situations but it is an idea which must be accepted, or it will
be imposed by regulation. As the scope of purchase of medicine by
government widens, the idea has greater implication for an ever larger
part of drug product distribution and sale.
PAGENO="0030"
1580 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
A product that does the job that modern drugs often do, affords
under any comparative value approach, a most unusual temptation for
the pricer to set his figure much higher than might otherwise be the
case. Classic theories of price established to get a share of a competitive
market are applicable only to a very slight degree. In fact, it often
might be said that each drug product can be established in its own
market by skillful promotion and exploitation of its own particular
virtues. Not only has the pharmaceutical industry been successful
in maintaining the conviction with many physicians and buyers that
not all drugs are alike, but it has even succeeded in persuading them
that all products are different, which is a much more effective argument
from a sales point of view. Leaving aside at this point the validity of
this claim, the mere fact that it has been frequently and effectively
established, and continues to be, even under the conditions of current
ôontroversy and attack, gives the pricer a unique opportunity to set
his figure without relation to any factor except what he believes the
market can bear. He, of course, will take into consideration, in a ge1~erah
way, the existence and success of comparable products, or products
used for the same therapeutic or diagnostic purposes, but not by defini-
tion any product exactly like the one to be priced because there is none.
It is well to remember here that differences for pricing purposes arise
iiot only from varying chemical or molecular. structures, but from
differing manufacturing sources or brand names and sales programs
as well.
Prices set in this way on prescription products lead to the very
situation that is so often criticized today; an enormous range between
trade or retail prices and those given to institutional or government
purchasers. The rigidities of prices at the trade level already de-
scribed tend to freeze the prices a.t the top level for at least a con-
siderable pei~iod of time during which the disruption of the institu-
ti onal pricing structure occurs, and the perplexing and dismaying
situation now deplored by retail pharmacists, legislators, and the
general public arise.
The precise and detailed approach that a pharmaceutical pricer
should use to avoid falling into the trap of an eventual "no-price"
for his product presents a complex and difficult problem. It amounts
to the better utilization of forecasting, and market research tools-
including computerization of much basic information-to establish
in advance the market utilization of each product and package of
product, and the pricing prerequisites of each market segment con-
tributing to the product's sales volume. Varying competitive forces
and their effect on prices in each market must be recognized and
evaluated in advance so determination of product dollar productivity
can be measured in realistic terms. By this means, formulas for pric-
ing can be set that do not lead in fact and theory-by actual use or by
company policy-to anything more than a reasonable price range for
a given product. It must be stressed that in setting up any such ap-
proach to pricing that due attention must be given to the new hard
facts of prescription product distribution-facts which the outside
world is imposing on an industry which is reluctant to believe and
acknowledge them and even more reluctant to change their policies
accordingly. Some such facts are:
PAGENO="0031"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1581
(1) That hospital and institutional volume of drug usage will be
substantially up in the immediate future. Some of this volume is now
due to more patient coverage through government and private insur-
ance, but some of it will come out of the retail pharmacy as more
niedicines are dispensed in association with hospital and other msti-
tutional treatment. As a result, hospital bulk packages of drugs will
now be used on outpatient prescription to a substantial degree in~
direct price competition with the retail pharmacist.
(2) That the governments, State and Federal, which will now pay
much of the Nation's drug bill, are going to insist on paying the gov-
ernment price for its purchases-not the much higher retail price.
(3) That the governments will see no reason to buy their drug
requirements. any way except direct from the manufacturer, thus
eliminating wholesalers, commission agents, and other middlemen
from this part of the market.
(4) The use of various means to permit substitution of brands for
those specified, or to require prescribing in generic terms, will con-
tinue to be urged by many influential persons and professional groups.
This will weaken market positions and set up new price pressures for
many products vulnerable to these trends. Facing up to this fact alone
is a difficult job and a major challenge for the pharmaceutical pricer.
Any new or revised approach to this job by the pharmaceutical
pricer will, of course, bring a substantially altered pattern of income
to his company. For this reason, this concept cannot be put into effect
with immediate or all-inclusive coverage of any company's products.
But, it can and should be adopted in new pricing, and also wherever
changing circumstances permit on existing products on which there
is need for one reason or another of pricing revision. Gradually this
pricing policy should be implemented to bring Order from chaos, and
public understanding and approval from bewilderment and hostility.
Before closing this discussion of the pharmaceutical manufacturer's
price schedule, two relatively minor aspects should be touched on-
first, the so-called service products and, second, the use of special deals,
temporary discounts, and other promotional activities that affect price.
This argument is made frequently by pharmaceutical companies
that prices are maintained high on some volume products in order to
permit the marketing of items which return no volume or profit at
all, but provide a needed service for physicians in treating rare ill-
nesses or for use in very specialized circumstances. Marketing of prod-
ucts at a loss does exist, and to the extent that money is found for these
products from the earnings of others, the argument appears sound.
However, the actual extent of such support is limited. If volume of
the service items is small, both the cost involved and the out-of-pocket
loss is small in terms of earnings elsewhere. Overall profits show this.
Senator NELSON. Mr. Squibb, may I interrupt a moment ~ That
buzzer is a rollcall. I have to go to the Senate. I will be about 10
minutes. We can recess or you could continue to read your statement.
I have some questions that I will want to get back to, but we can
proceed in whichever way you prefer.
Mr. SQUIBB. I can wait for your return, if you wish.
Senator NELSON. All right, I willbe back in 10 minutes.
(A short recess was taken.)
PAGENO="0032"
1582 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. Mr. Squibb, would you mind interrupting your
presentation? I would like to ask you about the press release of the
Pharmaceutical Manufacturers Association commenting on your
statement. Have you had a cha.nce to read it?
Mr. SQUIBB. Yes; I just received a copy a few minutes ago.
Senator NELSON. Do you have any comment you would like to make
about the release?
Mr. SQUIBB. I think I look at this more in sorrow than in anger.
Senator Nr~soN. I did not understand that.
Mr. SQUIBB. I say I am looking at this more in sorrow than in
anger. This, I think, just points up what I have said today. I certainly
have not commented anywhere that I know of that the prices have
not in fact declined. They have declined and they are still declining.
That does not change the burden of the argument that the industry
is thought to be still high priced. Of course, they have declined. Over
on the second page, I certainly do not derogate research. Quite to
the contrary. This points up the fact which I bring out later, that it
is very difficult to talk rationally about research in the industry with-
out somebody getting a little bit overexcited about the emotional im-
pact of killer diseases and so forth. And I know of nowhere in the
statement where I even mentioned the words "they are cutting equiva-
lency of drugs." I do not even discuss this at all. This seems to me
to be sort of setting up a straw man and then blowing it down with a
lot of wind but it does not really cover the points that I have been
trying to make here.
Senator NELSON. On the second page it says: "As to the value of the
research which Mr. Squibb derogates." In your statement you did not
derogate research.
Mr. SQUIBB. No, I did not derogate it at all. I indicated there might
possibly be within the scope of industry research some which is not
wholly productive or wholly important. Just this one slight suspicion
or remark ca.st on research brings forth this extraordinarily strong
answer which is the point I am making. It is very hard to stand off
and objectively talk about research. It is almost a holy man in the
industry. I think the research should be examined factually without
derogation, without criticism, like every other part of the business
of which it is an important part, one of many parts.
Senator NELSON. I think that the statement rather dramatically
emphasized the comments in the opening of your statement in which
you say that the only answer that the pharmaceutical industry has
given to date is to, "tell its story," to "improve its image", to "explain
its contributions to medical knowledge and the health of the people."
I think their statement dramatizes the cogency of your observation
about their kind of response to any suggestions or criticism.
Mr. SQUIBB. It says I speak for myself. Of course, I speak for my-
self. I only hope some day that the words I am saying here will be
adopted by t.he industry. That is my wish.
Mr. GROSSMAN. Mr. Squibb, I wonder if I might ask you in the same
area, do you feel that there is room for dissent in the industry and
that there are others who feel as you do and perhaps do not come
before us for any reason?
Mr. SQUIBB. Oh, surely.
PAGENO="0033"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1583
Mr. GROSSMAN. You do feel that?
Mr. SQUIBB. I am certain of it, oh, yes.
Mr. GROSSMAN. Do you feel that there are any undue pressures placed
on any of these people?
Mr. SQUIBB. No, I do not think there are undue pressures. It is a mat-
ter of judgment, even accepting the accuracies or the truth and the
basic reliability of the opinions I am expressing whether it is advisable
in tactic to express them at this time.
Mr. GROSSMAN. Are any of the dissenters in positions of leadership
in their companies?
Mr. SQUIBB. I think so, yes.
Mr. GROSSMAN. Why have not any of these dissenters who are leaders
taken-
Mr. SQUIBB. Because I think they feel that it does not prove any-
thing or do any good in developing the idea that they have to expose
them in any type of a public hearing.
Mr. GROSSMAN. For example, even without a public hearing, why,
for example, have not they done some of the things that you suggest?
Mr. SQUIBB. I think there are discussions and things underway in
certain companies all along the lines that I suggest.
Mr. GROSSMAN. Yet, there has not been much action.
Mr. SQUIBB. There has not been much action yet. The pressures are
mounting now and they are sharper than they have been for a large
number of years or ever. I `think the pressures are right now on the
industry for these reforms for obvious reasons. They are very heavy.
I think that there is some careful reexamination company by com-
pany going on of many of their pricing policies. I think that this
is indicated by the testimony of several of the presidents of these
companies you have had here. They say: "WTe are not going `to price it
like this any more," and that is almost an exact quotation.
Mr. Gnossi~rAN. I think that the heavy-handed nature of these press
releases implies that it is going to be very difficult for `others to come
up here. I think when `they say here in, the third paragraph that the
PMA will present "testimony of distingu'ished economists of unim-
peachable integrity and reputation," that certainly would seem to me
to reflect that they migtht `think otherwise of you.
Mr. SQUIBB. Perhaps they do, but `they do not say so, and it has been
interesting to me on my previous statement that there has been agree-
ment generally. No one h'as ever `said in `public that they disagree with
what I have said. As a matter of fact, I d'o not see any disagreement
here with what I have said, because I have not said anything along
the lines that they outline here in this press release. They have just
missed the point because they do not or will not see it. The fact that
wholesale prices have d'eclined 14 percent in 17 years and prices of
patent products declined and other products rose, I accept. I do not
see what difference it makes to the point we are talking about, which is
the public view of the business.
If all these things have happened as `outlined in the release, it has
not done the pharmaceutical industry any good with its legislative
critics or its public critics. And `this is the pickle they are in t'od'ay.
I am suggesting there is a way out of this pickle. It is not by simply
repeating that price's are going down or' that all i's well. You know God
is in his heaven and all is well with the world.
81-280-68-pt. 5-3
PAGENO="0034"
1584 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. GROSSMAN. I would just hope. that others of the leadership that
you mentioned and I mentioned would be free to come up here and
talk to us and let us know what they think, without this kind of press
release staring them in the face.
Mr. SQUIBB. That is t.heir problem. I do not think you are going to
get very many here right now.
Mr. GORDON. Mr. Squibb, isn't this a kind of intimidation that this
is what anybody can expect from the industry if he comes up to testify
along these lines; that PM1A is going to put out a press release attack-
ing him?
Mr. SQUIBB. I assume so. I am not intimidated and I do not think
they figured I would be. But, I think it is too bad that they cannot try
to make something good out of what I sa.y, even if only a crumb.
I think we were on page 16, talking about the loss of the service
products. A better answer to this situation is to try to price the losers
more appropriately; they can carry a considerable premium because
of the very rarity and urgency of their use, or to dispose of the product
to some other firm which beca.use of more suitable facilities, lower over-
head, or other advantages can better produce the item. In any case, it
does not appear that the practice of providing the medical profession
with a few under-priced service items is any real or significant burden
on overall earnings even giving no credit at all to the "goodwill" or
"good public image" created by the practice and exploited with en-
thusiasm by public relations departments.
The pharmaceutical industry has historically been partial to special
deals of the most extraordinary nature on its products-deals, of
course, offered openly to all competitive classes of customers. Perfect.ly
legitimate deals as far as Robinson-Patma.n or antitrust are con-
cerned, but cleaTs which are greatly disruptive of whatever price
structure of a formal nature, and there is not much of it in any case,
that may exist. For example, deals often go far beyond the one-free-
with-a-dozen offer frequently found in case-good pricing, to such levels
as five free with five, and even two free with one. It is not uncommon
f or an important pharmaceutical product of a strongly competitive
nature priced at, let us say, $10 to the retailer to be offe.red by the
salesman for a limited period of time at one free with each purchased,
bringing the cost to the retailer for that purchase to $5. The use of
this material on a refill prescription for an original which was filled
with $10 material opens up all sort of intriguing pricing questions.
Diversion, overbuying, substitution, special sales pressuies for price
alone-and many other questionable practices-are practically guar-
anteed by this sort of pricing. And, behind it all lies the question, what
is the real price of the product?
Other types of dea.ls offered seasonally for "inventory protection,"
or as stimulation for salesmen's selling activities, or as devices to move
potentially short-dated merchandise, or as lures for improved dealer
cooperation and support, all may or may not accomplish the intended
purpose, but each without question seriously confuses the price basis
for the product and contributes importantly to the feeling that the
original off-deal price is somehow inflated and not really necessary.
Deals add greatly to the difficulty of outsiders in understanding pha.r-
maceutica.l sales operations, and even have the unfortunate result of
PAGENO="0035"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1585
adding to the burden of public suspicion unfairly. Sometimes a coin-
pany has a published price for a product which is a fine target for
"high-price" critics, but rarely, if ever, sells the item at this price
because of the frequency and generosity of its deals on the product-
deals which the retailers wait for with the certain anticipation of a
bargain price. In the meantime, the published price remains for all to
see and none to use, and as a perfect~ subject for critical analysis by
industry's detractors. Manufacturers should take a new and hard look
at their published retail prices compared with their average selling
prices based on unit sales divided intO the dollars received. Deals are
importantly disruptive of orderly pricing introducing an element of
imcertainty into the market that often confuses the manufacturer as
well as his customer. They have been: overused in fact, misunderstood
in principle, and wrongly analysed in relation to their effect on profits
often enough to make them an area for immediate and determined
study by any management which believes in pricing refonn as a pre-
requisite to the solutions of many of the industry's critical problems.
Mr. GoiwoN. When the PMA brought up the problem of generic
equivalency a couple of weeks ago, one of its witnesses presented data
to indicate that its antibiotic produced higher blood levels than the
products of two other manufacturers, perhaps three others, I do not
recall. Now, in your opinion, how meaningful is this type of a study?
Mr. SQUIBB. It is very meaningful if you are the manufacturer of
one of those four products concerned. However, I think the generic
equivalency argument which certainly has not yet been developed, is
self-limiting, because you are working only with specific products,
whether it be 10, 12, or 50, and you prove the equivalency of those prod-
ucts. Outside of what you are testing, you have no data whatsoever, and
the fact that product A differs or is equivalent to product B, C. and D,
or vice versa, does not cover F, G, H, I, and J at all. And in this vast
field, whether you are coming in as a: supplier of product A or product
D or product F makes a big difference. The situation varies from where
you approach it.
Mr. GORDON. You would say that no broad generalizations can be
made?
Mr. SQUIBB. You cannot make generalizations at all. These are
specific examinations of specific products for a specific purpose, and
you cannot deduce a result in another area because product A and B
are different. You cannot say that product A and F are different unless
in fact you test product A and F, and so forth.
Senator NELSON. Let me inquire about that a moment. The asser-
tion continues to be made by some representatives of industry that
there is no such thing as generic equivalency. If that statement is to
be meaningful at all, it seems to me, it can be argued technically that
no two drugs are identical because no two things really are identical.
But is it not correct :that the assumption upon which formularies
are established in this country is that using the clinical information
that a therapeutics committee has, they can select drugs for their
formulary, and although there may be four or five brands, brand
names or generics, that purport to do the same thing and be the same
compound, based upon their experience the formulary committee
selects one.
PAGENO="0036"
1586 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. SQUIBB. That is right.
Senator ~ELSOX. And the same day across the Nation another clis-
tlnguished forniulary committee selects another company, and on the
same day another formulary committee in another part of the Nation
selects another. I notice in the Medical Letter of June 2, 1967, they
tested 22 preclmsones, and concluded, on the basis of clinical advice
and chemical testing, that all of them were equivalent.
So, if they are correct, you could have 22 hospitals using 22 pred-
nisones, and the assumption all these hospitals are going on is that the
one they select is going to produce as good a result as any of the other
21. They then have a chance to test it in the hospital, and if they get
the kind of result, the successful result they are seeking, they have a
chance to compare the result they are getting with their drug as
against another formulary in another part of the country.
I conclude from that that there are many drug products made of
the same compound by different companies that are, for all practical
purposes, therapeutically equivalent in the result they produce. Do
you have a comment on that?
Mr. SQUIBB. I think you are absolutely right. You have to arrive in
the process of making a decision to put one item in inventory that
this product has an equivalency. To say it is in fact identical to every
other prothict without testing is making an assumption that I do not
think has to be made. I do not think it is made by any responsible
formulary committee. You outlined it very well. They are assuming
that this will do the job for which they want this type of product in
their formulary. They made a selection because they have to make a
selection, and they make it on the best information that they can get,
whether they have to depend upon somebody else's analysis in the
Medical Letter or by outside laboratory testing or whether they can
make their own analysis. Whether they are simply doing it from the
basis of t.heir experience in the practice of medicine and the fact that
they have used a given label or a given product or for some other
reason, I think that we are beating around too much the question that
there might be or might not be identical products. Nothing is identi-
cal, as you say. I think that the equivalency has to be determined by
responsible men, because they have to in fact, for practical reasons,
select one product. They do that, and price is becoming now a factor
in this more than it used to be. I think that testimony in hearings
such as this, conversations and discussions in the press, and in medical
circles all around the country are beginning to stir up this issue more
than it has ever been stirred up before.
Senator NELSON. Thank you.
Mr. SQUIBB. I will proceed on page 18. So much for pricing-
Senator NELSON. Are you saying that the cost of the detail man is
much greater than all of the other costs of promotion, including ad-
vertising in the medical journals, free samples, and all other aspects
of legitimate promotion of the drug?
Mr. SQUIBB. This varies company by company. Obviously some com-
panies believe in large detail forces and others manage to get along
with relatively few. But basically speaking I think that in the case
of any house that hires detail men, the expense for detail men and
their equipment is the largest single factor in their sales budget. If
PAGENO="0037"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1587
you added all the other many factors tOgether you may come up with
an amount equivalent to the detail man, but this is the largest and
most important single item in many budgets.
Senator NELSON. How significant is the detail man's role in the
sales end?
Mr. SQUIBB. Very significant. The detail man is the best and most
effective means of producing a sale or prescription on your specialty,
no question about it. He can be made much better and that is one
of the points I make here, but it still is a fact that the detail man
working with problems he did not cr~ate himself and tlia~ are part
of the medical climate today, working in a. doctor's office, with a doc-
tor whom lie knows from previous calls over the years, has had re-
markable results in terms of sales produced out of this personal
contact.
Senator NELSON. Do you think the effort of the detail man to deter-
mine which products the doctor will use is a more significant influence
on the doctor than the literature lie will receive?
Mr. SQUIBB. Oh. yes, yes, I do. Some physicians, for their own rea-
sons, do not like to see detail men or spend much time with them, other
physicians perhaps are not contacted Often, for various reasons, by
the detail man, `but where there is a strong detail man's pieseiltation
to a physician who is known to be a good producer of prescriptions,
this physician will be more influenced by the detail man who works
with him, sees him once every month or so at least, than by any other
promotional force that the drug business can produce.
Senator NELSON. I am sure that there are many very intelligent
and conscientious detail men, and I do not mean to reflect upon them
at all. However, a detail man, no matter how intelligent or conscien-
tious lie may be, is not a professional man in this field, insofar as
having the qualifications to really judge what drugs a patient should
have, although he may be very expert in what his own company's
drug is and what they say it will do. So here you have a situation of a
doctor relying upon a man who is not a pharmacologist, not a pharma-
cist, though some may be, to help him make decisions about what kind
of a drug he should use in the treatment of his patient.
Mr. SQUmB. Well, perhaps because I have, been a detail man myself
and have done this work quite a hit, I feel that the quality of a detail
man's work for the reputable companies, although you can always find
examples unfortunately that are to the contrary, is excellent. They
know the area in which they are working. They know their product.
They are well trained. They follow the instructions which are given to
them in no uncertain terms as to what they can do, the limits within
which they can operate. The detail man is, generally speaking, working
in a marketing situation and in an information situation which was not
of his own creation. You understand the doctor is there often without
information. He has not gotten it, either from his training or back-
ground, or he has not any time to get it any other way. The detail man
moves into this situation with skill, with discretion generally speak-
ing, and is a fine means, although expensive, of transmission of this
information. I think he is the best because he is armed with the pro-
nlotional material he is going to talk about, he is armed with the ad-
vertisements, the samples, the promotional literature, and the refer-
ences. He has all of those with them and he can discuss them because
PAGENO="0038"
1588 COMiETITIvE PROBLEMS IN THE DRUG INDUSTRY
he has been trained to discuss them within the limits of that product.
He is not practicing medicine. He is talking about this one particular
product and the other products that relate to it. I think that this is the
finest practical way that has been developed today to bring informa-
tion to the physician. The physician knows that this man is not a
physician. He knows or should know the limits that the detail man has,
and it might come to a point in every detail conversation where that
conversation must stop and the question referred back to the home
office for further expansion.
I think generally speaking, and I know there are horrible examples
in your own record of detail men who have done wrong, this is an ex-
cellent method for the transmission of product information which is
one of the problems that we are faced with.
I think the fact is that sometimes the detail man is criticized for
things that are not his fault if he has too many samples, if he has
literature which is too flamboyant. He did not create that himself.
These are the policies of the house that sends him out. The detail man
as a transmitter of information, as a part of the drug world today, is
a very excellent though expensive method of doing the job that is
needed.
Senator NELSON. I come from a profession where you had to try
your matters in open court with an adversary who tests every con-
ceivable proposition that is advanced.
Now, if you had all detail men who handled the same drug before
the doctor, and the physician could hear all the arguments about
which one is best, and then make an evaluation, it would be quite
another matter. But here you have a situation where the doctor has
listened to one man from one company with one or several drugs and
only one purpose: to sell his firm's product. And even though he
may very well know that there is another drug equivalent or better
at a lesser price sold by another company, he will never disclose this
fact to the doctor. So the doctor is listening to a kind of a one-
dimensional presentatioi~, and I observe at this point that this bothers
me a little bit.
Mr. SQUIBB. Yes. That of course is the limitation. A man is presenting
one product with all the skill and knowledge that lie has, and that
is his job, to present that one product in that one field.
Senator NELSON. I am not critical. That is the job he has.
Mr. SQUIBB. That is the job he has, but as a matter of practical
operation another man will be along shortly to present his product,
a competitive product, and lie will do the best he can with that. I
think that these things are not perfect. These are far from perfect
ideal situations, but given the diffiei~ilties of a multiproduct market,
and the complex and technical problems that are involved here, the
detail man and his acceptance by the physicians has evolved, has
developed to such an extent that it works, and it works as well or better
than anything else now being used. This is where I think physicians
look for their information. These are detail men that they know and
accept. They will not see some other detail men. Perhaps, they do not
like them or they have made mistakes in their office. But generally
speaking the teclmique is the best one that the pharmaceutical indus-
tries have come up with for the purpose of simple transmission of
information on its products to its customers.
PAGENO="0039"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1589
Senator N~soN. That method certainly is not the equivalent of an
equality method of presentation, using a distinguished therapeutic
or formulary committee in a hospital, even though of course detail
men do go to hospitals.
Mr. SQUIBB. Yes; and sometimes the committee will ask a detail man
to come in and describe a given product to the therapeutic committee.
Senator NELSON. Yes.
Mr. SQUIBB Just within those limits and to make a presentation on
his product, they might ask several in. Then they have to make their
judgments.
Senator NELSON. As a matter of fact, I am familiar with cases where
that is the single assignment of the detail man. He spends all day
every day at the hospital. However, under that circumstance vis-a-vis
a sole practicing physician, you do have in that hospital an expert in
blood dyscrasia, an expert in heart disorders, internists, and others plus
pharmacologists and pharmacists. So you can benefit from the experi-
ence of people who have tried all these various kinds of drugs, or read
the literature or have their own expertise in some aspects of medicine
in which they know the drugs used. This is quite a different case from
the situation where a single detail man is giving some information to a
lone practitioner out in the country someplace.
Mr. SQUIBB. Yes, I do not mean this disrespectfully but I found
from my own personal experience that when you are talking to a
given physician in his office and telling him about your product you are
talking to the world's greatest expert. This is his attitude. He wants to
know for himself. He is not interested in other experts. You can bring
in reprints but you have to be careful in dealing with an individual
physician. You are talking to a man who is perfectly capable of making
judgment on products and does so as a result of your presentation and
information. Doctors do depend on skillful, accurate, careful detailing
of products. With all of the drawbacks which you have very well
pointed out, it has grown up over the years, and I think it has grown
up because there is nothing else, not because it itself may be the best
thing that ever will be. Under the conditions that have prevailed
certainly in medicine in the last generation, this has been the best way
to work it~. Now, with the growth of. hospitals, as I have pointed out,
it is going to put a new burden on the detail procedures, a need for
reexamining these detail procedures. There will be new requirements
for the detail man, how he is trained and what he does. This is the
point~. I think the burden that the detail man will carry in 3 or 4 years
from now will be quite `different from what he carried 3 or 4 years
ago, `because the medical world is changing. The companies that see
this first will get to work on their detailing studies, the effectiveness of
their detail staff, and the money spent for it, and they are going to `be
that much ahead in getting the business out of `these therapeutic
committees that are springing up and controlling the hospital business.
`Senator NELSON. Go ahead.
,Mr. SQUIBB. I am in the middle of page 19. The analysis of this cost,
that is the cost of the detail man, in terms of time spent to produce a
dollar not of `sale.s but of profit is urgently called for, frightfully
difficult to obtain, and very painful to apply. But it must be done, and
when, as and if it is done, there c'an only be a large reorganization,
reapplication in terms of assignments and resultant substantial
PAGENO="0040"
1590 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
changes in the product mix and customer service patterns of a given
company. Efficiencies obtained from such action, however, would off-
set much of the loss of income from price reform as already described.
Prices are inflated, and by some sort of Parkinsonian law, sales activi-
ties are inflated within the limits set by those inflated prices.
A better understanding of the precise source of the prescription
specification-who wrote it, why he wrote it rather than something
else, and where and at what price it was filled is essential to build a
Irnowledge of what to do about getting the next specification for your
label. The reason it was received yesterday, may not and probably
will not, be the reason it comes in tomorrow.
The industry does have at the present time much statistical data on
the number of prescriptions written, the products specified, and even
the prices charged. However, this data is slim on precise motivation
behind each selection of drug prescribed, and it is difficult to trace
back to the basic cause for the choice. Besides this failing, the shifting
trends insofar as numbers of prescriptions written under circumstances
other than those of the traditionai private patient in office practice
which a few years ago was literally the sole source of drug specialty
sales volumes are not properly reflected in data form. Improvement
in available information on prescriptions from every aspect will be
forthcoming soon as studies now being set up are completed, and fuller
utilization of computers is realized. Progress in this area has been
slow, for even although a pharmaceutical company is so often quick
to spend several hirndred thousands of dollars in research efforts to
develop the best possible formula for a drug product, it is more often
than not quite unwilling to spend even half that amount to determine
the key patterns of the ever-shifting marketplace for the product when
it is finally perfected.
Even more specifically than this, however, the pharmaceutical
manager needs to determine what activities on the part of his sales
staff, produce which results in his sales area today. Generally speak-
ing, lie is not succeeding in getting this information. Partly this is
because it is a very difficult thing to do, but partly it is because there
have not as yet been real pressures as yet to require it. Behind it all
is the knowledge that if he does acquire it, it will be an overwhelming
job to implement his new knowledge with a completely revised and
reorganized and reassigned staff.
The shifting pattern of medical care centered around the growing
importance of the hospital-its accident rooms-outpatient depart-
ments, diagnostic facilities, and its expansion to include nursing home
and convalescent care in its scope of influence will change the job of
the pharmaceutical detailman in the years immediately ahead. This
will be true not only to the extent of where he works, but also how
he works. It will reduce him in numbers, but no doubt will increase his
individual cost to his employer. It will place new burdens on his train-
ing, and new strictures of regulation and control on the performance
of his duties. All of this calls for immediate reexamination of every
facet of this part of the sales procedures of the industry. The phar-
maceutical detail man must be, like Caesar's wife, above any suspicion.
Economies and readjustments which will result will contribute to the
improved climate for pharmaceutical industry operations.
PAGENO="0041"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1591
A similar close look must be given to the costs of advertising and
product promotion outside the detail man. Criticism of the industry
in these areas has been mostly concerned recently with the quality of
the advertising material in terms of~ full disclosure and balanced
presentation of product information. The results of these regulatory
efforts will be, of course, of real significance to the industry, but the
question of the effect of advertising on prices is the consideration here.
Again we have a difficult problem of balance for the pharmaceutical
manager to solve. As always he has a special responsibility to avoid
waste, to keep within generally acceptable limits, and to persuade his
potential customers that his advertising efforts are useful. To the
degree he fails to do these things, and parenthetically he thinks that
this is going on, they are failing to do these things today to some
extent, to too great an extent, these expenses will be deemed an
unnecessary force and contributor to the upward trend of prices.
Advertising in an industry is vulnerable to critics who see in it
many sinister implications contrary to the public interest, but espe-
cially in the drug industry is it suspect of exerting influences which
are not only productive of undesirable medical practice, but actually
harmful to the heaith of the Nation both in a physical and an economic
sense. The pharmaceutical advertising manager finds himself in the
same sort of position we have seen before in relation to his counter-
parts in other industries. His rules should be different, his responsibili-
ties for careful,, accurate. and nonprejudicial work are greater, his field
of act~ivity is more restricted by regulation and by public policy, and
any deviation will not only be measured in terms of communicative
quality, but also in economic effect.
Senator NELSON. A week or 10 days aoo the representatives of the
Pharmaceutical Manufacturers Association appeared before the com-
mittee. We showed them an example of a drug for which the Food and
Drug Administration in this country requires in all the advertising of
the firm who makes the drug a very speciflc warning to the physician
that fatal blood dyscrasias have resulted from the use of this drug and
that it should not be used for minor diseases or illnesses.
In February of this year the company ran an ad in the AMA Journal
in this country, February 20 1 believe, in which they made a full dis-
closure as required, in a paragraph with a black line drawn around it.
It mentioned all the dangers and warnings very specifically.
WTithin the same month, in fact a few days earlier, they ran an ad
for the same drug in the British Medical Journal in which they did
not include a single word of warning. When I asked them why, they
said they simply comply with the laws of the country in which they
advertise. I said, "Aren't you worried, concerned about the British
people?" I said, "What about the underdeveloped countries where
there are no standards, no laws defining required quality. They do not
have doctors who are qualified or do not have enough doctors to set up
any testing procedures."
Well, their only response `to that was "We comply with the laws of
the country we are in." So what they were saying was that their stand-
ard of quality in `advertising here is as high as it is required to be by
law and they really aren't worried if some doctor is misled in England
or in some other country where the law does not require full disclosure.
PAGENO="0042"
1592 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Do you have an observation you would want to make about that?
Mr. SQUIBB. I think that is horrible. I think to the extent as you
have described it., if that feeling exists in the industry, it negates the
special social responsibility which I think anybody dealing in danger-
ous drugs has, because all drugs, even the greatest life-saving drug,
has dangerous potential. I believe that every pharmaceutical manu-
facturer should without limitation or selection at all times disclose
completely those things within his own knowledge of his product,
making them available to everybody who is a potential customer. The
fact that this might be in a country where there is no law as you say
whatsoever, and you sell off the back of a wagon, this is no different.
If you take any other attitude today, it is just disastrous, it is immoral,
it is unethical, it is stupid. I thjnk it is perhaps even worse than being
stupid, and a number of other things which you can be, because it is
bound to come back on you and bring retribution as soon as your drug
is inevitably used by ignorant or uninformed people. In the long term
this is not what you can live with. I think it is a horrible story that
you recite there. I cannot see how a responsible company can fall back
on "the law" to limit its burden for proper disclosure of things it
knows about its products, good and bad both.
Senator NELSON. Thank you.
Mr. SQUIBB. Now we have come to the problem of research, maybe
the stickiest of them all. We have seen this point brought up here
already today.
Research costs are used more often than any other one factor as an
explanation for drug prices whenever the 1)harmaceutical industry is
replying to its critics-or even when it takes the initiative in one of
its "image building" or "storytelling" phases. They are certainly part,
and an important part, of the industry's operations. They are in the
overall sense essential to keep the medical armamentarium of drugs
moving ahead. However, a great deal more light needs to be shed on
the general subject of pharmaceutical industry research and the costs
thereof-both as to its own internal research organizations, programs,
and procedures and as to the research projects it supports outside its
own w-alls.
It is most difficult in industry circles to criticize "research," or even
to attempt to stand off and look at it objectively. To many, drug re-
search is almost a religion, and to question it in any practical way is
to speak heresy. However, it must be said that there is a poor, waste-
ful, extravagant, unproductive, unimaginative, and pointless work
being done under the category of "research" by the industry.
It would be perfectly astonishing if that were not the case.
The degree of this among all of the productive and sound research
operations carried on is certainly open to argument. Most probably it
varies from time to time and from place to place, but it always repre-
sents a sizable part, and a significant one because of the amount of
money involved. There are several questions that should be asked by
every pharmaceutical manager and in turn by his research director, and
answers given before research objectives are set and funds to carry
them out are committed. These questions involve such matters as the
proportional size of research budgets both minimum and maximum
in relation to basic operating costs or even to sales; number of projects
PAGENO="0043"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1593
to be supported at any one time and in any one field; relation of basic
research budget to product and developmental research expenses;
similar or overlapping efforts elsewhere; and last but certainly not
least the need actual or potential of the anticipated output of its re-
search activities after the forecast costs have been incurred, and after
market trends now readily discernible have evolved into significant
new requirements for drugs from both the therapeutic and economic
standpoint.
In other words, I am saying let us look where you are going be-
fore you get there.
It will be said by every pharmaceutical manager and research di-
rector that they do all these things; that they could not do their job
if they did not; and that this is simply a restatement of what is clone
every time a research dollar is budgeted or a project entered into.
1-Jowever true this claim may be in theory, it leaves much to desire in
practice. From the standpoint of cost, the research operations of the
industry are not yet properly integrated into the whole pharmaceutical
operation, and their relationship with other parts-~articular1y sales,
are not clearly worked out for many reasons, most of them internal
and peculiar to each individual company. Before acceptiiig the idea
that all is well in this particular segment of his responsibility because
merely to do research is in itself a creditable and self-rewarding ef-
fort, a pharmaceutical manager should take a hard look at his research
expenditures in relation to his project accumulative, new prod1uct fore-
casts, his competitive position, and the state of his own pocket book and
arrive at a private conclusion that he apply the same principles of
cost control to his H. & D. programs as he does to manufacturing and
sales.
Research expenses can and should be closely reexamined by the
pharmaceutical industry, and the resulting adjustments need to be ex-
plained to, and understood by, the scientific community generally.
Objectives, and general routes toward these objectives, should be dis-
cussed openly so that better information is available to shape all con-
current research programs to make certain of better utilization of
the total research dollar. It is not enough for the industry to point with
pride at the size of its research budgets, or even to their past research
achievements. It is now essential that they explain their research pro-
grams in terms of their place in the overall scheme of medical progress,
their purposes in terms of public need, and their costs in terms of effi-
cient procedures with a minimum of duplication and waste.
Such are some of the things that the industry can do to answer the
critics now pressing for reform enforced by legislation and bureau-
cratic control. There is really no choice as things stand today so far
down the road have matters gone. These things must be done by the
industry or they will be done for it by others.
Senator NELSON. Everybody, of course, favors research and develop-
ment of drugs and other commodities, and research-oriented companies
have made great contributions to the drug industry and to the health
of the Nation, as have other such companies. With the great continuous
emphasis placed upon research in the drug industry as a justification
for high prices, interestingly enough a substantial portion of the
medical profession is convinced that this factor is very significant. In
PAGENO="0044"
1594 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
fact, a good many doctors will say "Well, I will prescribe the higher-
priced drug from the research company for my patient because they
are entitled to have more money to do more research." It is difficult to
find out how much is research and how much is not.
Mr. SQUIBB. That is a judgment somebody has to make in the account-
ing procedure.
Senator NELsoN. We have not had very much good testimony about
industry research except the assertion from the industry that they
have spent tremendous amounts. Is there any tendency so far as your
knowledge is concerned for the industry to inflate the figure as to the
amount of money they spend on research, or to include in it, items that
really are not genuinely cost or are not genuinely research?
Mr. SQUIBB. There is always the possibility, and this is generally
the result of internal arguments between the two men involved. The
transfer of funds from the sales program to the research budget by the
sales manager versus the research director or from manufacturing or
fro~n quality control. I think there is always a tendency, when you are
talking about research, to want to talk about doing a lot of research,
so you want that figure to he high when you are talking to bodies such
as this. You want to tell how many millions of dollars of research that
you did. So you want to include anything that might conceivably be
thought of as research.
I do not think that there is any inflation in terms of misrepresenta-
tion of research. I think you can judge yourself favorably or unfavor-
ably and make the judgment decisions on a perfectly honest basis
without anything subversive about it.
Senator ~ELSOX. The Harvard Business School did a study called
"A Note on the TJnitecl States Drug Industry" and as to research the
Harvard report said:
In the drug industry research and development is interpreted somewhat broadly
to include not only the search for new compounds with medicinal effects but also
the costs of elaborate testing procedures designed to determine the safety and
efficacy of the drug and elaborate quality control procedures designed to secure
purity and uniformity in manufacture.
Is there any standard which companies use to determine what should
be included as research and what should not `be included as research?
Mr. SQtTIBB. No, there is no standard, and I think there is a great
dccl of dispute among the various research personnel in the different
companies as to what it should be. I know that some companies quot-
ing another company's research will answer it by `saying, "Oh, well,
they throw a lot of things in there that we don't," but they do not
really know this. This is an individual decision in each company, and
reflects the philosophies of the management in charge at that moment,
and might vary even from administration to administration. I do not
think that this is a large enough factor, this judgment variant here, to
really be a confusion or a misunderstanding in the job on research that
the industry is doing. This is a.n internal matter of accounting proce-
clures which would make it vary a little bit, but not really a great `deal.
Mr. GORDON. Do you think that the wider use of the formulary sys-
tem would have a tendency to increase the quality of research?
Mr. SQUIBB. `Well, yes, to the extent that duplicative products exist.
Inasmuch as only one of a kind of a product `is all that you can put
into a formulary, if you have a line that is entirely duplicative of some-
PAGENO="0045"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1595
body else's line, you are going to have trouble getting it into a formu-
lary situation. From that point of view the pressure will be on research
to come up with products that are `different, either different physically
or with different therapeutic purposes. That is quite right. If only one
selection of products is `going to be made for a given formulary, the
pressure will be in every company to come up with some new, some
fresh, some unique products for the line so that they can be justified
for formulary inclusion. I think this will be an important factor in
directing research in the future. It will be a pressure on research
departments to come up with unique and new, not simply imitations
of successful products.
There are many approaches that a legislative body, State or Federal,
can take to attack high prices for pharmaceutical products. In all of
them it is suggested that they be studied from a more objective, long'-
ranged and balanced view with less emotion, prejudice and p01it~ca'1
consideration. The economics of health care, a potential financial bi~r-
den to the Nation that can be and is grossly underestimated, needs the
thoughtful attention of the best brains that can be devoted to it. There
should be no off-limit areas for investigation all the way from the man-
ufacturer through the last distribution point for the product. There
should be no preconce~vecl theories of merit based on real or imagined
evil pract~ces of the past. There should be no temporizing with the best
standards o-f quality and safety for all those w~ho are to benefit in the
future from new concepts of public health.
At the State level legislatures can reexamine the meaning and intent
of their drug substitution laws, in view of current welfare programs,
local, State and Federal. These laws passed 10 or 12 years ago served
quite a different purpose then, compared to serving a Government-fi-
nanced public assistance program. There are ways to permit utilization
of the finest drugs in these programs, or any specialties that the pre-
scribing physician may desire without freezing into the State welfare
budgets retail level prices. The State formulary system carefully and
thoughtfully adjusted to the practical needs of the patient can be com-
bined and coordinated with pharmacy laws concerned with subst~-
tut~ons.
Senator NELSON. What do you mean by a State formulary system
for welfare?
Mr. SQUIBBS. State formualries such as a good many States already
have and all States will have I am sure, setting forth the products which
that State will pay for under its welfare programs.
Senator NELSON. How many States now have a formulary system?
Mr. SQUIBB. I do not know exactly. Many of the major States. Cali-
fornia hafls been a notable leader in this, getting in first. I think it is
more general in the Western States than it is in the Eastern. But again
as the States come up with welfare programs and budgets and allocate
or appropriate money to spend for drugs, they have got to impose
some ground rules, some limits on what they will spend it for. Other-
wise the end is off and they are broke inside of 2 or 3 months. We begin
to see some of these things developing~ in the medicaid programs that
have been launched without careful consideration of the ultimate cost.
I think when a State, just like a hospital, has to pay a welfare drug
bill and contracts to do this, they are going to have to prescribe and set
forth the area in which they will pay. This work is most difficult, and it
PAGENO="0046"
1596 coMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
has to be done with all kinds of considerations. But it is their judgment,
absolutely their own judgment.
Mr. GORDON. Am I correct, Mr. Sqthbb, in assuming that what you
are saying here is that the original reason for passing the antisubsti-
tution laws no longer exists?
Mr. SQUIBB. No. The original reason does not exist as the primary
reason.
Mr. GonioN. Primary reason?
Mr. SQUIBB. The primary reason. Of course, you see when the sub-
stitution laws were passed 10 or 12 years ago there wasn't any con-
sideration that the State itself would be a purchaser of drugs to any
substantial degree at all, and the laws were to control individual retail
operations in terms of substitution. Now, when you come and set your-
self up with a formulary, deciding which drugs are going to be pre-
scribed and what price limitations you will recognize, and then you
have a substitution law which forbids the retailer to make changes, he
is in quite a box. He is in an impossible situation. He has to supply a
drug, at a price level at which he cannot buy it, or else break the sub-
stitution law.
Mr. GROSSMAN. Mr. Squibb, may I ask you along the same line, what
would you think of repealing these laws, in view of the position that
if a doctor wants to prescribe a certain drug, that he should be entitled
and he should lrnow that his patient will get that particular, even if it
is a brand name, drug? In other words, don't you think the doctor has
every right to make this determination?
Mr. SQUIBB. There is no question about it. The doctor should make
and must make the determination, and nobody else.
Mr. GROsSMAN. And that there should be no substitution, if that is
what he decides that he wants?
Mr. SQUIBB. There should be no substitution if he decides that is
what he wants. Unfortunately there were examples when the sub-
stitution laws were passed that this was not always carried out.
Mr. GROSSMAN. So then t.here is still a place for antisubstitution
laws?
Mr. SQUIBB. I think there is still a place for the substitution laws.
Unfortunately these are similar to police laws. Any professional man
who holds the title and license `of a pharmacist in properly carrying
out his profession does not alter the prescribed medicine without per-
mission. On the other hand, a law seems to be required because there
unfortunately seems to be those who will do it, for reasons of either
laziness or greed or other purposes.
Senator NELSON. Just one point here. In a hospital it is a clear-cut
case where there would not be any substitution, or rather the
formulary will be used by the physician as agreed in the hospital.
Then as I understand it all of them permit an exception when some
physician says, "I want to make an exception for this reason," and they
perimt it. The exception is rare because they do use the formulary.
But when you have a formulary for all practical purposes the vast
majority of the dri.igs that are used have been decided upon by a coin-
mittee and the doctor selects from that list?
Mr. SQUIBB. That is right.
PAGENO="0047"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1597
Senator NELSON. So you cut the size of the formulary and you
remove that element of decisioninaking by this particular physician
unless he has a strongly held view about some particular brand name?
Mr. SQUIBB. That is right..
Senator NELSON. The same would be true in the State formulary
for the welfare program, would it not, except in the case where a
doctor may decide that he will not select a drug from that formulary
although the State will only reimburse for the cost of the particular
drug on the formulary, and the patient will have to pay the difference,
if there is a difference?
Mr. SQUIBB. That is right. I think ~ye should not let consideration
of the exception cloud over the real purpose of the formulary. You
still put in an opening like that because the practice of medicine is
the uncertain thing that it is, or the nonspecific thing that it is; I think
is obviously is essential to leave a loophole like that.
While these two bodies of drug regulation are now generally separate
in the States, they must become parts of the same problem beca.use
sheer economic necessity has made it so. Under title XIX of the social
security law State welfare and medical programs have been made
a most important part of the administration of public meclica.l care
in the future. Each State is required to set up its own program in this
area for I-JEW approval before Federal funds are allocated to share
the cost. States should take this opportunity right now to review all
their pharmacy laws to make them suitable for maximum utilization
of funds available without lessening of the quality of medical care
afforded insofa.r as use of drug products is concerned.
At the Federa.l level patent laws ca.n be restudied with realistic
and dispassionate attention to the proposals made in the Kefauver
report several years ago. Time limitations on drug patents, systems
of compulsory licensing at uniform royalties, and the pace of Gov-
ernment-paid-for-research in patient procedure all need specific study
now for the pharmaceutical industry alone, without relation to
other industries who may read into such efforts a threat to their own
patent protection laws. Consideration of the importance of medical
care from both a social a.nd economic standpoint would seem to set
apart the patent rules for prescription medicines from `those for other
commercial operat.ions. Patents can provide an umbrella. for high
prices and t.o that extent at least their value for medicines should be
(luestionecl.
Senator NELSON. Once a patent is granted. the company that secures
it has the exclusive right in this country to manufacture and sell that
drug for 11 years; is that not correct?
Mr. `SQUIBB. That is right.
Senator NELSON. At the end of the 17 years, there is only one com-
pany in America., unless they license somebody else, which has devel-
oped complete knowledge of how their Particular drug `is formulated.
So there is one expert in `the whole United States.
Now the patent runs out and the company has `had the benefit of `the
umbrella, so to speak, protection that the law intended apparently to
give them in order to recoup their re~earcll costs and to reward the
company or the `individual for some creative contributions to society.
PAGENO="0048"
1598 COMPETITIVE PROBLEMS IN TUE DRUG INDUSTRY
1-Ioweve.r, when the patent runs out you have a situation as I under-
stand it, w'here neither the FDA nor the manufacturer discloses every-
thing they know about that drug. So a company that then wanted to
go on to the market would not be able `to get the bene.fi~ of all the
information that was accumulated by the company that had the legal
protection for 17 years, and this second firm could not come onto the
market competitively and have the benefit of all the experience and
knowledge that is accumulated by the original manufacturer.
This system is defended by `industry on the grounds that, "Well,
these are our trade production secrets and there is no reason why we
should give them to anybody else." On the other hand. the only reason
they got the 17 years' protection is that the people of this country gave
them the prote~.t.ion. Do you have a viewpoint about what the require-
ments ought to be in terms of disclosures to the public of all informa-
tion a company accumulates about the manufacture of a. drug?
Mr. SQUIBB. You mean the New Drug Application file Oil such a
drug after the patent has run out?
S~na.tor ~EL5ON. After the patent ha.s run out and a New Drug
Application is filed.
Mr. SQUIBB. I think there are some very interesting experiences
being accunmlated along these lines right, now as these major patents
run out in the tetracycline and chloramphenic.ol area. You see these
thins going on right now. I think you should not take the NDA
bundle of material in any seijarate consideration from the patent itself.
You have got to require, if you wani~ to soften the length of the patent
by making the drug patents say for only 5 years, that the NDA situ-
ation Should `be tied into that.
I do not think, having gone through `the patent in 17 years that
now you should ride for the next 17 years in effect on your NDA.
Here you are adding something again from the Government to the
patent law. In other words, it is not a Valid patent any more but has
t.he same effect. You have got this vast expensive bundle of NDA which
the Government will not disclose to your competitors, so you can ride
along on this.
Senator NELSON. Just so that the record will be understandable to
those who read it, would you explain in a little more detail? At the
end of 17 years the New Drug Application is then filed by the
company?
Mr. SQUIBB. It has been filed, the New Drug Application, and you
have been running on your patent.
Senator NELSON. Excuse. me, filed at the time they got the patent?
Mr. SQUIBB. Perhaps, but not necessarily at the time you got the
patent..
Senator NELsoN. And that New Drug Application contains all the
material, research results, and everything else that was required?
Mr. SQUIBB. At great expense.
Senator NELSON. To be supplied to the Government?
Mr. SQUIBB. Of course, the Food and Drug can call this a new drug
or an old drug. This has been covered in your testimony, particularly
I think with Mr. Stetler at some length. But I think we ought to con-
sider how long a drug product covered by a patent or a New Drug Ap-
plicat.ion in the public interest should be kept exclusively to one com-
PAGENO="0049"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1599
pany. I think we should examine this. I think very likely the period,
because of the nature again of the industry we are talking about,
should be shorter. Compulsory licensing at a reasonable-
Senator NEr~soN. You say the patent period should be shorter?
Mr. SQUIBB. The patent period, yes, the patent period. Let us talk
about the 17 years first.
Senator NELSON. Yes.
Mr. SQUIBB. That is running first, and there is no point in talking
about the NDA disclosure as long as the patent is in effect, so you have
got 17 years given by the people for that drug. Traditionally in the
Squibb family we think that patents are not desirable in drug lifesav-
ing situations like this, but I think that perhaps that is too extreme a
position in the current situation. Patents might be shortened. So when
you are shortening the patents and giving perhaps a patent for 5 years
exclusively plus required compulsory licensing after 5 years or after
some determined period of time, I think then you reexamine your
NDA in terms of what you do under the patent law, because you have
to break the NDA if you are going to ireak the patent in practical ef-
fect. There is no point in saying the patent only has 5 years, if you
have essentially a patent produced by your New Drug Application,
which only applies to your product, and is too expensive for anyone
else in a competitive situation to try to duplicate. It is often millions
of dollars of work and time.
Senator NELSON. If you had comp~ilsory licensing, are you saying
that it would require the company holding the patent to disclose every-
thing it knows about the product?
Mr. SQUIBB. Yes; and give a permission of the licensee to use the
documentary protocols and terms that have been put in that NDA.
Senator NELSON. How much information is in there? For example,
the argument is consistently made, which no doubt is correct, that even
if you have the same chemical compound and the same proportions in
a tablet, the mechanical formulation of it or the coating of the tablet
or what have you may still affect the degree of physiological availa-
bility. Does the New Drug Application information disclose all the
specific techniques for compounding and making the particular
product?
Mr. SQUIBB. Yes; but I think that is the minor part of it. That is
the least difficult part of it. NDA includes the background of the
clinical testing, the toxicology, and the studies that have been made
preliminary to the release of the product and the claims that are pre-
sented for the product are based upon the material that has been
accumulated over a number of years and organized in the NDA. It may
be $1 million or $2 million of expenditures involved there. There is a
great deal in there which the new licensee coming into the situation
would find it impossible to duplicate as he is faced with the oppor-
tunity to go into the market. If he had to do that, he just would not
do it. I think that the consideration is whether you should or should
not treat drug products, professional drug products on which a patent
has been given, any differently from any other coi~j~ercial item for
which a 17-year patent is granted. If you ~e-go1ng to say it should
be shortened, then you have got to effe~ti~ely change your rules on
NDA. This is needed so that you make ITt available in not just a token
81-250-OS-pt. ~-4
PAGENO="0050"
1600 COMPETITIVE PROBLEMS IX THE DRUG INDUSTRY
gesture, you make the product available, in fact, on a compulsory and
fair licensing basis to whoever wishes to get into the market.
Senator ~\ELSON. It might be well to shorten the patent period, but
in any event, let. us assume for the moment that the patent period
covers 17 years and that is not shortened?
Mr. SQUIBB. It is long.
Senator ~ELSOX. I mean is not shortened by Congress. I suppose it
is likely in the near future that this period will not be cut. So you have
17 years. You still have two other things you could deal wit.h. One,
compulsory licensing; the other, disclosure by the FDA of all the
material in the New Drug Application.
As I recall from the testimony, and you may have to correct me on
this, Dr. Goddard stated that there is no law now on the statute book
that forbids the FDA from disclosing the information, from making
public the information in the New Drug Application at the expiration
of the patent period. But that since it has been the practice, the
administrative decision of the agency over a period of years not to
do so, he, personally, would be reluctant to change the policy without
some guidance from the Congress; that is, that the Congress pass a
statute making it. the law. Is that correct?
Mr. SQUIBB. That is exactly correct. I think he feels that that should
be the procedure to be followed, because it is obviously a major ma.ttei
of giving several million dollars of information to competition. I
think he feels that. this is such a major decision, and that it would have
such a major impact. upon the industry, that lie should get, and I think
quite properly, a specific direction from Congress before doing it.
Senator NELSON. If the patent laws continue to give protection for
17 years and no legislation is passed to require compulsory licensing,
do you think that. at the expiration of 17 years companies holding
patemits have had adequate time to get the benefit that they should get,
and that. the law should require that t.lie New Drug Application be
made public
Mr. SQETIBB. Yes, but. you have to remember it is not always 17 years
because a product does not necessarily get on the market the day the
patent is issued, and you have to qualify this with some knowledge
and some interpretation. Sometimes you oiily have 2 or 3 years actually
of sale of a product. before the patent expires. So that that is iiot always
the case. But. generally speaking I think that the two things have got
to be looked at together, and without the arbitrary preconception of a
certain number of years. The market and the lifesaving nature of
particular products varies too, amid I think you have probably got to
give some judgment amid discretion, wlieii I say "you" I mean the
Congress. back to the Food amid Drug Adlministra.tion or the HEW,
to make decisions as to how certain products should be handled. Some
of them are much more significant than others in terms of their bene-
fits. Some are routine.
Senator ~\ELsOx. Do you believe that. compulsory licensing ought
to be required by law?
Mr. SQUIBB. I think compulsory licensing is the fairest amid best ap-
iroachi to this problem.
Senator NELSON. Amid undler compulsory licensing, t.lien, the owner
of the patent would receive a royalty?
PAGENO="0051"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1601
Mr. SQUIBB. Would receive a royalty, a fair standard royalty.
Senator NELSON. That was to be my next question. Would you estab-
lish the royalty by law or how would you do it?
Mr. SQUIBB. I would establish the royalty, I think, in the law.
Senator NELSON. And would you require compulsory licensing un-
mediately after the drug becomes available?
Mr. SQUIBB. No. I think probably some type of a period for ex-
clusivity is necessary to stimulate or to keep going with research. I
would have such a period a good deal shorter than 17 years, maybe 5
years or 3 years of marketing before anybody who asked for a license
would be required to be given one.
Senator NELSON. So if the law remained as it is, with a 17-year
patent period, do I understand you to be saying then once the drug
is introduced on the market, some other period of exclusivity, 2 years,
3 years, or 4 years ought to be established, to be followed by com-
pulsory licensing?
Mr. SQUIBB. That would seem to be the most equitable and fairest
approach, serving the public on the one hand and the inventor on the
other.
Senator NELSON. Thank you.
Mr. SQUIBB. The notion that without the inducement of a 17 year
patent with no licensing requirements all medical research would
die seems quite an extravagant and unrealistic claim. To the extent
it can be modified drug prices will come down.
Laws that now are concerned with competitive forces in the dis-
tribution system for medicine must also be reexamined, particularly
the effect of Robinson-Patman. The rigidity of drug prices in retail
pharmacy where this law is `applicable, and quite the opposite situa-
tion in hosiptals where it is not, have already been discussed. It has
been publicly suggested by Dr. Apple of APhA that consideration
should be given to enforcing Robinson-Patman on prices to institu-
tional purchasers. This would ceratinly tend to bring the same price
rigidity to this area of drug distribution now benefiting from unregu-
lated price competition among suppliers. It is worth at least as much
consideration to suggest the removal of Robinson-Patman from ap-
plication to the pricing of prescription drug prices to the retail phar-
mnacist. This would undoubtedly bring drug prices to the consumer
down, but it would also bring down many retail pharmacists now
benefiting from percentage markups of manufacturers prices. To the
degree that the professional fee replaces this markup system however-
pharmacists would survive. This is `a difficult problem, but one for
which solutions can be found if the public welfare is given first con-
sideration as certainly it should be.
As `a matter of fact and as prelude to any possible revision of Robin-
son-Patman or other laws bearing on pricing of drug products, a bet-
ter general understanding of the economics of the entire drug dis-
tribution system is essential. If real benefits are to be obtained in
the form of lower drug prices without loss of product quality or mcd-
ical progress in the form of new drugs, `a full realization of the costs
that now go into a patient's prescription dollar is obviously called
for. Attacking one small piece of this cost which even if completely re-
moved would not appreciably `help solve the problem at hand would he
PAGENO="0052"
1602 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
a very minor and even useless victory. Distribution costs of all kinds
have to be weighed-their benefits and their burdens. The possibility
that we are extending life today to a drug distribution Pattern that
should have been made obsolete long ago if not for the fear of outrag-
ing special interests should be given real objective consideration in the
light of today's requirements, not those of a general or more ago.
Among the approaches that could be taken by Government to the
pharmaceutical industry is that of the public utility concept by
which operations and profits are subject to close scrutiny and regula-
tion as are those of electric and telephone companies, and common
carriers. Here rates are subject to approval by Government. commission,
and profits controlled within generally approved limits. Price ap-
proval could similarly be given for drug products with the entire
financial operation of the industry subject to review and control to
establish levels of performance. This is a. most drastic alternative
to possible reforms, but one which is by means unrealistic or incon-
ceivable.
Thirty-five years ago the~se words were published:
The manufacture and distribution of medicines, because of their intimate
relation to the health and welfare of a community or nation, partake of the
nature of pu~~lic utilities. In view of the shifting of control from professional
to financial hands, manifested by recent developments in the drug industry,
the public interest may require "regulation" of the industry, through the
guarantee of a fair return to investors and the limitation of prices to be charged
to consumers.1
They are even more applicable today.
Senator NELSON. Who said that and where was that published?
Mr. SQUIBB. That was published in a book by the University of
Chicago Press, "The Cost of Medicine" by Dr. Fisc.helis and Dr.
Rorem in 1932, analyzing the costs of medicine at that time which
were certainly an entirely different situation than the bedrock costs.
today. Yet they saw this problem of special responsibility for an
industry which at that time was just a fraction in terms of the sales
dollar of what it is today.
Senator NELSON. What is the total drug sales today?
Mr. SQUIBB. Well, the prescription sales in 1932 were probably
one-twentieth of what they are today.
Senator NELSON. What are they today?
Mr. SQUIBB. They are $2 billion or some such figure, and the figure
back in 1932 might have been $30 million at the most. It is one of
the most astonishing growths in industrial history. The impact of
this growth of the prescription business in this country since the war,
since 1950 really, is the thing that has brought to the industry all
of its headaches. It is inevitable when you become very important
in terms of dollar volume that new regulatory problems be picked
up, and yet the industry does not always appreciate this. I think
this is interesting that way back in 1932, 30-some years ago, two
men looking at the industry from a philosophical point of view arrived
at this particular conclusion.
These are some of the principal approaches governments can and
should consider to the problem of drug pricing. The extent that any
of them are necessary depends upon the immediate reaction of the
`"The Costs of Medicines," by C. Rufus Rorem and Robert P. Fischells, pp. 233-234.
PAGENO="0053"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1603
industry to the pressures it is now undergoing from pubJic and legis-
lative criticism. It is probable, the nature of the modern competitive
corporation of no matter what industry being what it is today, that
some direct Government action is necessary and desirable now. What-
ever is done should be pointed toward tequiring the industry, and drug
wholesalers and retailers as well, to reorganize and reform their own
procedures, and not he restricted to the mere imposition of new regula-
tions and controls of outward and basically superficial parts of the
drug distribution system. Major economies are possible in getting our
present drug products `into the hands of the public who needs them,
changed only in price and not in quality. With the facts of basic costs
as they are today, it is clearly false and stupid to say that this cannot
be done. It is to be hoped that industry will take the leadership to do
it-if not, others will.
Senator NELSON. Are you familiar with the book written by Mar-
garet Kreig called "Black Market Medicine"?
Mr. SQUIBB. Yes, I read that hook.
Senator NELSON. We haven't taken any testimony on this question.
Do you have any opinion about how big a Problem black market
drugs are?
Mr. SQUIBB. Black market drugs are a problem. They are a problem,
and you cannot say that it is a negligible problem as long as there is
even one opportunity for black market drugs to get into normal chan-
nels of medicine distribution. rfliis is because of the nature of drugs
taken by people who are ill, people who are in no condition to resist an
improper drug. I think it always has been an important problem, and
always will be an important problem as long as human nature is what
it is, because it seems there is always gOing to be somebody who is going
to chisel and cheat and make false, dirty, substandard drugs. One of
`the burden and one of the `special responsibilities of this industry
among others is to fight this particular, battle. This is a more important
matter in the case of drugs than it is in several other industries, because
more `harm can be clone here. To say that there is only a minus fraction
in terms of the number of dosages given that might be substandard or
black market medicine begs the point, if it happens to be one of the
doses that you swallow, it is a major `problem, and it always will be.
Those of us who have been around the drug business for a while and
have been in the drugstores and dispensing points know the constant
danger of this, and the actual existence also of sedan peddlers and gyp
artists of one type or another who purvey `these black market drugs.
To the extent that they are resisted by ethical, upstanding pharmacists,
and to the extent that they are turned down in their efforts to distribute
these products, so much the better, but I `think the industry itself
should take a very strong opinion that this is an important problem
and always will be an important problem.
Senator NELSON. Is there any educated guess about the volume of
black market drugs or is it unknown?
Mr. SQUIBB. The `only people I have `talked `to who kn'ow much about
this are `the people in the Public Health Department of the State of
New Jersey, who are very acutely aware of it because in `the industrial
areas of the east, the built-up `areas of'the east, it is frequently seen. I
think as you go out into the Midwest `and the scattered population cen-
ters which are not so large, I think it is' much less `of a problem. I `think
PAGENO="0054"
1604 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
in the New York City area, including New Jersey, with its type of
pressures and the type of area we have there, I think it is a major
problem. There have been many recent examples of stolen goods and.
substandard goods, in and around New York, to indicate a very sub-
stantial activity of this sort. But I think it probably diminishes as.
you move away from New York City. It is probably a major problem
in Chicago. I am sure it is in `Los Angeles, but in small areas it is not.
Senator NELSON. Is it more of a problem to a brand name firm than a
nonbrancl name company?
Mr. SQUIBB. Yes, it is more of a problem to a brand name.
Senator NELSON. Why is it more of a problem there?
Mr. SQUIBB. Well, in the first place he is probably the target, the
brand name man.
Senator NELSON. Pardon?
Mr. SQUIBB. The brand name man is probably the target of the
black market medicines, as they generally are counterfeits of a given
brand name in `a pre-established market or it is an imitation of it or it is
some stolen goods of the brand name, or it is some kind of a deception
practiced on a brand name. Therefore the man with `the brand name
product has to take leadership in this. Nobody is trying to imitate
a generic product. They could not have the monetary rewards from
imitating a generic phenobarbital that `they get from Lluninal, the
brand name product. I think the misrepresentation in this area is a
direct `attack on a brand name product which is producing a sales and
price level which is worth while for the criminal `to submit his own
product.
Mr. GRoss~L~N. Mr. Squibb, I would like `to ask you first of all `as to
your concept of treating the industry as a public utility, do I `take
it that you would `advocate `this if `they did not take steps, corrective
steps, or do you advocate it a't this time?
Mr. `SQUIBB. No, I do no't advocate it at this `time. Obviously this is
the last resort to be imposed upon `an irresponsible industry which
shows that `it cannot police itself. I do not advocate it. I would be very
dismayed if that were necessary.
Mr. GROSSMAN. What would be the methods short of using `this public
utili'ty concept `that you would advocate now?
Mr. SQUIBB. I think that the industry has to restore its reputatiOn
among the general public and among the legislative `bodies, Congress,
State and Federal, that it is operating within reasonably normal eco-
nomic limits. It has a sense of responsibility which I have tried to
stress through my presentation here, and I believe it can do th'a't.
Mr. GROSSMAN. I do not know, I presume you are aware that the
FTC has just published new figures which show that the dlrug mdus-
try is now the No. 1 industry in terms of profit..
Mr. SQUIBB. Yes, I saw that. It was in the paper, I believe,
yesterday in the New York Times.
Mr. GROSSMAN. It now surpasses the automobile industry.
Mr. SQUIBB. Yes. In 1966, drugs are the most profitable industry.
I think this will probably be an occasion of congratulations internally
in the industry, c'\~Te are. really going great and stockholdlers will like
this," and all kinds of things and the WTa,ll Street analysts will greet
this and the stock will turn up. I think that our problem is what this
PAGENO="0055"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1605
sort of thing produces. I think we have a special difficulty with profit
when someone takes our product who does not want to take it in the
first place.
Mr. GRoss~rAN. What were the figures, I think 21 percent?
Mr. SQUIBB. 21.1 percent, a rate of return in 1965 had been 21 percent.
Mr. GORDON. After taxes?
Mr. SQUIBB. This is with an after-tax return of 21.1 percent.
Mr. GROsSMAN. Could you suggest to the committee what you think
would be a fair return?
Mr. SQUIBB. I think a fair return, as I tried to say before, is the
average return of comparable consumer goods industries that have a
broad base of utilization. As a matter of fact if you want. to pick out
a figure, I think it is ~rouncl 12 percent instead of this 21.
Mr. GROSSMAN. Do you think that some type of limitations on profit
might drive out some of the smaller companies?
Mr. SQUIBB. No.
Mr. GRossMAN. Leaving the field to the majors?
Mr. SQUIBB. NO.
Mr. GRoss~rAN. You do not think so.
Would you say that any action to reduce profits would necessarily
result in lower prices to the consumer?
Mr. SQUIBB. Yes. You don't make an action to reduce profits per se.
You do these things that I have outlined and some others, and you
find what happens then, what you come out with. You plan and you
budget ahead a 10- or 12-percent profit.. You reconstruct your indus-
try in a believable or a credible fashion for public acceptance rather
than leaving the situation as it is now which is out of control.
Mr. GROSSMAN. I was just thinking for example of the steel industry
which has been hurt very badly by imports and their profit margin has
gone down, and we are all aware recently that their prices did not go
down, but rather they have just gone up.
Mr. SQUIBB. Well, you have got a different situation here. I am say-
ing the industry should be concerned with the opinion that. outsiders
have of it, because it is such a prime target for regulations. Specific
regulation by people who can do it at legislative State and Federal
levels. So these people will do it, unless we in the industry take in our
hands to correct what are not really. abuses legally, but evidence of
good work done in the past. To get yourself to a 21 percent profit return
you have to be pretty smart, you have to be creative, you have to have
done something right to reach that position. But in reaching that posi-
tion, you have opened yourself up, because of the nature of your in-
dustry, and this alone, to real serious problems.
Mr. GROSSMAN. The social responsibility.
Mr. SQUIBB. That is the point, and it is not to the point that your
prices have gone down or other things in so doing, or that you have de-
veloped any drugs that are for killer diseases and all this kind of thing.
That is beside the point, too. We have got to maintain ourselves in an
atmosphere of extraordinary goldfish bowl credibility or visibility
expected of all serving the health area, not only of the industry which
we are talking about here today, but they expect it of everybody who is
taking part of the health dollar. We want to be sure that the public's
dollar, whatever is spent for medicine or hosipt.al services or medical
PAGENO="0056"
1606 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
care and doctors' offices or diagnostic purposes, that all of this is prop-
erly utilized.
Mr. GROSSMAN. May I just ask one final thing. You talked about the
detail man, saying that it was the best practical method thus far de-
vised, and you also talked about advertising. I was wondering what
are the chances in terms of practicality of devising some type of ob-
jective formulary or compendium with FDA approval, maybe using the
FDA, maybe using the Physicians' Desk Reference with the FDA, and
would this work? In other words, do you think it would be valuable to
the doctor? Could he use it? Is there a more objective way to get this
information to the physician without having him rely on the adver-
tising and the detail man exclusively?
Mr. SQUIBB. That is a tremendous question you have opened up
there. WTe could start over again on that. Of course, there are two
things in your question. Compendium and forniulary. They are not
the same thing, nor for the same purpose. You have to separate them.
I think the compendium idea is feasible. YOU can do it. I think it
would be a waste. I do not think it is practical in concept of use.
Mr. GROSSMAN. You (10 not think that the doctors would really
use it?
Mr. SQUIBB. 1-le would still have to continue to depend upon his
other sources for information even though it is there.. It is hard to
take that position because it seems von are sort of against knowledge,
not against the compendium. I just do not think in practice it is worth
the millions it would cost..
A formulary system for the Government, of course, is something
else again. And here, as I have said before, I think a formulary sys-
tem is inevitable for the Government, now that it has decided to pay
for medicine to the extent it has. I think the Government will pro-
duce, if not now in preseilt legisiatlon, in subsequent legislation, a
formulary. I think it will be a rather large formulary in terms of
scope. It has to be because it covers a large area. It is not just a local
problem or even one major hospital problem. But I think it has to be.
WTl-iei~ von take money and von decide to spend as much money as the
Government decides it is going to spend, you have to set down ground
rules, and a formi~ilary is simply the ground rules for spending the
money.
Mr GORDON. Mr. Squibb, are you acquainted with the PMA adve,r-
tisemnent in the November issue of the Reader's Digest?
Mr. SQUIBB. I received one.
Mr. GORDON. Have you seen it?
Mr. SQUIBB. Yes, I have seen it.
Mr. GORDON. I wonder if you would give us your comments about it.
What is our opinion of it in terms of its technique and its content?
Mr. SQUIBB. That is a big question, too. I think the technique is old-
fashioned and very disappointing. The one case history, as an adver-
tisinsr manager I would not have approved that technique. I think it
could have been iunch smarter and more effective in carrying out the
message. I think the technique of mailing something out~ without
1rol~er examination, and I have read the testimony, of course, and
have heard the arguments about trying to or leaving off the advertis-
ing designation, is most unfortunate. I think this is a most unfortu-
nate thing, because again the industry should take special precautions
PAGENO="0057"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1607
to avoid any of that type of suspicion. Now it may be that this was
simply carelessness or oversight or some low-level clerical mcompe-
tence. However, it was done and I think this is most unfortunate that
that was done. This is a personal opinion only, of course, but I do
not think that the advertising that the PMA is getting in that form
is worth, what was it Mr. Stetler said, $250,000 a crack. I surely do not.
Mr. GORDON. How about the content?
Mr. SQUIBB. Well, that is what I mean, the content of the advertis-
ing, using a one-case story. It is old hat, an old-fashioned way of doing
this. I see what they are trying to say. They are trying to tell the in-
dustry's story. I think they could do it in a nlucJI more modern effective
way than that particular one. This same type of thing has been done
many times in the past. I think, as a personal opinion, that this is
rather wasteful. And again I think this is the type of thing that does
not impress `the doctors, who say, "Well, you have spent all that money,
a million dollars for this, why don't you cut `the prices?" This is one of
the nastiest comments an industry man gets to his advertising and I
tried to cover that in my advertising discussion. You must be sure that
your advertising looks believable and effective so that somebody does
not say "nut that out and cut the prices clown." This piece seems to me
to be a waste, poorly conceived, and inefficiently presented for the dol-
lars spent.
Senator NELSON. I want to thank you, Mr. Squibb, for your testi-
mony. Your statement `and comments, I think, are among the most
thoughtful and refreshing contributions to the dialog about the
problems that these hearings are being conducted on of any witness
that `we have had. I am sure there will be `some in the industry who
will not agree with everything you have said, but your testimony com-
ing from a person with long and distinguished experience in the in-
dustry, your comments which are obviously intended to `help guide the
industry `to improve itself and avoid regulation of some kind or an-
other that might come, your testimony with its constructive criticism
evinces a dedicated concern for the industry rather than an attack on
the industry which the Pharmaceutical Manufacturers Association in-
terprets your remarks as being. I think it is unfortunate that the asso-
cia'tion would make an attack in `a manner which weakens their position
even more. But in any event, you have made a `significant ~ontribution,
and if the leadership in the induStry would give serious thought to `the
observations you have made, it could be of great value `to the industry
itself as well as to the health of the Nation. You have made a great
contribution and the committee is appreciative of it, and I am sure
that those `~h'o understand what this is all about and who are con-
cerned about the problem across the Nation, will be appreciative of
your public service in `coming here tod'ay `and expressing honestly,
intelligently, and creatively your concern as well as offering your sug-
gestions for improvements in an industry which i's a great in'dustry
and which can avoid getting itself into serious problems if it is willing
to impose some self-discipline as you suggest. Thank you very much.
Mr. SQUIBB. Thank you, Senator, very much.
Senator NELSON. That will conclude the hearings until Tuesday, the
19th `of December in the Caucus Room. rphank you very much.
(Whereupon, at 2 :30 p.m., the subcommittee was adjourned `to re-
convene on Tuesday, December 19, 1967.)
PAGENO="0058"
PAGENO="0059"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TUESDAY, DECEMBER. 19, 1967
IJ.S. SENATE,
MONOPOLY SUBC0MMIT1~EE OF THE
SELECT COMMITTEE ON SMALL BuSINESS,
Washington, D.C.
The subcommittee met, pursuant to ball, at 10:10 a.m., in room 318,
Old Senate Office Building, Senator Gaylord P. Nelson (chairman
of the subcommittee) presiding.
Present: Senator Nelson.
Also present: Benjamin Gordon, staff economist; James H. Gross-
man, minority counsel; Susan H. Hewman, research assistant; and
Williaiyi B. Cherkasky, legislative director, staff of Senator Nelson.
Senator NELSON. The hearings of the Senate Subcommittee on
Monopoly will come to order.
We will have a presentation today by a group of distinguished
economists representing the Pharmaceutical Manufacturers Associa-
tion.
Since the economists are here today, I want to make a brief state-
ment to raise a question `which I would, be happy to have them respond
to, sometime during the course of the testimony.
Last spring we had testimony put ii~ the record about comparative
prices paid by cities and municipalities around the United States for
drugs. There was one error in the st.atement with which the witness
had nothing to do, but the statement was attacked by the Pharmaceu-
tical Manufacturers Association, particularly by Mr. Stetler in a
number of speeches around the country saying that the committee
had been misled.
So I had the subcommittee do its `own survey, and so that there
may be an opportunity, while the econOmists are here, for the industry
to comment on this survey, I will just read my brief `statement and
during the course of the hearing we will be happy to have the com-
ments of any of the economists on the statement.
Early this year the New York City Department of Purchases con-
ducted a survey of the prices paid by 12 municipalities, cities, and
counties, for a list of 10 drugs. The survey revealed an extraordinary
disparity in the prices paid by these municipalities for the same drug.
There appeared to be no rationale to account `for the differential prices
paid for them.
On May 15, 1967, William F. Haddad, president of the New York
Citizens Committee for Metropolitan Affairs, presented the results
of the survey in his testimony before this subcommittee.
A great deal of criticism of the survey subsequently was heard from
the drug industry, and Mr. Joseph Stetler, PMA president in par-
1609
PAGENO="0060"
1610 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ticular. In a speech given in Chicago on September 23, 1967, Mr.
Stetler stated that Mr. Haddad's testimony was false, and asked,
"Why did the Senate witness mislead the Nelson subcommittee ?" Mr.
Stetier's criticism ~was directed in particular to the alleged excessively
high prices paid by Fulton County (Atlanta), Ga., for various drug
products. Although the implication in Mr. Ste'tler's speech was that
there was an attempt to intentionally mislead the committee on the
part. of Mr. Haddad, it is clear to me that no such deliberate falsifica-
tion was intended.
Mr. Martin Gersten, acting commissioner, New York City Depart-
ment of Purchases, had written t.o Fulton County requesting the prices
paid by the county for certain drugs. The data requested was furnished
in writing by Mr. F. T. Fa.rran, purchasing agent for Fulton County,
to Mr. Gersten. This information was then submitted to the committee
by Mr. Haddad exactly as he received it from the New York City
Department of Purchases. i wifl submit for the record the letter
requesting price data sent by the New York Department of Purchases
to Fulton County as well as Fulton County's response. it is clear from
this correspondence that there was no intent by Mr. Haddad or any-
one else to mislead the subcommittee. in fact., our own cross checking
of the New York City survey indicated that it wa.s correct with a few
minor exceptions which do not in any way invalidate the general
import of the survey.
However, since Mr. Stetler's criticisms may have cast doubt upon
the general validity of our record in the minds of some people I asked
my staff to conduct a similar survey of drug purchasing by a repre-
sentative group of cities and counties.
Letters were mailed from my office to a total of 77 cities and counties
with public hospitals-other than those cities surveyed by New York
City-using a list provided by the National League of Cities. In other
words, I want to make it clear that none of these cit.ies included in
this survey were surveyed by the New York City Department of
Purchase, and I used a list provided by the National League of Cities.
A total of 38 cities or counties responded. Of these, only 29 purchased
a sufficient number of drugs, or in sufficient quantities, to make their
price figures meaningful. The Defense Supply Agency and the Ve.ter-
ans' Administration are also included, as well as price's charged to the
retail pharmacies.
I also requested similar data. from Fulton County to verify the in-
formation given to the committee. The reply received by me from
Fulton County did indicat.e that the most appropriate source for in-
formation on Atlanta's drug purchases would have been Grady Mem-
orial Hospital, not the county, since it wa~s the largest purchaser in
the county. The request for price by Mr. Gersten in New York City
was made in good `faith-a misunderstanding clouded the interpreta-
tion `of Fulton County's response. I will submit for the record the
letter sent to me by the Fulton County Board of Coun'ty Commis-
sioners.
The 29 cities included in this survey by my office represent a cross
sect.ion of cities from all parts of the Nation. It is more extensive than
the New York survey, and, furthermore, it strongly reinforces the
PAGENO="0061"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1611
evidence of widespread and shocking price differences paid by pub-
lic agencies, with taxpayers' funds, for prescription drugs. The dif-
ferences cannot be explained by the quantity of the purchase because
in some instances the highest price is paid by the largest purchaser
and in others the difference in quantity is not sufficient to justify the
huge price differential. Neither does the source of the drug explain
the differences since the same supplier frequently charges different
prices to different buyers of the same drug. In the main, I think the
obvious conclusion to be drawn is that the price charged to public
agencies is whatever the traffic will bear in each particular instance.
For example, while Grand Rapids, Mich. (Kent County), was paying
$160 for reserpine, .25 milligram, 5,000's, Chicago (Cook County) was
purchasing the same drug for $2.09. That is $160 versus $2.09, a dif-
ference of 7,655 percent. Grand Rapids was buying on a trade-name
basis, while Cook County's order was for a generic product supplied
by a generic house. The quantity purchased does not explain the dif-
ference since the Chicago purchase was $338.58 worth and Grand
Rapids $160 worth. Another dramatic comparison is the Chicago price
of $2.09 for 5,000 tablets as compared with the Defense Supply Agency
price of $4.50 for 5,000 tablets-while, DSA paid more than twice as
much per unit for reserpine, its purchase totaled $31,410 compared
to only $338 for Chicago.1
In the case of dextroamphetamine sulfate tablets-S milligrams,
1,000's-while Des Moines and Newark were paying $22.60 for this
drug on a trade name basis, the city of Los Angeles was purchasing
the drug generically for $0.53, a striking differential of 4,264 percent.
And although the price paid for phenazopyridine-0.1 gram,
1,000's-by the cities of Erie, Indianapolis, and Winston-Salem was
$48, Phoenix, Ariz., purchased the drug for only $4.13, less than one-
tenth the price.
Further, the survey generally indicates that the total dollar value
of the contract involved is not a significant factor in price differentials.
For example, in the case of chlordiazepoxide-25 milligrams,
5,000's-although Los Angeles bought quantities amounting to $32,-
625.86, they were paying $25.50 per unit while Philadelphia, which
bought only $12,210.11, or one-third the quantity, was paying only
$18.50.
it is interesting to note, in this case, that both cities were supplied by
the same firm.
In the case of chloramphenicol-250 milligrams, 100's-the high-
quantity purchasers are not the cities buying at the lowest prices. Al-
buquerque paid $13.50 per unit and bought quantities amounting to
only $718.20, while Cleveland, Ohio, paid $24.99-almost 50 percent
more-and bought $2,249 worth, which is over three times the dollar
value of Albuquerque's purchase. Both cities again purchased from
the same firm.
An even more shocking example is the case of phenazopyridine
tablets-0.1 gram, 1,000's. The Veterans' Administration, a large Fed-
eral purchaser, while buying 80 times the quantity of Phoenix-that is,
1 See charts beginning at p. 1745, Infra.
PAGENO="0062"
1612 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Veterans' Administration bought $31,648 worth as opposed to $413
worth bought in Phoenix-the V.A. is paying eight times as much as
Phoenix for this product-$32.16 as opposed to $4.13. Further,
although the Veterans' Administration and the Defense Supply
Agency contracted for almost identical total dollar amounts for this
drug, the VA paid close to three times as much as the DSA-$3'2.16/
$11.60-per unit, and yet both purchases were by the Federal Govern-
ment, simply different arms of the Federal Government.
Similarly, in the ca~se of chlordia.zepoxide, the VA purchased, in
dollars, 35 times as much as the city of Philadelphia-strangely
enough, Philadelphia got a better price-VA, $21.75,; Philadelphia
$18.50-than the large Federal purchaser.
There are numerous other examples of gross price differentials
which defy any logical explanation. I have no doubt that any random
sample of 29 other municipalities would produce the same results
that both this and the New York survey have demonstrated.
I doubt whether any amount of explaining can justify this quixotic
pricmg structure. Furthermore, I do not see any evidence as of this
date that the industry is even concerned about reviewing it~ pricing
practices as suggested last week by George Squibb, former vice presi-
dent of the Squibb Company.
Obviously the cities and other public purchasers are going to have
to review their purchasing practices if they expect to get a fair share
for the taxpayers' dollars spent on drugs.
The committee will be pleased to hear testimony from the PMA
at any time on this subject matter.
(Letters referred to previously follow:)
THE CITY OF NEW YORK,
DEPARTMENT OF PURCHASE,
New York, N.Y., April 24, 1967.
Mr. J. FORREST GEE,
Purchasing Agent,
Atlanta, Ga.
DEAR MR. GEE: The New York City Department of Purchase is currently
engaged in an intensive survey of prescription drug prices. We have selected your
agency to represent the geographic area where you are located-along with 14
other areas across the United States.
Attached to this letter is a list of 25 drugs that have been chosen by my
staff as widely used and as pricing barometers. I would greatly appreciate
your providing prices for each drug, in the dosage form specified, from your
current contracts. If you do not purchase a certain item, or if you purchase
in a different dosage form, please substitute an appropriately similar item and
indicate that you have done so.
Because of our urgent need for this information, we have ~ieen forced to
undertake this project on an expedited basis. One of my staff members will be
calling your office in a day or two, and we hope that with the list we have sent
in front of you, you will be able to supply us with as much information as
possible over the telephone.
I apologize for the short notice we are giving you. I wish it were possible
to collect the pricing information over a longer period of time. I can promise
you, however, that whenever you require similar information for price corn-
parisons my office will cooperate with yours to the utmost.
Thank you in advance for your help.
Sincerely,
MARVIN GERSTEN, Acting Commissioner.
PAGENO="0063"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1613
FULTON COUNTY,
PURCHASING DEPARTMENT,
Atlattta, Ga., April 28, 1967.
Mr. MARVIN GEIOSTEN,
Acting Commissioner, Department of Purchase,
City of New York, New York, N.Y.
DEAR MR. GERSTEN: Mr. J. Forrest Gee, Purchasing Agent for the City of
Atlanta, Georgia, has referred to me your letter of April 24, requesting certain
drug prices.
Enclosed is the list of drugs which you furnished, llllecI in with Fulton County's
prices. None of these prices goes pack further than the winter of 1965. Most
of the prices are current. The items left blank are drugs not used by the County
and for which we cannot substitute an appropriately similar item.
I am glad to furnish you this informatiän and hope it will be helpful.
Yours, very truly,
F. T. FARRAN, Purchasing Agent.
Item Your price
1. Chlordiazepaxide HCL (Librium) capsules: 25 mg., 500 in bottle $45.00
2. Diazepam (Valium) tablets: 5 mg., 500 in bottle 38. 00
3. Meprobamate (Miltown or Equanil) tablets: 400 mg., 500 in bottle 31.20
4. Chlorpromazine HCL (Thorazine) tablets:
50 mg., 5,000 in bottle 270. 00
100 mg., 5,000 in bottle `7. 50
5. Prochlarperazine HCL (Cnmpazine):
Tablets, 10mg., 500 37.34
Tablets, 5mg., 500 28.79
Injection, 5 mg/cc., 2-cc. amp. 500's 240.00
6. Diphenylhydantoin (Dilantin) in ectable: 250-mg. vial 2. 27
7. Diphenylhydantoin (Dilantin Infantabs): 13.i gr 2 4.00
8. Sulfadiazine tablets: 0.5 mg., 100 in bottle .93
9. Sulfisoxasole tablets: 0.5 mg., 1,000 in bottle 8.75
10. Triple sulfa suspension (trisulfapyrimidines oral susp., U.S.P.) 16-oz. bottle 3. 18
11. Streptomycin sulfate solution: 5 gm./12.5 cc. per vial .88
12. Chloramphenicol (Chloromycetin):
Capsules, 250 mg., 100 in bottle 21.00
Succinate, 1-gm. vial 7. 50
Ophth., 25-mg. vial 1.00
13. Tetracycline HCL capsules: 250 mg., 1,000 in bottle 59.00
14. Penicillin G potassium in.: 1,000,000-unit vial . 20
15. Ampicillin:
Anhydrous or Trihydrate capsules:
250 mg., 500 in battle `18.50
500 mg., 500 in bottle
Oral suspension: 250 mg./cc., 80 cc. in bottle 3.75
Sodium tar in ection: 1-gm. vial 3. 85
16. Nitrofurazone (Furacin):
Soluble dressing: 1 lb 6. 12
Vaginal suppositories: 100/box 2. 10
Solution: 16 oz 5.57
17. Tolbutamide (Orinuse) tablets: 0.5 gm., 500/bottle 2 59. 24
18. Trivalent polio vaccine: 10-dose package 7. 20
19. Isoniazid tablets:
100 mg. 100/bottle 1.29
50 mg., 1,000/bottle
20. Sodium Para Aminosalicylate:
Tablets, 0.5 gm., 750/battle 3. 00
Powder, 25-lb. drum
21. Prednisone tablets: 5 mg., 1,000/bottle 10. 50
22. Diphenhydramine HCL (Benadryl) capsules:
25 mg./1,000 7.00
50 mg./1,000 3.90
23. Reserpine tablets:
0.1 mg./1,000 in battles 33. 21
0.25 mg./1,000 in battles 13. 14
24. Dextroamphetamine tablets (Dexedrine): 5 mg./1,000 in battles 72. 60
25. Phenuzopyridine HCL(Pyridium): 0.1 gm., 1,000 in bottles 48.00
Per hundred.
2 Per thousand.
2 Box of 12.
PAGENO="0064"
1614 cOMPETITIVE PROBLEMS fl~ THE DRUG INDUSTRY
FULTON COUNTY,
CoMMIssIONERS OF ROADS AND REVENUES,
Atlanta, Ga., May 30, 1967.
Subject: Prices paid by Fulton County for purchase of certain drugs.
Hon. GAYLORD NELSON,
Chairman, Monopoly Subcommittee,
Committee oa Labor and Public Welfare,
U.S. Senate, Washington, D.C.
DEAR SENATOR NELSON: This will acknowledge your letter of May 19, 1967,
requesting certain information on drug purchases made by the Fulton County
government. Before answering your questions, I want to supply some back-
ground information about the drug prices presented to your Committee.
Fulton County is not a major purchaser of drugs. In 1966, which is a typical
year, we spent only $22,282.43 for drugs for five County agencies-Haven Nurs-
ing Home, the Department of Public Health, Fulton County Jail, Fulton County
Juvenile Court and Detention Home, Fulton County Public Works Camp. We
do not purchase drugs for Grady Memorial Hospital, which is the major hospital
for indigent persons in Metropolitan Atlanta.
The request for information on drug prices came from Mr. Marvin Gersten,
Acting Commissioner, Department of Purchase, City of New York. It was sent
to the Purchasing Agent of the City of Atlanta. Since the City purchases prac-
tically no drugs, the questionnaire was sent to Fulton County. Actually Grady
Memorial Hospital should have received this questionnaire since they are large
users of drugs and purchase in large amounts. This would have given you a
comparison between large drug purchasers. Grady Memorial Hospital is operated
by an independent Authority which is financed by Fulton and DeKalb County
governments.
When our Purchasing Agent, Mr. F. T. Farran, received the questionnaire,
he noted the urgent nature of the request. To quote from Mr. Gersten's letter,
"Because of our urgent need for this information, we have been forced to under-
take this project on an expedited basis. One of my staff members will be calling
your office in a day or two, and we hope that with the list we have sent in
front of you, you will be able to supply us with as much information as possible
over the telephone." A copy of Mr. Gersten's letter is attached.
Since it came to us through the City of Atlanta, we received it on the same
day that the information was desired. A check of the 25 drugs contained on the
questionnaire accompanying Mr. Gersten's letter revealed that Fulton County
had purchased only 6 in the last 18 months. We compiled information on prices
paid on these 6 drugs from our records and then, in order to complete the
questionnaire, the Purchasing Agent called one of our local drug wholesalers
and asked for quotations. No bids were taken, no purchases were made, and no
public monies were expended as a result of these telephone inquiries. As it turned
out, the prices received on the telephone were not, in most cases, competitive
and obviously would not have been accepted by the County. Therefore, most of
the information presented to you by Mr. Haddad did not represent actual prices
paid by Fulton County for drugs purchased and, therefore, should not be used
by your Committee as a basis for action.
To turn now to the specific information requested in your letter, of the ten
drugs listed in the attachment to your letter, Fulton County purchased only
two:
Tetracycline Hydrochloride 250 mg (1,000) .-On January 31, 1967, Fulton
County accepted a low bid price of $35 per 1,000 for 7,000 tablets. The price of
$59 per 1.000 shown on your tabulation is incorrect. The drug was supplied by
Premo Pharmaceutical Company, which as we understand it manufactures the
drug supplied. Bids were requested on a generic basis.
D'iphcnhydramine Hydrochloride 25 my (1,000).-On January 31, 1967, Fulton
County accepted a low bid price of $7.00 per 1,000 for 5,000 tablets. The drug
was manufactured by Parke-Davis. Bids were requested on a trade-name basis
(Benadryl).
In answer to your other questions:
(1) Does your city buy on a generic or a trade name basis?
Fulton County buys drugs on both a generic and a trade name basis. A re-
view of recent requisitions from County physicians indicates that about a's many
drugs are requested on one basis as another. As our institutional physician,
PAGENO="0065"
COMPETITIVE PROBLEMS IN THE D'RTJG INDUSTRY
1615
Dr. William Reid explains, in certain situations he feels that certain `trade-
name drugs are superior to their generic equivalents and in his judgment trade-
name bidding is essential. In certain other instances, there is no generic equiva-
lent for some of the drugs we purchase. This was obviously not true of the list
attached to your letter.
(2) What is the trade name and manufacturer of the drugs purchased?
See information above.
(3) Were sealed bids solicited or was purchasing done through negotiation?
All drugs were purchased on the basis of sealed bids requested after public
notice and advertisement.
(4) What complaints were received regarding any drugs purchased during
the past year?
We have received no complaints regarding drugs purchased.
(5) What was the volume of drugs purchased during the past twelve months?
As noted above, during 1966 Fulton County spent $22,282.43 for drugs for
County agencies.
It is unfortunate that the purpose for which the information was requested
was not made known to us. If this had been done, we would not have obtained
information over the telbphone but would have used only bid information from
our files. In addition, the time element was totally unreasonable. We were re-
quested to supply information on the very day that the request was received.
I apologize for the length of this reply but unfortunately there has been a great
amount of misleading publicity regarding drug prices allegedly paid `by Fulton
County and because of this I felt that a full explanation was justified. I would
apreciate it very much if you would make this information available to the
other members of your Committee and to the representatives of the press who
attend the Sessions of your Committee in order that the record may be made
clear regarding Fulton County purchasing practices.
Sincerely,
JAB. H. ALDREDGE,
Chairman, Board of County Commissioners.
Senator NELSON. As I said in the beginning, the New York City
survey was attacked vigorously in various parts of the country by the
Pharmaceutical Manufacturers Association.
In reviewing the New York City survey, I have concluded that the
attacks upon it were not valid with the exception of some minor points
and one mistake in Fulton County which was not the fault of New
York City and not the fault of the witness before this committee, Mr.
1-Iladdad. The rest of the survey stood up very well in our cross checking
directly with the cities involved, but in order to avoid any suspicion
of the survey at all, we conducted this survey out of my office with
29 cities and municipalities. I think that since the first survey has
come under such attack by the Pharmaceutical Manufacturers Asso-
ciation, they are entitled to, and ought to, respond to this survey.
And if they are not prepared to respond today, we would be happy
to set a date for the Pharmaceutical Manufacturers Association to
respond.
STATEMENT OP LLOYD N. CUTLER,, SPECIAL COUNSEL, PHARMA-
CEUTICAL MANUPACTURERS ASSOCIATION, WASHINGTON, D.C.
Mr. CUTLER. May I make a brief comment, Mr. Chairman?
\Ve just received this schedule and~ your statement when the hear-
ing began, and clearly we cannot comment now on whatever factual
issues there may be between the committee study and our own evidence.
Perhaps two of our economic witnesses, however, Professor Mark-
ham and Professor Firestone, may be able to answer questions you
wish to put to them as to the conclusions, some of the economic con-
Si-280-68-pt. 5__-_5
PAGENO="0066"
1616 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
clusions, you draw from this data, assuming that there is no factual
dispute between us.
I think I should mention though, sir, that the table that you have
submitted is not a comparison of prices paid for the same drug in the
sense of the same drug product offered by the same manufacturing
source to two different buyers.
Senator ~\ELSON. Several instances are.
Mr. CUTLER. They may be. but. the table does not disclose that. The
table simply gives a generic, name of a drug and then lists prices.
Senator Nni~sox. Just so the record will be clear, let me remind von
that I recited in my statement. those cases where the same seller, the
same manufacturer, sold drugs for a dramatically different price to the
two different buyers.
Mr. CUTLER. I don~t recall any such statement in the 1)1eI)ared state-
ment von lust. delivered. Senator Nelson. You make a aeiuerai stale-
ment which is doubtless correct that. at the bottom of page 3. neither
does the source of the drug explain the difference since the same sup-
pont frequently charges different prices to different buyers of the
same drug. That may well be. But. the cases you give are cases in
whicl~ von identify a generic, name for a drug, which may have been
offered by different buyers. In fact you indicate in some of the illustra-
tions tha.t they were from different sources.
Senator NELSON. That is correct.
Several cases are pni~c1iases from different, buyers and in the very
dramatic one where it was S0.51 versus $17 or thereabouts. one was
generic and the other one was a. tradename drug. But I did read
this sentence and I will read it again
"For example. in the case of chlordiazepoxide (25 lug., 5,000's) al-
though Los Angeles bought quantities amounting to $32,625.86," tli is
is on page 5. `worth, they were paying $25.50 per unit while Pliila-
delphia, which bought only $12,210.11 worth or one-third the quantity,
was paying only $18.50. It is interesting to note in this case that 1)0th
cities were supplied by the same firm."
Mi'. CUTLER. rfliat is correct, sir.
Senator NEIsox. Now. there are several examples of that iii the
survey we had. There are limitations on how much information von
can put on a chart.
What. the survey indicates, among' other thing's, is that in several in-
stances the cities have bought from generic companies at a dramatical-
lv cheaper price than from brand name companies. It has ln(licatesl
that in many instances the same supplier, in the same year, will sell
his drug, brand or generic, to different buyers at a dramatically differ-
ent. price with no relationship to quantity pui~ciuased. In fact, in the
instance I just read, the purchaser of three times as much paid a sub-
staiutial percentage more. What it indicates is that there doesn't seem
to be any rationale to it. We hadl the Defense Supply Agency buy-
ing two or three hundlred thousand dollars worth in less than a year's
tIme paying more than a city that bouigltt a very minor amount. The
survey confirms what the first. demonstrated: that is, that. there seems
to 1)0 no rationale at all in the pricing practices of the dm2 companies.
But. I will be happy to have any explanation for these differences
from the. drug industr. The economists have addressed themselves in
PAGENO="0067"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1617
great detail to the intricacies and economics of this industry, and I
would appreciate some explanation from them as to this situation.
Mr. CUTLER. Some of today's witnesses may be able to comment on
the conclusions you draw but as to any factual dispute between us we
would like an opportunity to study this chart and to respond to you at
our next session.
Senator NELSON. Fine.
As I understand it, Mr. Cutler, you desire to have Prof. Paul Coot-
ner, Mr. Plot.kin, Mr. Gordon Conrad, and Professor Markham on a
panel at the same time?
Mr. CUTLER. Yes, sir.
Senator NELSON. And would you identify the witnesses for the
record?
Mr. CUTLER. Yes, I shall.
We have three groups of witnesses for you today, Mr. Chairman, to
testify as to various economic issues that have arisen iii these hearings.
The first group consists of the four gentlemen you have just named
who will testify to the relationship they find as economists between the
existence of above-average risk and the existence of an above-average
rate of return in that industry, and to the presence of above-average
risk in the pharmaceutical industry.
These gentlemen are, first, on my immediate right, Prof. Paul Coot-
ner who is a professor of finance at the Massachusetts Institute of
Technology. To his right, Mr. Irving Plotkin, of the Arthur D. Little
Co., in Boston, the consulting economist. To my far left, Mr. Gordon
Conrad, also of Arthur P. Little, and to my immediate left, Prof. Jesse
Markham, professor of economics at Princeton.
We would like to begin with a brief statement of the theory of the
relationship between risk and return by Professor Cootner.
If we may, since all of these initial four statements are quite short,
Mr. Chairman, we would like to have them delivered as a group, and
then have the entire four submit to questions rather than have ques-
tions put to one which could better be answered by another. But we
will, of course, abide by your wishes in the matter.
Senator NELSON. I have tried this so many times. It means we have
to go back and reread and recheck the statements and if I don't have
time to write down a question that occurs to me at the time the witness
is testifying, I can't remember it later. But, if I ask a question of any
one of the witnesses which someone else is better qualified or prepared
to answer, I am prefeetly happy to have that person respond to the
question, and if that question is more applicable to a subsequent witness
I would be glad to set it aside. But I would rather have the record show
the questions and answers in the text as I go along, because it makes it
easier to read the record and it makes~ it possible for me to ask the
questions which subsequently I may forget to ask.
Mr. CUTLER. All right, sir.
~Senat9r NELSON. If someone else can answer the question just let the
witness identify himself for the record so that the record will show
who is answering the question.
Mr. CUTLER. Right.
We would like to begin then with Professor Cootner, of MIT.
PAGENO="0068"
1618 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
STATEMENT OF PAUL H. COOTNER, PH. D., PROFESSOR OF FINANCE,
MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE,
MASS.
Senator NELSON. I am very pleased to welcome this distinguished
panel of economists. The committee recognizes these gentlemen as
men with distinguished professional credentials and we are pleased to
have you before the committee today. Professor, your short biographi-
cal sketch will be printed at this point.
(The biographical sketch of Dr. Cootner follows:)
BIOGRAPHICAL SKETCH OF PAUL H. 000TNER, PH. D., PROFESSOR OF FINANCE
Born: May 25, 1930.
Degrees: B.S., Florida, 1949; MA., Florida, 1950; and Ph. D., Massachusetts
Institute of Technoiogy.
Field: Finance, Economic Theory.
First appointed: 1959.
Professional experience:
1955-56: Ford Teaching Intern, Brown University.
1956: Assistant Professor, Northwestern University.
1956-59: Research Associate, Resources for the Future, Washington, D.C.
1959-62: Asst. Professor of Finance, Sloan School of Management, M.I.T.
1962-66: Assoc. Professor of Finance, Sloan School of Management, M.I.T.
1966 to date: Professor of Finance, Sloan School of Management, M.I.T.
Honorary and professional societies: American Economic Association, Phi
Beta Kappa, Phi Kappa Phi, Econometric Society, and American Finance
Association.
Publications:
"Speculation in the U.S. Government Securities Market", Commission on
Money an4 Credit, 1963
"Competition in U.S. Commercial Banking", Commission on Money and
Credit, 1963
The Random Character of Stock Market Prices, Edited by P. H. Cootner,
M.I.T. Press, Cambridge, Massachusetts, 1964
Water Demand for Steam Electric Generation (with G. Löf), Resources for
the Future, 1966
"Social Overhead Capital and Economic Growth", The Takeoff into Su-
stained Economic Growth, 1964
Mr. GROSSMAN. Excuse me, Dr. Cootner, I notice that in looking over
the statements of the other witnesses that they all, in their first para-
graph or so, mention that they are either on retainer or employed by
the PMA and I don't notice this in your statement. Are you appearing
independently?
Dr. COOTNER. No, I am employed by the PMA.
Mr. GROSSMAN. You are retained by the PMA?
Dr. COOTNER. Yes, sir.
Actually it is not exactly true. I am employed by Arthur D. Little.
Mr. GROSSMAN. Who is in turn employed by the PMA?
Dr. COOTNER. Yes.
Senator NELSON. Does Arthur D. Little have a contract to represent
the PMA?
Mr. CUTLER. Yes. The PMA has consulted Arthur D. Little, Inc., to
prepare economic studies relating to these hearings and in the course
of that work with our approval they have in turn retained as consult-
ants Professors Cootner and Markham who are about to testify here.
Senator NELSON. Is there any objection to informing the committee
as to the amolmt of the consultant fees of the witnesses?
PAGENO="0069"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1619
Mr. CUTLER. We will be happy to submit to you whatever you would
like in that regard; yes.
(The information referred to, subsequently received, follows:)
PHARMACEUTICAL MANUFACTURERS AssocIATIoN,
Washington, D.C., February 9, 1968.
Hon. GAYLORD NELSON,
Chairman, Monopoly Subcommittee,
Senate Small Business Committee,
U.S. Senate, Washington, D.C.
DEAR SENATOR NELSON: At page 2695 of the transcript of the hearing `before
the Subcommittee on Monopoly of the Senate Small Business Committee on De-
cember 19, 1967, YOU asked that we provide YOU with infor~mation concerning the
consultant fees received by the economic witnesses who presented testimony
on that day at the request of the Pharuiaceutical Manufacturers Association.
The fact that the PM'A compensated these distinguished economists for con-
ducting studies and for preparing and presenting testimony is, of cour~e, legit-
imate public information and was fully disclosed by PMA kt the outset of the
hearing. We do not, however, see any legislative purpose to be `served by dis-
closure of the amounts paid to particular individuals. Nevertheless, in view of
the request you m1ade `at the hearings, the information is set forth below.
Mr. Gordon Conrad and Mr. Irving H. Plotkin `of Arthur D. Little, `Inc., Pro-
fessor Paul H. Cootner of the Massachusetts Institute of Technology, and Pro-
fessor Jesse Markham of Princeton University were retained as economic wit-
nesses through Arthur D. Little, Inc. We are advised by Arthur D. Little that
the fees billed for preparation and `delivery of the testimony, and for the exten-
sive `studies prepared and presented by the witnesses were computed at the nor-
mal rates generally charged for consultation by these economists. Mr. Conrad
and Mr. Plotkin worked for a number of months on the preparation of `the
detailed studies they presented at the hearings. Professors Cootner and Markham
also spent a silbstantial a:mount of thne in reviewing the `studies and other per-
tinent material and in the preparation of their testimony.
The a'm'ounts billed for the work described above were as follows: Mr. Gordon
Conrad, $13,100; Mr. Irving H. Plotkin, $10,250; Professor Paul H. Cootner,
$6,450; and Professor Jesse Markham, $4,500.
On December 19, PMA also presented testimony by two kdditional economic
consultants, Professor John M. Firestone of City College of New York, and Pro-
fessor Simon N. Whitney of New York University. Professor Firestone spent
several days and Professor Whitney several weeks in preparing material for the
Association and his testimony. The a~noiint's `billed for their `services were as
follows: Professor John M. Firestone, $875 and Professor Simon N. Whitney,
$4,250.
We `are in the process of obtaining certain other information you asked for
during the course of the December 19 hearing. We cannot, however, furnish the
comments you requested at page 2688 of the transcript on the comparative price
study you introduced at the beginning of the hearing until we are permitted to
examine the data in your possession from which the study was prepared. I am
enclosing a copy of my February 6 letter1 to you on this subject and most earnest-
ly request that you reconsider your decision and instruct your staff to make the
data available to us.
Sincerely yours,
C. JOSEPH STETLER.
Senator NELSON. GO ahead, Professor Cootner.
Dr. COOTNER. Mr. Chairman, and members of the subcommittee. My
name is Paul H. Cootner and I am a professor of finance at the Mas-
sachusetts Institute of Technology. My, specialty at MIT is the study
of the effects of risk and uncertainty on financial decisionmaking of all
kinds. I am here to speak generally on research conducted by myself
and others on the effect of risk on the rates of return required to induce
investment.
1 Retained In committee files.
PAGENO="0070"
1620 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
The relevance of a discussion of risk and rate of return to a discus-
sion of pricing in the drug industry may not be immediately apparent,
but I think it soon becomes so if we note that any reduction, voluntary
or enforced, in the prices charged by the drug industry without simul-
taneously reducing its costs will flow through to profits immediately.
If prices are reduced across the board by administrative fiat, profits
will also be reduced.
Now, I do not appear here as an expert on the drug industry, either
with regard to its pricing policy or the riskiness of its investments. It
requires no such expertise, however, to establish the elementary prop-
osition that for an industry selling an item necessary to life and fac-
ing an essentially inelastic demand, changes in price lead immediately
to changes in profit in the same direction. It is the drug industry's role
as a supplier of necessities that apparently occasions this committee's
interest in its pricing policy and it is that same role that establishes the
close relation between price~s and profits.
Senator NELSON. May I raise just a point here?
You stated that it requires no expertise to establish the elementary
proposition that for an industry selling a product necessary to life
and facing an essentially inelastic demand, changes in price lead im-
mediately to changes ~ p~t in the same direction. I wonder if
that can really stand as an unqualified statement? For example, if a
drug is discovered which does something that no other drug does, al-
most immediately-that is, after some advertising, some understand-
ing on the part of the medical profession-that drug then, if it is a
valuable drug, will he in demand in every single case where a patient
needs that type of drug. In fact, drugs are more in demand than other
products, because health as a general proposition is a prime factor
motivating people in pain.
So the demand may be inelastic to the extent that every single person
who needs a drug is willing to pay for it. It is not true to automobiles
or houses or clothes. So I don't quite follow the statement that the
demand is inelastic.
Dr. COOTNER. Well, as a matter of fact, the description you gave
is exactly what I would consider inelastic, that is, that the demand is
there regardless of price, and I think we are in perfect agreement on
the facts.
Senator NELSON. I thought you were using the argument to prove
that this was some kind of a handicap to the industry. Whereas, as
matter of fact, if there is any other industry that faced this kind of
inelastic demand it would be considered a great asset to the industry,
wouldn't he, his product?
Dr. COOTNER. Well, that depends, if he had an inelastic demand and
he had monopoly power lie would certainly like that sort of situation,
but inelastic demand in and of itself is no advantage. For example,
many people believe, and I think it is probably true, that the American
farmer faces essentially an inelastic demand for the same reason, he
does not consider that an advantage because it means, as a matter of
fact, if the supply increases he is forced to take a substantial cut
in price.
Senator NELSON. Not if lie had a patent on potatoes. We will get into
that some later, but I misunderstood what you were saying al3out it.
PAGENO="0071"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1621
An inelastic demand in the case of a valuable drug means a total de-
mand by every single person who needs it. That is not true, to my
knowledge, of any other single product in the marketplace in America.
Dr. COOTXER. But all I am simply saying is that the drug industry,
the total sum of all drugs is very much like the total food industry,
total sum of all food. Any individual product, of course, any individual
food product is not necessary to existence, but all food products taken
together are necessary. But I don't think we are in any substantial dis-
agreement.
Senator NELSON. I see.
Dr. COOTNER. Drugs are necessities, and it is precisely because the
drug industry is a supplier of necessities that I suspect that this sub-
ject is of such great interest to the committee.
Senator NELSON. But there are all kinds of alternative choices in
food. There is such a great variety, you can be a vegetarian or not, you
can consume any number of a thousand different items. However, if a
drug is discovered that performs an important health function, and it
is patented, there is an inelastic demand in the sense that every single
persoi~ in the United States who needs it will want it and will pay for
it, and it is shielded from competition for 17 years. So I just want to
get that sharp and, I think, significant economic distinction into the
record at this time, unless you disagree with me.
Dr. COOTNER. No.
I will continue: My role in these hearings is to relate this price-profit
nexus to its implications for industry's ability to engage in risky, in-
novative activity and continue to attraCt capital. I will be establishing
the meaningful relationships that empirically exist between industry
risk and the rate of return required by investors. It will be left to sub-
sequent witnesses to establish the drug industry's role in this process.
In an economy such as ours, which is largely directed by a large
number of firms and households, all responding to opportunities and
hopes for financial gain, it is inevitable that some will prosper and
others will fail. When opportunities arise, men will jostle each other to
seize them and it is unlikely that they will all share the gains equally.
Some, hopefully, due to skill will choose a better or faster route to
satisfying a need and earn a better return in the process.
In our relatively unregulated capital markets, investors are free
to back the enterprises that they judge will be most successful. In doing
so, they must inevitably make judgments about the management of
the enterprise, the likelihood of success, and the prospective rate of
return. They not only evaluate the particular enterprise but they must
compare it with all other relevant alternatives.
Now clearly, an investor would like to earn a good rate of return on
his money. It does not follow, however that he will always put his
money to work at the highest rate of return available. The U.S.
Treasury is grateful for this "quirk" of investors, for while Govern-
ment bonds offer very high yields by historical standards, those interest
rates are substantially lower than investors could get from buying
the bonds of corporations and still less than those available from
bu ving corporate common stocks.
Investors, for the most part, do not accept these lower yields out of
charity or patriotism. They buy U.S. bonds because those bonds offer
PAGENO="0072"
1622 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
them something that corporate bonds cannot-immunity from the
fear of bankruptcy. In short, such investors are willing to sacrifice
gain in order to avoid risk.
When an industrial entrepreneur forms a new firm, or finances a
new investment, he must attract equity and debt capital from either
his own resources or those of the public. He must convince himself, or
other investors, that the prospect rate of return on the capital will be
adequate. In this industrial investment, as in the purchase of bonds,
the adequacy of the rate of return must be judged in the light of the
risks involved. A man will not prospect for gold or oil at the same rate
of return that suffices for an electric company, nor will he finance the
development of some exotic new technology for the same return as on
shoe manufacturing.
Senator NELSON. May I interrupt here for a moment?
Dr. Coo~n. Certainly, sir.
Senator NELSON. Your statement that a man will not prospect for
gold or oil at the same rate of return that suffices for an electric
company; nor will he finance the development of some exotic new
technology for the same return as he could get on shoe manufacturing,
he would do it for the same return if the risk were the same, wouldn't
he?
Dr. COOTNER. If the risk were the same, yes, sir; that is right.
Senator NELSON. So, do you address yourself to the risk question?
Dr. COOTNER. Right. I chose the particular cases because I thought
and, perhaps, wrongly, that they were prima facie cases where the
risk was greater than in the alternative. That is, I assumed that
prospecting for gold was an obviously more risky, alternative than
electric utility investment.
Senator NELSON. The real test then it seems to me is the consistency
of profits, and the consistency of the companies in the industry, their
capacity to stay in business and make a profit, is it not?
Dr. COOTNER. That is right.
It is the question of the relative risk; yes, sir.
Senator NELSON. Are you going to give us the statistics for the past
few years on the profits in the drug industry and also examples of the
drug companies that have gone out of business because the risk was
so high?
Mr. CUTLER. Mr. Chairman, those sul'jects are going to be covered
by the witnesses who follow immediately after Professor Cootner dem-
onstrates the relationship between risk and return.
Senator NELSON. I read all the statements and I didn't see anything
in any of them that related to the speci~c question that I have raised
here as to how many companies have gone out of business in the past
10, 15, or 20 years, a.nd how their profits have compared in the past 10,
15, or 20 years with profits in industry in general in the country. I saw
the statements, but no specific examples, and I raise the question at
this time so that the other economists will have a chance to think
about it while Dr. Cootner is testifying.
Dr. ~OOTNER. I will proceed. It can be empirically demonstrated
that this is so, but it is more than a passing fancy of this year's crop
of investors, or even this century's crop. This attitude toward risk is
deeply imbedded in our legal code. Thus, while trustees of an estate
PAGENO="0073"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1623
are enjoined to invest the funds entrusted to them as profitably as they
can, they are judicially warned against seeking return at the expense
of taking "imprudent" risk. And the U.S. Supreme Court, in 1909,
stated the problem clearly and forthrightly for the first, but not the
last, time with respect to regulated corporations.
"The less risk, the less right to any unusual returns upon the in-
vestments. One who invests his money~ in a business of a somewhat haz-
ardous character is very properly held to have a right to a larger
return, without legislative interference ~ *
As a matter of fact, the U.S. Supreme Court is rather a new-
coiner to the discussion of economic risk. More than two centuries
ago, Bernoulli, a great student of probability and risk, noted that
the ordinary man would usually reject a highly favorable bet if that
bet required him to take risks which were large in relation to his
capital. Like the genius he was, Bernoulli used this observation to
construct the first technical tools for the evaluation of attitudes to
risk, tools that have been added to by many of the important econ-
omists and statisticians who have concerned themselves with dynamic
problems of economic growth.
The list of economists who have~ concerned themselves wit1h the
question goes back to Adam Smith and reaches forward to embrace
almost all of the important economists of this century of yin ually
every persuasion: Joseph Schumpeter, Irving Fisher, Frank Knight,
John Maynard Keynes and including such diverse contemporaries as
Paul Samuelson, Kenneth Arrow, Milton Friedman, and George
Stigler.
While all of these writers have dealt with this problem, two of
them, Keynes and Schumpeter, deserve special note. Keynes' major
contribution from an economist's point of view was his theoretical
results, but he was also a very practical man, ~ho made (and lost)
several fortunes as a practicing investor. As a practical man, Keynes
was the first that I know of to make an empirical measurement of atti-
tudes toward risk. Like most first attempts, it was a crude attempt, but
he concluded that merchants carrying large inventories of raw mate-
rials required a premium rate of retUrn of 10 percent of capital, over
and above normal rates of return, because of their exposure to the risk
of price fluctuations. Surprisingly enough, work that I and others have
done in the last few years, using more sophisticated mathematical tools
suggests that Keynes' crude estimate is not far off at all.
Schumpeter's work was of a much different type but of direct rele-
vance to the questions faced by this, and other investigative commit-
tees. Schumpeter was deeply concerned with dynamic problems of eco-
nomic growth, and the role of risk taking and entrepreneurship on
innovation and investment. It was he, more than any other writer, who
stressed the fact that the willingness, to innovate played a vital role in
the rapid growth and efficient operation of Western economics. He also
was quick to draw the concomitant conclusion: that any regulations or
legislation that inhibit risk taking, also inhibit innovation and the rate
of economic growth.
Now I hasten to add, that neither Schumpeter nor I, nor. indeed any
responsible economist, will argue that industrial abuses should not be
corrected, when found. The important elements are (1) that one should
PAGENO="0074"
1624 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
not be surprised to find large average profits in risky enterprises or to
take such profits as evidence of abuse of market power, and (2) if
abuses are found, one must take care to eliminate only excessive, and
not necessary returns on investment, if one does not wish to strike at
the well springs of risk taking and growth.
Senator NELSON. I don't know whether this is an abuse or not, but I
will give you the facts and ask for your comment.
Rhone-Poulenc, a French firm, discovered chlorpromazine. They
licensed a company in the United States to produce it, and they licensed
a company in Canada to produce it, each of them with the exclusive
market in their respective countries. So neither the company in the
United States nor the company in Canada spent any money on
research. It was just a question of each one of them having an exclu-
sive market, both in the same continent with adjoining borders.
The price charged by the U.S. licensee for 25 milligram tablets to
the Defense Supply Agency was $32.62 a thousand. The price of the
Canadian Licensee to Canada's Department of Veterans Affairs was
$2.60 a thousand.' Ca.n you explain to me the economics of that?
Dr. COOTNER. I obviously can't, sir. I haven't done any study of the
drug industry. These figures, of course, are very much `like the figures
you presented at the beginning of the statement~, and I trust that the
PMA will deal with this subsequently, but I have no expertise in the
drug industry and I couldn't comment on it.
Senator NELSON. I didn't think I was asking a question that re-
quired expertise. This is general economic theory. Under what eco-
nomic theory would you justify vast price discrepancies between two
licensees of the same product, neither of whom did the research, both
on the same continent, with adjoining borders. Why is the Canadian
Government paying $2.60 a thousand for tablets, and the American
Government $32.60 a thousand. There must be some economic theory
under which each of these firms can succeed in charging their respective
prices and both still makea profit.
* Dr. COOTNER. Well, If I hadto comment I wOuld have to speculate
but, of course, there is always the possibility that there were differ-
ences in costs of the two firms or differences in quality. If there were
no differences in costs or quality you would have to find something
else to explain it if you could. But the point is I cannot comment on
it without knowing more about the difference in costs or quality. If
they are identical products then I think it would require some ex-
planation.
Senator NELSON. WTell, they are licensees for the same product, and
used all the background research and information that Rhone-Poulenc
had. Could the differential be explained by the fact that each one of
them has exclusive control of the market. There is no competition in
either `country, and yet one charges almost 15 times as high a price
as the other. The industry keeps saying, among other things, that,
"We need money for research." Well, they didn't do the research on
this product, and I assume both the Canadian and the American com-
panies are making a profit.
See chart, p. 1746, Infra.
PAGENO="0075"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1625
Mr. CUTLER. Mr. Chairman, Professor Cootner is clearly not quali-
fied to answer on this and perhaps I am not either. But I do know some
of the factual assumptions put into your question are not correct.
I do know, for example, that Smith Kline & French did a very large
part of the research needed to prove that Thorazine would be useful
as a mental illness drug. That Smith Kline & French did a very large
part-in fact all of the U.S. work, necessary to prove the safety' and
efficacy of that drug and put it through the new drug approvals of
the Food and Drug Administration.
There are differences. I am not competent to say how deep they
were or how significant they were, but~ this witness isn't either and if
you want further explanation on this point, we will be glad to sup-
ply it.
Senator NELSON. Well, the research you are talking about was the
clinical testing and other studies required of them to qualify for a
New Drug Application.
Mr. CUTLER. I believe it goes deeper than that. I believe, although
I am not specifically informed on this subject, that Rhone-Poulenc
had a chemical entity which it patented, but that its value in mental
illness was largely the discovery; certainly it wits perfected by Smith
Kline & French in this country.
Mr. GORDON. In a study made for the Pharmaceutical Manufac-
turers Association by Paul DeHaen, which was printed in the Amer-
ican Professional Pharmacist in July of 1964, Mr. DeHaen listed
Thorazine, originator, Rhone-Poulenc, France, marketed by Smith
Kline & French in the United States.
Also in the Kefauv.er committee's hearings, part 16, pages 9024,
9025, and the following pages, Dr. Heinz Lehmann, of the Verdun
Protestant Hospital, Montreal, Canada, clearly states that the new
uses for the drug were developed there.
Mr. CUTLER. Mr. DeHaen, as I understand it, has another column
next to originator called developer.
Mr. GORDON. That is right.
Mr. CUTLER. And that column has Smith Kline & French, but
neither of us really has the facts at hand. Why don't we get the facts?
My understanding is that Smith Kline & French did a very large
part of the work required to prove the value of this d.rug in mental
illness.
Mr. GORDON. Incidentally, it may be interesting to know that I
called Mr. DeHaen and asked him why lie changed the word "mar-
keter" from his earlier study to the words "marketer and developer"
in the later study, and he said it was done at' the request of the
Pharmaceutical Manufacturers Association.
Mr. CUTLER. It may be a more accurate description.
Senator NELSON. Well, if there is an explanation of why the same
drugs are being sold, one of them for almost 15 times as much as the
other, the committee would appreciate having it.1
Go ahead.
Dr. COOTNER. Thank you, sir.
For, if one reduces the rate of return below that commensurate with
the risk involved, that industry will no longer be able to attract the
1 See app. III beginning at p. 2129, infra.
PAGENO="0076"
1626 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
same volume of capital or investors will seek to reduce their exposure
to .risk by a greater conservatism or sluggishness in pushing down
uncharted paths.
But if I tell you that it is important to distinguish between exces-
sive rates of return and those necessary to attract venture capital, I do
you no service unless I can tell you how to make that distinction. The
fact that the U.S. Supreme Court has ruled that return must be com-
mensurate with risk in a whole string of important cases has not ended
the regulatory debates on rates of return. Federal Judge 1-larold
Leventhal has published a masterful, and lengthy summary of the legal
problems in the regulatory context in the Yale Law Journal of May
1965.
My earlier quotation was from 1Viilcox v. Consolidated Gas Com-
pany. In 1923, in a case involving Southwestern Bell Telephone Co.,
Justice Brandeis explicitly referred to the need for a firm to receive
"an allowance for risk incurred" if it were to attract capital. In Blue-
field Water Works (1923) and Hope Natural Gas (1944) the Court
again insisted that returns are to be concomitant with risks undertaken.
To quote from the Blue field Water TVorks case:
A public utility is entItled to such rates as will permit it to earn a return on
the value of the property which it employs from the convenience of the public
equal to that generally being made at the same time and in the same general part
of the country on investment in other business undertakings which are attended
by corresponding risks and uncertainties-
That is emphasized by me-
but it has no constitutional right to profits such as are realized or anticipated in
highly profitable enterprises or speculative ventures.
But despite these injunctions, courts have had difficulty implement-
ing them because there was no work done to make the concej~ts opera-
tional. To properly address this question requires an expansion of the
quantitative approach of Lord Keynes on a more sophisticated basis.
It requires, first, the measurement of risk and, second the establish-
ment of a relationship between that risk andthe required rate of return.
Economists have made attempts to solve these problems although
most of these attempts have been of recent origin.
First of all, this work was hindered by the slow development of the
necessary statistical tools. But even after those tools were developed,
further work had to await the general availability of the large amounts
of statistical dat.a necessary to establish propositions in which probabil-
ity and risk are involved. For example, one of the earliest of the re-
cent studie.s was a test by PrOf. George Stigler, of the University of
Chicago, which, despite some very imaginative attempts to extract in-
formation from inadequate data could not extract enough to get statis-
tically significant results. Good data first became available not for
manufacturing industry but for financial markets. In 1959, Lawrence
Fisher explained interest rate differentials on corporate bonds by risk
varia.bles, thus measuring the extra rate of return asked by bond in-
vestors for investing in more risky companies. In 1960 1 did related
work on holdings of commodity inventories related to the earlier work
of Keynes.
By the time Prof. Daniel M. Holland and I turned to the ques-
tion of industrial risk in 1962, we had unusual access to a large body
PAGENO="0077"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1627
of statistical data and could derive such significant results. Since our
study the availability of data has radically improved and studies
have i'nultiplied. Fred Arditti of Rand Corp., Prof. Shannon Pratt of
Indiana University and Portland State College, and Irving Plotkin
and Gordon Conrad who you will hear after me, among others, have
all completed major investigations into this question since 1964.
One of the surprises of all these recent studies-or at least it might
surprise those who have heard economists disagree in public testi-
mony-is that they all reach agreement on the broad outline of their
results. The rate of return does rise as risk increases, and even the
quantitative relationships show only minor differences. There are few
settled empirical truths in economics, and I am sure that differences
will arise and refinements will be made as time goes on and work
is reviewed, but the competition is no longer between different "be-
liefs" or "feelings," but a comparison, of competing hypotheses and
empirical tests.
Since the Conrad-Plotkin study is very similar to Professor Hol-
land's and mine through quite independently conceived, and will be
presented in some detail, I am only going to discuss some of the
main problems and implications of this type of work before you hear
from Mr. Plotkin.
One of the greatest obstacles to a scientific study of the relation
between risk and return is the difficulty of establishing a meaningful,
objective definition of "risk." Consider an investor planning to pro-
duce and sell a particular product. In choosing the size of the invest-
ment he must estimate, among other things, the size of the market, his
level of costs, and the competitive response of others in the industry.
In making his plans he will draw on whatever experience he has or
on any market research he can perform. Such analysis may lead him
to some "best guess," or it may permit him to place upper or lower
limits on the potential demand, but there will always remain a large
number of possible outcomes which have at least some probability of
occurring. There will be similar uncertainty about the levels of costs
that he will experience, or the price or production response of his
competitors. In making his final plans, he must weigh all these pos-
sibilities in estimating the profitability of the investment; whether lie
does so by "intuition" or "judgment," or by mathematical calcula-
tion. Finally, after making that estimate he must decide whether the
venture is worthwhile.
The whole theory about the relation between risk and rate of re-
turns on this final decision. In essence, it turns on the belief that if our
investor should examine investments in two different industries with
identical return, he would prefer the investment which was least likely
to result in a substantially reduced return if adversity befell the proj-
ect. If this is so we will find that the more risky project will not be
undertaken until the need becomes so pressing that the investor can
anticipate a significantly higher rate of return than he could on the
less risky project. If returns are out of. line with risk, the flow of
capital into the riskier industry will be cut back until profits return to
the desired relationship. In that case, whenever capital is flowing into
both industries, we will observe the more risky industry earning a
higher rate of return. Conversely, if the more risky industry fails to
PAGENO="0078"
1628 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
earn a higher rate of return it will not be able to attract new capital
and will decline or stagnate.
Mr. GORDON. You are not saying that because an industry earns a
higher rate of return it is necessarily an indication that it is a riskier
industry?
Dr. COOTNER. No, sir; I am not saying that.
Mr. GORDON. All right. I only wanted to get that clear.
Dr. COOTNER. This is only a hypothesis and it could only be tested
if we quantify the "risks" I just outlined in words. Now it turns out
that words are much less precise than mathematics, but much easier to
understand. When we turn to implementing this verbal definition we
find that there are a variety of precise formulations of the result. Some
of these were used in my study with Professor Holland and others are
described in detail in the Conrad-Plot.kin paper you will soon hear
about. No one of these acceptable measures is perfect or ideal, but what
is encouraging is that the relationships which are observed are all very
similar-that is, they are not affected by the particular definition used
and all lead to similar conclusions.
This basic conclusion is that as risk rises so does the required rate of
return. I should stress that this is an average relation-a statistical re-
lation that is subject to chance variation. While we develop a mathe-
matical expression which relates the two concepts, every industry is not
described perfectly by the relation. But subjecting this relationship to
all the normal satistical tests confirms the result. The indicated rela-
tionship is a real one; it is not spurious or accidental. Finally, I should
stress that this is not a complete theory of corporate profit. There are
other risks besides the ones I have investigated. Just to name one, the
marketability of an investment in Government bonds is substantially
greater than one in plant and equipment and this alone would require
a greater return for industrial investment even if it was equally risky.
Furthermore, there are market imperfections which can influence
particular observations. But none of these things affects the validity of
the partial relation measured here.
1-laying stated these results, I would like to spell out their implica-
tions. First of all, the relationship is an "average" or "long-run"
phenomenon. It is particularly true in risky enterprises that things may
not turn out as one expects and this accounts for a good deal of the
scatter of points about our line of best fit. Low returns which arise
from the impact of such risks will not immediately affect investment
any more than the outcome of any particular flip of a coin changes
the outlook for future flips. However, if one were to decide, by admin-
istrative action, to reduce the average rate of return in a risky, competi-
tive industry without at the same time reducing the risks, we would
find an immediate impact on that industry's investment policy.
Mr. GORDON. You say on page 9 that if one were to decide by ad-
ministrative action and so on and so forth. First of all, practically
all the testimony before our Committee indicates there is very little
price competition in the drug industry. Is that your understanding?
Dr. COOTNER. I am not qualified to comment on that;
Mr. CUTLER. Could that question be addressed to Professor Mark-
ham, Mr. Gordon, when he testifies as soon as we finish on this risk-
return relationship?
PAGENO="0079"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1629
Mr. GORDON. Well, this is important to what Dr. Cootner is say-
ing, but let me go a little further then. He says he doesn't know. Let's
accept that.
Now, the clause "without at the same time reducing the risk," it
seems to me misses the whole point. The risks in drug marketing al:e
largely derived from existing market institutions and if these are modi-
fied the risks will be reduced by the action of the same forces. If prices
are lowered, say, by introducing vigOrous price competition or by
Government action and the spread between cost and price is reduced,
U lien there will be less money spent on advertising and promotion.
This will in turn reduce the risk of obsoleting certain drugs.
This particular point was discussed more fully by Prof. Henry
Steele of the University of Houston in his presentation on February
14, 1964, before the Special Committee on Drug Costs and Prices
of the Canadian House of Commons.
I ask, Mr. Chairman, that pages 2482 to 2484 of Professor SUeele's
brief be included in the record at this point.
Senator NELSON. They will be included in the record.
(Data referred to above follows:)
[Excerpts from Canadian House of Commons, first session, 27th Parliament,
Special Committee on Drug Costs and Prices, Feb. 14, 196,7, pp. 2482-2484]
(By Henry B. Steele, Associate Professor of Economics, University of Houston,
Houston, Tex.)
Almost any drug will sell, if promoted intensely enough, at least for a while.
Drug firms complain of the high rate of obsolescence of drugs, and argue that
such risks justify high profit rates. The argument is not irrelevant under present
circumstances, but the risks of obsolescence are not inherent but result from
the way in which drugs are developed and promoted. High risks do not justify
high profits in this instance because the risks and profits are both symptoms of
the same disease: sales promotion rivalry substituting for price competition.
The chief reason for the high turnover rate among drugs is, I suspect, to be ex-
plained along these lines: advertising alone can sell physicians on a drug, if
intensive enough, but any number can play at the advertising game, especially
when brand names can be used to obscure the relationship between or even
the identical nature of nominally unique substances. The greater the accumulated
experience with a given drug, however, the more likely it is that its untoward
actions will become known. However, `if the rate at which new products is intro-
duced is as great as the rate at which publicity is given to the mischief caused by
existing products, the sales of the new products W'jil increase as that of the old
products declines, so `that the total cash flow need not `suffer.'
On time other hand, as any businessman knows, advertising rivalry e~n sub-
stitute-perhaps entirely-for genuine price competition. Price competition is a
good servant to the consumer, but a harsh master to the producer. 1-lence sellers
tend `to avoid it as much as possible under normal circumstances, and it generally
prevails only where it is forced upon them by the structure of the market:
numerous small sellers, none dominant; no collusion; no barriers to entry of
`new firms or expansion of existing firms. Where sellers are fewer and larger;
w-here barriers exist to entry by new firm; where legal devices exist to facilitate
a community of interest in price and production policies-under these circum-
stances, the forces which compel producers to undertake active price competition
will be so weakened that rival firms will attempt to maintain or enlarge their
share of various product markets by raising costs instead of lowering prices.
Advertising is inherently less destabilizing an arrangement than price corn-
petition. Some segments of the market may be loyal to a given brand even in
1 In the United States, brand-name sellers had to be compelled by law to give proper
prominence to generic names in advertising. But brand-name sellers do have their uses for
generic names. A firm, for example, may advertiseby brand name, but issue warnings under
the 5enerie name only. Pfizer and Wyeth adopted this opaque tactic for a triacetyloleando-
mycin warning.
PAGENO="0080"
1630 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the absence of advertising; other segments can be induced to prescribe only
by increasingly provocative sales appeals. In general, sales can be increased by
increasing advertising coverage, attracting new buyers while retaining the old,
and perhaps even reinforcing their allegiance to the product. In the case of
price competition, however, even though there may be a substantial segment of
the market which is not highly price-sensitive and would buy the product at
relatively high prices, in order to attract additional and price-sensitive custom-
ers, the prices which all customers pay must be reduced. Under such conditions,
the controlling considerations relate to the price sensitivity, or price elasticity,
of the total market demand for the product, and the expected price elasticity of
the net demand schedule which the prospective price-cutter estimates that lie
will face after all his rivals have consummated their reactions to his price
reduction. Only if demand promises to be quite sensitive, or relatively elastic, in
response to price reductions, will a particular rival feel justified in gambling
on a price cut. Even so, one or two moderate price reductions will ordinarily
be sufficient to traverse the region of sufficiently elastic demand and hence to
exhaust the possibility of further consumer-benefitting price reductions by the
rivals.
The major difference between the two strategies is therefore that price com-
petition benefits consumers through lower prices and higher output, while it
reduces the profit levels of producers to competitive rates-an outcome consistent
with maximum efficiency of resource allocation in an economy. But rivalry in
extravagant marketing campaigns raises costs and prices, benefits of advertising
media at the expense of consumers, and possibly also at the expense of company
profits, and keeps the total consumption of the products of the industry at rela-
tively low levels. In fact, the effects on profits of the two strategies may be
the same in the long run: initially high rates of return on investment serve
as a stimulus to efforts to increase output and market share; price reduc-
tions will directly reduce profits to equilibrium competitive levels; increased
advertising budgets, which are mutually offsetting in the same way as com-
petitive price reductions, except that they do not reduce costs to consumers
and increase quantities produced and consumed, may eventually reduce profits
to no more than competitive levels. Hence, monopoly prices may not necessarily
mean monopoly profits, but simply excessive sales promotion budgets. A monopo-
list does not always make monopoly profits-he does so only to the extent that he
is efficient, and one of the great attractions of monopoly is that it reduces or
largely eliminates the penalties which a competitive market imposes on in-
efficiency.
It should be noted in passing that while price competition benefits consumers
and while advertising rivalry may benefit no one except to the extent that it
attracts more resources into the advertising industry, it is not suggested that
monopolistic rivals are motivated by the desire of private gain at the public ex-
pense, while competitive producers are motivated solely to serve society. The
producers' motivations do not differ. Competition is always a competition in the
hope of establishing a monopoly, but where the structure of the industry rules out
the possibility of monopoly, the ambitions of competitors must fail of fulfillment.
The task of public policy is to adapt market structures in such a way as to pre-
serve the vigor of competition while securing the public against the dangers of
monopoly power on the part of the too-successful competitor.
Mr. GORDON. Let me just read a few sentences from Dr. Steele's
statement.
He says:
Drug firms complain of the high rate of obsolescence of drugs and argue such
risks justify high profit rates. The argument is not relevant under present cir-
cumstances because the risks of obsolescence are not inherent, but result from
the way in which drugs are developed and promoted. High risks do not justify
high profits in this instance, because the risks and profits are both symptoms of
the same disease, sales promotion rivalry, substituting for price competition.
Do you want to comment on that?
Dr. COOTNER. Well, first of all, if I can call your attention back again
to the sentence which we are discussing, my statement referred spe-
cifically to a risky competitive industry.
PAGENO="0081"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 1631
Now, the statement that you are making makes all kinds of empiri-
cal judgments which I am not qualified to evaluate. As I say, if you
can show that you will reduce risk at the same time that you reduce the
rate of return that is defensible, but that is an empirical question.
Mr. GORDON. That is right. I made an~ empirical statement. You are
making an assumption which apparently is not necessarily grounded
in reality; isn't that correct?
Dr. COOTNER. Well, it is possible that either my statement is not
realistic or your statement is not. That is, I can't-
Mr. CUTLER. Professor Markham would be glad to take you on right
now.
Dr. MARKHAM. I am not hankering to take you on, sir. I think I
would quibble with you that the statement you have made is an em-
pirical one. It is the statement of an opinion. What we are talking
about in this presentation largely are matters of fact. I did not see
the statement you read, any empirical evidence that says that risks and
rate of returns empirically are derived from the same symptom or
same disease, whatever the statement was.
Mr. GORDON. I didn't say it was an empirical statement.
Dr. MARKHAM. He did.
Dr. COOTNER. I made the statement that it was a hypothesis, if the
hypothesis held then we ought to find evidence from the data and if we
can't then it is a different proposition that I am stating. I think the
conditions you referred to are clearly met in my statement which says
if the industry is risky and competitive then you cannot lower the
profits by administrative action without at the same time reducing the
risks, if you can propose such a technique for reducing risks and
profits at the same time that is a matter that can be discussed empiri-
cally and tested and established in connection with evidence.
Mr. GolmoN. Let's go a little further.
You say here that the high risk-high profit relationship is an aver-
age or long-run phenomenon. Do you mean by this that short-run
losses or great gains may well occur, but over the long-run the ex-
traordinary successes and failures will balance out?
Dr. COOTNER. That is correct.
Mr. GoiwoN. All right.
Now, isn't it true that this balancing out seems to be absent in the
high-profit drug industry?
Dr. COOTNER. Well, that is an empirical question.
Mr. GORDON. Can you tell us, say, for the past 10 years about 1 year
in which the average profits of the drug industry fell below the aver-
age profits of industry as a whole?
Mr. CUTLER. Mr. Gordon, Professor Cootner is trying to testify as
to the theory of the relationship between risk and return as an eco-
nomic principle. We are then going to present to you, as you have seen
in the papers we have here, through the immediate succeeding witnes-
ses, the measurements of risk we have tried to make in the drug indus-
try, that is empirical data which you may disagree with.
As to the conclusions which you would draw or we would draw from
that data on risk in the drug industry Professor Markham is prepared
to testify about that. But I think we would have a more logical, orderly
discussion if we could first establish the theory and then go on to the
81-280-68-pt. 5-6
PAGENO="0082"
1632 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
empirical demonstration and then the conclusions that each of us may
draw as to the reasons and the absence or presence of risk in this
industry.
Mr. Goirnox. All right. Dr. Cootner, let's go back for a moment to
the second paragraph on page 2 of yourprepared statement:
111 our relatively unregulated capital markets, investors are free to back the
enterprises that they judge will be most successful. In doing so, they must inevit-
ably make judgments about the management of the enterprise, the likelihood
of success, and the prospective rate of return. They not only evaluate the particu-
lar. enterprise but they must compare it with all other relevant alternatives.
Now, these relevant alternatives to the investors you mentioned cover
all industries as well as firms within the drug industry, is that correct?
Dr. COOTNER. That is correct.
Mr. GORDON. In other words, the range of choice involves all firms
and all industries. Comparisons as to investment opportunities are
not confined within one industry. Am I correct in that?
Dr. COOTNER. That is correct.
Mr. GORDON. If I were to suggest, Dr. Cootner, that since investors
invest in firms, their main concern is with the profitability of the firm
as a unit, not with just one item in that firm's catalog. Would you
accept that as a reasonably accurate statement?
Dr. COOTXER. Sure.
Mr. GORDON. If the firm is the relevant business entity, the high-
risk firms are those that show more extensive fluctuations in profits
than do the rest of the firms in the economy, sometimes doing better
than average, sometimes doing worse, but over the long run doing
perhaps even better tha.n average to compensate for uncertainties. Is
that a reasonable interpretation of your view?
Dr. COOTNER. That is a reasonable interpretation of my views if
we agree on the yardstick. Let me make the following hypothetical
proposition. Say that we set up a company to explore for oil, and on
our first try we were to hit a major field. That would establish a very
high rate of return on our rather small initial investment.
Now, we might have large fluctuations in the return on subsequent
investments without any necessity for the company's average return
on all of its assets to fluctuate above and below the average rate of
return earned by all industry. The additional investments continue
to show the same high risks, but the average rate of return remains
above the average.
It is just as if you have a run of luck in a flip of a coin, you have
had that run of luck and you are relatively well off, it is behind you.
You will still face the same fluctuations henceforward and yet you may
maintain a. higher rate of return. With that proviso, I agree per-
fectly with your statement.
Mr. GORDON. Now, in the study that was made by Mr. Conrad and.
Mr. Plotkin on page 11, the top of the page, you emphasize-
Mr. CUTLER. Mr. Gordon, if we can go two more pages Mr. Plotkin
is going to testify on that. Wrouldn't it be more logical to bring the
question up then?
Mr. GORDON. Fine. There are some questions that have been brought
up by Dr. Cootner here which have relevance to the drug industry,
but. we will bring them up then.
PAGENO="0083"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1633
Mr. CUTLER. He is not going to go away. He will be here ready to
answer any question, but at that point you will have had this state-
ment, this study, put into evidence, and Mr. Plotkin will have had a
chance to explain the theory of that study and answer any factual
(1uestions you may have about it.
Mr. GORDON. In other words, Dr. Cootner is not going to speak
about the drug industry at all; is that correct?
Mr. CUTLER. If you ask him questions about this study we are about
to present on the drug industry as to its theory and methodology he
can certainly speak to that as an economist. But as to his knowledge
of the drug industry from other economic studies, he doesn't have
that. Professor Markham does have that to a considerable degree.
Mr. GORDON. All right.
Senator NELSON. Your presentation and the division of the labor
is so sophisticated that the line between what each of the witnesses is
prepared to answer isn't perfectly clear to the committee.
Mr. CUTLER. We understand that very well, sir, and this is why we
suggested that we have a chance to put in the four papers and then
all four can answer questions, whichever is best suited according to his
specialty.
Senator NELSON. I wouldn't object to the witnesses answering any
question addressed to any other witnessif it would help.
Mr. CUTLER. Fine.
Senator NELSON. Go ahead.
Dr. COOTNER. I am on page 9 and we just discussed the effect of
administrative action on the profits of an industry without reducing
its risks at the same time.
Such an impact could evolve in two different, but interrelated direc-
tions. If a company had no control over the risks it faced, it would
simply find investment less attractive and reduce its level of capacity.
If the reduction in return were great enough, investment might virtual-
ly cease, except for the outlay of funds necessary to protect existing
investment, and the industry would gradually decay.
To find the extent, however, that the firm can control the riskiness of
its investments, it has another alternative. It will try to reduce its ex-
posure to risk. A natural gas exploration company, for example, might
shift from looking for new fields to the more intensive development
of existing fields. A research-oriented áompany would try to reduce its
investment in more risky basic research in favor of the development
of variants of existing, well-researched products, or toward improved
or increased marketing of existing products. A manufacturing concern
might find itself less willing to risk building capacity in anticipation
of demand, or to be less quick to introduce new products. If none of
these risk-reducing alternatives were available, the lower profits would
result in a drying up of capital sources and a gradual reduction of the
industry's activities.
I would be foolish to claim that profit restrictions would lead to the
demise of industry, but there is no question but that such limitation-
again without a concomitant reduction in risk-would lead to a dif-
ferent kind of economy than Americans have been used to, one that was
less aggressive in anticipating demand and less prompt in accepting
innovation and change.
PAGENO="0084"
1634 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
In stating these implications of profit restriction, I wish to empha-
size the point with which I introduced my testimony: That one cannot
arbitrarily alter the prices of an industry without altering its profit.
If excess profits exist, price reduction may do no harm, but the ques-
tion of prices cannot be discussed in a vacuum. If present pricing
produces profits no larger than those required by the risks of the drug
industry price reductions will lead to reduced investment.
It is for the witnesses who follow me to discuss that point.
Mr. GROSSMAN. May I ask a question?
Dr. Cootner, I might ask you in general terms, if risks are not borne
out by some losses after a certain period of time, can't we change our
estimate of the industry somehow? For example, one of your, one of
the gentlemen who will testify today, Dr. Whitney, has said "20 years
are too few in any case to deduce absence of risk from continuing prof-
its. Look at what happened to the railroads or for that matter meat-
packing, sugar, and textiles."
Would you go along with that, we have to wait 20 or 40 or 60 years,
year after year of high profits, and we're still to believe that this is a
high-risk industry?
Dr. Cooa~ER. Well-
Mr. GROssMAN. I am not asking you in terms of the drug industry,
but in general economic terms.
Dr. COOTNER. I understand. I would certainly say I would take all
new information as it came and use it to adjust my a priori expecta-
tions. If you were a classical statistician you always want to wait
until all the information is in. I don't think that you have to do that
when you are making decisions as time goes on.
On the other hand, I do agree with Professor Whitney's implicit
statement which is that there are a lot of risks which are virtual-
which may not show up for a long time and yet which influence pros-
pective action.
Let me give you an example which I do know something about. If
you look at the rate of return, for example, on IBM, on its assets, it is
quite high, and the rate of return on its stock has been also quite high.
If an investor at any one time anticipated that the rate of growth of
profits of IBM was going to continue at the same rate in the future
that it had in the past, he would find that the stock is actually worth
a lot more than it is presently selling for. One of the reasons this
extraordinary rate of profit continues is because there is always the
expectation that its predominant share of the market may not continue.
Now that may be an incorrect expectation and if it is anybody
within the sound of my voice can profit from that.
Mr. GROSSMAN. is there any period of years, for example, he says 20
years are too few, I mean specifically, can you say there is any term of
years that we should wait to watch and decide?
Dr. COOTNER. No, I don't think there is any particular term. I
think-f or example, I used in my study, about 17 years, and Mr. Plot-
kin and Mr. Conrad used 16. I do not take the point of view that you
cannot ever or you must wait 50 years in order to make a decision that
the risks are too high. I think if you could really state that in 20 years
there was no evidence of risk then I would be willing to say that that
was sufficient to decide that there was no risk.
PAGENO="0085"
COMPETITIVE PROBLEMS IN ¶]HE DRUG INDUSTRY 1635
Mr. GROSSMAN. Do most industries like, for example, the insurance
industry, build up reserves to meet these risks? Don't most industries
build these up?
Dr. COOTNER. Well, some risks are more insurable, that is predict-
able than others, and-it is easier for a company like an insurance
company to foresee the potential magnitude of risks and plan for
them. As a matter of fact, it is much more difficult for a manufactur-
ing company which cannot retain earnings in excess of current needs
without running into some difficulty with the Internal Revenue Serv-
ice to establish such reserves. It is abstractly possible, but I think in
practice that isn't the way it is done, except perhaps in banking and
finance, where the risks are a little more predictable.
Mr. GROSSMAN. I think you mentiOned once the words excessive
profits. Could you explain basically how you interpret excessive profits?
If you could apply it to the drug industry it would be helpful. I don't
know if you could.
Dr. COOTNER. Well, excessive profits in the abstract sense in which
I was discussing it would be profits in excess of that required to
attract capital, given the existing risks in the industry. That is if
you could show, for example, that a company earned very high profits
and did not in fact have more risk than any other company then
those extra profits would be excessive.
Mr. GROSSMAN. If a company was showing a profit over 15 or 20
straight years would you have to look in terms of exceessive profits
even if the profits were to, say, one Of the leaders of all industries,
if they were the highest?
Dr. COOTNER. There is where we have a problem and it is the same
problem that I raised in the discussion with Mr. Gordon.
For example, suppose that you start with a very high initial rate
of return, and subsequently you face risks on new investments. The
average rate of return on all your assets is going to remain high. It
will not necessarily fall below zero or even below the industry aver-
age and yet it will show a continuing diminution or fluctuation
because of the varying rates of return on new investment.
Now, again if you allow for that phenomenon, then I agree with
your statement that you .make, that is if a company shows that it
never has any substantial risk on additional investment, that addi-
tional investment takes place at a high average rate of return that
*is very stable then I would agree that that is not a risky company.
Mr. GROSSMAN. One final question: Could you tell me as an econo-
mist, do you think it part of your job to assess an industry in terms
of its social responsibility at all.
Dr. COOTNER. Well, it is a very difficult question. Certainly an
industry should be aware of its social responsibilities. The problem
is that in the last analysis the industry has to attract capital from
investors who are relatively unconcerned and who in fact get no return
from accepting the social responsibility. A stockholder cannot bask
in the knowledge that he is performing a social service.
Now, within the limits imposed by the requirements for raising
funds, I think any industry should pay attention to its social
responsibilities.
Mr. GROSSMAN. You would say though it is a factor?
PAGENO="0086"
1636 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. COOTNER. Yes, sir, I would agree.
Mr. GROSSMAN. Thank you.
Mr. GORDoN. Dr. Cootner, I just want to ask you one thing: You
talk about prices being lowered as a result of Government flat, Gov-
ernment regulations. But you also assume that the costs are not regu-
lated or controlled. Isn't it more reasonable to assume that if prices
are controlled, costs also would be controlled?
Dr. COOTNER. Well, I don't know that that is the practice that exists.
Mr. GORDON. Let's take the American Telephone and Telegraph
Co. as an example.
Dr. COOTNER. Yes.
Mr. GORDON. Only certain items approved by the Federal Com-
munications Commission can be worked into the rate base.
Dr. COOTNER. Oh, yes.
Mr. GORDON. Not everything that they spend can be iut into the
rate base.
Dr. COOTNER. That is true. It is certainly true, for example, that
they are not allowed to-certain kinds of advertising are not allowed
in the rate base. But whether that is an advantage to the firm or not
depends upon what the efficacy of the advertising would have been,
and it may be in fact that you actually reduce the profitability of an
industry by preventing it from incurring certain costs, for example,
during the war, as part of a clear view of the Government about the
social priorities we prevented automobile manufacturers from expand-
ing capacity to produce private civilian automobiles. Now that didn't
help their profits although it certainly reduced the costs they incurred.
Mr. GORDON. But isn't it more reasonable to assume, if you are
going to make any assumption at all, that if prices are going to be
regulated by the Government, costs also would be?
Dr. COOTNER. Well, you know, wages are not. I am not saying that
you can't regulate some costs, but it would only affect my argument
if the regulations of costs in fact resulted in a rise in profits. If you
lower the profits by reducing prices and then increase profits by
reducing costs then you haven't lowered profits in which case there
is no problem with anything I say.
If you can reduce prices and costs simultaneously for an industry so
its profits are not changed you obviously have influenced the investors
not to view the company with any more disfa.vor than you did before.
Senator NELSON. Under your definition of a high risk company, that
is the part of the definition involving stability on the profit side for 17
or 20 years, a very high percentage of all American industry is high
risk industry then, is it not?
Dr. COOTNER. Well, all American industry has risk of some degree.
Obviously, only roughly half of the industries have above average risk,
I mean that is a property of the median. There a.re a lot of risky Amer-
ican industries. I am not trying to be fussy, it is just not clear what you
mean by high risk. Certainly risk is compa.rative and certainly many
industries are more risky than many others. Risk is not a property
solely of the drug industry or of any other single industry.
Senator NELSON. The reason the question comes before us is that the
industry asserts continuously that this is a special situation, a special
industry, that is very high risk and, therefore, very high profits are
PAGENO="0087"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1637
justifiable. Since the assertion is continuously made, it would be help-
ful to the committee if the economists could place before us a genuinely
convincing case to prove the claim.
Mr. CUTLER. That is just what we hope to try to do, Mr. Chairman.
Senator NELSON. Well, I don't know whether you are the one to
respond to this. If not I would want to ask one of the later economists.
Last week we had as a witness Mr. George Squibb, former vice presi-
dent of the E. R-. Squibb Co., now a consultant to the Squibb Co., with
30 years of experience in the company. On this question he states:
The industry cannot cite its risk from the uncertainty of research results or
from product obsolescence as the reason for the price differential-for some of it
no doubt, but not much because year after year the profits of the industry are
far above the average for other major industries, and currently appear to be
improving rather than worsening. The risks of the pharmaceutical business seems
to those outside it to be pretty ephemeral in view of the impressive profit per-
formances of the last two decades. Indeed, it is the concensus of the industry's
critics that more risk is needed to make it more sensitive to the normal influences
of competition.
Then I asked a question of Mr. Squibb,:
My I interrupt at this moment? Some of the testimony we have had from the
industry, statements from the industry and the medical profession, as a matter
of fact, have asserted that in order to attract capital for the purposes of expan-
sion and for research these high profit rates are necessary because of the great
risk and so forth. Do you have any comment about that?
Mr. SQUIBB. In terms of risk, I commented on risk earlier in my earlier paper.
I do not feel that the pharmaceutical industry itself under the definition of the
word "risk" as I understand it-is a risky business. It is a profitable business, one
in which companies have long engaged, and most of the companies have many
years of profit history behind them. And while it is true a given product in their
structure may come and go, or be relatively short-lived, however the risk, the
financial risk, it seems to me in my experience, has been very little in the pharma-
ceutical business. A well-managed pharmaceutical company has as good an
investment, is as good a place to put your hard-earned dollars, as safe a spot as
any I know. I do not think there is a risk involved here. In terms of what is com-
monly thought of as a risky speculative sort of business, the drug business is not.
Here is a rather clear, sharp statement about a di~tinguished mein-
ber of the pharmaceutical manufacturing industry. -
Dr. COOTNER. That is certainly an empirical proposition and I am
sure it is really going to be a matter of whether Mr. Squibb or the
witnesses who are going to discuss this question here are more con-
vincing to you. -
I don't necessarily think just because somebody has been in an
industry he necessarily is more convincing than an economist, but that
is perhaps because I am an economist and not a member of industry.
Senator NELSON. I will buy that because a lot of people in the in-
dustry don't think anybody outside of it is qualified to comment on it-.
If we will both agree that those outside are qualified to comment as
well as those inside I will accept that. )ou do not wish to respond to
that? -
Dr. COOTNER. I think, I am in no-~ -
Senator NELSON. Fine.
I didn't know whether somebody else would.
Just out of curiosity, as a former member of the Henry George
Society, what is the Econometric Society? Is that a group of econo-
mists, professional economists of a special kind?
PAGENO="0088"
1638 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. COOTNER. It is an association of economists whose particular
specialty is in mathematical and statistical approaches to economics.
Senator NELSON. That is the new school as contrasted with the
classical economists?
Dr. COOTNER. That is one way to define it.
Senator NELSON. Who is the next witness?
Mr. CiTTLER. The next witness, Mr. Chairman, would be Mr. Plotkin.
Mr. Plotkin and Mr. Conrad of Arthur D. Little have tried to make
some empirical, if I can use an economist's word, measurements of risk
in the drug industry, and have tried to correlate the relationship be-
tween their measurement of risk and return in that industry with an
array of other industries across the entire spectrum. We would like to
have Mr. Plotkin present that to you right now.
Senator NELSON. Both of these?
Mr. CirTr~Eu. That is the booklet entitled "Risk and Return in Amer-
ican Industry-An Econometric Analysis."
Senator NELSON. That is the case in which charts are used by both
witnesse~s?
Mr. CUTLER. That is correct.
Senator NELSON. All right. We are glad to have you here, Mr. Plot-
kin. Your biographical sketch will be printed at this point. You may
proceed to give your statement.
(The biographical sketch of Mr. Plotkin follows:)
BIOGRAPHICAL SKETCH OF IRVING H. PLOTKIN
I am an economist specializing in econometrics and operations research. My
academic training includes work at the Wharton School of Finance and Com-
merce (University of Pennsylvania) and Massachusetts Institute of Technology
Sloan School of Management, Department of Economics (for which I am present-
ly com~i1eting my doctoral dissertation). I have served as a consultant in prep-
aration of econometric Studies for several government agencies (Department of
Justice, AEC, FDIO) as well as some of the nation's largest corporations. My
work has been primarily concerned with decision making under uncertainty
and the measurethent of risk.
STATEMENT OF IRVING H. PLOTXIN, CONSULTANT IN ECONO-
METRICS AND OPERATIONS RESEARCH, ARTHUR D. LITTLE, INC.,
CAMBRIDGE, MASS.
Mr. PLOTXIN. Thank you, Mr. Chairman and members of the com-
mittee.
I am Irving H. Plotkin, an MIT economist and a consultant in
econometrics and operations research. As a part of their research for
the Pharmaceutical Manufacturers Association, Arthur D. Little, Inc.,
has retained me to perform econometric studies of the risk/return
pattern of the American economy.
Our full statement of the findings of this research and the rele-
vance to the pharmaceutical industry is contained in the report "Risk
and Return in American Industry-An Econometric Analysis",1 just
referred to by Mr. Cutler. I request that it be printed in the record as
part of my testimony.
1 The report "Risk and Return in American Industry-An Econometric Analysis," begins
at p. 1746, infra.
PAGENO="0089"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1639
I propose, with the aid of a few charts, to summarize here, by the
use of simplifying charts I have brought along, the key methodological
and quantitative results of this study. However, we must first be con-
cerned with the question of relevance.
Professor Cootner has shown that one cannot meaningfully discuss
prices without considering profitability. Likewise, one must consider
the risk/return pattern formed by the economy as a whole before any
particular industry's profitability can properly be evaluated. There-
fore, the body of research I am here to discuss today is a general in-
vestigation of risks and returns in all American industry.
I would like to underscore this point, that the first part of our pre-
sentation is addressed to the overall pattern of risk and return in the
economy as a whole. It defines the measures that can be applied to vari-
ous industries. The results and implications of this analysis have been
accepted for publication in the Harvard Business Review.
I, too, as did Professor Cootner, do not appear before you as an
expert on the drug industry. My expertise concerns the measurement
of relations between risk and rate of return. Other witnesses will dis-
cuss the specific nature of the risks faced by the drug industry. I can,
however, say that our research has established a valid measure of the
industrial risk/return pattern for the :American industry. Further,
based on our investigation, there is no reason to conclude that the drug
industry's performance in the statistical sense is out of line with the
statistical relationships we have measured.
Attached to my statement are three charts which illustrate the phi-
losophy and methodology we have followed in measuring the risk/
return relationship. They show how our risk measure concentrates on
the uncertainty or lack of predictability of marginal profits and, fur-
ther, how the economy's actual risk/return pattern can be econometri-
cally quantified and the position of individual industries isolated and
evaluated. I, therefore, with the aid of these charts, propose now to
summarize the findings which are spelled out in detail in the written
document. As Professor Cootner indicated, it is my task to show that
one can meaningfully use statistical theory and econometric analysis to
quantify the risk/return pattern; that is, to raise the argument from
one of mere beliefs and contentions to quantified results and compet-
ing hypotheses.
But before comparisons between competing industries can be per-
formed, it is necessary to specify definitions of the terms upon which
we want to compare industries.
Therefore, we have to define for this study two important concepts.
The first is a definition of rate of return, which can be applied across
many diverse industries: industrial as well as financial; those engaged
in manufacturing basic necessities as well as those engaged in manu-
facturing of luxuries.
More difficult, however, is the construction of some measure of the
risk environment which can measure the impact of, without isolating
the individual courses of uncertainty arid lack of predictability within
the wide sprectrum of American industries.
May I emphasize that point, Senator. We do not at this stage in the
study try to say why each industry, one by one, for the 59 industries we
have studied, is risky. We construct a statistical surrogate, a measure
which corresponds through the force of our theory and measurements
PAGENO="0090"
16 ~O COMPETITIVE PROBLEMS IN THE DRUG INDTJSTRY
we have taken to the lack of predictability or uncertainty that faces
persons making investments in the various industries.
We then see what type of pattern is established, and whether the
drug industry shows evidence of being out of line with the risk return
pattern.
Senator NELSON. May I interrupt a. moment? Is that the chart which
you show with the clots on it, and a. circle around the drug industry?
Mr. PLOTKIN. They finally end up on the chart with the clots. But
my first charts go to show the definitions behind each clot, the definition
of risk and the definition of the return. It is a. very straightforward
demonstration without mathematics.
Senator ~\ELSOX. And the graph line on that chart. represents the
average for all industries?
Mr. PLOTKIN. No, sir. That represents a statistical relationship be-
tween risk and return which, I think, can be most easily explained
after the other definitions are put forward.
The basic unit of concern in this analysis, as any risk analysis must
be, must be the individual company. Mr. Gordon has properly pointed
out that that is the unit of decisionmaking.
It is within the individual company that the balance between ex-
pected returns and expected risks is struck. However, this does not
imply that the. only or best source of information about risk ex~ecta-
tions is historical company information. Let us pause here, it is the
expected risk on a venture that we ary trying to measure, not some-
thing a.bout the past, but from the past gain information about in-
vestors, opinions about how risky are particular ventures.
Senat.or NELSON. In context, what do you mean by the word "ven-
ture"?
Mr. PLOTKIN. A venture, I am using it in it.s general sense, can be
viewed by an individual as a. gambling proposition, the rolling of dice
in the game of "craps" or the drawing of cards in the game of black-
jack; it ca.n be viewed in the industrial undertaking as being a new
biological active agent in a research and development program, or the
design of such as a new car, which could be either a Mustang or an
Edsel.
Senator NELSON. You are talking about a risk venture, a high-risk
venture?
Mr. PLOTKIN. I am talking about any venture, and I want now~-
Senator NELSON. Any investment venture by anybody?
Mr. PLOTKIN. Yes, sir. I want to see whether I can find a pa.ttern
among industries that show some of the ventures undertaken by certain
industries seem to be less risky than the ventures undertaken by other
i ndlustries.
I believe this goes to the heart of the question you are asking before,
to show the results that have been claimed by various witnesses.
Management, it is our contention, forms its risk expectations not
solely in the past profit history, of its own company, but also on the
diversity of histories that other companies undertaking similar "ven-
tures," to use the term again, other compames within its industry
undertake.
Likewise, entrepreneurs considering entry into an industry, will
assess the general riskiness of the industry by examining the range of
corporations active, and most importantly or previously active, in that
PAGENO="0091"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1641
industry. It is, therefore, not unreasonable to ask for a measure of
expected risks based on historical industrial performance.
I would now call your attention to the fact that the noted economists
Professor Cootner has referred to have offered very numerous defini-
tions of uncertainty or risk. The common element in all these defini-
tions, because they were all looking to the future as it unfolds before
us, was the uncertainty or lack of preclictabihty of the outcome of the
venture.
The uncertainty involved in predicting the outcome of a particular
investment or a particular investor's venture is, in general, what makes
the process risky. The greater the uncertainty about final outcome of
the event, the greater the risk for those investing in that venture, that
future event.
With that rather simple proposition in mind, consider the two alter-
native propositions as though you are being offered two gambling
propositions, one, the game of blackjack, the other a game involving
dice; alternatively the choice may be between an offer to invest in the
shares of companies making computers and the shares of an electric
utility company.
May I call your attention to the chart, sir?
Senator NELSON. You are not talking just about high-risk ventures
now?
Mr. PLOTKIN. No, sir. We considered the spectrum of American
industry.
I might at this point make a statement about the data source. I think
it is appropriate.
We did a pilot study where the data were very, very carefully looked
at to make sure that industries were closely defined. There were only a
small number of industries and a small number of companies, about
13 industries and 70 companies. The pilot study was just to test the
hypothesis where we understood something about the risks of these
industries to see whether we would get the line you referred to before,
the upward sloping relationship between risk and return.
However, for a public and general presentation, the presentation to
the scholarly and academic community to whom this study is also
addressed and, therefore, being published in scholarly journals, we
have taken the Standard & Poor's data~ source, something available to
everyone, so everybody can repeat it, the data source, and Professor
Cootner referred to this in his testimOny, is the Compustat Annual
Industrial Tape prepared by Standard Statistics. We have not
changed one iota of their data, including their industry classifications
as to which companies belong to which industries.
In one case we have some reasons to believe that the data might be
less than fully accurate, although not having a great effect on the
outcome of the study, I would just note, because we are particularizing
on the drug industry, S. & P. have included in that industry the Gil-
lotte Corp. When I called them they admitted there was an error on
their part, but that was in the tape, alid it only went to, in fact, it
greatly went to, increase the average rate of return in the drug industry
as so reported, I have left that point in there, and we have done so in
our analysis so as in no way to bias the standard authoritative source,
the Standard & Poor organization.
PAGENO="0092"
1642 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Well, let us stay with the chart. Let us consider the distribution of
possible outcomes of two ventures, venture A and venture B.
(The chart referred to follows:)
DISTRIBUTION OF POSSIBLE
OUTCOMES OF TWO VENTURES
Mr. PL0TxIN. Here-the "a," axis-is the value of the outcome of
either venture, and here-the "y" axis-is what is known as the proba-
bility of occurrence of that value. So as you see, the higher you are
here, that means this is the most likely event to occur, having most of
the probability. Any other value has somewhat lower probability.
Now, my claim is quite a simple one. It is not highly sophisticated in
its mathematics. It merely says that faced with venture A and venture
B, one is more certain of how he will wind up if he says, "I will ride
curve A," than if he says "I will ride curve B", because curve A is the
least dispersed, less diverse in its nature. Curve B is more spread out.
This characterization is a very good characterization of various
games of chance, some of which, like roulette, have very widely dif-
fering outcomes on very small probabilities, and very high gains.
Others of them-I am not a good enough gambler, so I am not sure-
are more concentrated as to what the value of the outcome will be.
I would, however, make one more point. The measure of risk does
not depend merely-the measure of uncertainty is not and I use these
words interchangeably, because that is the common thread through all
definitions-on the extremes of these distributions because theoreti-
Value of Outcome
PAGENO="0093"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1643
cally, and practice has shown this-the~r run out ~tll the way to the end
of the positive line and to the end of the negative line with very, very
small probabilities.
It is a measure of the weighted distribution, so that you get a picture
of the center of gravity, where most of the events occur; not just what
is the highest and the lowest. The initial scholarly investigation which
tried to rank industries by the highs and lows got those atypical com-
panies which would make the industry look very high or very low, and
did not find any answers that correspond to commonsense.
Here we are talking about gambling ventures or stock market ac-
tivity as viewed by an investor. How does this help us qualify indus-
tries, individual industries?
(The chart follows:)
DISTRIBUTION OF RATES OF RETURN
EARNED BY FIRMS IN INDUSTRiES
A AND B IN A PARfiOULAR YEAR
Rates of Return
Mr. PLOTKIN. Consider now, and I have not done any fancy foot-
work, these are precisely the same curves, I just relabeled the axis,
because I wanted to use the analogies, these are the distribution of
rates of return in two indusi~rie,s.
Here we standardize for the different size that companies might
retain by relating the return, the amount of money you get back, to
the amount invested. The chart shows rates of amount of returns in
hypothetical industries A and B in a particular year. Here we have
A
.~
B
PAGENO="0094"
1644 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the percent of the firms which earned a given rate of return, and here
we have the rate of return which they earned.
Distribution A might be considered typical of the utilities industry
where, because of regulation the final rate of return earned is very
similar from firm to firm. It is narrowly distributed, centered around
6 percent, and if you earned 8 percent one year, the next year at a rate
hearing, you are pulled back down. In any one year, given that you
are making a physical or stock market investment in utilities, you can
be fairly certain of what your rate of return will be.
In another industry, like the computer industry, the rates of return
are indeed very dispersed. They have a high average, but there are
some earning a good deal of return, and there are some earning less
return. In the same sense that venture A was less risky than venture
B9 I would like for you to think of these two industries here char-
acterizeci as industry A being less risky than industry B. Industry B,
if you are selecting a company at random, is more difficult to predict
than is industry A under these terms.
\Ve have now defined the basis of our measure of risk; that is, the
dispersion of the results of individual companies based on some
measure of return about the industry average.
I would like, returning to the desk there, to concentrate for a
moment on the problem you asked us to keep in mind, the definition
of the rate of return for that, too, must be specified before we can
measure the two quantities, the relationship between risk and return.
However, I do not wish to underestimate the difficulties with this
approach to risk measurement.
For example, the indicated amount of risk in the particular industry
will depend on how carefully the industry is defined or for some
paiticular industry,investorsmay not have foreseen the actual degree
of risk that really did occur. But if we define our variables carefully
and utilize a large number of industrial classifications, and they
standardized outside source, we can find a stable relationship with a
high degree of statistical confidence.
Furthermore, a number of different studies were undertaken by us
at the Arthur D. Little organization, as well as other scholars, to
compare the effect of using a number of different periods and different
industrial classifications.
It is encouraging, as Professor Cootner said before, and I reiterate
here, that the results were not found to be highly sensitive to the precise
definition or the precise number of years or indlustriail classification.
Therefore, as a scholar and technician, we can be sure what we are
seeing unfolding here is not a fluke due to chance and a particular
set of definitions, but is objective and verifiable and can be constructed
by any interested scholar.
As with the measure of risk, we have followed the same approach
in the measure of rate of return. The main variable used, and the one
we believe to have the most significance for this purpose, is simply the
rate of return earned by companies on their long-term investment at
book value, since it is this rate of return which is the most commonly
usedi and is the one considered by most investors and economists when
they are at all concerned with the overall profitability of the industry.
When the economists is charged with answering the question: "Is a
sufficient amount of capital flowing into that mdlustry to providle the
PAGENO="0095"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
1645
public with the services it needs?" The answer turns on the total
capital and the total earnings of that capital.
But, however, to placate common practice, we have used a number
of other measures in order to assure that our results are not sensitive
to the particular definition, all these measures are reported in detni 1
in the study which has been presented to you.
Having defined that the rate of return will be the rate of return
on the total asets invested, we must now state in precise terms the
statistical hypothesis we seek to test.
Formally, we can say we are constructing an equation. that was the
way of describing the line that you have referred to that says, "indus-
try return," the rate of return earned by an industry, is some linear
function-as a first approximation-of "a" plus "b" times industry
average risk: Industry return a + b (industry risk).
We expect the coefficient "b" to be positive such that higher returns
are associated with higher risks.
One should be cautious here that we are not just wrapping around
and meeting our own tail in the end. We are not saying that because
a firm has higher returns it is, therefore, risky.
Now, using two separate measures, one on the riskiness of the
industry and another on its return, we test to see whether there is
any such relationship. Once again the chart can help us.
(The chart referred to follows:)
RISK/RETURN PATTERN
~ MB
VA
Industry Risk (Variance)
VB
PAGENO="0096"
1646 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. PLOTKIN. In this particular hypothesized two industries here,
it happens that the average rate of return in industry B, is higher
than the rate of return for industry A. This is not ~ mathematical
necessity because this curve could have been centered here, and curve
A centered there. But it also turns out in this hypothesized case, not
necessarily true in reality, the spread of risk of B is greater than A.
Once again, this need not be true.
We can summarize-and here, sir, is where I get into a definition
of that line, I think, that will be most clearly understood-the fact,
that the mean, the average, return of A is below the average return
of B, and also the fact that the variance of A is below the variance
of B on a two-dimensional graph where we plot the industry risk or
variance here-x axis-and the industry average return here-y axis.
Now, we see where the risk or variance of A and variance of B
come, and we put dots at their intersection.
WTe have traced out a straight line connecting these points. It is the
hypothesized relationship, that an industry that has a higher return
will also have a higher risk and vice versa. It may not be true in fact.
Here is the way it happened in this hypothesized case. The statistical
question, when we put a similar dot for each of our 59 industries,
comprising 783 firms, will go through a scatter of dots, by which we
put the line of best fit and, therefore, it will not hit every dot. It will
show an upward sloping relationship, a downward sloping relationship
or no relationship at all.
Given we have established what the slope of the relationship is,
we may then ask where does any one industry fall. Does it have evi-
dence of being above or below the normal pattern?
Now, in measuring the rate of return to protect ourselves from the
sensitive of using any one year, which may be economically abnormal,
we have considered the period 1950-65, a 16-year period, which cer-
tainly allows for all post-World War II adjustment and conversion.
In each year we constructed the industry average return and indus-
try variance, so we had 16 yearly returns and 16 variances for each
industry.
PAGENO="0097"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1647
We then further averaged those numbers to take out any sensitivity
to a particular year and, therefore, the "average variance" and the
"average return" are numbers that contain 16 years' worth of
information.
From a statistical sense, and for Mr. Gordon's benefit, I say this is
a highly rich statistical body, so that if we have a significant statistic,
we can be quite confident in it.
We have tested besides the book value rate of return another very
important subsidiary rate of return, that earned by the stock market
investors should they decide to invest in the industry through the
mechanism of this stock and bond market, rather than investing hard,
cold cash in machinery, or research and development. We know in the
stock market there is certainly a continuum of returns that are r&ated
to risk. Government bonds yield the lowest of most taxable bonds.
Then come AAA corporations, Ak corporations, going up to mort-
gages, and third mortgages, each with higher and higher yields, and
more and more risk.
The same with common stock. The blue chips have low yields as
opposed to the speculative shares and computer shares which have
on average, for an extended period of time as an industry group,
high yields.
Therefore, we felt it necessary to test the validity of our definition
of variance, our definition of risk, in the stock market first.
The chart presented in the study on page 20, I believe, is a test of
the relationship measured at market value. The fact that it has a high
statistical significance, the fit is very tight, and forms an upward
sloping line, is to us a true validation of our measure of risk, that we
have not quantified a stable measure of risk.
We may then proceed to ask the second question which concerns the
real economy, the second chart found on page 19; that is, what is the
pattern really dependent on, do rate of returns and risks go hand in
hand, and if they do, what can we say about individual corporations.
I must correct myself, not individual corporations, but individual
industries.
(The chart follows:)
81-280-68-pt. 5-7
PAGENO="0098"
Industry Return (Net Income + Fixed Char~es)/ToiaI Capitalization
Industry Risk Average Inter-company Dispersion
Cl
0
ENDISTRY
RATE
of
20
RETURN
~rj
0
02
8*'
t:rj
Li
Li
2
ci
02
I I I I I _.__~ .-~
10 20 30 so 50 60 70 80 90 100 110 120 130 ~
INDWTRy RISK
MODEL I: ThIDUSTRY RISK/RETURN PATTERN MEASURED AT BOOK VALUE
NOTE: (~)REPRESENTS POSITION OF PHARMACEUTICAL INDUSTRY
PAGENO="0099"
COMPETITIVE PROBLEMS IN THE DRUG ~DUSTRY 1649
Mr. PLOTKIN. This is a reproduction Of the chart which appears on
page 19 of the report you have on hand. It has a striking similarity to
the chart for market value. You have this upward sloping relationship.
Higher rates of return are associated with higher risks in fact, not
in hypothesis, and not because it is a necessary mathematical relation,
but because it seems to be a validation of an economic truth.
This is, in essence-you asked me before about the econometric~
society-this is in essence what an econórnetrician does, takes a hypo-
thesis where he has an expectation of a certain relationship, goes to
the real world, measures the data and tries to determine whether or
not that is a valid hypothesis. This is on upward sloping relationship
with a good fit, the dots are for this purpose tightly clustered around
the line, tightly relative to normal statistical tests to which this rela-
tionship has been exposed.
Senator NELSON. This chart is through 1965?
Mr. PLOTKIN. Through 1965; 1950 through 1965, each point repre-
sents one industry, an average of over 15 companies per industry, but
this number varied because some industries were smaller and some
larger. So there are 783 companies, 59 industries, for 16 years.
Senator NELSON. How many companies did you use in the red dot,
in the drug industry?
Mr. CONRAD. Twenty-nine.
Mr. PLOTKIN. Twenty-nine, as classified by the Standard & Poor
Corp. It included Gillette with its fantastic return on total assets, of
about 30 to 40 percent.
We included precisely all those coiñpanies included by Standard
& Poor's in every industrial classification.
Senator NELSON. Is my eye correct-is the red dot representing the
industry with the third highest rate of return on that chart, or the
second?
Mr. PLOTKIN. It is the fourth highest in risk, and it is the third
highest in rate of return.
Senator NELSON. If it were brought up to date for this year then,
would not the drug industry be first on rate of return?
Mr. PLOTKIN. No, sir. I realize what the FTC report said, but that
only puts in 1 more year, so it would then be a 17-year study.
Senator NELSON. This is for a period?
Mr. PLOTKIN. This is for the 16-year period. It is unfair to ask a
statistical test to establish a relationship based on just 1 year's observa-
tion. That is why before one can have confidence in an econometric
analysis, it must be very rich in data drawn on many years' observa-
tions, so as not to be unduly influenced-
Senator NELSON. This past year's return would just raise it, but
you do not know by how much?
Mr. PLOTKIN. It would just raise it, but I can tell you that its effect,
this return is about 3 percentage points greater than its 16-year aver-
age, and would count one seventeenth of 3 percent-perhaps Pro-
fessor C'ootner can help to figure out mental arithmetic here.
Senator NELSON. I cannot tell what the difference, percentagewise,
between the three which are above it is, so I do not know how much it
would change it.. Would it change its position at all?
Mr. PLOTKIN. It would change it, but not perceptibly.
PAGENO="0100"
1650 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. It would change it to third or second?
Mr. PLOTKIN. It could change it to be equal to third or to second,
but it would not materially affect these results.
Senator NELSON. What a.re the three industries which are higher?
Mr. CUTLER. Mr. Chairman, you can see all those figures on page
78, industry by industry.
Senator NELSON. Seventy?
Mr. CUTLER. Seventy-eight.
Senator NELSON. That risky gold-mining business is right up there
in third place, I see.
Mr. PLOTKIN. I am very glad you asked about it, and perhaps this
is a subject of many senatorial and congressional investigations, these
days.
Gold mining was one of the industries which allowed us to have a
good deal of fit in this relationship. We know gold mining does not
operate under the normal economic pressures that confront companies
competing for the consumer's dollar in the marketplace.
The price of gold is artificially established by administration fiat,
independent of the demand for gold. Some question may be raised
whether that fiat will be maintained, despite the demand for gold.
However, as an industry, it does not return a rate of return com-
mensurate with the risks involved, which simply means that there is
a. distortion of the amount of capital that has flown into that industry.
This is something to be expected when the Government is primarily
controlling the market. It is not functioning along, and competing
along, with other industries.
Senator NELSON. Did you operate on the assumption in your evalu-
ation here that there are really no significant differences between the
drug industry and the other ones that you have compared it with?
Mr. PLOTKIN. No, sir.
Senator NELSON. You say there is a difference in the gold-mining
industry because its price is regulated. Did you take into consideration
the fact that there is a high degree of control by the individual seller
of the drug product in the market, and that he may very well be the
highest priced with a product that is no better than one that sells for
one-tenth of the price, and that there is no other industry that I know
of in which that is true?
Mr. PLOTKIN. Let me trace through my logic of the gold industry
much more carefully.
I, in no way, tried to determine the peculiarities of the individual
industries ahead of time. I set up a standard of measurement, a defini-
tion of return, and a definition of risk which was reasonable-reason-
able to a large body of scholarly economists-through which definition
I looked at all industries with the same set of glasses, and then I
looked at where they fall on this pattern.
After the analysis is done, I noticed that gold mining is seemingly
out of the pattern, is one of the outstanding ones to be out of the
pattern.
I then see whether I know of any reason which could cause it to
be out of the pattern.
The statistical question, and this is why they call economics the
dismal science, but perhaps statistics is equally dismal, and the statis-
PAGENO="0101"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1631
tician always starts out with a null or negative statement, a null
hypotheses. We are examining into the discussion-although it is
not what the purpose of this entire investigation was-to see whether
or not there is evidence, based on these definitions, alone, sir, that the
drug industry is outstanding, is out of the pattern, that is, for the
risks encompassed.
So, measured by our definitions, it does not yield a nonnormal return
to its investors.
The answer comes from a statistical sense, not from a judgmental
sense, by saying how far is that versus how far is something down
there, but applying standard statistics, and the tests, there is nothing
to suggest in fact, that leads us to reject he hypothesis that point 4
is outside of the distribution traced out by the rest of the points here.
But this hypothesis can be applied to any of the points, and this test
can be applied to any of the points.
When applied to point 3, there is doubt that it is within the normal
pattern traced out, this upward rise in pattern. That is the limit of our
abilities as statatisticians and econometricians to make the statement.
I would not want to seem that my testimony is making any other
statement but that on a measure of risk, which has shown to have
economic viability, making good commonsense and good practical
sense when viewed against the stock market, and applicable across
all industries, no matter what the cause or the source of risk, the drug
industry does not seem to violate that pattern.
This does not say that everything is rosy and perfect, and it does
not even say such an analysis cannot be pushed to say how much
should the rate of return be.
When an economist catches a solid repression line he really means
a gray area within which he cannot distinguish, and above or below
which he says points fall outside of the pattern. There is no evidence
that the drug industry is not in this hazy, gray area. This is not my
assertion, but any econometrician or any statistician can do the tests
for himself and, using standard confidence intervals, 95 to 99 percent
would be led to the same conclusion to, reject the hypothesis that the
drug industry is out of the normal pattern.
Senator NELSON. In your judgment, what are the most important
factors that go into your conclusion that the drug industry is a high-
risk industry?
Mr. PLOTKIN. The most important factors that go into my conclu-
sion are that I believe that I have a good and workable hypothesis
of risk and return in the American economy, that higher returns are
given with higher risks.
I believe, and we have submitted this to arbitration of scholarly
members of my profession, in MIT, Chicago, Stanford, the Rand
Corp., that there is a workable definition of return and risk proposed,
and, fitting that pattern, there is no evidence of a violation of the nor-
mal pattern.
I do not know, have no expertise, have no prior beliefs, about why
this point should be there and not down here or not here. I am speak-
ing only from the evidence.
I am raising this to a contending hypothesis which someone else
might come up with the alternative quantifiable measures.
PAGENO="0102"
1652 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
If economics has made any progress in the last 20 years, it is our
ability to take things out of the field of judgment, out of the field of
authority and whose textbook you read, and into objective, verifiable,
repeatable experiments.
Senator NELSON. The first students of aerodynamics came to the
conclusion that the bumble bee could not fly because in proportion
to the body the wing size was not big enough. But he flies pretty well.
You used 29 companies to represent the drug industry?
Mr. PLOTRIN. Twenty-nine companies.
Senator NELSON. How many of these companies have lost money in
this period; is it 17 years?
Mr. PLOTKIN. Sixteen years, sir.
Senator NELSON. In the 16 years that you used, did any one of those
29 companies have a loss year?
Mr. PLOTJIIN. Yes. The answer is yes. These facts are in the report.
May I once again-I submit that. we do have a rather subtle division
of labor, but I think that the committee benefits by the expertise of this
division.
Mr. Conrad, of the Arthur D. Little organization, is primarily in
charge of the relevant information with respect to the pharmaceutical
industry, which firms were comprised in it, how it reflects on it, and
how his additional studies, together with those of Professor Markham,
go to sa.y why this point is to be expected, why the risk should be high;
and I would respectfully ask that these questions be deferred and be
integrated into his testimony.
Here I am just prepared, at this point, without any expertise other
than a personal opinion which I have eschewed, discussed the general
validity of such measurement.
Mr. GORDON. Mr. Plotkin-
Mr. PLOTKIN. Yes, sir.
Mr. GORDON (continuing). In the study that you have presented to
the committee on the proposed theory of risk and return, on page 12
you say:
In order to perform these statistical tests, it was necessary to construct a
quantitative measure of the industry risk (or uncertainty). We selected measures
of the dispersion of individual companies' rates of return about their industry's
average rate of return for a given year.
Mr. PLOTKIN. That is correct. It is so worded.
Mr. GORDON. Now, it seems to follow from your definition that if
there were two industries-let us take two industries, A and B, each
with two firms, A-i and A-2, B-i and B-2. Now, A-i earned 25 per-
cent on its investment, or its assets, after taxes.
A-2 earned 15 percent on its investment, on its assets, after taxes.
B-i earned 9 percent. B-2 earned 7 percent. Maybe you ought to write
that clown.
Mr. PL0TKIN. Fifteen, twenty-five, seven, nine.
Mr. GORDON. A-i earns 25 percent, A-2 earned 15 percent, B--i
earned 9 percent, B-2 earned 7 percent.
Now then, industry A, according to your definition, would be the
high-risk industry; is that not correct?
Mr. PLOTKIN. That is not correct, because you would have to further
hypothesize that this was the case year-in and year-out., and also-
Mr. GORDON. All right. Let us take it on a year-in and year-out basis.
PAGENO="0103"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1653
Mr. PLOTKIN. And also that those two firms-I am wondering if this
is part of your hypothesis-were the entirety of the industry, not a
sample thereof, but the entirety of the industry. And if you have only
two industries in your economy, the entirety of the economy~
Mr. GORDON. All right. Let us assume that I said there were two
industries.
Mr. PLOTKIN. I wanted to have~-
Mr. GORDON. And two firms in each industry.
Mr. PLOTKIN. I wanted that clarified.
Mr. GORDON. Let us say, year-in and year-out they earn 25 percent,
15 percent, 9 percent, and 7 percent. Industry A, according to your
definition, would be the high-risk industry; isn't that correct?
Mr. PLOTKIN. Sir, this is correct. This is a case, and this is typical of
cases we considered in drawing up this definition.
Mr. GORDON. All right.
Let me go just a little further. It would also mean that the better
firm A performed, all other profit rates remaining the same, the riskier
industry A is compared to B. That is also correct, is it not?
Mr. PLOTRIN. Defining risk as the uncertainty of taking a blind
choice of getting firm A-i or A-2, and knowing \~hat your outcome
will be, that is correct.
Mr. GORDON. It also means, does it not, if A-i and A-2 maintained
their profit rates, but B-i and B-2 kept changing back and forth, say,
from seven to nine, industry A is still riskier, is it not?
Mr. PLOTKIN. Yes; under these hypothetical situations, that is true.
Mr. GORDON. All right.
Now, Mr. Plotkin, are you going to tell us if you are assured of
profits of between i5 and 25 percent, even though your exact rate of
return may be uncertain, that this is a riskier use of your money than
putting it where the return may be between 7 and 9 percent?
Mr. PLOTRIN. I am going to tell you that, to the extent that one can
give me that assurance. In the hypothetical economy which you have
drawn, reminiscent of the Senator's hypothetical bumble bee of the pre-
vious example which could not fly, that would be true.
Our definitions were not meant to cover cases not relevant to the
economy as we know it. This is not a textbook exercise. It is true
they break down like any `other econometric definitions when carried
to the extreme, to reductio ad absurdum.
We performed specific tests against the robustness of these de~
uctions, against just the type of hypotheses you mentioned, to see
whether they hold true in the real world, and whether they would
affect those definitions. We have found they would not.
I would like, though, to ask Professor C'ootner's comment especially
on the point, or Professor Markham, that there is no possibility in
industry A of earning less than industry B. The possibility of that
describing a real world phenomenon, and what the requirements of
that then would `be, I would like to ask them to comment.
Mr. GORDON. Mr. Plotkin, it seems to me that the record of the high
and stable profitability of the drug industry will speak for itself on
the matter of evidence of risk.
Mr. CUTLER. Mr. Gordon, we have `asked ourselves these very same
questions that you are putting now, to see whether the range, the dis-
PAGENO="0104"
1654 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
persion, of rates of return in the drug industry data that are included
within this study, goes frequently below either the drug industry
average, or indeed, the all-industry average, and Mr. Conrad has
made a statistical analysis of that that he would be glad to tell you
about. I think we will show you that your hypothesis A is not the
hypothesis of the drug industry.
Mr. GORDON. I would like to continue with Mr. Plotkin. He is the
one who is making the statements, and he is the one who presented
the study. We will get to Mr. Conrad in just a few- moments.
Mr. CUTLER. I believe we will let Mr. Conrad-I mean, Mr. Plotkin,
continue, but I believe the chairman said that if we thought a par-
ticular witness can answer a particular question better, we can have
that witness do so.
Senator NELSON. That is correct. Anybody who wants to answer,
may answer.
Mr. CONRAD. My name is Gordon Conrad, from Arthur ID. Little,
Inc.
In looking at the data bank that we have for the drug industry,
and attempting to see what kind of fluctuations actually occurred in
the data over the 16-year period, we ran an analysis w-hich took each
data point for each company, for each year in the study, and analyzed
the distribution of profit performance that relates to that chart for
the total number of companies, and related this to the average re-
turn on investment for all the industries shown on that chart.
The average return for all those industries comes out to about 10.75
percent for that 16-year period.
Now, for the 16-year period, about 24 percent of the drug industry
points fell below that average. Four percent of those-
Mr. GORDON. Excuse me. Would you repeat that once again?
Mr. CONRAD. Twenty-four percent of all drug industry data points
out of the total number of data points for the 16-year period again
24 percent fell below the all-industry average return on investment
level of 10.75 percent.
Mr. GoRDoN. Are you saying that 25 percent in the last 16 3-ears,
the average profit for the drug industry fell below the-
Mr. CONRAD. No.
Mr. GORDON. What are you saying, precisely?
Mr. CONRAD. If you take each company and each year and give it
a weight of one, and add up all the company years, 400 points, it
comes out to-in other words, you have 400 actual data points over
the 16-year period, and of that 400 data points, 24 percent fall below
the all-industry average rate of return on investment of 10.75 percent.
Saying this another way, in 24 percent of the years by company
there were returns that were lower than the industry average, the
all-industry average.
We are not relating this at all to the drug industry's average return
value which is shown on the chart. So, yes, then we are down-
Mr. GoRDON. Can we get the raw material on which you have based
this study?
Mr. PLOTKIN. You have the raw material in the report. The num-
bers are spelled out. Each company year value is given. There are a
number of negative values given, also, we would point out.
PAGENO="0105"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1655
So that is why I answered your question as I did, that it is true,
one can posit situations under which the situations would break down.
Whether there is such gain in such positing depends on how relevant
both situations are to the real world.
W7e submit not through judgment or opinion, but through analysis,
they do not represent real world situations and do not affect the va-
lidity of the results.
Senator NELSON. Risk is a relative thing, is it not?
Mr. PLOTKIN. We say the risk/return relationship is relative, one
industry versus another. It is not an absolute quantification, but it is
a measure such that when it is higher, the risk of investing in that in-
dustry is higher than the risk of investing in an industry when that
is low-er. It does not say that it is worth $10 or $20, or something like
that.
Senator NELSON. But you could project your chart which now cov-
ers 16 years to 50 years, and if you came out with exactly the same
result, would you come to exactly the same conclusions?
Mr. PLOTKIN. I would, sir, because what the chart seeks to do, does
not matter as long as you have some minimal number of years so you
are certain that it is not a perturbation that you are measuring, the
chart seeks to look at the ex ante, if Imay use the technical word, the
risk that appears to a person before he takes the plunge, he makes that
decision based on past information.
Let us see where is the relevance. Before a drug company president
can commit capital to research and developing a bioactive agent, he
must ask himself whether that is the most profitable use of his capital
for the risks involved in that venture. He repeatedly asks himself that
question, or would he be better investing in IBM stock, buying the
telephone company, or, as some of our insurance companies have
come to conclude, buying out credit card companies, rather than in-
vesting in insurance.
If he feels that the prospective risks of the individual ventures are
not justified by their rate of return, he will not commit capital to those
ventures.
This analysis addresses itself to quantifying the prospective risks
based on the simple hypotheses, given those two ventures, which one
is more risky, which one is more uncertain. This is not a complete
theory of profitability. It is not-I would be both naive and boastful
to claim that I have quantified here the entire American economy.
I have taken two important dimensions of it and shown its relation-
ship. It is based not on mathematical concepts that a businessman
w-ould not be operating under such as-~-
Senator NELSON. Of course, when he is making that decision as to
where to put his capital, he may very well put it into a company in
whmhi the profit is lower than it is in his own business, perhaps for
purposes of diversification, or he may do it for a tax loss. He may do
it for any number of reasons-noneconomic reasons.
Mr. PLOTKIN. The reason you mentioned yourself, the diversifica-
tion, diversification is the other side of the coin of risk. It is a way of
getting away from risk.
Now, what we have to do here is isolate the risk of the particular
industry, the drug industry, to see whether capital will flow into there
PAGENO="0106"
1656 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
or will the drug companies become like the fire and property insurance
companies, merely holding companies of capital, not actively expand-
ing their own markets, but going into other ventures which are, may
be, less profitable on the average, but have less risk and which they
feel more confident in.
This is the thing which worries economists when there is regula-
tion proposed. If there is excess profit, profit above the risk level, then
there is no harm to the normal flow of capital in reducing the profit.
But if the profit is concomitant with the risk level, the harm is a
potentia.l threat to capacity in that industry, and that is the only rea-
son we commend such a study to your attention, pointing it out not as
proponents of specific points of view, but a proponent of a more general
point of view that regulation must take into effect and into account,
not only the rate of return but the risk in the industry, and asking
only that that judgment be made.
I am critical, personally, and admit this is a personal point of view-,
about the FTC year-in and year-out ranking industries by rate of
return, and indicating by implication that ones at the top of the list are
bad for the economy. This is a false statement, and cannot be justified
by even the most elemental of the textbooks. They should also have
another column ranking the riskiness.
Senator NELSON. I have not followed it, obviously, as closely as you
have. Did the FTC issue a report that those industries which ranked
so high are somehow doing damage to the economy, or is that your
opinion?
Mr. PLOTKIN. It is my interpretation, not from reading only the
entire report but from the news release, the way they point things out,
there is an implication.
In discussing it with my students at MIT, they have read the news
release, I have received questions as to why is it. bad to be profitable.
Senator NELSON. Did you read the actual report of the Federal
Trade Commission yourself?
Mr. PLOTKIN. I read the whole, the entirety of the report. But I feel
as far as influencing Iublic opinion, the news release is sometimes more
potent, and I concentrate heavily on that as well, because I know that
not many people are professional economists and will not read the
whole report. That is always t.lie danger of epit.imization and summary.
Senator NELSON. But I am just trying to get this straight in my
mind. Are you saying that the Federal Trade Commission says some-
thing derogatory about high profit companies, or are you just saying
the very fact that they released the high profit companies somehow
is derogatory by implication?
Mr. PL0TKIN. I am saying that they release information in such a
w-a.y that it has been interpreted, to my knowledge, by not very so-
phisticated students of economics as implying some derogatory
function.
I am not saying that the FTC does imply it, but I would think it
would be useful if in addition to listing the companies in order of re-
turn, they also listed the companies in order of riskiness.
Senator NELSON. Do they have sufficient. information to list them in
order of riskiness?
Mr. PLOTKIN. They have the finest economists who certainly could
create a measure of risk the way they have created it at MIT and
PAGENO="0107"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1657
Arthur D. Little in their research, aiid any other university in the
country.
Senator NELSON. You will concede, however, that there is a large
judgment factor in your chart and your own theorem, won't you?
Mr. PLOTKIN. I readily concede that and refer back to Professor
Cootner's point that no matter how one changes the definitions of risk
or the precise definition of rate of return, the basic relationship of
higher profits with higher risks, cannOt be killed. This is one of the
few times we see economists in agreement of any persuasion, Paul
Samuelson and Milton Friedman, for example.
Senator NELSON. What I am getting at is, you are asking the Fed-
eral Trade Commission to come up with their own theory about what
high risk is, and then if a group of you economists disagree, you will
start pounding the Federal Government for making that conclusion.
Mr. PL0TKIN. I am sorry if my remarks have led to that interpre-
tation. I only meant it is sometimes implied that just because one has
high returns, he may be doing a disservice to the economy. I was hop-
ing this analysis disproves the implication.
Mr. CUTLER. Mr. Chairman, unless you get any impression that the
Federal Trade Commission's economists would differ from the meth-
odology that was followed here, both Professor Markham and Profes-
sor Whitney, who follows in turn, were both chief economists at the
Federal Trade Commission at earlier stages of their profession.
[Laughter.]
Mr. PLOTKIN. I certainly have no quibble with my colleagues.
Senator NELSON. Being one who majored in economics when I was
in school, and having listened to the arguments in those days, I just
learned enough to discover that you could find any point of view you
wanted from a set of economists any place in the TJnited States, and
from my discussions with my friends tOday, I find that is still true.
Dr. MARKHAM. If I may intrude on the dialog, Senator, I think
you are quite correct when you went to school, and that was true when
I went to school, too.
If this analysis means anything, it means, I think, what precisely
Professor Cootner and the rest who have testified, have said here, and
that is as Professor Plotkin has said, this takes the issue really out of
the area of opinion, it puts it in the area of quantitative fact.
One may have some difference of opinion as to what is the very best
measure of risk or, indeed, for that matter, what is the very best meas-
ure of the rate of return. But it, at least, reduces the dialog to the
meaningful definitional questions.
The factual analysis stands on its own feet, and, as has been pointed
out, when you have picked someone who may be extraordinary highly
sympathetic to Government regulation, of which we may name some
economists, or someone who is very much opposed to Government
regulation of the economy, and here I can be specific, Prof. Milton
Friedman, those economists of different views, political views and dif-
ferent opinions, about this sort of thing would still have to, I think,
indeed already have, agreed that this establishes the statistical rela-
tionship between risk as measured and rate of return as measured, and
both the rate of return index and the risk index are very sensible in-
dexes that most people would agree measure these respective aspects
of the economy.
PAGENO="0108"
1658 COMPETJ~fIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. Before we recess for noon, I would like to make
just one point. This is an industry evaluation of risk and profit, is it
not?
Mr. PLOTKIN. That is true.
Senator NELSON. By industry.
Mr. PL0TKIN. The unit of observation on the chart is the industry.
Senator NELSON. So it is no doubt true that. you can reach into one of
your industries on the chart, which appears to be a relatively low-rate
return industry, and within that industry select, out a very fine com-
pany that would rank very high in rate of return.
Mr. PLo~I~IN. You could not do that with any assurance, blind-
folded. If you looked in with a lot of foresight and particular exper-
tise, you might have a chance of doing that. In fact, that is part and
parcel of the variance measure that all the companies within the in-
dustry, because of the undertaking that they do pursue, do not have
the same results. You are perfectly correct, Senator.
Senator NELSON. So what I am saying is that what we are looking at
here is a chart that measures, according to your theory, the rate of re-
turn and the risk of an industry. But within one of the industries
which will rank relatively low in rate of return and high in risk, you
can select out a company that. has a long record of profit and that
would rank high in rate of return and yet would rank low in risk; is
that not. correct?
Mr. PLOTKIN. It is possible to do that. It does not necessarily mean
that there is one in the particular sample which we have used here.
Senator NELSON. Well, can't you reach into one of those industries
that ranks relatively low and select out one of the companies that does
rank high in profit and low in risk?
Mr. PLOTKIN. I believe-
Senator NELSON. These are averages that you have given.
Mr. PLOTKIN. Yes. The answer is-
Senator NELSON. Is it true, or is it not? Who knows the companies,
who knows the answer?
Mr. PLOTKIN. Well, your assertion is right in respect to company
years.
Senator ~\`ELSON. With the knowledge that you have, since you put
the cl~art together and selected the industries from Standard & Poor's,
can't you say there is within a higher risk, low rate of return industry,
a high-profit, low-risk company?
Mr. CONRAD. One point here.: the industry risk factor is an overall
factor for that industry, and is part of the riskiness that any company
exposes itself to. It is not the total riskiness that a firm exposes itself to.
So while it is true that. you can pick from any one of these imiclustry
points, a firm that may have high profits for an extended period of
tinae, this does not speak to the validity of the average industry risk
factor that we have calculated. Again that is only OIIC component of
the total riskiness that firm experiences.
Now, there will be others. We have not addressed ourselves to the
individual firm risk in this particular part of the analysis. There are
other analyses that could be `conducted which would then assign par-
ticular company risks to each firm. But that does not relate to what
we have done here.
PAGENO="0109"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1659
Senator NELSON. So you measured each individual company, did
you not?
Mr. CONRAD. Yes. The whole analysis is put together from indi-
vidual company measurements.
Senator NELSON. And then from the individual companies, you
came to a conclusion about the industry as a whole which represented
the averaging out of those who were higher risk, those who were lower
profit, lower return; did you not?
Mr. CONRAD. Well, the actual mathematical measurement is to
measure the distribution of performances-rates of return or total
investment-of the companies around their own industry average
return. So each firm~s weight is taken into account in calculating a
total risk factor.
Senator NELSON. Let me get back to my question. Perhaps I am
not being clear. All I am saying is that., based upon the statistics that
you gathered, you took 29 drug companies. So, if you took the drug
industry, couldn't you select the poorest rate of return company in that
industry and nome up with one that is below the average?
Mr. PLOTKIN. Yes, sir. But it is true not of the average return over
the history of the company. This is something that addresses itself
to the year-by-year return. We can select a company in particular
years whose return is below the drug industry average and is below
the economywide average as well.
Senator NELSON. Well, let me go back to my original question again.
Take one of your companies on the chart which ranks at about the
average, 10.5, for industry in rate of return, and there sits the industry
ranking right in `the middle.
Now, within that industry you have, say, listed 30 companies. Can't
you take, examine, those 30 companies, and over the past 16 years pick
one out that over the past 16 years has consistently had a high rate of
return that would put it, if you used that sole company, put it up at
the top of the rate of return companies; isn't that correct?
Mr. PLOTKIN. No; that is not the case. You can find One 111 the past
1(~ years whose rate of return has been both above and below. About
tIre company's 16-year average, we `can make no stateme.nt~, sir.
Senator NELSON. Why not?
Mr. PLOTKIN. Because that is not what the analysis is addressed to.
it is addressed to the profitability of ensuing ventures, the marginal
profit, what will next year's return be..
Senator NELSON. What you have given us is an industry average.
All I am saying to you i's that the average is made up of X number of
companies, maybe 30. In that grouping. you are going to have a corn-
Iany with strong management, strong sales, goodi promotion, good
research, good recognition in the market place, and regularly high
profits.
In the same group you are going to have one with average manage-.
ment, average sales, average research, average everything else, and
below average return `in the market; are you not? And you are going
to have one that has a 50-year history of good returns and no losses, or
very few losses, and one with very lOw average rate of return; are
you not?
It is true of every other industry.
PAGENO="0110"
1660 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. CUTLER. You can undoubtedly find companies, Senator, in the
data we have given you that had a consistently high rate of return,
drug companies, year after year.
Senator N~soN. Yes.
Mr. CUTLER. But I think the fallacy of what you are trying to say
is best illustrated by the New York Yankees, who won the pennant
9 years out of 10. On your theory, they became a low-risk enterprise.
It was easy for them to win the pennant. But in the 11th year, they
finished ninth. A company which has a consistently high rate of
return in a high-risk industry has the same risks on its new invest-
ments every year, even though it has batted out, let us say, a 25-percent
rate of return in each of the 10 preceding years.
That is the point these economists are trying to make.
Senator NELSON. I am not trying to work out a theory. I am just
saying that when you collect 30 companies that have been in business
for 10, 20, 30, 40 years, you can find companies that have consistently
clone very well, and you can find companies that have been consistently
below your curve; isn't that correct?
Mr. PLOTKIX. No, sir. I do not, depending how it is-
Senator NELSON. Let me give you General Motors as an example.
Compared to the rest of the auto industry, they have done well. If
you took the years when Chrysler was falling apart and went from
23 percent of the market to 12, and American Motors went from four
or five down to less than one, and Ford dropped off from 23, and
General Motors occupied about 50 percent of the market, can't you
have a case that General Motors consistently, compared with all the
rest of the industry, has done well in the marketplace?
Dr. MARKHAM. I think, Senator, the answer is, yes. Could you, in
looking at General Motors now, their investments, infer that it is
goingto make 20 percent as a rate of return from that new investment?
The answer is, you cannot infer from the past history, given from these
past indexes, that General Motors will make 20 percent on, say, a new
model coining out the next year.
Senator NELSON. You cannot do that in any industry?
Dr. MARKHAM. You cannot do that in any industry.
Senator NELSON. Then it is irrelevant.
But my point is, what you are talking about here, and it may be a
very fine theory, I am not qualified to judge it, but I am just a layman
listening to your testimony. I am saying you come in and take an
industry in which there is good and bad, and then you plot it out
on a curve. Let us assume that you are correct, and that all knowledge-
able people will agree with you.
All I am saying is that within the industry selected there are in-
dividual companies that have a record that is good enough so that
investors are going to say, "That is where I will put my money", and
then when you talk about the drug industry, if that is the kind of
company you have, a firm that is doing all right in research and
development and is making a good rate of return. I have never heard
of anybody investing in an industry. I have heard of people investing
in companies within an industry. That is the only point I am trying
to make. This theory may have validity as to what the whole industry
represents, but there is some questionable validity about it as applied
to a particular company within an industry.
PAGENO="0111"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1661
Mr. CONRAD. I think we certainly have been cognizant of this
problem and we are sensitive to it, and I think the following testimony
that we have, both the report that I am putting in and the joint work
that Professor Markham and I are doing, speak directly to this issue,
and will take a bit of time to explain. But I think we will clarify even
in an industry where you have good performers and poor performers,
the good performers may well be subjected to the same risk influences
that the poor performers will be subjected to.
I think we are going to try to illustrate that point for you here.
Mr. CUTLER. Mr. Chairman, Mr. Conrad's and Professor Mark-
ham's statement is just two and a half pages long. If we got that in
before lunch, we would be through with this group except for your
questions.
Senator NELSON. I will have a few more questions. Why don't we
just break for lunch right now.
We will adjourn until a quarter to 2.
(Whereupon, at 12 :50 p.m., the subcommittee recessed, to resume
at 1 :45 p.m., this same day.)
AFTERNOON SESSION
Senator NELSON. Come to order, please.
Mr. Plotkin, your report does list the companies involved in each
one of the industry selections, is that correct?
STATEMENT OP IRVING H. PLOTKIN, CONSULTANT IN ECONO-
METRICS AND OPERATIONS RESEARCH, ARTHUR D. LITTLE,
INC.-Resuined
Mr. PLOTKIN. That is correct, sir; and the number of companies in
each industry.
Senator NELSON. What was the basis of the selection?
Mr. PLOTKIN. It was derived from the Standard Statistic Corp., a
subsidiary of Standard & Poor, Compustat Industrial Tape. This is
a magnetic computer tape which contains, in essence, balance sheet,
income statement, and stock market data on approximately 800 cor-
porations in 59 industries. The period was 1950 through 1965.
Senator NELSON. Did you use the same number of companies in each
industry?
Mr. PLOTKIN. No, sir; I used in every point all the data that were
available to me on the tape. This gave a different number of companies
in each industry. In general, these companies are the larger, more
successful, actively traded corporations because it is in the interest of
Standard & Poor to collect them.
Senator NELSON. So in the auto industry, you used all four of the
auto companies, is that it?
Mr. PLOTKIN. I think since the period went back to 1950, we may
have as many as five in one field.
Mr. CUTLER. The number of companies in each industry, Mr. Chair-
man, appears on pages 34,35, and 36.
Senator NELSON. On what basis does Standard & Poor list the
companies?
PAGENO="0112"
1662 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. PLOTKIN. They have experts in their organization in each indus-
try using the Department of Commerce SIC, standard industrial clas-
sification numbers, based on the type of product sales of the individual
companies, who will assign, admittedly arbitrarily, companies to
various industries. Because this is a standard reference source, we have
elected to use it rather than make up our own industries, which could
be somewhat objectionable or at least questionable as to whether the
results were influenced by our own selection.
As I admitted earlier in my statement, I have some serious qualms
about particular items-should one company be here or there. But we
have done careful studies to show that the results are not sensitive to
any one company moving in or out of an industry.
Sir, if this would help, I might add that these data sources have been
used for similar studies in a number of university campuses-MIT,
Chicago, Stanford-by diverse scholars for many purposes, always
relying on these classifications as a general, widely accepted usage.
The First National City Bank, in their monthly letter, often lists in-
dustry by industrial classification, and it is quite similar to the ones
appearing on Compustat tape.
Senator NELSON. You used 29 companies in the drug industry?
Mr. PLOTKIN. Yes, 29, sir.
Senator NELSON. Is that all the drug companies listed in Standard
& Poor?
Mr. PLOTKIN. To the best of my knowledge, it is all the drug com-
panies listed in Standard & Poor. It even contains a company of
exceedingly high profitability which, by their own admission, Standard
& Poor now say should not be in there. That is the Gillette Co. Rather
than take the chance of influencing the outcome by my personal selec-
tion, and Arthur P. Little take a chance by its exercising judgment,
we decided to go by what is an industry standard and used in all
academic circles.
Senator NELSON. In what groirnds does Standard and Poor's select.
them? Eliminating Gillette, are these the 29 largest companies?
Mr. PLOTKIN. To some extent, they are the largest in that these would
be the actively traded, the ones that information is most readily ob-
tainable on. If I can paraphrase the Standard and Poor's description,
they try to select that group of companies actively traded on the New
York and American Stock Exchange which represent the overwhelm-
ing majority of assets or sales in any given industry. Obviously, this
necessitates a different number of companies as industries are more or
less concentrated. In the autos, with four, you have 100 percent of
them.
Senator NELSON. Are any of the large drug companies closely held
and not actively traded?
Mr. PLOTKIN. Squibb, I believe, is closely held. It certainly is not.
actively traded, and for the Standard and Poor's standard, it is not
included in here. Mr. Conrad, who is an expert in the chemical field,
in trading in that field, might be better able to elucide on this point.
Mr. CONRAD. Standard and Poor's data here includes whatever they
could gather that was publicly held and also what they could back up
in discussions with the companies and through SEC registration state-
ments, proxy statements, and so forth. There are eliminated from this
PAGENO="0113"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1663
a number of larger closely held pharmaceutical firms where there is no
data available.
Senator NELSON. Which are the large, closely held pharmaceutical
firms?
Mr. CONRAD. Well, E. R. Squibb is part of Olin Mathieson. Hoffman
LaRoche.
Senator NELSON. Ciba?
Mr. CONRAD. Yes, Ciba.. They are not in the sample in the report.
There are many that are not. We could not take other than what the
data source allowed.
Senator NELSON. Schering? I don't see that.
Mr. CONRAD. Yes, Schering is included.
Senator NELSON. Well, do you have a. list of the substantial com-
panies that are not actively traded and therefore are not included in
the list?
Mr. CONRAD. I do not have tha.t at my fingertips, but we could put
that together for you.
Senator NELSON. Could you furnish that to the committee?
Mr. CONRAD. Surely.'
Senator NELSON. If my memory is correct, the PMA states that their
members produce about 95 percent of the prescriptions used in this
country. Is that correct, Mr. Cutler?
Mr. CUTLER. Yes, sir.
Senator NELSON. What percent of the drugs produced is represented
by the companies listed here in your report?
Mr. CUTLER. I do not have that information, sir, but if we took a
guess, say, in the range of 75 to 80 percent., I am sure we would not be
more than 3 percent off.
Senator NELSON. Would you have the figure?
Mr. CUTLER. No.
Senator NELSON. I mean do you have it available?
Mr. CUTLER. No, we have not added the figure. I think we could com-
pile a figure as to these 29 companies.
Senator NELSON. Would you supply that for the record?
Mr. CUTLER. Yes, we will.' I think the only unincluded companies
are the American subsidiaries of Swiss companies and a few others,
like Squibb, which are part of a much larger conglomerate, whose
overall results do not really reflect the earnings of Squibb. Squibb's
earnings are just not available to us. But they were given to this coin-
mittee, as I recall, by `Squibb's president.
Mr. PLOTKIN. An important reason why we did not go beyond the
confines of the Standard & Poor's tape just for the drug industry
was our desire to be evenhanded in treating all industries which appear
on the chart. Not having access to nonpublic information would mean
an unfair comparison. Therefore, whatever `bias appears in the drug
company, the same bias is present in all other companies, so the rela-
tive distribution does not change.
Senator NELSON. To get back to the question we were discussing be-
fore recess, if a company were included in an industry grouping that
was high profit and high risk, but that company itself had a long,
1 See Appendix III beginning at p. 2129, Infra.
S1-280---OS-pt. 5-S
PAGENO="0114"
1664 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
long record of high profitability, without any significantly bad periods,
is that still called a high-risk company?
Mr. PLOTKIN. That is a company still participating in a high-risk
industry whose new investments and future investments are subject
to the same vicissitudes as those undertaken by any other company in
the industry, sir.
Senator NELSON. As I recall the testimony this morning, whoever
w-as responding to the question, if a company had a 20-year record of
stability and good profitability, the conclusion by one of the witnesses
was that then at some period, say after 20 years, it would be con-
sidered a low-risk company.
Mr. PLOTKIN. I think it is Professor Cootner's statement to which
you refer and I will turn the mike over to him at this point.
Dr. COOTNEP~. The statement I made was that a 20-year period was
sufficient for discussing the riskiness of an industry. But I did not
make reference to the return of any individual company, I was saying
in other words, the 20-year record of the drug industry was a sufficient
basis on which to make your decision.
Senator NELSON. But you do not say that about an individual
company?
Dr. COOTNER. I think I can give you
Senator NELSON. You would not draw that conclusion about an indi-
vidual company within a high-risk industry after 20 years of stability
and high profitability?
Dr. COOTNER. Well, again, I want to refer back to my previous
testimony, wthere I pointed out that a company could have a high rate
of return for 20 years but still show substantial volatility in changes
reflecting the very variable rate of return it had on its marginal in-
vestments. Just looking through Mr. Plotkin's study, I can give you
a specific reference-for example, the Kendall Co., which has a greater
return, which goes from 14 percent steadily down to-I am afraid the
number is not completely clear, but it looks like 7 or 8 percent, year
after year.
Now, the rate, if you looked at the variability of rate of return from
year to year, `it is rather small. But the sharpness of `that decline over
the entire period suggests that the marginal investments were relatively
unprofitable.
Senator NELSON. What is a marginal investment?
Dr. COOTNER. That is an additional investment beyond the invest-
inent you have at the beginning of the period.
In other words, these companies will be substantial companies at
the beginning of the period. The new investment that they
undertook-
Senator NELSON. In the same product line?
Dr. COOTNER. In the same total product line. That is, I am not talk-
ing about any specific product, but I mean their total investment in
the drug industry.
Senator NELSON. Well, then, supposing you have a company with a
20-year record of stability and good profits, consistently above the
average in all industries, maybe a fair degree above it, you still would
not conclude after 50 years that that company was a low-risk com-
pany? Is that what you are saying?
PAGENO="0115"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1665
It would seem to me if you look at some company over a period of
time, perhaps a 50-year period, it would be difficult to conclude that
either the investor or anybody else can assert that a stable company
consistently making profits above average for all industries is not a
low-risk company, apart from the industry as a whole.
Mr. CONRAD. One answer to-Professor Cootner, do you want to
take that first?
Mr. PLOTKIN. I lust have a point as it relates to the industries and
then to the companies.
The statement about 20 years being sufficient time in an industry,
and we are not quibbling about the precise number, 16 or 20, was not
by looking at the time pattern for 20 years, but if we average 20
individual years' variances that would be sufficient to get a risk idea.
We never claimed that we could look at whether or not the industry
average fluctuated over 20 years and learned anything about risk.
In fact, referring back to my testimony, we explicitly rejected the
hypothesis that the best place to look~ for risk information is in the
time pattern of industry or company results. This is something that
is very commonplace and has been used sometimes in economics, but
without good results. It seems to not be able to explain in the sense
that this line explains in rates of return the true phenomena in
economics.
Now, if you would like an additipnal memorandum on this, we
could submit it to you why in economic terms the temporal variance,
as it is called, the time variance, is not a good measure of uncertainty,
does not let us extrapolate into the future with any certainty.
Mr. Conrad can address himself to the individual industry now.
Mr. CONRAD. To answer the point you raised for this company that
has had 50 years of high profit, high and steady profitability, I think
the correct statement that one might make about that would be that
the company in its past history was in a relatively low-risk situation.
But that does not necessarily allow you to say anything about that
company's future possibilities. That company may today be in a
condition where it faces extremely high risks, even though for 50 years,
it has not. You cannot, a priori, say that because those 50 went by
and the company earned a high profitability, that tomorrow it can't
turn around and either lose money or suffer a significant decline in
profitability.
Senator NELSON. Let's take an industry which does not require a
high rate of technical skill to enter. Now, I will mention an industry
in a moment. I do not mean to say that those who stay in it and are
successful do not have lots of skill, but anybody can get into it and
tens of thousands of people do-that is, highway construction. Now,
everybody who can buy a truck and a grader or a ditchdigger can
enter the industry. As you probably know, tens of thousands of people
are in business who have just one piece of equipment. It is a ditch-
digger. It costs $25,000. You can borrow the money to get into it. They
may not have capital behind it; they may not have adequate bidding
knowhow, or perhaps they do and they go into big business. But
most of the construction people I am familiar with, and I am famil-
iar with quite a few, started out on their own, with their own truck,
or if it is an old company, went in with their fathers. Now, there are
PAGENO="0116"
1666 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
loads of very fine highway construction outfits with a long, long stable
history in every State in this Union-highly profitable, highly suc-
cessful. But if you took a. look at that industry as a whole, you would
have to, I would think, by your measurement or any other measure-
ment, say it is a. high-risk industry. A lot of people get into it who do
not belong there.. But it is not. a high risk for the people who have the
capital and the knowhow and get into the. business. Now, how do you
evaluate that factor?
Mr. CONRAD. One point I would like to make is that we hope to
show by the testimony that will follow now that in the drug industry,
at the point we are at, regardless of performance in the past., the finns,
the highly successful firms, are subject to a set of risks which could
put them into a great deal of trouble.
Senator NELSON. A set of risks that. are not common to or comparable
with other types of risks, in many other industries?
Mr. CONRAD. Correct. They are somewhat, in some sense, peculiar
to the pharmaceutical industry. Those firms, large or small, successful
or unsuccessful, are all subjec.t to these same risks in the pharmaceu-
tical industry.
Senator NELSON. I read the list of factors-some seven of them, I
believe, that you list-
Mr. CONRAD. Five, I think.
Senator NELSON. Well, whatever it is.
But anyway, I did not see anything in your evaluation. You listed
what you considered to be peculiar risks, some of which, when we
get to them, I have doubts about whether they are peculiar risks. I
think they are a security factor in the industry. But did you consider
other factors of special insurance, so to speak, in the marketplace that
no other industry has when you made this evaluation? Such as the
fact t.hat a drug established after 17 years in the marketplace, with its
patents, still coiitinues to hold a high volume selling at several times
the cost of comparable drugs? Did you consider that protective factor
which is not found, to my knowledge, in any other product area in
the marketplace in this country? Did you consider that?
Mr. CONRAD. We did and we will address ourselves to that.
Mr. CUTLER. Sir, if we could go to Mr. Conrad's and Professor
Markham's paper right now, I think these questions are germane to
that paper.
Senator NELSON. All right, but be.fore we do, the biographical
sketches of these gentlemen will be inserted in the record at this point..
(The biographical material follows:)
BIOGRAPHICAL SKETCH OF JESSE W. MARKHAM, PH. D.
Present position: Professor of Economies, Princeton University.
Education: A.B., University of Richmond, 1941; AM., Harvard University,.
1947; and Ph. D. Harvard University, 1947.
Professional experience:
Has served on the faculty of Vanderbilt University, Harvard University
and Princeton University.
Acting Director and then director of the Federal Trade Commission,
Bureau of Economies, 19~3-55.
Has acted as consultant to the Federal Communications Commission, the
Brookings Institution, the U.S. Department of Justice, the Tennessee Valley
Authority and the Office of the Special Representative on Trade Negotiations.
Participant in a Ford Foundation inter-research program on the economics.
of technological change and economic growth.
PAGENO="0117"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1667
Author of numerous publications, articles and reviews in the area of
public policy towards business. Among others, his books include: The Fertil-
izer Industry: Study of an Imperfect Mar1~et (1958) and The American
Economy (1963). _______
BIOGRAPHICAL SKETCH OF GORDON H. CONRAD
I took my academic training in chemical engineering at New York University
and later continued with business administration at Columbia Graduate School
of Business. Since 1956 I have been employed in the Industrial Economics Section
of Arthur D. Little, Inc. conducting market research, economic feasibility and
long-range economic planning studies in the chemical and allied process indus-
tries. During the last two years I have been in charge of a number of economic
studies relating to the operations of the phamaceutical industry which have been
sponsored by the Pharmaceutical Manufacturers Association.
Dr. MARKHAM. Mr. Chairman, and members of the subcommittee,
the short statement I have is to report on the study that Mr. Gordon
`Conrad of Arthur D. Little and I have been conducting for the past
month or two, one of us has to read it, since it is not very appropriate
to do this in the form of a duet. So we tossed, and I am going to put it
in the record.
STATEMENT OP JESSE W. MARKHAM~ PH. D., PROFESSOR OP ECO-
NOMICS, PRINCETON UNIVERSITY, PRINCETON, N.J.; AND GOR-
DON R. CONRAD, SENIOR STAFF ASSOCIATE, ARTHUR D. LITTLE,
INC., CAMBRIDc+E, MASS.
Dr. MARKHAM. My name is Jesse W. Markham and I am here
largely as a consultant to Arthur D. Little Co., Inc., a company with
which I do some consulting from time to time.
We appreciate the opportunity to appear before your committee
and report on some additional studies we are making at Arthur D.
Little, Inc., concerning the relationship between risks and profits in
the ethical drug industry. The study just presented to you indicates
that there is a strong positive relationship between risks and profits
in American manufacturing industries generally. Since this committee
is inquiring into the economic functioning of the pharmaceutical in-
dustry, the question naturally arises as to what specifically are the
elements of risk the pharmaceutical industry confronts. We have been
engaged in a study `designed to shed light on this question. While the
study is far from complete, some of the preliminary findings should
be of interest to this committee.
* The risk/return study just presented shows that the pharmaceutical
industry ranks high among manufacturing industries in terms of risks.
Just what are the components of these risks? Our current study indi-
cates that among the major risks a pharmaceutical firm confronts are
those unforeseeable events beyond its control that could, in a relatively
short period of time, wipe out a substantial portion of its total profits
and total revenues. We have, with reasOnable certainty, identified the
following as major risks somewhat peculiar in this regard to the
pharmaceutical industry:
1. The development of a complete product superior to one of the
company's major products which causes virtually complete replace-
ment of it in a short time.
PAGENO="0118"
1668 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. May I interrupt a moment? Do you consider this
risk higher in the drug industry than any other industry?
Dr. MARKHAM. Not higher than in all other industries; no, sir. I
elaborate on this a little bit later on. I can anticipate that or `go on
to it, but I do have a paragraph that tries to tell whether or not this
is fairly unique of the drug industry in terms of the incidence of the
impact of a new product.
Senator NELSON. Do you consider this risk of t'he development of a
competing product superior to one of another company a poor risk
factor in the drug industry?
Dr. MARKHAM. Yes, I do.
Senator NELSON. Can you give me a list for the last 15 years of the
drugs that have been replaced by a superior competing product.?
Dr. MARKHAM. I do not irnow whether I could do this or not over
the past 15 years. We can certainly give you some fairly striking
examples. I do not even have an exhaustive list of those with me now,
Senator.'
Senator NELSON. I understand you to be talking about the develop-
ment of a competing product superior to one already marketed by
another company, as one of its major products, which causes virtually
complete replacement of it in a. short time.
I do not `think an example here and there proves anything. I am
assuming that you came to the conclusion that this was an important
risk factor after a detailed study of examples of where that happened
and where the company marketing the original product as a conse-
quence suffers substantially. What I am interested in for the record, in
addition to your conclusion, are the factors which caused you to reach
the conclusion. For the record, I would like to see the name of the
drug, the company involved, and I would like to see the impact of
this particular case upon the earnings, the profit of that particular
company. Because if it did not have any impact or a substantial impact
on their profits, I cannot see that it means very much.
Dr. MARKHAM. I am delighted that you want it and so do we. Let
me elaborate. We have not finished this study yet. We have some
preliminary results. That particular aspect of the study has been
formulated. We have not gotten yet the amount of data that would
make it worth while to submit it at this time. What we do have, how-
ever, are the data that you will find in "Trends in Market Shares for
Ethical Products," which, in an overall sense, documents the point
pretty well, that products are knocked off the leading list very
frequently.
If you will take a look at what firm ranks No. 1 in a given therapeu-
tic class, 2, 3, 4, and 5, and then look at where these firms rank 5 or 6
years from now~
Senator NELSON. When you say a. given therapeutic class, you are
talking about antibiotics or~
Dr. MARKHAM. Diuretics, things of that type, yes.
Senator NELSON. But you again reach your conclusion as to item
1 without having an example of a series of compelling specific cases
that is large enough to prove the case.
1 See Appendix III beginning at p. 2129, infra.
PAGENO="0119"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1669
Dr. MARKHAM. No, we have some examples, Senator. For example,
Merck in production of Diuril hit very hard all the producers of
diuretics and it hit it instantaneously at the time it was introduced. We
have not, let me make it clear, built a complete catalog of all such
instances. I referred, however, to the interchange in range of firms
which largely goes at the rank of their product in that particular
therapeutic class. You could-evidence presented here and contained
in the hearing before the late Senator Kefauver's committee as well
shows a good deal of turbulence among the various products. The
only way you can, for example, be dropped from one to where you
do not even appear in the top five is that some product came and
knocked you out of that list. When you see firm B, in other words,
displacing firms A, C, D, E, and F, then one has to infer from this
that a product has knocked another product off of the list of the top
five.
Senator NELSON. Well, when they have been knocked off the list
of the top five, you are talking in term of total sales?
Dr. MARKHAM. No, I am talking in terms of sales of that particular
product, therapeutic class, yes.
Senator NELSON. Are you talking about amount or dollar sales?
Dr. MARKHAM. I am talking only about rank.
Senator NELSON. Rank in total amount of products sold? Did you
consider when you used this factor that you can have less than 50
percent of the marketplace and substantially more than 50 percent
of the dollar value of the sales in this~ business, this industry? Did
you consider that?
Dr. MARKHAM. I have not considered it. There is a study-I do not
know whether it is in evidence or not, Senator Nelson-that puts this,
I believe, in terms of dollar value and in terms of-
Mr. CUTLER. Mr. Chairman, perhaps I should offer now this study
entitled "Trends in Market Shares for Ethical Pharmaceutical Prod-
ucts," and the accompanying statement of Mr. Gordon Conrad, which
you have before you, which we did not plan to read but which we
planned to testify about. For each of the pharmaceutical classes, we
do break down the sales by product on a prescription basis and on a
dollar basis so you may make those comparisons.'
Senator NELSON. On the number of prescriptions and the number
of dollars involved?
Mr. CUTLER. That is right; for the sale price.
Mr. GonioN. But this deals only with products. It does not deal
with firms; is that not correct?
Dr. MARKHAM. I think we have some additional evidence also on
firms in terms of how they rank among the leading pharmaceutical
manufacturers. I assume this is put together in terms of total revenue
as well.
Mr. GolmoN. Just a second. It does not tell us, for example, how much
investment was put into these. Suppose a drug is made obsolete. The
firm producing the drug may have recouped their money in 2 or 3
years. It may have had very small investment. Now, it seems to me that
what we ought to have is an exhaustive-not just a sample, but an
1 The prepared statement of Mr. Conrad and accompanying study, "Trends in Market
Shares for Ethical Pharmaceutical Products," begin at p. 1784, infra.
PAGENO="0120"
1670 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
exhaustive-history of those drugs which have been superior to one of
a company's major products and caused virtually complete replacement
in a short time. These are the conditions given in your statement.
Dr. MARKHAM. Well, one slight modification to that "causes virtu-
ally complete replacement." I am aware of the fact that many drugs
that lose their commercial position for the firm are still held on. They
appear among the drugs that are offered for sale.
The point that I am trying to go into here is that this is an industry
in which you get rapid product innovation. I do not think anyone
denies this.
Mr. Goiwox. Dr. Markham, we are talking al)out this particular
one risk that you have mentioned as risk No. 1.
Now, in addition to the exhaustive list which I hope you will supply
us with, as the chairman suggested, that you will supply us with the
investment for that particular product which has been made obsolete;
what the investment for that particular product recouped; how much
money was lost, if any; how much money was spent on research and
development by the firm for the product; as the chairman asked you,
what happened to the profits as a percentage of net worth or on invest-
ment of the firm; whose work was made obsolete.
Dr. MARKHAM. I think that last we can give you.1 I acknowledge the
relevance of what you are asking for. I think I can categorically state
for you you can never get that information. The amount that is spent
on research and development on a particular product is philosophically
impossible to determine as well as from an accounting point of view
impossible. The number of times you make an experiment and you come
back with the answer no, for example, the number of compounds that
ou test and find that they are of no immediate therapeutic value, are
nevertheless a part of the research and development cost that in a
sense should be charged to some of the successful products. I do not
think that any set of books will reveal for you or to me in the terms that
I would like to have them all of the investment, including resea.rch
and develooment, technical research and development, including
clinical testing and all the other costs that go-I just do not believe
that you will find, in other words, profit-and-loss statement by product
line. The companies have at least convinced me they do not `have these
data. I am going to, in a moment, come to some possible estimates that
might be relevant to this point, but you are just not going to get those
data and I do not think-I would be less than honest if I said I would
try to get them, implying that I could get them for you.
Mr. GORDON. In my opinion, Dr. Markham, that is very essential
data, because a drug can be made obsolete after the company has made
a considerable amount of money on it., even after its investment has
been rec.overed. I would not call that a. risk.
Dr. MARKhAM. I think you would concede-let us take this last set
of data you asked for. If he can show, for example, that a particular
product by a rival, contributed to the effect, I think you would take
this, although it is not specific data, as meaning it would have at least
nn impact on the drug company's revenues.
Senator NELSON. That is true, but you are using a 17-year period to
test the risk of your company. These companies are all in the same
1 See Appendix III beginning at p. 2129. infra.
PAGENO="0121"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1671
business. Doe.s this knock them out of bus;iness? Does this knock them
way down? Every company that is any competition has this risk. I
was just looking at the reports on the manufacturers of sugar pops and
Wheaties and corn flakes, and so forth. Almost every month, General
Mills or Kellogg or somebody else is coming out with a new cereal and
somebody's frosty flakes are knocked out by somebody's sugar pops. It
is going on all the time. When you loOk at the stuff they are putting
out now as compared with what they did when we were kids, you can
see that they are doing it right along. It is all in the same business and
going up and down just like the drug companies. But the company,
General Mills, is doing right well; so is Kellogg.
Dr. MARKHAM. Let me make it clear that I am not making any
presentation here, Senator Nelson, whatsoever, that the drug industry
is the only industry that confronts this kind of risk. Later on, I try to
show the unique magnitude, or rather the magnitude of this risk and
that it does have a certain element of uniqueness when applied to the
drug industry, but it certainly is not a risk that the drug industry
alone confronts. No one is making that claim here.
Senator NELSON. That is what I want; to make clear for the record.
Dr. MARKHAM. Nor am I trying to assert here that the drug industry
is the only risk industry in the economy. The chart you have just had
given to you there will document in my mind fairly conclusively that
other industries have high risk as welL The point that has been docu-
mented is that, in general, the rate of return and the industry risk tend
to be strongly correlated. But that is not only the drug industry. There
are 59 points on that chart from which that particular line was made.
And there are many other industries that fall on the risk scale as welL
Senator NELSON. I was listening to your statement. That is why 1
raised the question about your point that "we have, with reasonable
certainty, identified the following as major risks somewhat peculiar in
this regard to the pharmaceutical industry."
The word "peculiar" itself means unique. It means somewhat rare
in my opinion.
Dr. MARKHAM. It is somewhat rare in that all these five, plus some
others, converge at the same time in this industry. I would not make
any presentation to you that the drug industry is the only industry that
suffers from product competition.
Senator NELSON. Then you will submit evidence showing the im-
pact of the introduction of new drugs upon the profits of firms manu-
facturing less effective but comparable older drugs. I do not think
select handful mean anything. If it is a factor that produces a high
risk industry, it must be a factor that is burdening them all the time
and has been in the past with considerable frequency. If it is in the
range of frequency for a number of other industries, I do not see that
it is peculiar to this industry.
Dr. MARKHAM. I am arguing that it has peculiar impact, and that
is one of the points I am going to try to document, Senator.
Again, I can simply repeat that I am not trying to urge on you here
that the drug industry is the only industry that suffers from or en-
counters severe competition from new products. I agree with you about
the cereal industry; occasionally also in new models in automobiles.
In response to your request that I submit to you additional informa-
PAGENO="0122"
1672 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
tion or cases where the introduction of new products has significantly
affected the profits or the revenues of competing firms, I have to defer
to Mr. Conrad, because this would be done at Arthur D. Little's. I am
just a consultant there. I do not frame the projects, nor do I execute
them myself. Perhaps he would like to answer that now.
Mr. CoxRAi. One point. \Ve, in constructing this analysis that Pro-
fessor Markham is presenting now, had the belief that the drug in-
dustry is subjected to the risk whether or not one can list over the last
15 years, 20 or 30 or three or four instances where this type of risk
took place. The fact that logically, it can take place, that the industry
is set up, the medical profession is set up iii such a way that it is pre-
pared to replace a product virtually overnight if it finds a superior
agent to treat a particular illness, the fact that this characteristic
exists in the pharmaceutical industry we see as an element of its spe-
cial riskiness. There are other industries that have product obsoles-
cence, but they do not have it with the peculiar character of product
obsolescence that the drug industry has.
Now-, we can cite examples. We have talked to some of the companies
in starting this study and we will be able to cite some examples where
product obsolescence has hit the company and had some impact on
its earnings. A list of 30 or 40 may not be possible. We do not know
yet just how many companies can respond.
But the point is that, regardless of how many, the risk from the
economist's point of view is there and it is real and the companies are
also facing it.
Senator NELSON. I suppose there is risk there, but it would seem to
me that if it is a. risk that is special to this industry, you ought to pro-
cluce evidence of its frequency and then also show the impact upon the
particular company within the industry in each case. If it is a tem-
porary impact, that is one thing. If it is a permanent impact, that is
another thing. But the fact is that the leading manufacturers in this
industry have been the leading ones for a long, long time. They have
identification in the marketplace, they have identification with the
physician. They are producing more than one product. They are all
doing research and so they may have one product knocked out by an-
other firm. It would seem to me that you would have to show some
impact upon a number of companies as a consequence of this type of
situation; otherwise you could not conclude that it is a very serious
factor in risk, it would seem to me.
Dr. MARKHAM. May I just cite one example, Senator Nelson, to il-
lustrate at least the point I am trying to make?
I did use diuretics as an example. If you were to take a look at the
market share of product No. 1 on the chart on page 42 in this docu-
ment that was just handed you, Table 11-A, Diuretics, if you take a
look at 1957 it had 53.5 percent of the market. In 1958, 1 year later, it
had dropped to 17 percent and by the next year, had dropped to 7.8.
Go to the bottom of the numbers in that table. Product No. 5 in 1957
w-as eight-tenths of 1 percent of the market.. By 1958, it had 72.1 per-
cent of the market.
I only submit that in spite of the fact that other industries encounter
product competition, this is an unusual shift in market share in that
product line. I would find it very difficult, and I know something
PAGENO="0123"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1673
about markets in manufacturing, I would find it difficult to find other
examples that come close in these magnitudes.
Senator NELSON. What were the names of the two companies?
Dr. MARKHAM. I do not know, I suspect No. 5 was Merck. It would
have to be. These are products and that is, I suspect, Diuril. But I
have no idea. of what the others are.
Senator NELSON. You say Merck is the one that had the 35 percent
in 1957?
Dr. MARKHAM. No, sir; it is one that had practically nothing in
1957 and 27 percent in 1958. The others, however, had their shares of
the market decline very perceptively.
Senator NELSON. If you will give me the name of the company that
had 35 percent rate of return in 1957~ and dropped to 16 percent in
1958. Does that column for the company run vertically?
Dr. MARKHAM. No, sir; it goes across the page. It dropped to 16 in
158 and to 7.8 in 1959, 6.6 in 1960, and by 1965, that same product
had dropped to 3.7 percent.
Senator NELSON. Now, then, it is company No. 5 here that went from
~o percent to 72 percent of the market in that product, is that it?
Dr. MARKHAM. Yes, product No. 5.
Senator NELSON. Now, could you give us the names of those two
companies? Merck is No. 1, is that it?
Dr. MARKHAM. No, sir; I am virtually certain that product No. 5
is Merck's Diuril. I do not know, but we can, I assume, supply data.
Again, I have to defer to Arthur D. Little, because I do not know the
particular conditions or safeguards that surround these data. But I
assume they can be gotten.
Se.nator NELSON. Do you have the names of the companies?
Mr. CONRAD. Not with me.
Senator NELSON. I must say that does not help very much. But we
have Merck Co. here in 1957-58. They went from 8/io percent to 72
percent of the market in this product. The earning assets column
shows that they went from 13.79 in 1958, 14.060 in 1959, and down to
12.393 in 1960. So the fluctuation based upon the one who got 72 percent
of the market was very tiny. If it were the same on the company that
lost the market, I do not call that much of a risk factor, either, do you?
Dr. MARKHAM. Yes; I do consider it a significant factor.
First of all, I think it ought `to be clearly understood that, take any
group of firms, particularly looking tOward the top, you are talking
about firms who have somehow met risks, somehow or another you
would expect that, by definition, the leading firms over a long period of
time will be leading firms in almost any given year. That does not in
any sense indicate that the risk is not there. Some firms undertake
satisfactorily and successfully risk. There also are some winners. There
are a few winners at Monte Carlo who win consistently. But that does
not alter the probabhistic outcome of the gambling game.
Senator NELSON. Well, maybe so far as your mathematics and eco-
nomic theory are concerned, it does not, but the cold, hard fact of the
matter is that you can look at the earnings statement of General Motors
and it will fluctuate 10, 15 times that much in a single year without
any additional competition, holding the same share of the market. So
now you say that a company loses a substantial share, a major share,
PAGENO="0124"
1674 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
of the market on one product and there is a fluctuation of a half, a third
of a percent and that, therefore, this is a. risk factor. I say to you I have
looked at General Motors over the years and they will fluctuate 10
I)ercent or more without any change in the marketplace at all. That is
true all the time.
You are saying here that obsolescence of a drug constitutes a risk
factor. I am just saying that you have areas of industry where there is
no obsolescence at all, a.nd just the ordinary play of the marketplace
causes fluctuations.
Dr. MARKHAM. You are only confirming the theory that has been
given you, Senator. The automobile industry ra.nks very high. It hap-
pens that they have a different kind of risk. Given a collapse in con-
sumer expenditures for durable goods, especially for automobiles, they
encounter risk, too. No one denies the automobile business is not highly
risky. You are quite right in singling that one out. General Motors
is an excellent case of a company that encounters risks, but different
from these. You would be hard pre.ssed-at least I would-t.o find a
single General Motors model that went from zero on to 70 percent. in
even a model in a wide price range in the market, or to find another one
that had been reduced from 35 to 11.
What you are saying is that the automobile industry has certain
kinds of risks; they are a cyclical nature. And indeed they do. And I
believe they show up in the three or four top industries in terms of risk
and the automobile industry is about the same as the drug industry.
Senator NELSON. 1ou say you would be hard put t.o suggest a com-
parable instance in the auto industry. I might point out that Corvair
sales in 1 year-if I am not mistaken-fell about 50 percent. I recall
that the drop was very, very substantial. I think around 50 percent..
So that is a peculiarity of the circumstances involving a.n automobile.
Dr. MARKHAM. I am sorry, sir, I did not hear the percentage you
were using.
Senator NELSON. I say what sticks in my memory is 50 percent. I
may be wrong. I will check it and correct the record if necessary.
Does Mr. Cutler know?
Mr. CUTLER. YOU are asking about the drop of Corvair's share. of
the market?
Senator NELSON. I do not know about the share, because I do not
irnow who they compete with.
Mr. CUTLER. It is true that the Corvair's sales dropped over a year
by a factor of about 50 percent, but all of Corvair's sales could not have
amonnted to 1 percent of 2 percent. or 5 percent. of the sales of General
Motors.
Sena.tor NELSON. We have not in the record here what percentage of
the sales were represented by the dramatic drug instance Dr. Markham
has suggested
Mr. CUTLER. Dr. Markham is about to give you some data or some
theory he has on this subject.
Senator NELSON. Before we get. to the theory, I would like to ask a
question on the point we were talking about a while back. If it is a risk
wh1ch ultimately affected a company~s earnings, and its general sit-
uat.ion as a. competing company, not. for 6 months or a year, but for a
16-year period, that would be much more. impressive to me than simply
PAGENO="0125"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1675
to say there are lots of products coming on the market in the drug in-
dustry. You are bound to have that. But it is `the actual impact that
makes it a risky business.
You cite a case in which one company picked up almost all of the
business but their earnings assets column barely changed. I would like
to have the name of the company that lost the business to see what that
earnings assets column looked like.
Dr. MARKHAM. I think we can furnish you some information on
that, but we are not prepared to give you a detailed list of that today.
Senator NELSON. All right, if you would submit them for the recoid.
Dr. MARKHAM. Fine.'
Senator NELSON. We will put it in the record at this point along
with table ha on page 42 of the document entitled "Trends in Market
Shares for Ethical Pharmaceutical Products."
(Table ha referred to follows:)
TABLE 11-A--DIURETICS
[Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 52.6 53.5 17.0 7.8 6.6 5.6 4.7 5.1 5.0 3.7
2 13.3 12.9 3.3 1.4 .9 .4 .2 .2 .4 .3
3 6.3 2.4 .2 .1
4 5.6 3.1 .6 .3 .1 .1 .1 .1
5 . 8 72. 1 61. 8 42. 3 33. 3 32. 7 23. 7 23. 1 25. 9
6 14.3 20.6 19.6 17.3 18.8 14.4 20.0
7 7.8 11.3 7.8 6.9 4.5 3.3 4.9
8 5.4 5.0 5.4 6.3 7.1 4.1
9 1.0 3.0 4.9 5.7 6.6 5.9
10 .2 2.2 3.2 5.7 8.7 3.3
11 , .1 3.5 4.1 5.0 5.1 4.7
12 .4 6.1 4.9 4.4 4.9
13 .2 2.6
Cumulative market
share of above
products 77. 8 72. 7 93. 2 93. 5 88. 5 80. 9 85. 6 79. 9 78. 4 80. 3
All others 22. 2 27. 3 6. 8 6. 5 11. 5 19. 1 14. 4 20. 1 21. 6 19. 7
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
Senator NELSON. All right. I `did not have any more questions on
the first item that you listed. You were going on to No. 2.
Dr. MARKHAM. Yes, No. 2. And we are enumerating the various
risks. Back again to the heading in the statement, the major risks,
somewhat peculiar in this regard to the pharmaceutical industry.
2. The discovery of unanticipated side effects of a drug which lead
to an immediate limiting of the physical `indications for which it may
be prescribed.
Again, an example of this might be Chloromycetin, which was with-
drawn because it was found to have certain-it gave rise to some blood
disorders.
Senator NELSON. You are talking about a situation back in
1950-something?
Dr. MARKHAM. I am sorry, I used the~ wrong word. It was limited
in its use and in a sense, withdrawn by prescribing physicians from
some uses because it was found to generate or be conducive to certain
blood disorders.
1 See Appendix III beginning at p. 2129, infra.
PAGENO="0126"
1676 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. Go1i~oN. But what happened to it, Dr. Markham?
Dr. MARKHAM. Well, now we can be a little bit more specific in the
risk and return of American industry you have. That was the volume
we were discussing before lunch.
I think I am correct on this. If you turn to page 60 of the Risk and
Return in American Industry, the disorder that that drug created was
discovered in, I believe, 1952. If you will take a look at Parke, Davis
& Co., the manufacturers of the drug-
Senator NELSON. Just a minute. We do not have the right one before
us.
Dr. MARKHAM. Its profits were reduced almost by 50 percent. Now,
they went from a 20 percent rate of return in 1952 to 11.4 percent in
1953. Let me make it clear, I have not, Senator, made the precise
analysis that can establish the cause and effect on this, but following
your own line of reasoning earlier, it looks as though that was a sig-
nificant blow to Parke, Davis & Co. in terms of what happened to
profits.
Mr. GORDON. Dr. Markham, even though, as I understand it, Parke,
Davis' profits had fallen, it was still 1.1 percent of net worth after
taxes or less than 3 percentage points below the average for all large
drug firms for that year.
Dr. MARKHAM. Well, that may be; I do not know where you get the
12 percent. My table does not show that. If you go to earnings on
assets-
Mr. GoRDoN. Well, I went some other place. I can put. that source in
the record, too.
But even then, when the profits went down, it was still less than 3
percentage points below the average for all large drug firms for that
year.
Dr. MARKHAM. Oh, sure. What one then has to ask himself is, sup-
pose two product lines had been hit. The probability may not. be high,
but suppose it had.
Mr. GORDON. And incidentally, do you know what happened to
Chloromycetin last year? Sales were $45 million in the United States
and over $70 million worldwide.
Dr. MARKHAM. It may be, but I see by the morning pa.per, that has
not solved all of Parke, Davis' financial problems. I do not want to
read this into the record.
Mr. GoRDoN. We are talking a.bout Chloroinycetin now.
Dr. MARKHAM. Yes. All I am saying is that even t.hough it may
have come back, disorders do get cleared up, I assume, the point is
really a very simple one. A single product that gets hit because of this
particular risk drops the profits by 50 percent. I do not think anyone
would argue that that is not a. significant risk.
Mr. GORDON. Even then they could still be making a high return on
their investment.
Senator NELSON. So the record will be clear, do you have the total
sales of Chloromycetin for the year 1952 versus total sales by the com-
pany? I would like to have that projected. We do have the figures
on the sales of Chloromycetin, I believe, for last year7 which were $45
million in the United States and $70 million worldwide and it repre-
sented a major factor in Parke, Davis' total sales.
PAGENO="0127"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1677
Dr. MARKHAM. I think we can supply this. You want the total sales
of Parke, Davis & Co. in 1952 and 1953 and also the sales of
Chloromycetin?
Senator NELSON. Yes, and for each ~year as a proportion of total
sales, to see what that factor was.
Dr. MARKHAM. Yes. I can only report we will all say we will try to
get it.'
Senator NELSON. All right.
In any event, discussing the risk factOr, that is, a risk, if that were
the factor that did cause the decline-I do not know. They were back
up within a matter of 3 years to within~ 3 percentage points and then
they jumped in 1957, more than 3 percentage points on earnings above
the 1952 figure. And, they stayed 4 years in a row above the 1950
figure.
All I am saying is that this may be one risk factor. There are all
kinds of risk factors and I would just point out that you could prob-
ably project a whole series of companies~that did not face any product
replacement, obsolesecence, competition, and still experienced the
same fluctuation.
Dr. MARKHAM. I will concede it did not drive Parke, Davis out of
business and that it did subsequently recoup. I do not think, however,
that that would lead one to the conclusion that this is not a risk that
confronts pharmaceutical manufacturers. I shall continue on, sir.
Senator NELSON. Go ahead.
Dr. MARKHAM (reading). 3. The discovery that a drug can be mis-.
used in ways that create a significant social problem which then leads
to limitations on its marketability or possible removal from the market..
I am told that meprobromate might be a good example of that.
Mr. GORDON. Has that been removed from the market?
Dr. MARKHAM. No, just its sales are subject to greater control.
4. Withdrawal or restriction of a drug by the U.S. Food and Drug
Administration until additional evidence of safety or efficiency is
produced.
I am told that MER-29 might be an example of that.
Senator NELSON. Are you going to b:e able to give us a series of
examples? Again as with the first question, if it is an important risk
factor, it seems to me we ought to see the number of examples of drugs
that either had to be limited in their use because of discovered undesir-
able side effects not known at the time of the New Drug Application,
or for some other reason. But I would think in order to prove it is
a risk factor, peculiar or somewhat peculiar, as you put it, to the indus-
try, we would need a series of examples to demonstrate that fact. If
you have concluded that that is the case, you must have some examples.
Dr. MARKHAM. Yes, we are in the process of compiling such a list,
Senator.2
Senator NELSON. Let me ask you a question: you referred to unantic-
ipated side effects or withdrawal or restriction of the drug on the
market.
Now, how is that risk, in terms of imposition of expense, more im-
portant, for example, than the risk of~ the automobile industry on
auto recalls?
1 The information had not been supplied at time of going to press.
2 See Appendix III beginning at p. 2129, infra.
PAGENO="0128"
1678 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. MARKHAM. Oh, considerably greater. I happen to have followed
rather closely the steering defect in Mustang. It does not affect the
market at all. What you do do is you incur some, the dealer or the
manufacturer-I am not so sure how they share it-you call in to have
this particular problem corrected right away. This does not happen
in the case of drugs. You cannot just call in the thousand bottles of
pills that you sold and say, we will try to correct this. It just does not
happen that way. It is not a mechanical defect, in other words, Senator,
that you can take the current output in inventory and say, well, we
incur some slight expense to eliminate this particular risk.
Senator NELSON. Do you know how many cars were recalled last
year?
Dr. MARKHAM. No, sir; I do not. A substantial number. But I would
say that is not the same kind of risk at all. I happened to take one
back myself.
Senator NELSON. I do not have the figure on costs, because I do not
know if it is available. But there were over 4 million cars recalled
in 1 year's time. I assume-maybe Mr. Cutler would have an answer
to that, too-I assume the manufacturer pays the dealer for handling
that aspect.
Mr. CUTLER. Just from memory on this, I am not sure of the 4 mil-
lion figure, but in any event, a figure of that magnitude is the number
of cars that had to be inspected to determine whether they might. con-
tain the particular defect. The number that contained the particular
defect that had to be repaired was far smaller.
Senator NELSON. Correct.
Mr. CUTLER. That is a part of the warranty cost that every company
incurs and factors into its pricing. Compared to the removal from
the market of a product such as MER-29, or a decision such as in the
case of meprobromate that it will be subjected to drug abuse, the effect
of the latter restrictions would be far greater. In any event, we concede
that the automobile is a very high risk industry, with almost the
identical rate of return as this industry.
Senator NELSON. Go ahead.
Dr. MARKHAM (reading). 5. The development of quality control
problem which necessitates withdrawal of a product from the market
until the problem is traced and alleviated.
I offer as just an example, Cutter's polio vaccine.
This list of risks is only illustrative and we make 110 pretense that
it is complete. Moreover, we do not mean to suggest that other indus-
tries do not confront risks. The study just presented to you provides
abundant evidence that American manufacturing industries confront
varying degrees of risks and that these risks are strongly correlated
with profitability.
Our study of indivithial pharmaceutical manufacturer's experience
reveals that these risks are not simply potential but that most have in
fact materialized in recent years. We can provide illustrations in suffi-
cient numbers to support this point should the committee wish us to
cite them.
The most important issue to which Mr. Conrad's and my study is
directed, however, is not the frequency with which such risks actually
materialize, but the magnitude of the loss in profits and revenues tile
PAGENO="0129"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1679
typical firm would suffer should one of these potential risks material-
ize. There may be several approaches to answering this question, but
the one we are pursuing involves the measurement of the extent to
which the major pharmaceutical firms' revenues and profits are con-
centrated in relatively few product lines. The logic underlying our
particular approach is quite simple. If the typical firm sells numerous
pharmaceutical products, no one of which accounts for more than a
small fraction of total sales, the above risks may not be significantly
larger than the usual risks American business firms encounter in their
day-to-day operations. The materialization of any one of these risks
would simply mean that a relatively small share of a firm's profits
and revenues would very likely evaporate-an event that almost any
business firm, confronting average risks, might experience, and ex-
pect to experience, in a dynamic economy such as ours. On the other
hand, if profits and revenues are highly concentrated in two or three
product lines the potential loss from the materialization of any one
of the enumerated risks might be quite substantial. The potential col-
lapse of a single product's sales accounting for a large percentage of
a finn's profits is obviously a much more serious risk than the collapse
of the sales of a product accounting for a very small fraction of a
firm's profits. Such a potential situation becomes even more uncertain
and risky in the case of the pharmaceutical firms when account is taken
of what they may have in prospect after the possible collapse of a
major product. The types of collapse we refer to do not offer hope,
in most cases, of subsequent recovery of the product's market position.
Senator NELSON. Do you have some examples of cases where there
has been such a product collapse and the company has not made a sub-
stantial recovery?
Dr. MARKHAM. Oh, yes, we have numerous products.
Senator NELSON. I mean companies.
Dr. MARKHAM. No, we are not saying that a company collapses,
although I think perhaps there may be data on that, too. I am saying
that a firm's product line, if indeed it does turn out that product lines
can evaporate and do evaporate, then the mere fact that a large per-
centage of the firm's sales or profits is concentrated in relatively few
product lines makes that potential risk much more of a risk than if
you produce, say, a hundred products and 1 percent of your profits
were just undercut by the introduction of a product.
As I have said, that may be a normal risk that any bilsiness firm
encounters.
Senator NELSON. Well, I certainly appreciate that you have a risk
here but the likelihood of it happening and having serious consequences
does not seem to be too great.
Mr. CUTLER. As an example, you might look at the Cutter Labora-
tory figure on page 50 of the risk and return tables. Cutter's problem
with its polio vaccine developed seriously in 1955 and now, in these
1965 figures, 10 years later, had not rea~ched the rate of return which
it had enjoyed before that particular event occurred. So in 10 years
Cutter has not yet fully recovered from the effect of this particular
incident relating to its polio vaccine.
Senator NELSON. But if you backed up to the year 1952 you would
see that in the year 1965, they are approaching four times the 1952
earnings.
81-280-68-pt. 5-9
PAGENO="0130"
1680 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. MARKHAM. But you cannot read that tahle without concluding
that it has taken Cutter a very long time indeed to recover from that
particular `incident and they are not back yet to where they were just
before it happened.
Senator NELSON. I am not sure that this is the real factor involved.
The first year after it happened they went down from 16.8 to 5.5, but
the next year they were up to 13.7.
Dr. MARKHAM. You are omitting 1 year. They went from plus 16 to
minus 17.
Senator NELSON. Oh, I see. Then they went up to a plus 13 2 years
later. Figures are figures. I do not know what that proves. I do not
even know how much they had invested in polio vaccine and what
other factors might have `been involved. All I am saying is that sure,
it is true that if a company has a lot of products and they each repre-
sent only a small percenta.ge of sales and one gets knocked out, that is
not going to be a. substantial loss. Then you go on to say if a firm has
only one or two products and one gets knocked out, it is catastrophic.
You have presei~tecl 29 companies here. They are manufacturing,
according to Mr. Cutler's educated guess, 75 to 80 percent of all the
perscription drugs in America. So they are the ones who are the `big-
gest in the marketplace. Can you show me any of those, anything sig-
nificant among several of those 29 that indicates the kind of risks you
are talking about that has seriously damaged them?
Dr. M~uKIr~r. If you are asking me as a professional economist, I
have a difficult time putting myself in the position of president of
Cutter in the year 1955. But as an economist standing on the sideline,
I have no difficulty whatsoever being impressed with a firm who goes
from plus 16 percent rate of return on his equity to minus 17 percent
because a product has encountered this risk. It seems to me fairly evi-
dent on its face. I do not need to look much beyond that.
Senator NELSON. Let me ask you the question, who are you im-
pressed as an economist with what happened between 1950 and 2 years
later, 1952, when they went from 19 percent to 3.2 and the drop did
not involve polio vaccine at all?
Dr. MARKHAM. That may be very true. I have not examined to see
what caused that drop in rate of return. I am impressed also with that
decline in the rate of return.
Senator NELSON. How do you know that the decline to minus 16.55
had anything to do with polio vaccine if you cannot explain to me what
caused them to go from 9.5 to whatever in a 2-year period?
Dr. MARKHAM. I made this statement earlier, Senator, I make it
again: I have not traced through the sales of any one of these products
into the revenue accounts of these firms. The timing here is so close
with the problem of the polio vaccine, I think the reasonable inference
is that that. was one of the risks that the firm encountered in the year
1955.
I have also made it clear I have not tried to catalog every single
risk that Cutter or any other pharmaceutical firm might encounter.
it is perfectly possible that a confluence of all five, seven, or 10 major
risks tended to drive its rate of return from 20 down to 3 percent over
the previous time that you have indicated. I have not studied that.
Senat.or NELSON. So as a scholar, maybe you would recognize that
the same fa.ctors were at work in 1955, when t.hey dropped to minus
PAGENO="0131"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 1681
16, as were at work from 1950 to 1952, when they dropped a total of
15 percent. And maybe it did not have anything to do with product
replacement or obsolescence at all.
Dr. MARKHAM. It may possibly be, and I am making no assertion
that it is. .Just to be again understood, ,I only say I am impressed with
the magnitude of the change, the timing with the polio vaccine prob-
lem is anmistakable. That is as far as I would care to go.
Senator NELSON. Well, you have given us financial data about 29
companies, which we all have studied very carefully. The drug indus-
try ranks as one of the highest in earnings of all industry.
Can you name for the committee among the 29, then, for which you
have shown dramatic losses, one that suffered any kind of a detrimental
effect as a consequence of drug obsolescence?
T)r. MARKI-IAM. I have not gone over these tables, Senator Nelson~
Senator NELSON. Oh, all right.
Mr. CONRAD. We have not gone over these tables in the detail, as
you have suggested, that would be required to say with firmness what
happened in the Cutter situation. We have not talked to the financial
people and the marketing people about the specific impacts on their
earnings from any of the drug changes that Professor Markham has
been suggesting. But if we can just quickly read through the list and
see where there were a number of instances of very significant change
in profit level for these 29 companies.
Senator NELSON. But if they substantially restored their position,
do you call that a risky company?
Mr. CONRAD. As long as the company is subject to significant fluctu-
ation in its profit levels over time, the fact that it restored profit levels
does not change the fact that it was a risky situation w~hen it oc-
curred. It was subject to the risk, the risk took place, the profitability
dropped, the profitability then went back up. The `company, through
diligence and through research and development and good market-
ing, was able to return itself to the better position. But it is still sub-
ject, even after returning to this better position, to the same set of
risks.
Senator NELSON. I just point out tli~tt even Cutter, which you recite
as the sole example thus far, has met the industry average 7 out of the
last 9 years, or even shown profits above it.
I)r. MARKHAM. No; I think you are possibly reading the wrong
column. The colunm that we used in the study and that we used when
we set the 10.75 percent all-industry ~verage this morning, that is re-
turn on total invested capital, the second column, earnings plus interest
over book capitalization.
Senator NELSON. You used column No. 2?
Dr. MA1uicIJAM. Column No. 2 is the column that the study uses and
the ones we referred to this morning when we calculated an all-indus-
try average return over the 16-year period of 10.75 percent. Using
that column, Cutter has been below that level in all except one year~
Senator NELSON. As a matter of fact, that is Cutter's position in
the marketplace. They have. only been above it t wiee in the last 1G
years. So you have a company that does not have an earnings record
that is at the average anyway. Right? They are low.
Dr. MARKHAM. And for 3 of the years prior to the polio thing,
they were 9.9, 9.2, and 2.9.
PAGENO="0132"
1682 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
So I do not think that proves an awful lot in terms of this big
risk you are talking about.
Mr. CUTLER. Sir, if we can get this information from Cutter and
if we can demonstrate to you that Cutter's decline in rate of return
was primarily attributable to the incident that occurred with its polio
vaccine and that it has not yet. recovered, would you agree that that
is evidence of a substantial risk of t.he type that. Professor Markham
is talking about?
Senator NELSON. I do not know how well they manage the com-
pany or anything else.
Mr. CUTLER. We will submit that.'
Senator NELSON. I~t may very well be a. factor. All I am saying is
that it does not impress me, quite frankly, very much. You have the
whole manufacturing mdustry and all the best economists in the
country available and we are quibbling a.bout one case which may or
may not support your thesis. All I am saying is that you are proving
tha~t these factors, such as obsolescence in the marketplace, constitute
~a significant risk factor beyond what normal industry would face.
Then let's take a look at the major companies and see how many of
them went into a decline and have not recovered a reasonable spot in
~the marketplace because of obsolescence of a drug. That is all. You
~would not go into court with less than that. You cannot go in with
one example to prove a case or use the exceptions to prove a rule.
Mr. CUTLER. It simply depends on the court, Mr. Chairman. The
question is will the judges of the court be economists or who will
they be?
Senator NELSON. I think any judge, economist or not, is going to
say, give me a pattern of evidence that proves the case. I am asking
for a pattern; you have given me one case. And you do not have
available before you the facts that prove whether or not that is a
case that proves your thesis.
You said you will find the information. Fine. I am just saying I do
not want the record to stand with these assertions that product ob-
solescence is a significant risk when you do not have the evidence to
prove a single case.
Dr. MARKHAM. No one has said, Senator-at least I have not. I think
I ha.ve made it quite clear from the beginning that we have not over
time traced these particular events through the revenue accounts of
the companies that a.re concerned. I thought some time. ago, we did
promise to give them to you, to the extent that we could get them to
put them together. The focus of our study, I just finished saying, was
on a different aspect of this. We have not tried to catalog the total
number of times this occurred or any of these risks occurred. We
have, however, assembled some da.ta on the extent to which the firm is
vulnerable by Virtue of the fact that its sales a.nd its estimated
profitability seemed to be fairly highly concentrated in a few, at this
point in time, successful products. We are making the very simple
thesis that when you have this kind of all your eggs in one basket, or
two baskets, so to speak, that firm is more vulnerthle to these risks,
however frequent or infrequently they may occur, than would be the
case if a firm went on all the time with its sales dispersed over eight
or so or 100 products, and fairly evenly.
1 See Appendix III beginning at p. 2129, infra.
PAGENO="0133"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1683
I have not come in here to attempt to prove to you that the polio~
vaccine problem of Cutter did indeed reduce its rate of return from
plus 16 to minus 16 percent in the course of 1 year. What I did do
was do what all economists standing on the outside have to do. You~
see the temporal coincidence of the two events and you at least advance
it as a logical and plausible connection~ between the two sets of facts.
And that is as far as I would care to take it. In fact, as far as I can
take it until I have access to all of the intimate financial records ot
Cutter.
But I was responding to the comment you made that I was at-~
tempting to prove the general theory with one case. I am not trying
to do that at all.
Mr. PLOTKIN. Sir, in this morning's testimony we purposely, did
not specify the causes of risks in individual industries. We specified
a measure which could go over many industries and evaluate the risk.
Those results are on the chart. As your example very well points up,
there are sources of risk in the automobile industry different but
whose causes may be the same in a sense as the causes of risk in the
drug industry. This morning an analysis, which was not particularized
to the automobile or to the drug industry, shows your intimate feel-
ing to be quite true, that the two industries are operating at about the
same level of risk and profitability.
This afternoon, we are saying, well, there is the drug industry,
fairly high in both risk and return. We now ask ourselves are there
possible reasons, without saying these are the only reasons, are there
possible explanations for why the drug industry has risks as high
as the automobile industry's? I think Professor Markham, taking the
reasonable man's point of view of what could cause risks, is now
enumerating these possibilities. So I submit these are two legs of the
same problem. Both of them are being~ addressed in our testimony.
Senator NELSON. Well, all I would say in response to that is, one,
I named the auto industry off the top of my head. This afternoon, I
mentioned the highway construction industry, although I have no
statistical basis at all, just the empirical evidence that I have looked
at for many, many years, in my State. I think you will find that there
are more people in and out of the highway construction business than
even are in this business. Nobody goes into the auto production busi-
ness. They are all going out.
But you also find that those who have the capital and know how
are very stable. They have a very long record. We have them in our
State for 50 years, very distinguished.
The reason I am raising these questions is that you tell me you do
not have available any specific statistics to prove the assertions made
here. The assertion on page 1 of Dr. Markham's statement is, "We have
with reasonable certainty identified the following as major risks, some-
what peculiar in this regard to the pharmaceutical industry."
Then when I ask for the specific examples to prove the assertions
which I question, I am told that you do not have them compiled yet.
As just an ordinary country lawyer, when somebody makes an as-
sertion, I say OK, you have drawn a conclusion. Then I ask, on
what basis did you reach that specific conclusion? That is all I have
been saying here.
PAGENO="0134"
1684 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. CONRAD. The assertions we make are that we have analyzed the
statements of the executives in the drug industry and we have analyzed
the drug industry ourselves in terms of the types of risks that are
inherent and operating in the industry. We are not claiming and we
did not come in to say that we are proving through examples of past
evidence that these risks have taken place. We are saying that the
combination of these risks in prospect and these risks that we have
enumerated, these five risks are present, I did not think anyone who
has studied the drug industry in depth at all would say that they
were not there in the drug industry. Those are reasonable examples
of risks that drug firms face.
The second point is that we are trying to link that group of risks
with a high concentration of profitability in the drug firms. If there
is a high concentration of profitability in drug firms among a few
product lines, then anyone of those risks could very significantly dis-
rupt that company's earnings for any length of time from 1 year up
to as long as it takes to introduce a new product into a line to replace
the revenues and profits represented by the product that was lost.
We in no way and nowhere in the statement, I think, can be inter-
preted as coming in here to prove these risks, on the basis of their
past existence. We have not attempted to do that, we have not done
it in our study to date. We further do not think it necessarily germane.
Senator NELSON. Well, OK. I guess we have covered that.
Dr. MARKHAM. Shall I proceed, sir ?
Senator NELSON. Yes.
Dr. MARKHAM. These risks, referring again to five or possibly
others, are not of the cyclical type where demand for a product dips
only to recover again when conditions affecting its position in the
economy improve. Most manufacturers whose profits are reduced
during a recession can in general look forward to a restoration of
profitability when the economy subsequently recovers. However, the
pharmaceutical firm generally must find a replacement for its lost
revenue and profits in the form of a successful new product. To the
extent that the research, development, approval, and introduction
of a new major product requires a prolonged period of years, the
firm may suffer long periods of significantly reduced profits.
In the study we are making at Arthur P. Little, Inc., we have sought
through requests to a representative sample of major pharmaceutical
firms to obtain the percent of total net profits and total gross profits
accounted for by each of their five most important product lines in
the 3 years 1961, 1964, and 1966. To facilitate this transfer of infor-
mation the returns are being coded in such a way that neither of us
can identify the responding companies.
We are prepared to discuss with the committee our interpretation of
what we believe the returns received to date show in terms of the mag-
nitude of the risks encountered by the firms who have submitted the
returns.
I would like to just make an addition to the written statement and
say, as you can pretty well appreciate, that firms who do not keep
their books, or at least do not keep profits, gross profits, or net profits,
in terms of specific products or product lines, are very reluctant to
step out to make meaningful statements of just the extent to which
PAGENO="0135"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1685
profits might be concentrated in a certain product line. However, on
the basis of the returns that we have, and bearing in mind that these
are estimates, pretty largely made by adjusting the impact of, say,
the withdrawal of a product on gross revenues or on gross margins, it
does turn out that in the limited number of returns we have already
received, it is not unusual to find that 50 percent of the estimated
profitability of a firm is accounted for by its leading product in any
one of these 3 years. And it is equally not unusual to find that the top
two leading products often account for as much as, say, 65 to 70 per-
cent of the estimated profit of the firm. This is only an arithmetic way
of saying that sales are not distributed Over a wide range of products
so that any one of them happening to encounter any of these risks,
the revenue impact would be slight. The fact is that they are concen-
trated and when a leading product, if and when it does encounter one
of these risks, the firm is at least vulnerable to a very significant shrink-
age in its profit or revenue, and this, therefore, seems to us to add up
to some reinforcement of the possible quantitative impact of these
potential risks.
Senator NELSON. Looking at your comments on page 2, we discussed
at some length the five items listed as peculiar, or somewhat peculiar,
as you put it, to the drug industry.
Then, in the middle of the page, you state:
The study just presented to you provides abundant evidence that American
manufacturing industries confront varying degrees of risks and that these risks
are strongly correlated with profitability.
Our study of individual pharmaceutical manufacturer's experience reveals that
these risks are not simply potential but that most have in fact materialized in
recent years. We can provide illustrations in sufficient numbers to support this
point should the Committee wish us to cite them.
That is what I was asking you a while back. Can you provide these
instances or are these the ones that have not materialized?
Dr. MARKHAM. Yes, sir; we are compiling lists of some of these.
I gave you some examples-or at least one example-of each of the five
as we went through. I am certain that we can provide you with, in
case you are interested in them, a much more complete list than this.
Senator NELSON. Well, that is entirely up to you. I would say quite
frankly, with what you have presented, the record is not convincing.
The record does not, in my judgment, as an ordinary lawyer, contain
specific items sufficient to support the five assertions you make. It just
does not.
Now, you do not even have to be a lawyer to conclude that. If you
want to prove your case, I think you have to come in with a pattern
such that you use to prove any situation. You would have to come in
with evidence that proves the case that these five characteristics are,
as you put it, somewhat peculiar to the industry and that they have had
adverse effects that makes this industry a risk industry in order to
develop a substantial case. Otherwise, quite frankly, I do not believe
it and I do not think anybody else reading the record will. You say
you can provide illustrations in a sufficient number to prove the point
should the committee wish you to cite them.
Then you tell us, if I understand you correctly, that you really do
not have them yet. The drug industry has argued those five points for
years, so anybody could have discovered them. But the drug industry
PAGENO="0136"
1686 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
has not yet supplied this committee with evidence and these witnesses
have not supplied the committee with such information either.
I am not trying to be querulous about it. I am just trying to say
honestly to you that the record you have made today does not prove
these assertions.
Dr. MARKHAM. I am not sure what you are asking me to prove,
Senator.
Senator NELsON. I am asking you to prove the five conclusions you
have drawn. Produce the evidence to support the conclusion. I can
give you examples in all kinds of industries, off the top of my head,
that would prove them to be high risk. For example, just look across
this country. Every single distributor of cigarette machines in America
is knocked galley west by the 101-millimeter cigarette, because none
of those packages fits the machines. I know a poor guy back in my
State who put all of his savings into a bunch of machines. They are
brand new, a.nd one change in the length of the cigarette knocks out
all the machines. There is no market for them, and he is bankrupt.
That is a very high risk business. The whims of the cigarette makers
to change the length of cigarettes by 1 millimeter or 2 erases all the
business.
All I am saying is that a single instance like this does not prove that
this is a risky business. But you draw five conclusions. I would like to
have a number of specific examples proving each of these conclusions.
Not one example to prove each conclusion. I can do that for any
industry you want to name.
Dr. MARKHAM. Perhaps you can, sir. I would not want to go into
that. The only thing I was trying to prove is their existence, not how
frequently they occur.
Perhaps I can use a specific example. If I were to take you to the
State of Michigan and show you an apple tree and it was indeed
bearing apples, I have proven to you that apples will grow in the
State of Michigan. However, I have said nothing about the importance
of apples in the Michigan economy. I think what you are asking me to
do is to try as best I can to tabulate and get together all of the instances
that we can get from the history of the drug industry, certainly the
recent history of the drug industry, to give you a better idea of the
frequency with which this kind of risk has occurred.
I wish to make it clear that my only point was, and I have proven
the case because I have shown you an example of the materialization
of everyone of the five, just as I would have proven to you that apples
can grow in Michigan if I could show you a live apple tree growing
apples.
What you are asking me for is a different level of proof and I am
saying I will try as best I can to amass that amount of information.
But I do not think it is quite fair to say that I have come in and tried
to prove a big case with one example. I only tried to prove one marrow
issue; that is that these risks not only are indicated in the history of
the industry, but that in fact, a rather striking case can be found to
illustrate all fi~re of them.
I have also attempted to indicate that I am not making any pretense
that this is a complete list of the risks that drug firms have. But I do
not think anyone can deny the fact that if Cutter's vaccine has to be
PAGENO="0137"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1687
withdrawn, this does document that it ôan happen. It does, because it
did happen.
What you are asking for, I think, is something else. You would like
to know, if I interpret you correctly, you would like to have me give
you now some evidence that this happens with some degree of fre-
quency and not just that I can prove it happens. But you want me
to prove the extent to which it happens. And that is a different issue.
That is the reason that we do not, or one of the reasons we do not have
the whole list compiled. If you would be interested in as complete a
list as we can get, and I think we were prepared to gather as complete
alistaswecan.
Senator NELSON. Well, I think the issue is that the industry asserts
that there is something special, very spdcial about the risk in the drug
industry, that does not exist for very many other industries. All I am
saying is that if you are claiming a special case, then we need I would
think, some substantial evidence to prove the special case. And we have
not received it.
Dr. MARKHAM. Let me just address myself to that too. We have not
said, Senator, that the drug industry is the only industry that encoun-
ters risks. What we have tried to do is to fit it in a pattern along with
the manufacturing industry generally. That was the presentation there.
Now, the purpose of this study is to try to specify the particular
kinds of risks that drugs encounter. We say that they are somewhat
unique to drugs. I believe we are right. But whether they are entirely
unique, each one of them, that you could not find some counterpart for
example if the rest of manufacturing enterprise, would not really do
very much damage to either the major risk study or to this one that is
attempting to show how risk comes to rest possibly on particular
product lines. There are many industries up there, as you can see, that
have varying degrees of risk.
I am sure if you were to ask some people knowledgeable in the auto-
mobile industry to catalog their particular risks, they would come up
with some risks and it may very well be that one or two of them would
slightly resemble or perhaps closely resemble one or two of the risks
I have enumerated here.
But I am arguing that this bundle of risks is fairly unique. I believe
you have made the statement that you could offer me any number of
industries that encounter the five. If yOu can, then my statement falls.
Senator NELSON. I am sorry, I did not intend to say the same kind
of risks, but industries that have risks that are just as significant to
the industry involved as these risks are to the drug industry. That is
what I am really getting at.
I do not doubt the scholarliness of the presentation by any means.
I know your credentials are distinguished. There is no doubt about
that. Perhaps everything you have said could stand unchallenged.
Bu.t if you can find a group of risks in other industries that are as
substantial, even if they are of a different kind, the claims of special
risk in the drug industry does not mean so much. This is the only
industry I know of that spends as much time, with as much money, as
it does propagandizing the Nation about how great a risk it faces.
I can take you to any medical society meeting and without planting
a word with them, just say, "What do you think about the profits of
PAGENO="0138"
1688 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the drug industry?" And they will say, "Well, it has to have high
profits because the drug companies do a lot of research and it is a
very risky business." They sound like automatons.
I had the very delightful experience, from my standpoint, of attend-
ing a reception at which there were about four doctors. They tackled
me on these hearings. I said "I will tell you what I will do. Just let one
of the doctors step aside and I will tell him what you are going to tell
me about the drug industry and we will come back together." And I
did anticipate his words in some detail; he was outraged. I said, "You
have `been receiving this education, if that is what it is, for some years,
you `believe it all."
This is one industry that insists upon this argument of risks. Now,
I know a couple of wealthy fruit growers, millionaires. Every once
in a while, they know a poor season is coming; they will have a frost
and be wiped out for the summer. But they are still millionaires and
they have been in the business for 50 years. Well, that is a high
risk industry and I can give you other examples.
Agriculture-I know of people in the beef-raising business. They
are millionaires. Once in a while a hoof and mouth disease or some-
thing else, although not recently in this country, will cause them a
substantial loss.
Pulpwood-I know of all kinds of pulp people in my State, probably
3,000 of them, who do very well for a while. All of a sudden, there
comes a surplus of pulp and they are in bad trouble.
All I am saying is that if you claim that a sharp distinction of great
significance to the drug industry is a special risk, that has an effect on
the industry different from other kinds of risks on other industries,
then fine. But to claim that you have some risk nobody else has because
nobody else manufactures drugs does not impress me.
Mr. CUTLER. No one has claimed that, Senator.
Dr. MARKHAM. I thought the pattern presented this morning sup-
ports, Senator, that we are taking. 59 industries, not just one. No one
has said the drug industry is the only risky industry. In fact, the chart
documents fairly carefully that at least three, possibly four-my vision
is not too good from here-industries ra.nk higher in risk on the chart
you have been given. And there are quite a number that just lean
over a little bit toward the left. That is all this is saying.
In fact, all it is saying is that the drug industry's rate of return
seems to fall in the ballpark it should fall within, given its risk,
in the way it falls with respect to the rest of the economy. And no one
is arguing that it is the only industry.
You see, I am not among those doctors. I have not been propagan-
dized.
I look at the figures and `they speak pretty eloquently to me that it
falls within the pattern of the rest of the industries of the American
economy.
Senator NELSON. Well, OK.
Is that the conclusion of these witnesses?
Mr. CUTLER. That is the conclusion of this group, Senator. We now
have Professor Firestone to talk to you about the downward trend
of the drug prices.
I would like to add one word, Senator, on this before we close the
subject.
PAGENO="0139"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1689
You have said that we have not come forward with evidence to sup-
port how products get displaced in the market by other products. We
have given you that evidence in the study that was handed to you
that is called, "Trends in Market Shares for Ethical Pharmaceutical
Products."
We have also assumed that this committee is fully aware of the
economic data that was presented in the Kefauver hearings by PMA,
by Professor Markham and by Arthur~ D. Little, in which we gave you
detailed product line by line breakdowns showing how each company
moved up, down, and sideways within each of the particular pharama-
ceutical classes. That appears starting at page 2526 of part 4 of the
hearing before the Kefauver committee back in 1961. We have not
tried to repeat any of that data. We didn't see any point in trying to
prove it again.
Mr. GORDON. In "Trends in Market Shares for Ethical Pharmaceu-
tical Products," where you have the table on various products, prod-
ucts 1, 2, 3, 4, could we have the names of those products? Could we
get the data, the basic raw material, upon which these are based?
Mr. CUTLER. Well, the data on which they are based comes from th~
proprietary studies made by the Gosselin line people of Pharmaceutical
Sales, by product. We are not at liberty to give you the names of the
products. Perhaps you can get them from Gosselin, which owns the
data.
Now, in the 1961 hearings, we did provide you, again in coded form,
with the name of the company turning out each of those products. If
that is any help to you, I suppose we could do that again.
Mr. GORDON. Since you presented this, we would like to get the raw
material on which it is based.
Mr. CUTLER. Perhaps you would ask Gosselin. We will join you in
that request to see if they can give you those names. We can give you
a set of coded names to go with the product codes, if that is any help
to you, as we did in 1961. That will show you which company was the
leader at any given time and how the company's fortune fared.
Mr. GORDON. We may want to duplicate this work to check it.
Mr. CUTLER. We have told you our source, Mr. Gordon. If you can
get them to give you their names, that is fine.
Senator NELSON. Let's take a 5-minute recess. I want to talk to my
office.
(Recess.)
Senator NELSON. Come to order, please.
Mr. OUTLER. Mr. Chairman, our next pair of witnesses is Dr. Joim
Firestone, professor of economics at the City College of New York
and Mr. Howard Binkley, vice president for administration and plan-
ning of the PMA.
Dr. Firestone has a statement describing the pharmaceutical prod-
uct price index, which he has compiled over the last many years. He
will just summarize the statement very briefly. We will then place it
in the record if we may, and Dr. Firestone and Mr. Binkley will an-
swer any questions you may have.1
1 The complete prepared statement of Dr. Firestone begins at p. 1702, infra.
PAGENO="0140"
1690 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. We are pleased to have you here, Dr. Firestone.
Your biographical sketch will be printed at this point. Please proceed.
(The biographical sketch follows:)
BIOGRAPHICAL SKETCH OF JOHN M. FIRESTONE, PH. D.
Education: B.S., College of The City of New York, 1930; M.B.A., College of
The City of New York, 1933; and Ph. D., Columbia University, 1955.
Present Position: Professor of Economics, College of The City of New York.
Teaching experience: Teaching elementary, intermediate and advanced courses
in statistics at: College of The City of New York, 1929 to date; Southern Meth-
odist University, summers of 1937 and 1938; Albany Graduate Program in Public
Administration (sponsored by New York University, Syracuse University and
the University of New York State), 1958; and Fashion Institute of Technology,
1954-56.
Government experience: U.S. Department of Labor, Bureau of Labor Statistics
Economist, New York Regional Office (part-time), 1955-64; Price Economist,
Chicago Regional Office, 1951-52; National Housing Agency, Chief, Cost Analysis
Section, 1946-47; Patent Royalties Commission, 1945-46; War Department, Chief,
Price Index Section, 1943-45; and Office of Price Administration, 1943.
Business research experience:
Consultant, Pharmaceutical Manufacturers Association, 1964 to date.
Consultant, various Litigating Groups of Public Utilities, 1963 to date.
Consultant, Anti-Trust Investigation Group of investor owned public
utilities, 1961-63.
Consultant, Merck, Sharp & Dohme, 1961 to date.
Consultant, Prentice-Hall, 1947-54.
Consultant, Controllers Institute of America, 1949-51.
Research Associate, National Bureau of Economic Research, 1950-51.
Consultant, McGraw-Hill Publishing Company, 1947-50.
Consultant, World Calendar Association, 1949.
Professional and other memberships: American Statistical Association; Ameri-
can Economic Association; National Association of Business Economists; Ameri-
can Association for the Advancement of Science; Social Science Research Coun-
cil; U.S. Chamber of Commerce; and United Cerebral Palsy Association of New
York, 1952-58.
STATEMENT OF JOHN M. FIRESTONE, PH. D., PROFESSOR OF ECO-
NOMICS, THE CITY COLLEGE OF NEW YORK, NEW YORK, N.Y.;
ACCOMPANIED BY HOWARD BINKLEY, VICE PRESIDENT FOR
ADMINISTRATION AND PLANNING, PHARMACEUTICAL MANU-
FACTURERS ASSOCIATION, WASHINGTON, D.C.
Dr. FIRESTONE. Mr. Chairman, gentlemen of the committee, my
name is John M. Firestone. I am a professor of economics at the Col-
lege of the City of New York where I have taught since 1929.
Much of my work as a consultant to industry and government in-
volves my principal field of interest-price statistics and indexes. In
this area I have served as a consultant to the U.S. Bureau of Labor
Statistics and have designed and developed price indexes for the War
Department, the National Housing Agency and for nongovernment
organizations.
I am a fellow of the American Statistical Association and of the
American Association for the Advancement of Science, and am a mem-
ber of the American Economic Association and the National Associ-
ation of Business Economists.
My testimony here is related to my activity ias a consultant for the
past 6 years to the Pharmaceutical Manufacturers Association. I have
PAGENO="0141"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1691
been requested to address my remarks to drug prices and particularly
to testimony on this subject which has been presented by earlier
witnesses.
First may I express my appreciation to the chairman and to mem-
bers of the committee for the opportunity to appear before you and
may I say that I sincerely hope that my testimony will contribute to
a better understanding of the factors entering into drug price trends.
For the measurement and meaning of changes in price levels is indeed
not a simple matter.
Price changes cannot be taken at face value, since very often the
prices compared are not strictly comparable though at first view they
may seem to be.
I should like to discuss briefly the ways in which wholesale and
retail price indexes are constructed, hOw they differ from each other,
what they mean, and as important, what they do not mean in respect
to drug prices. I should also like to examine the concept of an average
prescription price, and point out the pitfalls in using this as a measure
of change in the price of drugs, particularly at the wholesole level.
Price indexes are desigiied to do only one thing. They measure the
change in prices over time for the universe of commodities which the
index is meant to represent. The index~ number does not, for instance,
show the effect on the housewife's market basket when she substitutes
cake mixes for the crude ingredients or frozen packaged vegetables for
a sack of potatoes. It does not tell us whether we are using more ex-
pensive or less expensive commodities than we used in the past. It does
tell us whether, once introduced into the index, they and the other
components of the market basket, have risen or fallen in price from one
period to the next.
Once the market in which prices are to be measured has been defined,
a sample must be taken of the commodities and of the specifications
to be priced. For prescription pharmaceuticals for example, a selection
must be made of package size, dosage form and strength, of transac-
tion size, of terms of sale and of points in time and place at which these
prices are to be taken.
In constructing a price index, logical weights (measures of the im-
portance of the products selected for the index) must be employed.
These weights should, as truly as possible, reflect the relative im~
portance of the commodities in the index. Changes in the prices of the
most widely and frequently used commodities should have greater in-
fluence on the index than changes of price in less important or in-
frequently used products.
These studies of relative importance Of commodities cannot be "once
and for all" studies. Like most else in our society, they become obso-
lescent. For, in a dynamic economy such as ours, new products are con-
stantly appearing on the market, many older products diminish in
importance and often disappear altogether. Also the use of existing
products changes continuously as custom and needs decree.
Du~ to the high cost usually entailed in collecting data, revisions of
indexes are infrequent, and the accurac~ with which the index meas-
ures average price changes is impaired.
Until 1967, as long as ~ years after the fact, it was necessary to em~
ploy 1958 weights for the BLS wholesale price index. The 1963 weights
PAGENO="0142"
1692 CO~ETITIVE PROBLEMS nc THE DRUG ThThUSTRY
could not be introduced until this year. Accordingly, when new weights
are introduced into the index, they are already 3 to 4 years old, and
at the time they are changed, they are nearly 10 years old. The Con-
sumer Price Index has undergone even more infrequent weight
changes.
In general terms, for there are no objective criteria available,
changes should be made as frequently and as early as possible in order
to more truly reflect current conditions. As a case in point, the British
Government kept the commodity list in its price index unchanged from
1913 to 1947. At this latter date, with some difficulty, women's high
button shoes, whalebone-stay corsets, and kerosene lamps (for light-
ing) were still being priced.
The BLS Wholesale Price Index is designed to measure changes in
the prices of all conimodities sold in commercial transactions in pri-
mary markets of the United States including Alaska and Hawaii.
All these commodities, taking account of grade, size and other speci-
fications number in the hundreds of thousands and the total number
of price transactions for them in a month are in the untold millions.
To measure what happens to prices, the Bureau collects actual price
quotations on a sample of about 2,300 representative items from man-
ufacturers and other producers.
Ethical pharmaceutical preparations account for about 0.5 percent
tf the total importance of conimodities in the WPI. A relatively
larger number of commodities are used to construct these index corn-
`ponents of the WPI than for most other product classes. There are
~1 drug and pharmaceutical materials, 55 ethical pharmaceuticals
divided into 15 subproduct classes and 24 proprietary pharmaceuticals
divided into eight subproduct classes making a total of 110 products.
A proportionate share, that is 0.9 percent of the 2,300 products used
for the entire WPI would be 21 which probably results in the drug
category being more representative than most other categories of like
weight. Even before the revision and updating of the pharmaceutical
indexes in 1961, 37 products were included. Thus, pharmaceuticals are
not glossed over in the WPI. But even this list of products leaves
something to be desired in a definitive study of drug prices.
Before the BLS had indicated that it was revising its drug index, I
had been asked by the industry to construct the index now called the
PMA Wholesale Price Index of Ethical Pharmaceuticals. To do this,
I went back to 1949, and with a more comprehensive list of drugs con-
structed an index which showed price movements since that date. The
construction of this index is described in a succeeding portion of this
paper.
It is pertinent to compare the BLS Drug and Pharmaceutical Index
with the PMA index for the periods 1949-60, the period preceding the
BLS revision, and 1961-66, the period after the revision. From table
I-A and chart I-A, one can see that the BLS index in the earlier period
showed practically no change in drug prices while the PMA index
declined 7.6 percent. The comparison for the later period (with no
change in the technique employed in constructing the PMA index)
shows that the movements of the two indexes were very similar, with
the BLS index showing an even larger decline than shown by the PMA
index. (See table I-B and chart I-B.)
PAGENO="0143"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1693
If you look at the period from 1949 to 1960, the BLS index is
practically stationary, whereas the PMA index declined almost
continuously.
From 1961 on, the BLS index showed a decline and the PMA index
continued to move downward as it had to that date. There is a com-
parison made in the last two columns of the movement since 1961, you
can see from this that the BLS index declined to 94.9 whereas the
PMA went down only to 95.8.
Mr. GORDON. Looking at page 10-A, we see that all of the drug
prices fell over the years. There are large drops in the early 1940's and
1950's.
Dr. FIRESTONE. Yes.
Mr. GORr~oN. Do these drops account for the full decrease which
occurred between 1940 and 1949?
Dr. FIRESTONE. Not entirely. There are a few years in which the
prices did rise.
Mr. GORDON. I am talking about the decline.
Dr. FIRESTONE. Well, there have been declines in practically every
year but the last four.
Mr. GoiwoN. They account for most of the declines, do they not?
Dr. FIRESTONE. Yes.
Mr. GOIu0N. Now, it is my understanding that the late 1940 and
early 1950 drops are accounted for by the price drops in penicillin in
the early forms and the early forms of steroid hormones which are
unpatented or at least freely available and can be supplied and sold
by many firms.
Is that a fair statement?
Dr. FIRESTONE. No, sir, it is not, because a little further on I discuss
a patented and nonpatented index. You will find the index on the non-
patented products show practically no change, whereas the patented
index accounts for practically all of the change in prices.
Mr. GORDON. You say the drop in penicillin prices has nothing to do
with these drops?
Dr. FIRESTONE. I did not say that. I said the drop in penicillin
prices was at a time when penicillin w~s still under patent protection.
Mr. GORDON. Penicillin was never under patent protection. The
Government owned the patent on that, sir. Maybe that is why that
index of yours on patented products shows a decline.
Dr. FIRESTONE. Yes, but some forms of penicillin did carry some
form of patent protection based on developments of the companies
that produced these products.
Mr. GORDON. But when did they come out? They were marketed
fairly recently, were they not?
Dr. FIRESTONE. No, this goes back into the 1940's.
Mr. GORDON. Penicillin G was not patented. It was freely available.
The U.S. Government developed and patented the fermentation proc-
ess used in its manufacture. The patents were put in the public domain.
Dr. FIRESTONE. The information I have is that these are covered by
patents.
Mr. GoRDoN. We will come to that.
Dr. FIRESTONE. We come to page 11.
Before detailing the construction of the PMA Wholesale Price In-
dex of Ethical Pharmaceuticals, I should like to state that I have
PAGENO="0144"
1694 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
always had complete freedom in designing and constructing this index.
This is the only way I would do it; the Pharmaceutical Manufacturers
Association never asked me to do it otherwise. As a result, I alone am
responsible for this index in its content and its accuracy.
The index measures the changes in prices of ethical pharmaceuticals
as reflected in the price quotations for those brand-name products
which are prescribed frequently enough to be recorded in national
surveys. The sources of the price data used are relevant editions of a
trade journal annual, the Drug Topics Red Book.
I am going to skip over most of this material, and at the middle of
page 12, refer to the weights for this index were derived from confi-
dential proprietary data. compiled by Medical Data Services, a market
research firm of Darien, Conn., from its prescription panel service and
provided to me on a confidential basis.
The weight used in constructing the index is the average dollar vol-
ume for 2 consecutive years ending with the year of reference. Thus a
product is on the market at. least 2 years with substantial dollar vol-
nine in prescriptiolls before. it is then included in the index for the
second year.
The result is that only products which become established in the
market can be included in the index. Initial exploratory prices dur-
ing the introductory period for the product would not generally affect
the index. Indication of the. rapid changes that have taken place in
drug usage may be seen from the fact that only 27 of the original 99
products are still included in the index. It should be obvious that fail-
ure to recognize and take account of such changes on a current basis
when constructing the price index would result in rapid obsolescence
of the index. Accordingly, the list of products and their weights are
reviewed each yea.r to make sure that no important item is excluded
from the index and to provide for the retirement of drugs t.hat have
declined to insignificantly low levels. I note that the Commissioner
of t.he Bureau of Labor Statistics in recent testimony before this coin-
mit.tee stated that the BLS has recently adopted this policy for its drug
index.
In the 17 years covered by the PMA Index o*f Wholesale Prices of
Ethical Pharmaceuticals, there has been a total decline of 14.1 percent.
Declines have occurred in 11 of t.he 17 yea.rs, increases in 4 and there
have been 2 years with a. change of less than 0.05 percent. Of the 17
therapeutic groups, 10 have shown increases a.nd seven have had de-
creases since 1949. However, the declines of more than 50 percent that
have occurred in two of the most. important groups, antibiotics and
hormones, more than offset increases of much smaller magnitude in
less important groups.
\~Then the drug produc.ts had been classified by their patent status,
patented products showed a decline of 24.8 percent (through 1966)
while the nonpatented products had risen by 1.1 percent in this same
period.
Mr. GORDON. Would you please supply us with the raw data and
your calculations.
Dr. FIREsTONE. What do you mean by raw data ?
Mr. Go~ox. Well, whatever information you had on which you
ba.sed this particular statement.
PAGENO="0145"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1695
Dr. FIRESTONE. Well, part of it is confidential.
Mr. GORDON. In other words, we cannot check it?
Dr. FIRESTONE. I would certainly have to get clearance from Medi-
cal Data Services and Gosselin on the weighting material.
Mr. CUTLER. Do you want the list of the patented and unpatented
products?
Mr. GoRDoN. `We want all information, including the calculations
on which this is based. We wa.nt to be able to check it.
Mr. CUTLER. Well, we can provide you with all of the data we used
except the weights-well, we can provide you with the weights we
used, but the sources of the weights, I take it, are confidential data
furnished by Prescription Audit Service, which has restrictions on
its ability to disclose the data.
If the committee would ask that organization to furnish the data,
perhaps they would do so, but we are not at liberty to do so.
1-lowever, the method, and, I believe, the sources, are the same as
those used by the BLS.
Mr. Gordon, could we satisfy you by supplying all of our raw data
in confidence to the BLS and have them satisfy you as to whether
they consider it was accurately constructed?
Mr. GoRDON. I would like to have it myself. I would like to do it
myself so I could know if this was the correct data.
Mr. CUTLER. Dr. Firestone will sit down with you and show you the
index. If he used data which we have received in confidence from Pre-
scription Audit Service, we would suggest that you ask the Prescrip-
tion Audit Service if they can give you the raw data. `We have no
objection to their giving it to you if they are free themselves to do so.
`We are not the ones placing the confidentiality on the data.
Senator NELSON. Let me ask you a question. On the last paragraph
on page 16, your sentence is that "When drug products were classified
by their patent status, patented products showed a decline of 24.8 per-
cent-through 1966-while the nonpaténted products had risen by 1.1
percent in this same period."
What is the period?
Dr. FIRESTONE. 1949 through 1966.
Senator NELSON. At what stage did the dramatic declines occur in
any particular product? Whe~n the patent expired?
Dr. FIRESTONE. No, sir; at that time they were placed in the non-
patent status. Once the patent expires.
Senator NELSON. So imniediately when a patent expires it is placed
for purposes of your study in a nonpatent status?
Dr. FIRESTONE. That is right.
Senator NELSON. So you are saying that during the course of this
period from 1949 to 1966, prices for drugs that were still under patent
declined 24.8 percent?
Dr. FIRESTONE. Yes, sir.
Senator NELSON. `Were any of these drugs ones that were patented
but were licensed to other companies to produce?
Dr. FIRESTONE. Probably.
Senator NELSON. So that might be one fa.ctor?
Dr. FIRESTONE. Yes, sir.
Senator NELSON. In other words, competition in the marketplace?
81-280-08-pt. 5-10
PAGENO="0146"
1696 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. FIRESTONE. That might be the factor; yes, sir.
Senator NELSON. Did your statistical analysis demonstrate that there
was a pattern of decline in patented drug prices as the expiration date
of the patent approached?
Dr. FIRESTONE. I have not examined that closely, Senator Nelson. I
would be happy to go back and take another look from this aspect at
tile index and see whether this is tile case or not. I do not think it is,
frankly.
Senator NELSON. I know of one case where that happened fairly
dramatically, so I wondered if that could be the pattern.
Dr. FIRESTONE. The interesting thing is when you have the most com-
petition, when the products are not covered by patents, you have rising
prices instead of falling prices.
Senator NELSON. Where what?
Dr. FIRESTONE. Where you have no patent protection and you have
the greatest degree of competition, prices do not fall.
Senator NELSON. That has been true recently of almost every prod-
uct in the marketplace, has it not? Food has no patent and it is going
up in price. Automobiles have no patent and they are going up. Clothes
have no patent and they are going up. The cost of production is going
up.
Dr. FIRESTONE. There is some question about the rise in price of
automobiles.
Senator NELSON. Not in the dollar cost.
Dr. FIRESTONE. Senator, one of the factors -that is often overlooked
is tile change in the quality of the product.
For example, you could hardly compare a 1950 automobile with a
1966 automobile. Let's talk about a Chevrolet.
The price of a Chevrolet has gone up. But a car comparable to the
1950 Chevrolet, perhaps Corvair, would be at a comparable price. And
it would be a better car.
Senator NELSON. I think you used the wrong example.
Dr. FIRESTONE. Well, perhaps. Or you might take the case of auto-
mobile tires. I remember when I paid $10 for a tire back in the 1920's
and I got 10,000 miles out of that tire. Now I buy a tire for $20 and
get 40,000 miles. Would you say the price had gone up 100 percent
or declined 50 percent?
Senator NELSON. You chose another unfortunate case. I happen
to have introduced the tire safety bill and we have statistics galore
about the quality of low-grade tires onthe.market. -
But do your statistics show that when the patent goes off and a
number of other firms go into the marketplace, the price of the
product drops?
Dr. FIRESTONE. It does not.
Senator NELSON. It does not drop?
Dr. FIRESTONE. Yes, sir.
Senator NELSON. Well, I would like specifically to have the list of tile
patented drugs that you used, the prices for each year, and then the
da.t-e the patent went off and what happened to the price then.
Dr. FIRESTONE. I should be happy to give you that.
Senator NELSON. If that is the case I wonder if it does not prove
something that I have suspected all along and that many people have
asserted.
PAGENO="0147"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1697
There is no competition in the marketiilace; that is why the price did
~not drop. In other words, once the drug has 17 years of patent protec-
tion behind it and the doctors in the country know only that particular
drug, they continue to prescribe it despite the fact that in the market-
place there may be a comparable drug for a half, a fifth, a 10th, or
even a 20th of the price, which indicates something peculiar about the
pricing structure.
I will give you an example and ask you to comment on it as an
economist. I have used this one before. I have here an issue of the
Medical Letter, which all the industry and all the distinguished physi-
cians in this country accept as a very fine publication. The Medical
Letter studied the case of prednisone. The original patent on predni-
sone was held by Schering, which called~ its product Meticorten, and as
of the time of this publication, June 2, 1967, Meticorten was selling for
$17.90 for 100 tablets. Merck was in the marketplace at $2.20 a hundred,
Upjohn at $2.25, Wolins Pharmacal at 59 cents, versus $17.90, and
several others were selling the drug-some at $1.84, $1.75. The Medical
Letter concluded that the great price spread among the tablets pur-
chased from different pharmaceutical companies suggests the desira-
Ibility of prescribing by generic name, and the Medical Letter urges, at
least for persons of limited means, that~ the prescription be filled with
low-priced generic tablets.
So, do you not have the case here where once the name gets estab-
lished, the company can continue to hold its share of the marketplace
despite the fact that there are competing drugs available at a fraction
of the cost and the same quality?
Dr. FIRESTONE. I am not technically qualified to comment on the
relative merits of Meticorten versus other forms of prednisone.
Senator NELSON. I just gave you that as an example.
Dr. FIRESTONE. Yes; but as an economist I cannot answer that for
you.
Senator NELSON. As an economist, hOw would you answer the fact
that here is a medical publication listing 22 drugs they say are of
equivalent value. They range in price from 59 cents a hundred to
$17.90 a hundred and still a substantial proportion of the market is
held by the drug selling for $17.90.
Dr. FIRESTONE. Apparently, for some reason, physicians did not
:agree with that Medical Letter.
Mr. CUTLER. Senator, I cannot help~ but suggest that as the testi-
mony before your committee shows, Schering's share in this corti-
costeroid market is down considerably and secondly, prednisone hap-
pens to be one of the six or seven drugs that Dr. Lloyd Miller of the
TJSP has informed you in writing on November 29, is a drug which
has been found to have differing therapeutic effect in different product
forms, even though all of them meet USP standards.
Senator NELSON. Not in the case of 22 companies that are manu-
facturing pTednisone according to the Medical Letter of June 2.
Mr. CUTLER. That I do not know and I do not believe Dr. Miller's
letter gives the source of his figures on prednisone.
Senator NELSON. Well, all I am saying is that here are 22 of them
and they vary in price from 59 cents to $17.90.
PAGENO="0148"
1698 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
I am just asking your comment as an economist. Mr. Conzen's testi-
mony, if I recall correctly, is that Schering had about 88 percent of
the retail market for prednisone.
Mr. CUTLER. I do not believe that is true, sir.
Senator NELSON. That was his testimony. I do not. know whether
he is correct or not. He may have been mistaken.
Well, go ahead.
Dr. FIRESTONE. We were discussing the declines in the index.
On page 17, skipping down to the first paragraph, one market
analysis available, a market analysis compiled by Lea Associates of
Ambler, Pa., shows among other data the distribution of prescriptions
for products by age group. One of these age groups is the "60 and
over" category, and these data are available back to 1959. An index
of wholesale drug prices for this age group was constructed, with 1959
as a base-equal to 100-and the total drug index was compared with
this. The results are shown in table IV.
An important reason for the difference in the decline of the two
indexes is that oral contraceptives, which have shown a marked de-
cline, are not included in the "over 60" index.
Medical care represents nearly 6 percent of the expenditures of wage
earners and clerical workers. Only 0.6 percent of the expenditures are
on prescriptions. Professional services of doctors, dentists, optome-
trists and nurses,hospital supplies and nursing home services, health
insurance and over-the-counter drugs and sundries account for the
remainder, and in all, are nearly 10 times as great as expenditures for
prescriptions.
It is only since 1960 that the Consumer Price Inde.x has included
prescription items that by current standards can be considered truly
representative. Prior to that yea.r the prescription index was based
on three iteins-APC tablets, elixir terpin hydrate and codeine, and
buffered penicillin-hardly a representative list of ethical pharma-
ceuticals. This index rose 27.8 percent between 1949 and 1960, but I
would be reluctant to accept this as an indicator of retail prescription
prices. The item list was revised and expanded in 1960 to include
seven thera.peutic groups, and just this year three additional groups
were added. Each group is represented by only one or two products,
with a total of 17 items to represent the 10 groups. Since the revision,
the index has declined 12.7 percent.
An index cannot tell us whether prices are too high or too low, or
whether more or less expensive products are in use today than yes-
terday.
Senator NELSON. It does not really tell us whether the consumer is
really paying more dollars or a larger percentage of his income~ for
the product, either, does it?
Dr. FIRESTONE. No; as a matterof fact, he is spending less.
Senator NELSON. The index does not tell whether he is spending
more of his dollar on drugs or on any other product, does it?
Dr. FIRESTONE. No; and it does not tell us what his income is, either.
Senator NELSON. So it has a very limited significance in that respect?
Dr. FIRESTONE. Senator, I would not accept that as a correct state-
ment, because whatever other measures we have of prices are not as
good as this.
PAGENO="0149"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1699
Senator NELSON. Well, that may be, but the fact is that you might
be paying twice as much for your prescriptions and perhaps twice as
much is being spent by the consumer in this country for drugs as was
spent, say years ago.
Dr. FIRESTONE. Senator, I will come to that very point very shortly.
Senator NELSON. And still the index does not reflect an increase.
Dr. FIRESTONE. I will come to that very shortly, Senator.
Mr. GORDON. Dr. Firestone, just one second. You say you cannot
tell whether prices are too high or too low. Prices which are lower
today than they were, say, 10 years ago, may simply be only, say, 85
percent as exorbitant as they were before. If they were exorbitant in
the past, they could be slightly less exorbitant as before.
Dr. FIRESTONE. Conceivably.
Mr. GORDON. So what the prices are, are just as important, too.
Dr. FIRESTONE. Comparing prescription drug prices with all others,
this is one group that has uniquely declined while practically all other
groups in the consumer price and wholesale price indexes have risen.
Mr. GORDON. Could that be a indication, by the way, that it is not
u competitive industry?
Dr. FIRESTONE. I should think not, because there is a wide range of
prices for these products, as you know.
Mr. CUTLER. You do not mean, Mr. Gordon, that it would please
you more if prices had gone up, do you ~
Mr. GORDON. It is peculiar that if everything else goes up, prices
and costs for all other commodities, and this is a thriving industry,
when prices are going down, there must be quite a~
Mr. CUTLER. You have convinced me, if I needed any convincing,
that no fact presented can make an impression on this committee.
Senator NELSON. Just a minute.
Mr. CUTLER. If the fact that consumer prices for drugs and whole-
sale prices for drugs have gone down proves a lack of competition in
the industry to Mr. Gordon, I wonder what the fact that they have
gone up means.
Senator NELSON. Well, Mr. Gordon is not the committee and not a
member of the committee. Some fact may not impress him, but do not
say that no fact impresses this committee.
Mr. CUTLER. I am sorry, Mr. Chairman; I withdraw that remark.
Mr. GORDON. May I comment on that?
Senator NELSON. I do not see anything fruitful in it. We have gone
through this with Mr. Ross and I will put in the record the colloquy
between me and Mr. Ross.
(The material referred to follows:)
IExcerpts from hearing May 16, 1967, Competitive Problems In the Drug Industry, Part 1,
pp. 201-2021
Senator NELSON. In your index, then, you don't include any statistics to show
what consumers are paying for drugs. You don't include any statistics showing
the differential in prices from year to year in prescription cost?
Mr. Ross. No, we don't.
Senator NELSON. Don't you think that would be of some significance? What
puzzles me a little is if people are paying 50 percent more per prescription now,
or 15 percent more per prescription now, than 10 years ago, that it has some
significance.
Mr. Ross. I don't challenge the significance of that for a minute. I merely say
that is not the nature of a price index of our type.
PAGENO="0150"
1700 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. But it is a price index of a type.
Mr. Ross. It is a price index of a type, the problem being that what it is.
measuring is open-ended.
* * * * * * *
Senator NELSON. I think that what is misleading is that your statistics are
used to indicate there is only-I think the interpretation put on your index
obviously is incorrect, that the price of drugs that the consumer is buying has.
risen 2.9 percent.
Mr. Ross. We have never interpreted it that way.
Senator NELSON. No, no; I say I think that the interpretation put on your
index, because it is not understood, whereas the fact is that the price per
prescription from the Lilly statistics has gone up 50 percent in 10 years and
not 2.9 percent, and I think it is very important to make a distinction between
the nature of your index-
Mr. Ross. Yes.
Senator Nui~sox. And this other index.
Mr. Ross. I fully agree that that is a very important distinction. I would
argue that our type of index, which measures change in prices for specified items,
is very important for many purposes, but as you say, it does not give you an
average price of the other type of index.
Senator NELSON. I would agree, and I am certain that from a statistical stand-
point, for the purpose that you have in mind, I have great confidence in the
validity of your index, but actually it was misleading to me, and I assume it is
misleading to a few other people anyway who read it.
Mr. Ross. Yes, sir.
Senator NELSON. The fact is that if the price of my prescriptions have gone
up 50 percent, I am paying out the money, my cost of living with respect to that
prescription has gone up 50 and not 2.9 percent.
Mr. Ross. Yes.
Senator NELSON. The index does not measure some things.
Dr. FmESTONE. May I comment on that?
Senator NELSON. Yes.
Dr. FmESTONE. At the bottom of page 19, average prescription
prices, a very simple statistic that has bean used at times to measure
what has happened to drug prices, is the average prescription price.
This, if not properly interpreted, can be a very misleading yardstick.
In an independent analysis of average prescription prices, Prof.
Robert V. Eva.nson. head of the Department of Pharmacy Adminis-
tration. School of Pharmacal Sciences, Purdue University, has arrived
at conclusions essentially similar to those presented in my discussion.
Mr. Chairman, I would like to call the committee's attention to Dr.
Evanson's pa.per as published in the Septe~nber 1967 issue of Tile and
Till, pp. 58-63, a.nd I respectfully request that this paper be made a~
part of the record of these hearings.
Senator NELSON. Do you have it with you? You did submit that
paper with your statement?
Dr. FIRESTONE. Yes.
Senator NELSON. That will he made a part of the record with your
statement..'
Dr. FmESTONE. Thank you.
In contrast with a price index, an average prescription price has a
number of variables other than the price of t.he drug. No a.ccount is
taken of the fact that the mix of prescriptions written and filled in a
given period of time may vary from the mix found at some other time.
Thus, the average prescription price can rise or fall even if the price
of each of the drugs included in the average has not changed by 1 cent..
2 See p. 1714, infra.
PAGENO="0151"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1701
Over the past few decades new drugs and medicines have made im-
portant contributions to the health of this country and are in part re-
sponsible for a declining death rate and a rising life expectancy. New
products have been developed to provide drug therapy where none had
been available previously, e.g., tranquilizers and drugs that are more
certain and speedy in effecting cures, the so-called wonder drugs.
Many of these drugs are higher priced than the products they re-
placed on a per dose basis. But this kind of comparison is no more
valid than a comparison of jet fares with railroad fares simply on the
basis of miles traveled. Just as it makes more sense to measure the
comparative cost of a trip from Washington to Los Angeles by plane
and train taking into account not only the fare, but all other incidental
costs such as meals, and time lost in travel, so is it more reasonable to
measure the total drug cost over the period of an illness rather than
the price of a prescription. But more than that, except for the geriatric
illnesses which come with prolonged lifespans, new more sophisticated
drugs shorten the duration of an illness; shorten the required period
of medical, nursing, and hospital care; reduce the pain and discomfort
suffered by the patient; put him back on his feet, and back to his job
more quickly, reducing his earnings loss due to illness.
I cite this, Mr. Chairman, not as argument, but as sound economics
which goes to the question of real cost. For some specific illnesses these
savings can be measures in dollars and cents. Before the development
of .antibiotics, for example, lobar pneumonia meant 5 weeks of hos-
pitalization, long convalescence, and several hundred dollars for doc-
tors, nurses, medicines, oxygen, and hospital care. Today pneumonia
means a week to 10 days of illness, usually at home, $15 to $30 for
drugs, and just two or three visits from the doctor; $15 to $30 for drugs
sounds expensive. What is generally overlooked is the astonishing re-
duction of all other medical costs, and the reduction in loss of time from
work.
There is a real need therefore, I submit, for some measure which will
show the true change in the total cost of treating an ailment. A some-
what higher drug cost may result in a considerably lower total cost in
effecting a cure. Even such an index could not evaluate the lowered
mortality rate, the shorter duration of an illness with the accompany-
ing relief from disability, pain and absence from one's occupation.
There have been occasional attempts to develop such a measure, but
to date these have not provided adequate data. A principal source of
difficulty is the unavailability of suitable data concerning medical,.
hospital, nursing, and medication costs in earlier years. The Depart-
ment of Health, Education, and Welfare might be the appropriate
agency to develop such measures.
In summary, I would say that average prescription prices cannot tell
us what has been happening to drug prices. There are too many factors
other than changes in drug prices that tend to distort this measure. In
fact, for nearly. two decades prices for ethical pharmaceuticals, as
measured by the most reliable price indexes, have declined, while the
price of an average prescription has been rising.
Neither average prescription prices nor an index of pharmaceutical
prices can tell us what is happening to our total expenditures for
drugs. The increase in our lifespans and the advances of medical
science inevitably must add to our drug bill. The continuing research
PAGENO="0152"
1702 COMPETITIVE PROBLEMS IN THE DRUG II~DUSTR1
discoveries in the chemotherapeutic management of ailments which
previously baffled the medical profession, have resulted in more ex-
tensive drug usage and an increase in our total drug bill. It is impor-
tant, therefore, that. while we note an increase in a patient's drug bill,
we do not overlook the tremendously increased benefit to his well-being
and the resulting positive contribution to our economy.
I hope, Mr. Chairman, and gentlemen, that these remarks have shed
additional hght on drug prices and will contribute to a better under-
standing of the elements properly entering into the measuring of price
changes.
Senator NELSON. Thank you very much, Dr. Firestone, for your
very thoughtful presentation. We appreciate your taking the time to
come here.
Mr. FIRESTONE. Thank you, Mr. Chairman.
(The complete prepared statement and supplemental information
submitted by Dr. Firestone follows:)
STATEMENT OF JOHN M. FIRESTONE, Pu. D., PROFESSOR OF EcoNoMIcs,
COLLEGE OF THE CITY OF NEW YORK
Mr. Chairman and Members of the Subcommittee, my name is John M. Fire-
stone. I am a Professor of Economics at the College of the City of New York
where I have taught since 1929.
Much of my work as a consultant to industry and government involves my
principal field of interest-price statistics and indexes. In this area I have served
as a consultant to the U.S. Bureau of Labor Statistics and have designed and
developed price indexes for the War Department, the National Housing Agency
and for non-governmental organizations.
I am a Fellow of the American Statistical Association and of the American
Association for the Advancement of Science, and am a member of the American
Economic Association and the National Association of Business Economists.
My testimony here is related to my activity as a consultant for the past six
years to the pharmaceutical Manufacturers Association. I have been requested
to address my remarks to drug prices and particularly to testimony on this
subject which has been presented by earlier witnesses.
First may I express my appreciation to the Chairman and to members of the
committee for the opportunity to appear before you and may I say that I
sincerely hope that my testimony will contribute to a better understanding of
the factors entering into drug price trends. For the measurement and meaning
of changes in price levels is indeed not a simple matter.
Price changes cannot be taken at face value, since very often the prices com-
pared are not strictly comparable though at first view they may seem to be.
For example, in my opinion there has been a tendency to compare prices which
are not comparable and there has been a failure in some quarters to properly
differentiate between wholesale and retail prices, and wholesale and retail price
indexes. All of these are appropriate measures that have their place in economic
analysis. but they can be used properly only if the measures are correctly under-
stood and interpreted.
I should like to discuss briefly the ways in which wholesale and retail price
indexes are constructed, how they differ from each other, what they mean, and
as important, what they do not mean in respect to drug prices. I should also
like to examine the concept of an average prescription price, and point out
the pitfalls in using this as a measure of change in the price of drugs, particu-
larly at the wholesale level. Finally, I should like to point out that just measur-
ing the dollars and cents changes in drug prices, without taking account of
the changing benefits derived from their use, may result in wholly erroneous
conclusions.
BL~ price indeves
The United States Department of Labor's Bureau of Labor Statistics (BLS)
has earned the well-deserved reputation of being the world's outstanding author-
ity on the subject of price indexes. The BLS indexes of wholesale and retail
PAGENO="0153"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1703
(consumers') prices are not equalled in quality, integrity or accuracy by any
other indexes known to me. Statisticians come from almost every country to
study the technique of index construction developed and executed by the
Bureau's Price Division, so that they may go back to their own countries and
try to develop price indexes modeled after ours.
Purpose of mdccv numbers 1
Price indexes are designed to do only one thing. They measure the change in
prices over time for the universe of commodities which the index is meant to
represent. The index number does not, for instance, show the effect on the
housewife's market basket when she substitutes cake mixes for the crude in-
gredients or frozen packaged vegetables for a sack of potatoes. It does not tell
us whether we are using more expensive or less expensive commodities than
we used in the past. It does tell us whether, once introduced into the index, they
and the other components of the market basket have risen or fallen in price
from one period to the next.
HOW INDEXES ARE CONSTRUCTED
While technical descriptions are sometimes boring it is necessary to under-
stand how indexes are put together and how they may be properly used if one
is to appreciate the true significance of the changes reflected by them.
Price data
Index numbers are designed to show us the amount and direction of price
movement in the aggregate for it is impossible to discuss the net effect of the
many and diffuse price movements of all items. The total number of price quota-
tions is almost infinite. Furthermore, there is a wide range of commodities, styles
and sizes. Therefore, it becomes a practical necessity to select a representative
sample of the many variables involved. These variables include, among others:
(a) the commodity composition of the category being studied (for instance, if we
are considering groceries, what items should go into the market basket) ; (b) the
grades, sizes, and specifications of the commodities chosen; (c) the classes of
buyers and sellers to be selected and their terms of trade; and (d) the geographic
areas to be covered.
Once the market in which prices are to be measured has been defined, a sample
must be taken of the commodities and of the specifications to be priced. For
prescription pharmaceuticals for example, a selection must be made of package
size, dosage form and strength, of transaction~ size, of terms of sale and of points
in time and place at which these prices are to be taken.
Often these decisions are limited by the availability of data, such as when~
how often, where and for which of the items prices are published. Obviously,
prices for the most important products, at the most usual terms of sale are
likely to be readily available on a continuing basis. But even for these commodi-
ties it is important to set rigid specifications so that variations due to changing
transaction terms or grades of commodities do not affect the price comparisons.
Weighing
In constructing a price index, logical weights (measures of the importance of
the products selected for the index) must be employed. These weights should,
as truly as possible, reflect the relative importance of the commodities in the
index. Changes in the prices of the most widely and frequently used commodi-
ties should have greater influence on the index than changes of price in less
important or infrequently used products.
This information for weighting is usually harder to come by than price data
and is frequently unavailable except through specially designed surveys. It is
relatively easy to ascertain the going price of an item at any time. It is much
harder to find out for any time period bow much of the product was produced,
sold, or consumed, particularly when there are an unknown number of producers
or distributers of the item. Too often this information is considered confidential
by the seller or not tabulated in a way that is useful to a price statistician.
1 For a fuller discussion of Index numbers, their purposes and uses as well as methods of
construction see, Mitchell, W. C., "The Making and Using of Index Numbers", Bull. #284,
U.S. Bureau of Labor Statistics; Fisher, Irving, "The Making of Index Numbers"; King,
W. I., "Index Numbers Elucidated"; Mills, F. C., "The Behavior of Prices"; Mudgett, R. D.,
"Index Numbers"; Persons, W. M., "The Construction of Index Numbers"; The National
Bureau of Economic Research, "The Price Statistics of the Federal Government".
PAGENO="0154"
1704 COMPETITIVE PROBLEMS fl~ THE DRUG fl~DUSTRY
The BLS, in constructing its Wholesale Price Index (WPI), relies on various
government and other advisory agencies to supply suitable data for weights.
BLS generally has little choice as to when or how often these data are collected
or as to how quickly the data are processed and released.
For its Consumers' Price Index (CPI), BLS cannot fall back on any agency
for data suitable for weights. It must therefore conduct its approximately
decennial Survey of Consumer Expenditures. Without these special studies of
consumer spending patterns, the consumer price index could not be constructed,
or revised to reflect changes in these patterns.
Changes in weights and commodities
These studies of relative importance of commodities cannot be "once and for
all" studies. Like most else in our society, they become obsolescent. For, in a
dynamic economy such as ours, new products are constantly appearing on the
market, many older products diminish in importance and often disappear al-
together. Also the use of existing products changes continuously as custom and
needs decree. Since such information can be obtained only after the fact, it is
somewhat obsolete even when first used (applying information for a past period
of time to a current situation) and it steadily becomes more obsolete with each
passing period.
The rate of change in the relative importance of commodities, and in the birth
rates and death rates of commodities varies in time and with the class of products.
In any event, such changes are universal. No price index in the present day can
long endure without an updating of the list of items for which prices are
measured and without a re-evaluation of the importance of the individual prod-
ucts and the classes to which they belong. Because changes in drug usage are so
rapid and steady, BLS has found it necessary to contact the American Pharma-
ceutical Association annually for information on weight changes.2
Due to the high cost usually entailed in collecting data, revisions of indexes
nre infrequent, and the accuracy with which the index measures average price
changes in impaired. In the 1920s and 1930s, a Census of Manufactures was
taken every other year. As soon as each of these censuses became available (once
~very two years), the Bureau of Labor Statistics revised its wholesale product
weights and commodity lists. Since World War II, censuses were taken only in
1947, 1954, 1958, and 1963. The data from these surveys were incorporated into
the Wholesale Price Index as soon as they were available. This, however, has
~ntailed a considerable time lag. Thus until 1967, as long as nine years after the
fact, it was necessary to employ 1958 weights. The 1963 weights could not be
introduced until this year. Accordingly, when new weights are introduced into
the index, they are already three to four years old, and at the time they are
changed they are nearly ten years old.
The Consumer Price Index has undergone even more infrequent weight
changes. The 1960-1962 weights still in use in this index were introduced in
1964. Prior to that, 1952 was the weighting base; and until 1954, the 1934-1936
Consumer Expenditure Survey was used (with some necessary interim adjust-
ments). One can readily see, therefore, how unrealistic consumption patterns
in the depressed mid 1930s must have seemed in the first post World War II
~1ecade.
In general terms, for there are no objective criteria available, changes should
be made as frequently and as early as possible in order to more truly reflect
current conditions. As a case in point, the British Government kept the com-
modity list in its price index unchanged from 1913 to 1947. At this latter date,
with some difficulty, women's high button shoes, whalebone-stay corsets, and
kerosene lamps (for lighting) were still being priced. Contrast that with the
U.S. War Department's Index of Contract Price Changes, maintained during
World War II, when commodity weights and lists were revised every six months
to keep pace with the changing character of the war and its requirements.
2 See statement of Commissioner Ross before this Committee May 16, 1967 (transcript
pp. 230-259).
PAGENO="0155"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1705
If data are readily available, and the cost of revision is not too great, annual
changes in weights and commodity lists should be made, particularly in a product
-area where changes are rapid and dramatic. Such frequent changes permit in-
dexes to reflect a more gradual transition to current conditions than when
changes are made at infrequent intervals. In any event, changes are inevitable.
They are unavoidable if the index is to have any meaning.
`Quality changes
The Price Statistics Review Committee, in its 1961 report to the Office of
Statistical Standards of the Bureau of the Budget, stated that "if a poli were
taken of professional economists and statisticians, in all probability they would
designate (and by a wide majority) the failure of the price indexes to take full
account of quality changes as the most important defect in these indexes. And
by almost as large a majority, they would believe that this failure introduces a
systematic upward bias in the price indexes-that quality changes have on
average been quality improvements." This is no reflection on the ability or in-
tegrity of index statisticians. It merely points up the difficulty of measuring in
dollars and cents the value to the purchaser of style changes, improved efficiency
of a new model of refrigerator, the more silent operation or greater durability
of a room air conditioner, or the speed of action and the sustained duration of
pain relief from a new drug as compared to its predecessors.
WHOLESALE PRICE INDEXES
The BLS Wholesale Price Indea'
The BLS Wholesale Price Index is designed to measure changes in the prices
of all commodities sold in commercial transactions in primary markets of the
United States including Alaska and Hawaii. A primary market is defined by
the BLS as the first commercial transaction for a product, so as to avoid con-
iusion with subsequent wholesale transactions. Commodities produced in the
United States are included, as well as those imported for sale. They include
manufactured and processed goods and the output of manufacturing, agriculture,
forestry, fishing, mining, gas and electricity. All these commodities, taking ac-
count of grade, size and other specifications number in the hundreds of thousands
-and the total number of price transactions for them in a month are in the untold
millions.
To measure what happens to prices, the Bureau collects actual price quotations
-on a sample of about 2300 representative items from manufacturers and other
producers. The actual number of price quotations collected per month numbers
nearly 7000, in order to represent various markets and multiple producers of
a given product. These 2300 items are classified into 15 major groups, 90 sub-
groups, 314 product classes, and 37 subproduct classes. Thus, "antibiotics" is a
subproduct class of the Ethical Pharmaceutical Preparations product class of
the Drug and Pharmaceutical -subgroup of the Chemicals and Allied Products
major group. -
Ethical Pharmaceutical Preparations account for about 0.5 percent of the total
importance of commodities in the WPI. If we add to this Proprietary Phar-
maceutical Preparations and Drug and Pharmaceutical Materials (the other
product classes of the Drugs and Pharmaceutical subgroup), the combined
importance to the economy is less than 0.9 percent. I point this out to demonstrate
that a relatively larger number of commodities are used to construct these index
-components of the WPI than for most other product classes. There are 31 drug
-and pharmaceutical materials, 55 ethical pharmaceuticals divided into 15 sub-
product classes and 24 proprietary pharmaceuticals divided into 8 subproduct
classes making a total of 110 products. A proportionate share, that is 0.9 percent
of the 2300 products used for the entire WPI would be 21 which probably results
in the drug category being more representative -than most other categories of like
weight. Even before -the revision and upclating~ of the pharmaceutical indexes in
1961, 37 products were included. Thus, pharmaceuticals are not glossed over in
the WPI. But even this list of products leaves something to be desired in a
definitive study of drug prices,
PAGENO="0156"
1706 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
At the time of the hearings on drugs conducted by the late Senator Estes
Kefauver eight years ago, the PMA and the pharmaceutical industry believed,
and quite correctly, that the drug and pharmaceutical component of the BLS
WPI did not measure fully, nor correctly, what was happening to drug prices.
This is not to say that the attention given to drug products did not represent a
commensurate share of the funds available for constructing the WPI, at a time
when the national spotlight had not as yet been focused on drug prices. With
the vast universe of commodities that had to be accounted for in the whole-
sale price index under a very limited budget, it is understandable that a group
of products that accounts for less than one percent of the value of all commodities
did not receive more attention, particularly with respect to the updating of the
list of pharmaceuticals priced for the index.
The revision of the wholesale price index of drugs by the BLS in 1961 went
far toward correcting the earlier deficiencies. Prior to this revision, the index
included a large number of such products as agar, citric acid, ethyl alcohol and
glycerin-old standbys of the pharmacist's trade, but hardly representative of
the products used even in the 1950s. The only relatively new products included
as late as 1960 were penicillin, streptomycin, and four sulfa drugs. No represen-
tation was accorded the hormones, diuretics, cardiovasculars, tranquilizers, corti-
cal steriods or the most important of the pharmaceutical groups, broad spectrum
antibiotics.
Before the BLS had indicated that it was revising its drug index, I had been
asked by the industry to construct the index now called the PMA Wholesale
Price Index of Ethical Pharmaceuticals. To do this, I went back to 1949, and
with a more comprehensive list of drugs, constructed an index which showed
price movements since that date. The construction of this index is described
in a succeeding portion of this paper.
It is pertinent to compare the BLS Drug and Pharmaceutical Index with the
PMA index for the periods 1949 to 1960, the period preceding the BLS revision.
and 1961 to 1966, the period after the revision. From table IA and Chart IA,
one can see that the BLS index in the earlier period showed practically no change
in drug prices while the PMA index declined 7.6 percent. The comparison for
the later period (with no change in the technique employed in constructing the
PMA index) shows that the movements of the two indexes were very similar,
with the BLS index showing an even larger decline than show-n by the PMA
index. (See Table I-B and Chart I-B).
TABLE 1.-COMPARISON OF THE BLS AND PMA WHOLESALE PRICE INDEXES FOR DRUGS AND PHARMACEUTICALS
1949-66 (BASE PERIOD 1949), AND 1961-66 (BASE PERIOD 1961)
Year
1949-6
BLS
(A)
6(194
9=100)
PMA
(B)
1961-66 (1961=100)
BLS PMA -
1949
100.0
100.0
1950
98.8
95.9
1951
102.2
96.2
1952
98.8
93.0
1953
99.3
93.0
1954
100.3
91.8
1955
99.2
91.4
1956
98.4
91.4
1957
99.7
92.7
1958
100.4
93.0
1959
99.6
92.7
1960
100.1
92.4
1961
99.9
89.7
100.0 100.0
1962
97. 5
87. 1
97. 6 97. 1
1963
96.4
86.2
96.4 96. 1
1964
96. 1
86. 0
96. 1 95. 9
1965
95.4
86.2
95.4 96.1
1966
94. 8
85. 9
94. 9 95. 8
PAGENO="0157"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1707
8F
1949 1952 1955 1958 1961 1964 1966
Chart I (B)
Base Period 196 1=100
~100
~.-BLS
PMA~~
90 _J_. L~
1964 1966
The PITA Wholesale Price Inclew of Ethical Pharmaceuticals
Before detailing the construction of the PMA Wholesale Price Index of Ethical
Pharmaceuticals, I should like to state that I have always had complete freedom
in designing and constructing this index. This is the only way I would do it;
the Pharmaceutical Manufacturers Association never asked me to do it other-
wise. As a result, I alone am responsible for, this index in its content and its
accuracy.
The index measures the prices of ethical pharmaceuticals as reflected in the
price quotations for those brand name products which are prescribed fre-
quently enough to be recorded in national surveys. The sources of the price
data are revelant editions of a trade journal annual, the Drug Topics Red
Book.3
While the Red Book shows prices by distributor for generic products (no brand
name), to my knowledge, there are no satisfactory available data on the quan-
tities of these products that are dispensed. In any event, indications point
to few, if any, generic producers selling sufficient quantities or with enough
market activity to prompt inclusion in the analysis. The items included in
the PMA index are volume sellers and, therefore, are significant. The survey
on sales volume, described here after, does not include data on generic products
by distributor. For example, there is information on the sales volume of gen-
erically dispensed tetracycline in the aggregate, but not by distributors. Since
Published by Topics Publishing Co. of New York City.
Chart I (A)
bLS and-PMA WHOLESALE PRICE INDEXES
for DRUGS and PHARMACEUTICALS
Base Period 1949=100
Index
105
100
~95
~90
~4~BLS
- PMA
1949
1952
1955
1958 1961
PAGENO="0158"
1708 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
there is wide price variation in generically produced tetracycline with divergent
price movements, it is impossible without quantity data to measure the effect
of these price changes. The 1907 Red Book shows 5 distributors(by generic
term only) increased prices for tetracycline, 2 had decreases, and 9 had no
price changes. Eight distributors dropped out of the Red Book listing and nine
new distributors entered.
In addition, it is doubtful whether any non-specialty generic distributor has
sufficient dollar volume for individual products to warrant inclusion in the
index under the criteria established.
It should be made clear that the prices used in the index which I have
prepared, are list prices as reported by the producers or distributors. There
are no adjustments for trade, quantity or cash discounts nor for free goods
since such information is not available. This deficiency would affect the index
only if there were trends which lead to effective increases, or decreases in dis-
counts off list, over the years. Short run changes, without such changes in trends
tend to counterbalance each other in the long run and probably have no lasting
effect on the index.
The weights for this index were derived from confidential proprietary data
compiled by Medical Data Services a Market Research Firm of Darien, Con-
necticut from its "Prescription Panel Service" and provided to me on a cosifiden-
tial basis. Medical Data Services describes its service as providing information,
from a scientifically designed sample of new prescriptions filled at pharmacies,.
"on national prescription movement in order that a client may determine
trends by therapeutic class by company, and by product form and strength."
The cumulative total of prescriptions for each product in the sample for an
entire year, multiplied by the average prescription price for that product, was.
used to determine the relative importance of each of the prescription drugs.
and to determine the total value of all prescriptions for a therapeutic category.
This information is available within a month or so after the year-end, and
provides reasonably good information on drug usage as reflected in prescrip-
tions actually filled. A parallel but independent service, that of R. A. Gosselin
& Company, of Dedham, Massachusetts, entitled "National Prescription Audit"
was used together with the "Prescription Panel Service" in the preparation of
the 1966 index to verify and reinforce the weights developed.
The two services were remarkably similar in the drug usage patterns they
revealed. The use of either service alone would have provided sufficiently
accurate information with which to design the weighting system, but to reduce
sampling errors to a minimum, the data from the two services were combined.
Commodity selection
From the outset, sixteen categories of drugs were selected for the index plus
an "all other" group to include the remaining drug products. Drug specialties in
each category were selected on the basis of their importance in their theraputic
class. In each of the years since 1949 a product was included in the index if
it was one of the five leading drugs in its group, so that all important specialties
could be included. Items were dropped from the index only when the dollar
volume fell below 0.1 percent of the group volume, to insure stability of the
product list. Table II shows changes that were made in the number of drugs
used in constructing the PMA index:
TABLE Il-DRUGS INCLUDED IN THE PMA ETHICAL PHARMACEUTICAL INDEX, 1949-66
Number of items Total
Year number
Added Deleted included
Number of items Total
Year number
Added Deleted included
1949 99 99
1950 18 1 116
1951 13 5 124
1952 46 7 163
1953 22 4 181
1954 21 7 195
1955 20 4 211
1956 14 4 221
1957 21 5 237
1958 14 7 244
1959 24 20 248
1960 18 19 247
1961 14 1 260
1967 2 0 262
1963 11 0 273
1964 8 1 280
1965 4 4 280
19661 189 106 363
1 In 1966 the basis of item selo:tion was changed somewhat to get better representation for the important products in
groups like hormonas and antibiotics which were previously excluded from the index because they were not among the
5 leading products in their respective categories. Items with $1,000,000 in prescriptions in each ot 2 consecutive years.
were added to the index. A product, already in the index that did not have at least $1,000,000 in prescription volume in
either year, was dropped except that, no therapeutic group is ever reduced to fewer than 5 products.
PAGENO="0159"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1709
The weight used in constructing the index is the average dollar volume for
two consecutive years ending with the year of reference. Thus a product is on
the market at least two years with substantial dollar volume in prescriptions
before it is then included in the index for the second year.
The result is that only products which become established in the market can
be included in the index. Initial exploratory prices during the introductory
period for the product would not generally affect the index. Indication of the
rapid changes that have taken place in drug usage may `be seen from the fact
that only 27 of the original 99 products are still included in the index. This is
considetably less than 10 percent of the total number of products included in the
1966 index. An index `based on these 27 items at 1949 weights would stand at
105.9 in 1966. This, however, would be a most unrepresentative list of the drug
items in use today. Such an index would be no more an index of drug prices
today `than an index based on horse and buggy~ items would be an index of trans-
portation today. It should be obvious that failure to recognize and take account
of such changes on a current basis when constructing the price index would
result in rapid obsolescence of the index. Accordingly, the list of products and
their weigh'ts are reviewed each year to make sure that no important item is
excluded from the index and to provide for the retirement of drugs that have
declined to insignifl'~antly low levels of usage. I note that the Commi~s'ioner of
the Bureau of Labor Statistics in recent testimony `before this Committee stated
that the BLS has recently adopted this policy for its drug index.4
Uha~im indexes
Whenever new items are added to the index `or old items are deleted, or when-
ever the relative impottance of the included items changes, the comparability
of the index with the preceding period diminishes. To avoid any distortion that
might follow from such changes, and to make sure that the only movements
in the index are the results of price changes, a new index series is s'tarted at
the time of each change. This series has, for both the current period and the im-
mediately preceding period, an identical list of commodities with identical
weights. Whatever percentage change is shown by `the new series is applied to
the index through the preceding period in order to measure the cumulative
changes in prices for the life of the index. This procedure is called linking or
chaining, and the index series is referred to as a "chain index". In this wa~ an
index series reflects price changes and price changes oniy. This is a standard
procedure that is followed in all indexes to preserve continuity. To illustrate in
a simple fashion what is achieved, let us assume `that our hypothetical index
as of 1965 stands at 90.0. In 1966, a new list of products (which normally would
contain an overwhelming proportion of the prOducts in the 1965 index) is com-
pared with the prices for this identical list in 1965 and it is found that the aver-
age price has declined 1.0 percent. The new index for 1966 would then be equal
to 90.0 (1965 index) reduced by one percent (of 90.0) or 89.1. Thus, we say in
effect that prices that, up to 1965, had declined 10.0 percent (to 90.0), had
declined an additional 1.0 percent in 1966 from the 1965 level, making the total
decline in the index 10.9 percent (to 89.1).
Sub-indeaes of the PIIIA price indev
In addition to the PMA Index of Wholesale Prices for Ethical Pharmaceuticals,
indexes for various reclassifications of relevent data are available. These are:
1. Therapeutic Group Indexes;
2. Indexes for the group of items covered by patents and those products
without patent protection;
3. An index of drug~ used by those who, are 60 years of age and older.
What the PMA inde~ves show
In the seventen years covered by the PMA Index `of Wholesale Prices of Ethica'l
Pharmaceuticals there has been a total decline of 14.1 percent. Declines have
occurred in eleven of the seventeen years, increases in four and there `have `been
two years with a change of less than 0.05 percent. Of the seventeen therapeutic
groups, ten have shown increases and seven have had decreases since 1949. How-
ever, the declines of more than 50 percent that have occurred in two of the most
important groups, antibiotics and hormones, more than offset increases of much
smaller magnitude in less important groups.
Statement by Commissioner Ross, op. cit.
PAGENO="0160"
1710 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
When the drug products had been classified by their patent status, patented
products showed a decline of 24.8 percent (through 1966) while the non-patented
products had risen by 1.1 percent in this same period. Apparently, the products
with no patent protection show little price change, and such price change that
occurred has been in an upward direction. At the same time, contrary to expec-
tation, the existence of `a patei~t has not operated against price declines. (See
Table III and Chart II, following pages.)
The National Disease and Therapeutic Index, a market analysis compiled by
Lea Associates of Ambler, Pennsylvania shows among other data the di~tribu-
tion of prescriptions fo~ products by age group. One `of these age groujts is the
"60 and over" category, and these data `are available `back to 1959. An index
of whosesale drug prices for this age group w~as constructed, with 1959 as a
base (equal to 100), and the total drug index was compared with this. The results
are shown in the Table IV:
TABLE IV.-PMA INDEX OF WHOLESALE PRICES OF ETHICAL PHARMACEUTICALS, 1959-66
[1959
=1001
Year
Total index
Over 60 index
Year
Total index
Over 60 index
1959
100.0
100.0
1963
92.9
95.8
1960
99.6
99.6
1964
92.7
96.7
1961
1962
96.7
93.9
97.3
95.9
1965
1966
93.0
92.6
96.6
96.5
An important reason for the difference in the decline of the two indexes is
that oral contraceptives, which have shown a marked decline, are not included
in the "over 60" index.
RETAIL PRICE INDEX
BL~ en edical care component of the UPI
Medical care represents nearly 6 percent of the expenditures of wage earners
and clerical workers. Only 0.6 percent of the expenditures are on prescriptions.
Professional services of doctors, dentists, optometrists and nurses, hospital sup-
plies and nursing home services, health insurance and over-the-counter drugs
and sundries account for the remainder, and in all, are nearly ten times as great
as expenditures for prescriptions.
TABLE Ill-WHOLESALE PRICE INOEXES OF ETHICAL PHARMACEUTICALS (PMA) AND CONSUMER NONDURABLE
GOODS OTHER THAN FOOD (BLS) 1949-66
[1949= 1001
Year -
Ethical pharmaceuticals
Consumer non-
durable goods
other than food
Patedted
Nonpatented
All
1949
100.0
100.0
100.0
100.0
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
92. 1
92.2
85.4
84.9
82.7
82.3
82.6
83.8
84.0
83.6
83.4
79.8
77.1
76. 1
75. 6
75. 6
75.2
100. 1
100.7
99.8
100.0
100.6
100.8
100.9
102.7
103.6
103.5
100.8
98.7
98.2
98. 1
99. 1
101. 2
101.1
95. 9
96.2
93.0
93.0
91.8
91.4
91.4
92.7
93.0
92.7
92.4
89.7
87.1
86. 2
86. 0
86. 2
85.9
101. 6
109.4
106.7
107.7
108.0
108.6
110.8
113.3
112.6
114.3
115.1
115.1
115.2
115. 5
115. 2
116. 6
118.8
PAGENO="0161"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1711
Chart II
WHOLESALE PRICE INDEXES of ETHICAL PHARMACEUTICALS (PMA)
and NON-DURABLE CONSUMER GOODS OTHER THAN FOODS (BLS), 1949-1966
Base Period 1949100
1nde~c
120 Wholesale Prices Non-durable
Consumer Goods Other T~ian Food
100
Ethical
90 ~bb., Pharmaceuticals
Patented Ethical
80 ~. Pharmaceuticals
70 ~ I I t
1949 1952 1955 1958 1961 1964 1966
It is only since 1960 that the Consumer Price Index has included prescription
items that by current standards can be considered truly representative. Prior
to that year the prescription index was based on three items-APC tablets, elixir
terpin hydrate with codeine, and buffered penicillin 5-hardly a representative list
of ethical pharmaceuticals. This index rose 27.8 percent between 1949 and 1960,
but I would be reluctant to accept this as an indicator of retail prescription
prices. The item list was revised and expanded in 1960 to include seven thera-
peutic groups, and just this year three additional groups were added. Each group
is represented by only one or two products, with a total of seventeen items to
represent the ten groups. While this is a rather thin basis to measure price
movements of the individual therapeutic groups, it should be reasonably reliable
to measure price movements for all prescriptions. The BLS retail prescription
index is available on a quarterly basis while the BLS wholesale index for ethical
pharmaceuticals is available on a monthly basis and thus can reflect price move-
ments in the short run.
Interpretation of price indexes
Price indexes are designed to show what changes have taken place in average
prices over time. It is therefore important, in constructing a price index, to have
price as the only variable. To do this, all other factors, in so far as possible,
must be kept constant. That is why in calculating price ratios between two
periods of time, the list of commodities and their relative importance must be
identical in the two time periods. If prices have changed, the index will reflect
the net effect of all price changes that have occurred in the interim. An index
Statement by Commissioner Arthur M. Ross before this Committee, May 16, 1967, p. 235.
Sl-280-CS-pt. 5-l1
PAGENO="0162"
1712 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
cannot tell us whether prices are too high or too low, or whether more or less
expensive products are in use today than yesterday. An index also does not
reflect quality changes introduced in the new drugs. If only half as much of
the new drug at twice the unit price does as much or more for the patient this
cannot be determined from the price index. A price index answers only one
question: for products on the market and included in the index, have prices
risen, fallen, or remained stationary and by how much?
An index of retail prices should not he used as a gauge of wholesale prices
nor should wholesale price indexes be construed as measuring retail price move-
ments. Time lag is one important variable. Also, additional elements determining
retail prices may be and usually are different from the factors which affect
wholesale prices. We find, for example, that consumer prices have a larger ele-
ment of wage costs, often in the form of services rendered the purchaser at
the retail level. These are more likely to be personal, such as waiting on a cus-
tomer, filling his prescription, delivering it to his home, providing credit, and so
forth. These -kinds of services are not readily amenable to automation or mech-
anization, so that rising wage costs cannot be offset by higher productivity in
the same degree as may be the case at the manufacturing level. This may be a
principal reason for the greater rise in consumer prices than in wholesale prices
during the past twenty years. Thus, since 1957-59 all wholesale prices have risen
6 percent while consumer prices have risen 15 percent.
Average prescription prices
A very simple statistic that has been used at times to measure what has
happened -to drug prices, is the average prescription price. This, if not properly
interpreted, can be a very misleading yardstick.
In an independent analysis of average prescription prices, Professor Robert
17 Evanson. Head of the Department of Pharmacy Administration, School of
Pharmacal Sciences, Purdue University, has arrived at conclusions essentially
similar to those presented in my discussion. Mr. Chairman, I would like to call
the Committee's attention to Dr. Evanson's paper as published in the September
1967 issue of Tile asd Till, pp. 58-63, and I respectfully request that this paper
be made a part of the record of these hearings.
In contrast with a price index, an average prescription price has a number of
variables other than the price of the drug. No account is taken of the fact that
the mix of prescriptions written and filled in a given period of time may vary
from the mix found at some other time. Since prices for pharmaceuticals differ
from item to item, we would expect the average prescription price to vary as
the mix greater or lesser emphasis on higher priced drugs. Thus the average
prescription price can rise or fall even if the prices of each of the drugs included
in the average have not changed by one cent. This usually occurs seasonally, as
the incidence of respiratory ailments changes, for example, but can also occur
in an irregular fashion with the incidence of epidemics, or even as a result of
new approaches in drug prescription for various ailments as the medical art
progresses.
More and more, the pharmaceutical industry is introducing sustained action
capsules, so that one dosage now does the work of two or more of the former
dosages. We can expect these new capsules to be more expensive than the former
single dosage, but often the cost for a given amount of medication is less. A.
single prescription of this new drug may therefore cost more than a single
prescription of the older drug, but at the same time may require fewer refills.
Thus the total drug cost during the course of an illness may be lower with the
new drug even though the single prescription price is higher.
The pharmaceutical industry has been developing combinations of drugs for
physicians who wish to prescribe two or three drugs to be taken simultaneously
by a patient. The combination naturally may be expected to cost more than any
one of its components alone, but it will probably cost less than the total of the
components purchased separately. Certainly at the retail level, there would be
only one pharmacist markup for filling the prescription for the combination
rather than the usual two or three for filling two or three separate prescriptions.
An average prescription price would not reflect this saving. On the contrary it
would show the higher price for the new single prescription.
Also, with advances in medical science, there tends to be an increase in the
number -of cases of chronic ailments (persons with chronic heart ailments,
diabetes and arthritis, for example). Doctors in such cases are prone to prescribe
PAGENO="0163"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1713
drugs in larger quantities to reduce the need for frequent refilling of prescrip-
tions and to give the patient the benefit of whatever lower prices might accrue
with volume purchases. Thus, even with lower prices for drugs (on a per dose
basis), the average prescription price rises.
Retail vs. wholesale prices
However, what has been most serious in the misuse of average prescription
prices is the use of these prices for measuring what has happened to manu-
facturers' prices. Since this Committee has concerned itself primarily with
manufacturers' prices, I respectfully submit that what should be examined is
prices charged by manufacturers before pharmacists' margins `have been added
to th'ese prices. A previous witness, for instance, testified before this committee
that "In Washington, D.C., one preparation was found to retail for 75 cents
at one store, and $2.45 at another".° How this difference can in any way be
attributed to the misdeeds of the' manufacturer of this product I cannot under-
stand. Even if this manufacturer should cut hi's price for this product in half,
this reduction would be almost completely obliterated by the markup of the
second pharmacist.
Past studies by government agencies and Congressional committee,s have been
directed at the growing spread between the prices received by producers of
wheat, cotton and wool, and the prices paid by consumers for bread and clothing.
Such studies have affirmed what economists are fully aware of; namely, that
distribution costs have been growing steadily and that this widening spread does
not of necessity reflect price gouging. Most costs have been and are still rising.
Taxes, transportation costs and rentals of commercial facilities have generally
been rising on an upward curve. Even sharper increase,s have taken place in
wages. Automation of retail pharmacy services is practically impossible so that
these increased costs cannot be absorbed by the use of more efficient equipment,
as has been possible, to a large extent, in manufacturing. These cost increases
must therefore be passed along to the ultimate consumer in the form of higher
prices and must of necessity be reflected in the higher average prescription price.
Total illness cost
Over the past few decades new drugs and medicines have made important
contribution,s to the health of this country and are in part responsible for a
declining death rate and a rising life expectancy. New products have been de-
veloped to provide drug therapy where none had been available previously, e.g.
tranquilizers and drugs that are more certain and speedy in effective cures, the
so-called wonder drugs.
Many of these drugs are higher priced than the products `they replaced on a
per dose basis. But this kind of comparison is no more valid `than a comparison
of jet fares with railroad fares simply on the basis of miles traveled. Just as it
makes more sense to measure the comparative c'ost of a trip from Washington
to Los Angeles by plane and train taking into account not only `the fare, but all
other incidental costs such as meals, and time lost in travel, so is it more rea-
sonable to measure the total drug cost over the period of an illness rather than
the price of a prescription. But more than that, except f'or th'e geriatric illnesses
which come with prolonged life `spans, new more sophisticated drugs shorten the
duration of an illness, shorten the required period of medical, nursing and l1os-
pital care, reduce the pain and discomfort suffered by the patient, and put him
back on his feet and back `to his job m'ore quickly, reducing hi's earnings loss
due to illness.
I cite this, Mr. Chairman, not as argument, but as sound economics which
goes to the question of real cost. For some specific illnesses these savings can
be measured in dollars and cents. Before the development of antibiotics, for
example, lobar pneumonia meant five weeks of `hospitalization, long convales-
cence, and several hundred dollars for doctors, ,nurses, medicines, oxygen and
hospital care. Today pneumonia means a week to ten days of illness, usually at
home, $15 to $30 for drugs, and just two or three visits from the doctor.7 Fifteen
to thirty dollars for drugs sounds expensive. WhOt is generally overlooked is the
astonishing reduction of all other medical costs, and the reduction in loss of
time from work.
From statement by Commissioner Arthur M. Ross before this Committee, May 16, 1967.
Kramer, Lucy, "Drugs and Medicine," Public Health Reports, Vol. 73, No. 10, October
1958.
PAGENO="0164"
1714 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
There is a real need; therefore, I submit, for some measure which will show the
true change in the total cost of treating an ailment. A somewhat higher drug cost
may result in a considerably lower total cost in effecting a cure. Even such an
index could not evaluate the lowered mortality rate, the shorter duration of an
illness with the accompanying relief from disability, pain and absence from one's
occupation.
There have been occasional attempts to develop such a measure, but to date
these have not provided adequate data. A principal source of difficulty is the
unavailability of suitable data concerning medical, hospital, nursing and inedi-
cation costs in earlier years. The Department of Health, Education, and Welfare
might be the appropriate agency to develop such measures.
Several additional observations are pertinent at this point. The introduction of
unit dosages in hospitals, for instance, has made available a dose in a sanitary
package, properly labelled with a space for the patient's name. That patient is
thus assured there is less risk of a mixup in the medication provided, and of con-
tamination from handling, and simultaneously the nurse can take care of her
medications in much shorter time. The higher dosage cost is more than offset by
the advantages that the unit dosage system makes possible.
Also an expenditure of relatively few dollars for tranqualizers now makes it
possible for a person who formerly had required hospitalization in a mental
institution to live at home with his family. And now chemotherapy is possible
with patients, who before the advent of these drugs, were not amenable to
psychiatric treatment. Our national drug bill must inevitably rise as such new
products are developed to cure previously incurable illnesses whether these
products implement or replace the use of other more expensive forms of medical
services.
As indicated previously, we cannot escape the conclusion that the reduction in
the mortality rate is accompanied by a rise in the morbidity rate. Many whose
lives have been saved by progress in medical care and drug therapy, must rely
on continued medical care and drug usage for long periods of time, and, in many
instances, for the rest of their lives. Their very existence may depend on contin-
uous drug therapy. It is thus inevitable that the quantity of pharmaceuticals
consumed in this country must rise, not only because of a growing population,
but because of our growing per capita reliance on drugs for protracted periods.
CONCLUSION
In summary, I would say that average prescription prices cannot tell us what
has been happening to drug prices. There are too many factors other than changes
in drug prices that tend to distort this measure. I fact, for nearly two decades
prices for ethical pharmaceuticals, as measured by the most reliable price indexes,
have declined, while the price of an average prescription has been rising.
Neither average prescription prices nor an index of pharmaceutical prices can
tell us what is happening to our total expenditures for drugs. The increase in our
life spans and the advances of medical science inevitably must add to our drug
bill. The continuing research discoveries in the chemno-therapeutic management of
ailments which previously baffled the medical profession, have resulted in more
extensive drug usage and an increase in our total drug bill. It is important,
therefore, that while we note an increase in a patient's drug bill, we do not over-
look the tremendously increased benefit to his well-being arid the resulting Posi-
tive contribution to our economy.
I hope, Mr. Chairman, and Gentlemen, that these somewhat lengthy remarks
have shed additional light on drug prices and will contribute to a better under-
standing of the elements properly entering into the measuring of price changes.
[Reprinted from Tile and Till, 53: No. 2 (June) and No. 3 (September), 1967]
WHAT Is AN AVERAGE PRESCRIPTION CHARGE?
(By Robert V. Evanson, Ph. D., Professor and Head, Department of Pharmacy
Administration, School of Pharmacy and Pharmacal Sciences, Purdue University)
An average prescription price, more properly referred to today as an average
prescription charge, is a well-known item of statistical information in community
and hospital pharmacy practice. Simply stated, it is an average value, expressed
in dollars, of the total value of prescription revenue divided by the total number
of prescription orders dispensed.
PAGENO="0165"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1715
A given average may result from operating statistics of a single pharmacy,
of several similar pharmacies in a controlled sample, of a number of pharmacies
selected purely by chance, or of a sample or complete population of pharmacies in
a specific geographical area. It is possible that the average of any of two or more
samples may be the same, but the possibility of such a result is reduced by many
factors which contribute to determining an average charge. For this reason, it
is appropriate to be mindful of these factors when we view trends in average
prescription charges, particularly over a period of years as related to actual
changes in costs of prescription products.
What is an average prescription charge? This might be better asked in the form
of "What is an average?" Webster's dictionary requires several hundred words
to describe the many concepts and possible applications of the "average." Statis-
tically, an average represents a mean value, medial sum, or quantity, made oufof
an array of unequal sums or quantities-i.e., an arithmetical mean. A group of n
numbers are added, and then the sum is divided by n to arrive at a figure which
should express or imply some central tendency of performance or value. In fact,.
as generally used, "average" refers to a loose approximation to central tendency,
but the centralness depends on the nature of the items being averaged. It is quite
possible for a few high values in an otherwise low-valued distribu.tion to pull the
average well above the true central tendency of the distribution.
In our context, perhaps the most applicable concept of the term "average"
is that of a ratio. Then it can be stated that an average is a mean value expressed
as a ratio of value or performance between one set of items and another set of
items. It may be expressed as a whole number, as a fraction, or as a dollar value.
However, as a ratio, it is not necessarily the value of any actual item or the
actual performance of any performing unit. An array of prices may produce an
average which does not exist as any one of the prices in the array. An average
time for a group of runners may not be the time at which any one runner ran the
race.
Because it is of interest for analysis and planning to review the various trends
of operating data and because it is difficult to analyze masses, average prescription
charges have been widely published and have been of much interest to community
pharmacists.
However, to have any value to a .system in which a particular item is used,
the item must offer some vestige of useful information or meaning, and it must
be considered in the context of its derivation. Useful information or meaning, in
the case of an avearge presciption charge, would include the reflecting, meas-
uring, or predicting of economic conditions, professional circumstances, inter-
workings of market actions, or merely simple trends for these facts or factors.
A brief review of factors involved in the determination of prescription charges
will indicate the questionable value and character of a single average charge as
a meaningful expression of prescription department economics or performance for
a single pharmacy. If the charge cannot be meaningful in this respect, it most
certainly cannot be so for any other aspect of the drug industry, including manu-
facturing, distribution, and the overall practice of pharmacy.
Cost factors
Two major cost categories are involved in every charge: material costs and
operating costs. Variability of material costs can exist in a pharmacy when the'
level of operation permits added investment in quantities to realize the advan-
tages of quantity-discounted cost prices. Some products are dispensed occasionally
and need be purchased and stocked only in small quantities to meet the existing
demand. In this case, the cost is well defined for the pharmacist. However, the
rate of dispensing other products is high enough to permit buying at regularly
offered quantity discounts. When this is possible, the pharmacist may use the full
wholesale cost for the smallest unit, the fully discounted cost for the largest
unit, or, at his discretion, any value in between necessary to `meet the needs of
his customers, to enter and remain in local competition, or to allow for special
considerations which are a part of his total community service.
The material costs problems is not avoided when an order requires compound-
ing. No one questions the fact that compounding involves time, effort, and the
use of technical skills. The question arises about adding the value of these factors
to the ingredient values in order to determine a cost (total value) for the final
product to he dispensed. In some pharmacies, these values are added to establish
such a total cost value; in others, they are assumed to be a part of the markup
or fee to be added to the ingredient costs. The average charge will be lower or
PAGENO="0166"
1716 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
higher, depending on the procedures followed to determine cost values for both
manufactured specialties and custom-made, compounded products.
Operating costs seem to present a different problem, since they can be de-
termined definitely by the simple process of adding all expenditures paid out
(or accrued) during the period. A small portion of the total can be added, by
means of a percentage markup or a fiat fee, to each unit dispensed. However, the
amount of operating costs to be charged against the prescription department is
not always easily or correctly assigned.
For the prescription specialty shop, little or no problem exists, since all ex-
penses accrue to conducting a professional practice. Only those incurred in
selling physicians' supplies or furnishing nondispensing services and supplies
(for example, to nursing homes) need be deducted to get a total for the dispensing
function. As soon as over-the-counter drugs, other health-needs items, and general
merchandise are involved, the assignment of expenses becomes more difficult.
Unless adequate records are kept to assign expenses properly, it is possible to
undercharge or overcharge the department for its share of the indirect expenses
which represent services rendered for the store as a whole.
The point here is that, unless some valid means of defining the actual expenses
incurred by the prescription department is devised, any markup or flat fee is
little better than a purely arbitrary guess. Since these, too, exert a direct effect
on individual charges, the resulting average charge is valueless even as an indi-
cator of trends for one pharmacy. It follows that an average is even less sig-
nificant for any sample of pharmacies, and significance decreases as the sample
size increases.
A final "cost" is that incurred by the business for the owner's investment-
usually referred to as "return on investment and/or net profit." As sales volume
increases, a set profit-return goal can be achieved with a small percentage return
on sales. Increased turnover rates will reduce the profit needed per turnover. The
economic status of customers, local competition, amount of welfare business, and
internal merchandising policies all affect the potential profit or lack of profit.
Each of these cost factors brings something to bear upon the potential varia-
bility or level of prescription charges both individually and collectively. If so
many variables exist in the seemingly simple procedure of determining a given
charge for a prescription, an average of several charges will be affected by all
these variables without reflecting or measuring any of them.
Pre$criptio~i miclJ
The strongest single control over the average charge for prescriptions is the
prescription mix, because it contains all the potential variables which affect the
charges represented by an average. Any given charge depends on four variables:
the specific product prescribed, the unit price per dosage form, the quantity pre-
scribed and/or dispensed per prescription order, and the number of prescriptions
dispensed for each product.
The selection of product is the function of the medical practitioner, and it
depends on his course of therapy and the inherent characteristics or values of a
specific product to treat specific diagnosed conditions. The incidence of certain
diseases or malfunctions of the human body, specialty practices, prescribing hab-
its or patterns, and many other important variables affect the choice of product.
Hundreds of products are available with different components, different dosage
forms, varying strengths, and different prices per dosage-form unit. The
selection of any one will automatically establish the dosage-form unit price.
All of the variables previously discussed as cost factors are immediately brought
to bear upon the prescription mix when the product of choice is selected and re-
corded on the prescription order.
The quantity dispensed per order has a direct effect on the total charge and
is dependent on the amount prescribed by the physician to carry out his treatment
pattern or-in some instances-by the amount the patient can afford. For exam-
ple, if 50 tablets at 10 cents each are ordered, the charge would be $5.00. If 100
tablets are ordered, the charge would be $10.00. However, if 100 tablets are or-
dered. but the patient cannot afford to pay for more than 50 at one time, two
half-orders - could be dispensed over the treatment period, each with a $5.00
charge. A similar effect on the average is exerted by the frequency and charges
for prescription renewals.
The number of prescriptions for each product will be affected by many different
kinds of. variables, including weather, time of year, the range of age levels among
a pharmacy's clientele, epidemics of specific diseases, the extent of preventive
PAGENO="0167"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1717
health care possible among different socioeconomic groups, prescribing patterns
of all the area's practitioners, and the incidence of chronic and acute illness. Any
emphasis on particular products or classes of products to meet demands; occa-
sioned by these variables will have a direct and immediate effect upon all those
charges as well as the average charge for any representative sample of pre-
scriptions.
To understand better the effects of prescription mix on average charges, look at
the following series of hypothetical examples that have been developed and
analyzed to illustrate the mix (Table 1). Five products are structured into this
example. They could be similar, competing products or different, noncompeting
products.
For purposes of analysis of both prices and, market fluctuations, all products
are asumed to be manufactured specialties prescribed and used to treat similar
or identical conditions. Each product is produced and distributed by a different
company. The addition of a product indicates its introduction into the market as
a new drug at the beginning of the year first shown.
Table 1 details specific data for five years as they might be found in the records
of a representative pharmacy. A few significant changes will be noted here
without any suggestion of the reasons for which the changes occurred. During
the period, four products were reduced in price by 5 to 37 percent, yet no similar
direct effect was exerted on the average charge. Contrast this with the many
instances in which public references to prescription charges state or imply that
drug prices or prescription prices are increasing because the average prescription
charge is increasing from year to year!
TABLE 1.-THE EFFECT OF PRESCRIPTION MIX ON THE AVERAGE CHARGE
Price/unit
Year Product dosage form
No. of
prescriptions
dispensed
Average No.
of units!
prescriptions
Total
prescription
revenue
Average charge
1 A
B
2 A
B
C
3 A
B
C
0
4 A
B
C
D
5 A
B
C
0
E
$05
.15
.05
.15
.30
.05
.12
.30
.20
.04
.12
.30
.20
.04
.11
.19
.19
.08
1,000
100
30
20
$1,500
300
1,100
`25
1,800
$1.64
500
500
500
30
20
10
750
1,500
1,500
1, 500
`20
3, 750
2. 50
250
600
500
350
30
20
20
10
375
1,440
3,000
700
1,700
`20
5,515
3.24
250
600
500
350
20
10
20
30
200
720
3,000
2,100
1,700
`20
6,020
3.54
200
550
600
300
150
20
10
30
30
40
160
605
3,420
1,710
480
1,800
`26
6,375
3.54
1 The averages given for average No. of units/prescriptions are raw averages not weighted for prescription frequencies.
As more products in any class become available, the total number prescribed
increases, but more significance is found in the effect of the shifts of each prod-
net's numbers. The great number of product A in year 1 kept the average charge
quite close to the $1.50 treatment `cost even though 100 patients paid $3.00 for
treatment with product B.
In each year, the total number dispensed had a different effect upon the average
charge by affecting individual market shares and volume shares. As the reader
PAGENO="0168"
1718 COMPETITIVE PROBLEMS IN THE DRUG fl~DUSTRY
can see, these effects are not directly noticeable in the value of the year's average
charge.
The average number of units dispensed per order also moved down for some
products and up for others with no apparent or predictable pattern. Although it
is obvious that increases and decreases in the number dispensed will affect
the individual prescription charges directly, nothing here indicates that a change
in the average charge will clearly reflect such changes within the sample being
studied.
Total volume per product accurately describes that product's dollar share of
the total revenue from all products. It is neither static nor absolutely predicta-
ble for any given period of time. It depends entirely on the three factors which
combine to establish individual charges, the sum of which make up the total for
each product. In the data given, one product decreases in volume share each suc-
cessive year, and one increases each year. Two products increase and then de-
crease. However, because the average charge is determined by using the total
revenue from all products, there is no way for it to represent or measure any
of these changes. In fact, the average charge for year 5 is the same as that for
year 4, even though significant changes occurred within each of the factors and
affected its ultimate value.
Average prescription charges have been determined for every imaginable sam-
ple of prescription performance for decades. They have been reported religiously
by every publication concerned with the economic status of the practice of phar-
macy. They vary every year from report to report by as much as 10 to 20 percent.
Characteristically, they have been increasing every year. They are quoted by all
manner of personages in pharmacy, government, consumer groups, and the press
to prove or disprove many opinions and concepts, including the biases and mis-
understandings of those persons who are using an average charge out of context.
The sample truth is that the "average prescription charge" is one of the most
overused and least significant ratios of performance in American pharmacy. It
is affected by everything, but it affects nothing. It is taken to indicate changes in
pharmacy's economic facts, but in and of itself it is not capable of reflecting.
measuring, or predicting any of these changes. What, then, is an average pre-
scription charge? An average prescription charge is a ratio of the total revenue
obtained from a specific number of pre~criptions to that number of prescriptions,
reduced to a relative valve per prescription and expressed in dollars and cents.
It is nothing more and nothing less.
Prescription ci? arges versus price indicators
The following list will serve to summarize the concepts most commonly re-
ferred to in discussions of prescription economics. It summarizes also, in the
order of presentation, the basic points made in the first part of this analysis.
Specifically, an average prescription charge is not:
L The price of any particular product dispensed or sold.
2. A charge for any professional service rendered.
3. An indicator of drug prices or trends in drug prices.
4. An average of all patient treatment costs.
Product price
An average prescription charge (or price) cannot be the price of commodities
dispensed on prescription. Although the products, as commodities in trade, may
be priced individually per bottle or per dosage-form unit, such prices are not
based on the total value of any of the prescriptions dispensed. The "retail price"
for each order dispensed will depend on the price per unit and the number of
units involved, with consideration of breaking bulk, special quantity discounts,
and certain other exceptions to basic pricing policies and procedures. It also will
depend on whether the product is prefabricated or compounded. Another factor
is the submersion of any product "price," other than that of the drug dispensed,
in the value of the total package of goods and services received by the patient.
A simple example is the added cost of a container when the manufacturers' orig-
inal. ~outainer is not used.
~erv~ce charge
An average charge is not a charge for any of the services rendered in connec-
tion with the compounding and/or dispensing of an order. In some cases, a simple
service package. properly labeled, is sufficient. In other cases, a pharmacist may
have to expend much professional effort and time to satisfy the patient's needs.
PAGENO="0169"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1719
The charge for each prescri~rtion may be varied to reflect the effort involved.
It may be included in a single markup or a fee added to' the product price. Or it
may be ignored entirely for certain persons who cannot afford to pay for prescrip-
tion service but require it to sustain their normal health.
The charge for a prescription includes the value of the total product dis-
pensed-that is, the price of the commodity plus the charge for the services
rendered. In some cases, these can be distinguished easily; in others, they are
inseparaby intertwined. Therefore, it is impossible to show that an average
charge is an average for either factor or for both together. It is clearly and
irrefutably definable, but the value of any single unit or groups of units that
comprise the average prescription charge actually remains indefinite until sep-
arated from the total array of prescription charges being analyzed. For this rea-
son, the average is not and cannot be a price or a charge for anything.
This fact must be emphasized because too often, in the press and in personal
statements, references are made erroneously to the "average price of drugs" in
a particular year or to the "average charge for a prescription in 1966." Then,
if the 1966 average is greater than that for 1965, the generalization is made that
prices of drugs or charges for prescriptions increased in 1966 over 1965 charges.
Such a generalization is false as well as misleading, and it cannot logically be
made from the two consecutive averages. Nor ~can it be made from the data from
which the averages were determined without a meticulous analysis of intragroup
items that would hold size, quantity, dosage forms, and services involved con-
stant, in order to measure comparable units.
Drag prices
An analysis of drug products sold over a number of years will bring out specific
facts about the prices of pharmaceuticals at various levels in the channels of
distribution, it will show that prices of many products have been reduced sig-
nificantly, that other prices have necessarily been increased when costs of produc-
tion and/or distribution rose as a natural consequence of economic conditions,
and that still others have remained constant for long periods of time even though
general conditions may have caused significant price increases for other com-
modities and services.
An analysis of the average prescription charge over the past twenty-three
years indicates a consistent but variable rate of increase dollarwise and per-
centagewise. Table 2 shows `the annual increases in cents and percentages, as
reported each year in the Lilly Digest from the 1944 average of $1.10 to the
1966 average of $3.59. These are averages based on the total sample.
TABLE 2.-ANNUAL INCREASES IN AVERAGE
Year Average Dollar Percent of
charge increase increase
Year Average Dollar Percent of
, charge increase. increase I
1944 $1.10
1945 1.20 $0.10 9
1946 1.33 .13 10
1947 1.41 .08 6
1948 1.51 .10 7
1949 1.60 .09 6
1950 1.77 .17 10
1951 1.90 .13 7
.1952 2.08 .18 9
1953 2.19 .11 5
1954 2.27 .08 3
1955 2.46 .19 8
1956' 2.62 .16 6
1957 2.85 .23 9
1958' 2.96 .11 4
1959 3.09 .13 4
1960 3.19 .10 3
196L' 3.25 .06 2
1962 3.32 .07 2
1963 3.39 .07 2
196L 3.41 .02 1
1965 3.48 .07 2
1966 3.59 .11 3
1 Percentages are rounded at the first whole number value.
The anual increases in the average charge range from $02 to $23 and from
1 to 10 percent. No definite pattern is evident, except for some apparent stability
in the last seven years at about 2 percent., Successive eleven-year comparisons
provide increases as follows: 1944-1954=115 percent; 1950-1960S0 percent;
1953-1963=55 percent; 1955-1966=46 percent. The total increase from 1944 to
1966 is 226 percent.
Taken at face value, these percentages offer erroneous evidence of the relative
movements of drug prices. Actually, the movements are not measurable by these
data. During this period of twenty-three .years, drug products have changed
progressively in form. A substantial number of medications that formerly were
PAGENO="0170"
1720 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
compounded extemporaneously have been replaced by an almost complete prod-
uct mix of prefabricated dosage units. There has also been continual change in
the kinds and classes of prescription products used. The ratio of renewals to
total orders dispensed (based upon the actual numbers) has increased from
39.5 to 53.6 percent. This represents an increase from 2,505 to 10,623 units, or
324 percent. Therefore, it is practically impossible to make comparisons of aver-
age charges for the purpose of demonstrating price change over time.
It should be apparent at this point that the average prescription charge can
in no way be an indication or evaluaton of drug prices at any level of distribu-
tion. However, as recently as March 1, 1967, the average price per prescription,
taken from a leading pharmacy journal, was cited in a government report for the
purpose of comparing prices of drug products. Furthermore, the report aserted
that the average prescription price is capable of reflecting changes in drug prices
as well as in quantities of drugs dispensed and the use of new drug products.
It is unfortunate that this simple ratio should have been given such official promi-
nence `to support contentions which it cannot possible support or conclusions
which cannot be derived from it.
Treatment costs
In the first part of this article, the discussion of treatment costs implied that
the average charge might be interpreted as an average treatment cost for drugs.
It is not possible for an average of all prescription charges to represent an aver-
age treatment cost for any illness or condition requiring drugs. It may be neces-
sary for the patient to have several renewals of a prescription before his health
is regained. He may also require two or more prescription products, with or with-
out renewals. He may need OTC drugs or other items to complete his treatment.
Therefore, a patient's treatment cost may be only $1.25 (for an inexpensive, once-
used prescription), or it may be $25 (for several renewals or more than one prod-
uct), even though the average charge in the pharmacy which dispensed his
medications was only the $3.59 reported for 1906.
Price indicators
Indicators in the form of carefully constructed, statistically significant price
indexes that validly measure price movements over time are available to the
public, the business world, and pharmacy. The Consumer Price Index (CPI)
and the Wholesale Price Index (WPI) are measures of monthly price changes,
prepared and published by the Bureau of Labor Statistics (BLS) of the 1J.S~
Department of Labor. The "PMA-Firestone Index of Prescription Specialties,"
a wholesale-level price index, is compiled by Dr. John M. Firestone, professor
of economics at `the City College of New York, and published by the Pharmaceuti-
cal Manufacturers Association. A brief description of these indicators will pro-
vide background for understanding the inability of the average prescription'
charge to indicate price trends or changes.
Consumer Price Inden
The CPI is computed by comparing, at different periods, costs of a fixed set of
goods representative of all purchases made by urban wage and clerical workers
and ranging from food to medical costs to automobiles. It is popularly called a
"market-basket index" or a "cost-of-living index." Technically, it is a price
index with fixed or constant weights. The weights are devised to reflect the rela-
tive importance in total spending for the particular goods and services which
are priced.'
The concept of price in the CPI is the amount charged for a particular specified
product, identifiable by style, number, or other designation, on a single sale in
an outlet at specific terms of sale. Specification pricing is used for a great
majority of commodities. In this case, "specification" is a detailed description
of the physical characteristics which determine an item's quality and influence its
price. Both producers and retailers assist in developing and making changes in
the specifications.
A complete tabulation includes average expenditures for 1,800 individual items
in the consumer expenditure schedule. Drugs included are such products aa
liquid tonics, multiple vitamin concentrates, cold tablets or capsules, cough
syrups, aspirin compounds, capsules of tetracycline, chiordiazepoxide hydro-
`U.S. Department of Labor, Bureau of Labor Statistics: The Consumer Price Index:
History and Techniques (Bulletin No. 1517), Washington: U.S. Government Printing Office,
1066.
PAGENO="0171"
COMPETITIVE PROBLEMS IN THE DRUG ~DUSTRY 1721
chloride, and sodium secobarbital, and elixir of terpin hydrate with codeine. Also
included in this index are tablets of sulfisoxazole, buffered penicillin G, pheno-
barbital, propantheline bromide, meprobamate, reserpine, crystalline digitoxin,
prednisone, and chiorothiazide.
Wholesale price ~ndes~
More important to an understanding and appreciation of drug prices and price
trends is the BLS Wholesale Price Index. It is structured to measure average
changes in prices of commodities sold in primary markets of all fifty states of the
union. Its universe comprises all commodities produced or imported for sale.
Commodities selected usually include those with the largest shipment values, as
shown by industrial censuses.2
The WPI is calculated as a weighted average of price changes in which the
weights represent the total net selling value of the commodities. The relative im-
portance of an item represents its basic weight used in the index multiplied by
the index of price change between the weight date and a later date. The result is
expressed as a percentage of the total for all commodities or for the major group.
Several major changes have occurred in the sample of drugs and chemicals
since the WPI was originated. The separation of drugs and chemicals, the re-
moval of cosmetics from the drug group, and more recently, in 1961, the break-
down of drug products into ethical and proprietary pharmaceuticals are exam-
ples of the kind of continual revision necessary to maintain a factual and realistic
analysis of prices. Because of the 1961 revision, it is possible to recognize relative
trends in drug prices that were previously obscured by the grouping. A total of
111 items were included in the class of drugs and pharmaceuticals. To show the
relative importance of the drug group to all commodities sampled as well as the
relative importance of items within the drug group, Table 3 gives these relation-
ships for 1961 and for 1964.
TABLE 3.-RELATIVE IMPORTANCE OF PHARMACEUTICAL PRODUCTS IN THE WHOLESALE PRICE INDEX
Subgroups
Relative impo
maceuticals
items
rtanc
to
e of phar-
all WPI
Percentage of importanceof
individual items within drug
group
1961 1
1964 1
1961 2 1964 2
Ethical pharmaceuticals 0. 534
Anti-infectives . 148
Vitamins .086
Ataractics . 057
Dematologicals . 036
Antiarthritics . 030
Cough and cold preparations . 027
Hormones . 023
Sedatives and hypnotics . 023
Cardiovasculars and antihypertensives . 020
Diuretics . 017
Hematinics . 017
Diabetics . 013
Antiobesity preparations . 012
Analgesics . 012
Antispasmodics and anticholinergics . 013
Proprietary pharmaceuticals . 236
Internal analgesics .069
Cough and cold preparations . 041
Antiseptics . 032
Laxatives and elimination aids . 028
Vitamins . 022
Antiacids . 021
External analgesics . 015
Tonics and alteratives . 008
0. 519
100. 00 100. 00
. 131
.087
. 057
. 038
. 030
. 026
. 023
. 022
. 021
. 017
. 017
. 013
. 013
. 012
. 012
27. 72 25. 20
16.11 16.76
10. 67 10. 98
6. 74 7. 32
5. 61 5. 78
5. 06 5. 01
4. 31 4. 43
4. 30 4. 24
3. 74 4. 05
3. 18 3. 28
3. 18 3. 28
2. 43 2. 50
2. 25 2. 50
2. 25 2. 31
2. 43 2. 31
. 241
100. 00 100. 00
.070
. 041
. 033
. 030
. 022
. 022
. 015
. 008
29.23 29.04
17. 37 17. 01
13. 56 13. 69
11. 86 12. 45
9. 32 9. 13
8. 89 9. 13
6. 36 6. 33
3. 38 3. 31
1 U.S. Department of Labor, Bureau of Labor Statistics Wholesale Prices and Price Indexes, 1963 (Bulletin No. 1513).
Washington: U.S. Government Printing Office, 1966.
2 Calculated.
2 U.S. Department of Labor, Bureau of Labor Statistics: Wholesale Prices and Price
Indexes, 1963 (Bulletin No. 1513). Washington: U.S. Government Printing Office, 1966.
PAGENO="0172"
1.722 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Pharmaceuticals accounted for 0.77 percent of all commodities listed in the
WPI in 1901 and for 0.76 percent of those to be included in the 1964 index. Anti-
infectives, vitamins, and ataractics accounted for 54.50 percent of the ethical
pharmaceuticals listed in 1961 and for 52.94 percent in 1964. Among proprietary
pharmaceuticals, internal analgesics, cough and cold preparations, and antiseptics
accounted for 60.16 and 59.74 percent of the products listed in these same years.
PMA-Firestone fndea
During the 1950's circumstances evolved a need for a better, more precise evalu-
ation of prescription drug prices and price changes. Because these were not re-
ported individually but as part of a larger group in the BLS Wholesale Price
Index, no factual data were immediately available to support the industry's con-
tentions that drug prices were affected by competition and economies in produc-
tion so as to be reduced over time and that they were decreasing while prices of
other nondrug products increased. Therefore, the PMA engaged the services of
Dr. Firestone to construct a realistic index of prescription drug prices.3
So that the index of prices would be current and meaningful, it was based on a
structured sample of prices for the year 1949. The index measures the changes in
prices paid by pharmacies for specialty drugs used in dispensing prescriptions.
It is computed by a method called a "chain index of weighted price relatives."
~Uhis provides flexibility for changing weights as often as may be necessary. New
drugs can be introduced into the index when they become sufficiently important;
drugs declining in importance can be dropped without invalidating the compara-
bility of the index over time, while continually reflecting the true current market
conditions.
The index was set up by listing the five specialties that led in dollar volume in
reach of the eleven years from 1949 to 1959. A. drug was and is maintained on the
list as long as it appears on the market. All dosage forms of a product are repre-
vented by the price of the most widely used dosage form. Prices and sales volume
data are derived from Drug Topics Red Book at its press time each year. A total
of 308 drugs appeared at one time or another during the initial period. From 99
products, the base for the 1949 index, the sample was expanded, in order to reflect
market conditions more accurately, to include 242 products in 1959. Table 4 indi-
cates the relative importance of product classes as it changed over the initial
period.
TABLE 4-PERCENTAGE SHARES OF PRESCRIPTION SPECIALTIES BY PRODUCT CLASS IN THE PMA-FIRESTONE
INDEX
Product class 1959 1949
Product class 1959 1949
Percent Percent
Antibiotics, oral 23 11
Hormones 11 10
Tranquilizers 9 0
Analgesics, narcotics 7 4
Antispasmodics 5 4
Cough and cold preparations 5 6
Vitamins 4 13
Cardiovascular agents 4 4
Sulfonamides 4 3
Sedatives, hypnotics 3 6
Dermatol3gicals 3 4
Percent Percent
Diuretics 3 0
Antihistaminics, oral 2 5
Hematinics 2 4
Central-nervous-system stimulants... 2 4
Antiarthritics, antirheumatics 1 1
--
Total 100 100
- -
Total number of products in base
sample 242 99
-__________________
Note: Data rearranged in 1959 order.
Indecces Show Change
The development and use of indexes has been necessary in all areas of endeavor
in which changes can be measured significantly by quantitative means. The in-
dexes described here have three key factors in common: (1) use of a value repre-
senting the price of a given item precisely defined (physically, chemically, or
otherwise) at a specified point in time; (2) use of weighting factors which recog-
nize the item's impact on the market in relation to all other products; (3) flexi-
bility of the sample to expand or contract, which permits inclusion of new
Firestone, J. 21.: A Price Index for Drugs, Drug & Cosmetic Industry, 87: 470 (Octo-
ber), 1060.
PAGENO="0173"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1723
products and dropping older or obsolete items to maintain a sample that accu-
rately reflects the products in current use.
Each index is based upon a model developed to meet its specific objective.
Because the model has both fixed and variable elements, any of these which re-
quire subsequent adjustment can be changed to correct the index so that it
reflects the actual circumstances in each time period. At the outset, the first
period's data represented the index base of 100. All subsequent values have been
reported in relationship to the base. When conditions change so much that the
base is no longer representative of normal circumstances, complete reevaluation
of the index numbers to form a new base can be accomplished by simple arithmetic
without changing the relationship of the numbers to one another.
The primary focus of such an index is on the prices of the products listed. An
index for one item would consider only the pricing of that product and not how
many units are produced or sold. Since a price index for groups of products must
measure the changes for all the items to show an over-all trend, the quantities of
each item affect the model only in proportion tO the importance (weights) of the
amounts produced or sold. This depends on the objective of the index: to measure
production, consumption, or merely changes in prices without relation to either
production or consumption. In other words, an index can be structured model-
wise to measure any specific kind of change or trend one wishes to measure.
Table 5 summarizes the indexes described abOve, using as a base the average of
values for 1957 through 1959 (the current BLS base time period) as 100. It is
easy to see that the price trends for all consumer items, medical care, and all
wholesale commodities were consistently upward. However, the price trends for
drugs and pharmaceuticals, when measured separately by the Prescription Drug
Subsection of the CPI and by the PMA-Firestone Index or when extracted from
the wholesale group, are definitely downward.
TABLE 5.-GENERAL AND DRUG PRICE INDEXES
11957-59=1001
PMA-
Year
.
Consumer price i
ndex
~
Wholesale
price index
Firestone
index
All items
Medical
care
-
Prescription
drugs
All corn-
modities
Drugs and
pharma-
ceuticals
-~--
Prescription
specialties
1949
83. 0
72. 0
80. 3
83. 5
100. 1
107. 8
1950
83. 8
73. 4
82. 4
86. 8
98. 9
103. 3
1951
90. 5
76. 9
86. 4
96. 7
102. 3
103. 7
1952
92. 5
81. 1
87. 5
94. 0
98. 9
100. 2
1953
93. 2
83. 9
87. 5
92. 7
99. 4
100. 2
1954
93. 6
86. 6
89. 2
92. 9
100. 4
98. 9
1955
93. 3
88. 6
90. 4
93. 2
99. 3
98. 5
1956
94. 7
91. 8
93. 2
96. 2
98. 5
98. 5
1957
98. 0
95. 5
96. 5
99. 0
99. 8
99. 9
1958
100.7
100.1
100.7
100.4
100.5
100.2.
1959
101.5
104.4
103.0
100.6
99.7
99.9
1960
103. 1
108. 1
102. 6
100. 7
100. 2
99. 6.
1961
1962
104.2
105.4
111.3
114.2
99.2
95.3
100.3
100.6
98.3
96.0
96.7
93.9
1963
106. 7
117. 0
93. 0
100. 3
95. 1
92. 9
1964
1965
108.1
109. 9
119.4
122. 3
91.8
90. 8
100.5
102. 5
95.0
94. 4
92.7
92. 9;
1966
113. 1
127. 7
90. 6
105. 9
94. 5
92. 6
Each of the indexes discussed above measures price changes annually and
accurately reflects trends over time. Should the quantities of products used in-
crease or decrease during a year, costs to the consumers would increase or
decrease, but prices would not be affected. If more or less expensive products
are substituted, total costs may change, but prices will not change. If less of a
drug is needed to cure a patient, treatment cost is reduced-but not prices. If
one pharmacy charges full prices while another charges discounted prices, total
sales of drugs or total costs to patients will be affected, but these indexes will
not be affected and will continue to measure only prices in the marketplace.
Unfortunately, this is not true for the average prescription charge. It is affected
directly by every charge made for every prescription. Although the price of
certain tablets may be ten cents each and may remain so for several years, the
PAGENO="0174"
1724 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
dispensing of increasing quantities will cause the average charge to increase.
Table 6 depicts the sharp contrast between the price indexes and an index of the
average prescription charge as reported in the hilly Dige$t.
Although the average prescription charge is an average of prices actually paid
by consumers for prescriptions, Table 6 shows beyond doubt that it is not a
measure of prices, price changes, or price trends at the consumer level. If it
cannot measure or reflect any kind of valid price or charge, and if its index of
change moves countercurrent to true price indexes, it can be a major source of
confusion and result in a liability rather than an asset to pharmacy when it is
taken out of context.
TABLE 6.-THE AVERAGE PRESCRIPTION CHARGE INDEX COMPARED WITH PRICE INDEXES
[1961 = 100]
Year
BLS-CPI
Prescription
drug subsection
BLS-WPI ethical
pharmaceuticals
subsection
PMA-Firestone
prescription
specialties
Lilly Digest
average prescription
charge
1961
100.0
100.0
100.0
100.0
1962
96.1
97.6
97.1
102.2
1963
93.8
96.4
96.1
104.3
1964
1965
92. 5
91. 5
96. 1
95. 4
95. 9
96. 1
104. 9
107. 1
1966
91.3
94.9
95.8
110.5
CONOLTJSION
The average prescription charge is not an indicator of drug prices or of
prescription charges. However, it can be a useful tool to the pharmacist in under-
standing his particular departmental operation. A case in point is that of a
pharmacist who decided to lower his drug prices. After a reasonable length of
time had passed, he was surprised to find that his average charge continued to
rise. This led him to seek the reasons. He found that lowered prices had en-
couraged the prescribing of larger quantities per order. His dollar volume re-
mained the same, but greater quantities per order had decreased the number
of prescriptions and renewals dispensed, which caused the average charge per
prescription to increase.
Many useful ratios are available to indicate standards for or levels of perform-
ance in pharmacies. They depend on data which may vary over time, but they
neither measure nor reflect changes or trends in prices. Their usefulness lies in
calling attention to an area and, when necessary, triggering further investigation,
a review of policies and procedures, and, finally, some specific corrective action
for improving operations.
As a ratio of prescription revenue to the number of prescriptions dispensed,
expressed in dollars and cents per prescription, the average charge has limited
value as an over-all measure of the economics of professional performance. Aside
from this, it has no other useful value in itself, in economic analysis, or in
pharmacy.
Senator NELSON. Dr. Whitney?
Mr. CUTLER. We have just one more witness, Mr. Chairman, and
that is Prof. Simon Whitney, professor of economics at New York
University, who has a statement on the subject of profits and prices
in this industry.
Senator NELSON. Dr. Whitney, we are very pleased to have you
come here today. You have been very patient to wait as long as you
have to he presented. Your biographical sketch will be printed at this
point. Please proceed.
(The biographical sketch follows:)
BIoGRAPHIcAL SKETCH OF SIMoN N. WHITNEY, P11. P.
Present position: Professor of Economics, University College of Arts and Sci-
ence, New York University.
Education: B.A., Yale University, 1924 and Ph. P., Yale University, 1931.
PAGENO="0175"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1725
Academic appointments:
Professor of Economics, New York University, 1907 to present.
Professor of Economics, Rutgers University, 1901-1967.
Lecturer, New York University, 1954-55.
Professor, New York University, 1949-54.
Associate Professor of Economics, New York University, 1948-49.
Dean, Deep Springs (California) Junior College, 1942-48.
Instructor of Economics, Yale University, 1926-28.
Business experience:
Director, Bureau of Economics, Federal Trade Commission, 1956-61.
Chief, Research Division, Twentieth Century Fund, 1954-55.
Associate Economist, Twentieth Century Fund, 1949-54.
Vice President, Econometric Institute, 1949.
President, Telluride Association (educational endowment foundation),
1930-31.
Principal Economics, Office Export Control and Board of Economic Wel-
fare, 1941-42.
Economist, O'Ryan Financial Commission to Japan, 1940.
Associate, Lionel D. Edie and Company (investment counselors), 1936-40.
Staff, Division of Research and Statistics, Board of Governors Federal
Reserve System, 1936.
Senior Economist, Research and Planning Division, National Recovery
Administration, 1934-36.
Assistant Economist, Chase National Bank of New York, 1931-34.
Economist, Antitrust Division, Department of Justice, 1928-29.
STATEMENT OF SIMON N. WHITNEY, PH. D., PROFESSOR, OF
ECONOMICS, NEW YORK UNIVERSITY, NEW YORK, N.Y.
Dr. WHITNEY. Mr. Chairman, in view of the lateness of the hour,
you will want me to abbreviate my talk. I can cut it by about one-third.
Senator NELSON. Thank you. Your statement will appear in full in
the record.
Dr. WHITNEY. Thank you. Then I will just skip here and there, re-
wording slightly.1
Senator NELSON. Fine.
Dr. WHITNEY. I am beginning about the middle of page 1, Mr.
Chairman.
More than one distinguished medical witness has charged that
branded ethical drug prices and manufacturing profits are "too high".
Fortunately, some industry witnesses and those today have anticipated
me in part;, but I shall speak as an academic economist who applies
the established principles of this science to the problems before you.
Economics begins with the truths that goods are not produced with-
out work and progress does not occur without saving and constructive
investment. Experience points to the desire to better one's self and one's
family as the most effective single motive. Reliance on self-interest
also minimizes compulsion, conscript;ion and penalties.
Clearly important to labor, people being eager to go wherever their
wages or salaries will be highest, it is the application to capital that
concerns us here. Self-interest tells people to invest where the out-
look for profits is best-after taking account of risks-and this in turn
is normally where consumer demand is greatest. Our standard of living
is proof that, on the whole, the process has worked. Countries which
condemn the profit motive are glad to draw from us some of the sur-
pluses which it produces.
1 complete prepared statement of Dr. Whitney begins at p. 1740, infra.
PAGENO="0176"
1726 COMPETITIVE PROBLEMS IN THE DRUG ~DUSTRY
How does this apply to the questions before the subcommittee?
Quite simply, drugs are too important to the public welfare for us to
withdraw or dampen the incentives which have created and multiplied
all our goods. Any university teacher of economics denying the con-
tribution of profit in attracting capital-and this might happen, since
a reputation for originality is prized in my profession-should be
asked what alternative system he recommends for inducing owners
of capital to put it where consumers want it and why this alternative
would work better.
A normal profit will not attract investment where it is most needed,
nor will it achieve the second fimct.ioii of profits, stimulating cost re-
duction and efficiency. These essential jobs are done by differences, not
uniformity, in earnings.
How has this worked in practice? The First National City Bank cal-
culated the average rate of return on net worth for all manufacturing
since 1950 as about 65 percent of that for drugs and medicines. As a
result, $3 of additional stockholders' money per dollar of net worth iii
1950 went into drug manufacturing for every dollar into all
manufacturing.
Mr. Gon1~oN. Dr. Whitney, can you tell us how much of this $3
was raised through the selling of new stock issues?
Dr. WHITNEY. Very little. Most of it was reinvested profits. If
profits had not been made, they would not have been there to invest.
Mr. GORDON. They would have had to go to the capital market,
would they not?
Dr. WHITNEY. If they had had poorer profits, they could not have
raised the money in the capital market.
About 25 cents additional went into mea~packing, the least profit-
able major manufacturing industry reported. Comparing drugs with
meatpacking, 31/3 times the rate of earnings brought 11 or 12 times the
rate of investment. The free market process-consumer demand, prof-
its, attraction of capital-has worked in drugs brilliantly.
Mr. GROSSMAN. Dr. Whitney, may I ask you, you seem to show a
great deal of concern here for the investor. At the same time, you
mention on page 2 of your original statement, quite simply, that drugs
are too important to the public welfare for us to withdraw or dampen
the incentives which have created and multiplied all our goods.
I spoke on this question earlier, this morning. What do you consider
the elements of social responsibility of the drug industry? In other
words I am concerned about the investor and the need for the investor
to build the industry, but are we not forgetting the other side of the
corn?
Dr. WHITNEY. The primary responsibility of the industry is to pro-
duce the drugs that are vital to our health. 1 believe they are doing it,
are they not? That is why we are investigating their practices.
As for the investor, I am not one myself. I have refused to invest in
this industry. I have considered it too risky. But the investor is im-
portant, because only through him can we get the industry going
and producing the necessary drugs.
Mr. GRossMAN. You say on page 2, "But such a profit will not attract
investment where it is most needed."
Dr. WHITNEY. That is an average profit.
PAGENO="0177"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1727
Mr. GROSSMAN. In other words, if we were to regulate. the drug
industry, you would be able to say assuredly that we would have a
serious diminution of investment, is that correct?
Dr. WHITNEY. Yes, if you reduce their profits, you certainly will.
They are the No. 1 industry in profits, in my opinion, over the last 17
or 18 years. The FTC just reported that the other day. Although they
have been criticized this morning or this afternoon, still they have
issued that, and they have been the No. 1 industry in attracting capitaL
Mr. GRoss~1AN. I do not pose to be an economist, but I wonder, on
page 2 also, whether you really believe-you said "Adam Smith spoke
of `an invisible hand' by which the search for profit leads to the gen-
eral good, and also wisely said that we expect our dinners from the
self-interest, not the benevolence of butchers and bakers." Do you still
believe that?
Dr. WHITNEY. Did you understand how I explained that? I will
give it to you again in different words. A higher profit, where consumer
demand is greater, attracts capital into making that product and sell-
ing to consumers. Thus the consumer gets what he wants Now, is it
clear?
Mr. GROSSMAN. He gets what he wants. He gets it at a better price,
you would say?
Dr. WHITNEY. The only alternative is to have perhaps the Govern-
ment make the thing. Under a free system-we have found that a free
system works well. Under our free market system, he gets it by paying
enough to induce the capital to go in. He pays more where he wants
the product most.
Mr. GROSSMAN. I take it you would agree with, I think, Mr. Plotkin,
who said earlier that the FTC reports give an improper image to the
profits of the drug industry?
Dr. WHITNEY. Oh, no. I am not saying that. I did not understand
all of `the argument to and fro, because some of it was put forth in a
low voice. I do not want to be in the position of criticizing the FTC
report.
Mr. GROSSMAN. In other words, they, say the drug industry is the
most profitable industry.
Dr. WHITNEY. Their figures show th~ year 1966, the latest figures,
that this is the most profitable. I ran that back a few years and found
that compressors and pumps and the ,` auto industry, were running
along on that level, too.
Mr. GROSSMAN. You say on page 3 `that new capital will avoid an
area with the continuing risks of this industry. Can you tell me where
you see the continuing risks? I know we have heard from other people
where they see continuing risks, but I wanted to get your opinion.
In other words, an industry that has made money over the last 16
years, where are the continuing risks? When does a risk become a con-
tingency and when does a contingency become so infinitesimal that we
are talking about a straw in the wind?
Dr. WHITNEY. It is that straw, whatever it is, that has kept people
like me from investing in the drug industry. `V\Te are afraid that al-
though science in the last 20 years has discovered wonderful new drugs,
it may not do it for the next 20 years. I can go on from that, but that
will be one point..
81-280-68-pt. 5-12
PAGENO="0178"
1728 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. GROSSMAN. Do you feel this is much more so than in comparable
industries? Is it not the same thing? Somebody could find out that if
I drink a Coca-Cola I am going to have gout, and the whole Coca-
Cola Co. goes down the drain. And they have only one product.
Dr. WHITNEY. That is a possibility. Do you consider that a serious
one?
Mr. GROSSMAN. I consider it as much of a risk as another industry
that has made huge profits for 15 years.
Dr. WHITNEY. The way I see it is the industry that invests the most
heavily of any industry in research and development incurs thereby
a tremendous risk. Their profits and so on are being risked in research
and development year after year.
Mr. GROSSMAN. Is there not anything that industry can do, for ex-
ample, to act to prevent these things from occurring? Are there any
steps that the industry could take, in terms of reserves against these
risks, that they could take to lower the risk, whatever the risk is?
Dr. WHITNEY. You mean, for example, if they reduced their dlvi-
*dends, put their money in reserve and if 5, 10, 15 years from now, they
stop being productive, they would pay the dividend out of that?
Mr. GROSSMAN. I am not suggesting-
Dr. WHITNEY. You are suggesting reserves.
Mr. GROSSMAN. I am asking is there any way to cut down on these
risks?
Dr. WHITNEY. How can you cut down on the risks of research?
That is something for the scientists to do. They do their best to get
results. They cannot guarantee them.
Mr. GROSSMAN. Are these risks not borne out by profits, then?
Dr. WHITNEY. Oh, I think the profits-we do not know. In the long
run, we will not know until the end of the 20th century. The profits
are high, but the risks are also high.
But by point is not that the risks are so high. This is my pred-
ecessors' point.
Mr. GROSSMAN. When does risk become contingency; when does
contingency become infinitesimal?
Dr. WHITNEY. In the railroad industry, I think there were no big
railroad failures in the 1900's. The New Haven failed in the 1910's, the
Milwaukee failed in the 1920's, and then only in the 1930's did it
appear wl1at a bid situation railroads had been in for decades, and
they had been considered a prune investment.
Mr. GROSSMAN. In the report we have been looking at earlier,
Searle reported profits of at least 20 percent since 1950. I do not see
the 1966 figures. You say you are not an investor. Would you be afraid
to invest in that company because of the risk?
Dr. WHITNEY. I certainly would.
Mr. GROSSMAN. Would you be afraid to invest in General Motors
because of the risks?
In other words, you are not an investor, I take it.
Dr. WHITNEY. You are wrong in that. I have not considered the
General Motors risk so severe as the drug company risk. It may be as
*severe; I have not considered it so. If the risks were as mild as you
make them out to be-
Mr. GROSSMAN. I face risks every day. Don't we all? I want to
know how n~uch more this risk is than the risk of walking across the
:street.
PAGENO="0179"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1729
Dr. WHITNEY. No one knows. It depends on what the investors
decide. If they are willing to put in the money, then you may assume
that the profit is covering the risk. If they are not willing to put it in,
you may assume that the profit is not covering the risk. The money has
gone in, but that is at a level of profit that has seemed to justify it.
My point, in any case, is not that the risks are high. My predecessors
have talked about that. But at only this profit level, whether you justify
it or not by risk, it is bringing in the money, and the money must go
in for the benefit of the health of the future.
Mr. CUTLER. Mr. Grossman, you picked a good example in Searle,
because Searle's income is concentrated very, very heavily in a partic-
ular type of oral contraceptive drug. As you know, oral contraceptives
and contraceptives in general are a field in which many, many drug
companies around the world are operating, and the possibility of an
entirely new method invented by someone else is quite strong.
Mr. GROSSMAN. Do you have any idea when this product was de-
veloped? Was it developed as far back as 1950?
I do not know. I am asking you.
Mr. CUTLER. I do not know the answer to that, either.
Mr. GROSSMAN. I believe it probably had not been.
But one point that I was looking for is, is there any attempt on the
industry's part, perhaps, to diversify their products so they would not,-
would not that cut down the risk?
Dr. WHITNEY. That would cut down the risk, and the investment
in lifesaving drugs. That is exactly the action we want to avoid.
Do you not want to avoid this industry putting its money into some-
thing else instead of into drugs?
Mr. GROSSMAN. 1 want to avoid this fantastic conjured-up risk.
Dr. WHITNEY. You are thinking of the investors. I am thinking of
the health of the people. Do you want the money to go into drugs?
Mr. GROSSMAN. Just `finally, did you or did you not answer when
the risk becomes a contingency. Is there any line that you see?
Dr. WHITNEY. No.
Mr. GROSSMAN. You keep giving me: answers, but I would like that
one just specifically. When is a risk no longer a risk?
Dr. WHITNEY. The railroads prospered for 30 years and it turned
out there was a risk after all underlying them. You cannot tell when it
will occur.
Mr. GROSSMAN. So we just have to wait and see?
Dr. WHITNEY. You can sit around waiting. But I want to urge
you not to throw a damper into the investment of mon:ey into research
in drugs and to force companies to diversify out of a field they are
doing so much good in. True, :they are making profit, but they are
saving lives.
Senator NELSON. Has anybody suggested forcing them?
Dr. WHITNEY. Yes; Mr. Grossman has suggested that they should
diversify out of drugs into some other less difficult lines.
Mr. GROSSMAN. I think I suggested just diversifying t'heir products.
That is all I said. I said cannot they diversify so that they `do not `have
one product that they rely on.
Dr. WHITNEY. Oh, into other drugs. All of them are trying to. They
are `still running the risk's.
Mr. GROSSMAN. Would this not lessen the risks?
PAGENO="0180"
1730 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. WHITNEY. They have tried that already. They do not need our
advice to do this. They are all struggling to do this. Parke, Davis
just led off this week-
Mr. GROSSMAN. You really think they are struggling?
Dr. WHITNEY. And you think it isso easy.
Mr. GROSSMAN. I certainly do not.
Dr. WHITNEY. Then we agree.
Mr. GROSSMAN. I know it is a struggle.
Dr. WHITNEY. I shall go on.
If profits are reduced to average, either new capital will avoid an
area with the continuing risks of this industry but no chance of a cor-
responding profit, and managements will diversify out of ethical drugs
and into products of less importance where profits can be earned with-
out drawing public or congressional resentment, or human nature
will be found to have changed and people to be employing their sav-
ings with the primary purpose of helping their fullow men. I would
expect the first.
What have the drug industry's reinvested profits accomplished?
First let me cite two statistical series. The public demand for durgs
has been met with a rapidly rising physical volume of production.
Consumer purchases of ethical plus proprietary drugs and sundries-
an increasing proportion, now over 60 percent, being ethical drugs-
expanded by 5.5 percent annually from the first postwar year, 1946,
to 1966, as compared with only 3.7 percent for all other consumer
expenditures.
The second series, research and development expenditures by the
industry, advanced from $50 million in 1951 to $416 million in 1966-
faster than either profits or net worth. Total B. & D. for 1957-66
seems to have been greater than dividends paid.
In 1965, according to the National Science Foundation, 95 percent
of this R. & D. was ~nanced privately, as against only 45 percent for
all of industry.
Many hundreds of new drugs, as documented by earlier PMA wit-
nesses, resulted from this profit-motivated research. What they have
done for health need not be detailed to this informed body. As an
economist, I shall merely mention that the monetary value of working
time saved through improvement of physical and mental health and
lengthening of life is probably in the billions.
We imperil further gains if we strike at so-called "excess" profits
which, in fact, have furnished the driving force. For a mere 6 cents
in dividends per dollar of sales of all types of products made by com-
panies classified in this industry, or less than a cent and a half in
dividends for each dollar of total medical care costs, the industry has
done its essential job. This is not primarily one of basic pharmaceutical
and biological research-though it happens to lead all industries in
percentage of B. & P. going to basic research-but one of discovery
and bringing into use of new medicines. Eliminate 2 of the 6 pennies
in dividends-make them average for all manufacturing-and investor
enthusiasm will decline to average also. Does anyone really expect
enough phychological satisfaction from insuring that this one industry
earns only average returns to warrant endangering future investment
in the search for cures of our dread diseases?
PAGENO="0181"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1731
It is implicit in my argument that prices are not "too high." I
shall review briefly what seem to be the three most common reasons
given by those who contend that they are: the size of profits, the con-
trast between some branded and unbranded drug prices, and in-
ability of low-income groups with chronic illnesses to pay.
1. To call higher than average profits a proof of excessive prices
overlooks the economic functions of prOfits, as I have already de-
tailed. It forgets also that in a free market earnings tend to vary with
risks. When privately financed R. & D. is several times that of other
industries in relation to sales, and when drugs become obsolete so fast
that over half the 1966 prescriptions were written for drugs not known
in 1956, there is plainly a high built-in risk factor. Any search for
the new is more hazardous than production of the old.
According to economic theory, a short to medium term higher-than-
average profit is understandable; but entry of new capital will even-
tually bring it to a long term level which is commensurate with risk.
In this industry, capital has entered. Rising consumer demand,
stimulated by the industry whenever it develops a new and effective
drug, and the inherent risks of research which have prevented even
more money from entering, are sustaining its earnings.
2. Prices of certain drugs of one manufacturer may be higher than
those of another producer which are at least theoretically chemically
equivalent. Prices could hardly be the same when one company must
cover the costs, and be rewarded for taking risks, of research, testing,
quality control, and original distribution, and when it makes all dos-
age forms available where and when needed; whereas the maker of
the other drug may only manufacture,~ and then probably only the
most profitable form.
Senator NELSON. On that exact point, that one company must cover
the costs of research, testing, quality control, and original distribu-
tion. Maybe you can answer the question I asked earlier about Thora-
zine, where research was done by Rhoné-Poulenc, a French firm, and
they licensed Smith Kline & French in this country, and they ii-
censed Bell-Craig in Canada. The firm in Canada sells it to the Gov-
ernment for $2.60 a thousand tablets. In the United States it is selling
for $32.60 a thousand.
As an economist, can you give me an explanation for this dramatic
difference? Can you explain why 15 times as much, almost, is charged
here for a licensed product in which neither the company here nor
the one in Canada did the research?
Dr. WHITNEY. Is that a fact, that Smith Kline & French did no
research on this product?
Senator NELSON. We do not have any evidence of research other
than the submissions for the New Drug Application.
Dr. WHITNEY. I happen to have read in the Kefauver hearings
about the work they have done. Perhaps not the volume that Mr. Gor-
don has, but elsewhere.
Mr. CUTLER. I did have a chance at recess to check with one of the
SFK people and what I said was a fact, that SFK did undertake a
very large proportion of the research wOrk necesary in proving Thora-
zine valuable in treating some forms of mental illness, and did under-
take to obtain the right to manufacture them and in prosecuting their
approval in this country.
PAGENO="0182"
1732 COMPETITIVE PROBLEMS IN THE DRUG TNDUSTRY
Senator NELSON. They did more of manufacturing in this respect
than the Canadian company; did they not?
Mr. CUTLER. I do not believe-the Canadian company did not even
begin until Smith Kline & French had done all the work and the
product was proven a success.
Mr. GORDON. Here is testimony from Dr. Hans Lehman, on page
9024, part 16, of the Kefauver hearings. Now, on the top of page 9025,
Dr. Lehman says that Rhone-Poulenc came up with such a drug and
that was Thorazine, which he said was used in anesthesia. A year or
two later, the same drug was used in the treatment of mentally ill
people, and doctors were very excited because it had these drug-pro-
ducing properties and they wanted to see what it would do in people
who needed to be sedated.
Then Senator Kefauver asked him if he and another physician at
Verdun Hospital were the first to put the drug to this use on the
Continent, to which Dr. Lehman answered, "That is right."
Then Senator Kefauver asked how much later it was when son~e-
thing was done about it in the United States. Dr. Lehman answered
that they had been working on it about the same time here, but not
as systematically, and not in this way. I have been asked, he said,
why we had to jump from the United States to Canada and why the
first English-speaking article came from us and not from England,
the people there being so much closer to France, and so on.
It seems to be pretty clear from this testimony before the Kefauver
committee, that it was in Canada that the drug was used for the first
time.
Mr. CUTLER. We are debating a. factual issue. Let us submit to you
Smith Kline & French's reports on the work it has done on this drug.'
Senator NELSON. All right.
Now, how much is charged by Rhone-Poulenc in Paris for that
drug?
Dr. WHITNEY. I do not know that fact.
Senator NELSON. We have a statistic on that. We think it is about
$10. If it is, would you explain the economics of that situation, where
the originating company, which obviously has the patent, charges
less by far-well, I guess we do not have the figures.
If you will submit what the other company did, I would like to get
to this later, because it is a most puzzling situation to me, in any event.
You do not think it has anything to do with the lack of competition
in the United States; do you?
Dr. WHITNEY. You mean there is more competition in France?
Senator NELSON. It is $10.80 a thousand in France, where Rhone-
Poulenc got the patent., and it is $32 a thousand here, and $2.60 a
thousand in Canada.
Dr. WHITNEY. The average Frenchman's income is about a third
or less than a third of the average American's income.
Mr. GORDON. You say, then, there should be no relationship to costs?
Senator NELSON. Excuse me. I have st.ated this incorrectly; I am
sorry.
The price to the druggist. in Paris for one hundred 25-milligram
tablets is $1.08. That would be $10.80 a thousand. That is in Paris.
1 See Appendix III beginning at p. 2129, infra.
PAGENO="0183"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1733
The price in the United States is $57.58 a thousand to the druggist.
Now, as an economist, could you give me any idea why in this coun-
try, where the research was not done-Rhone-Poulenc obviously did
most of the work and got the patent-it sells for $57.58 a thousand,
and in Paris, it is $10.80 a thousand?
Dr. WHITNEY. If all goods were priced according to cost, you could
throw my paper in the wastebasket. The law of attraction of capital
by profits would not be allowed to apply. You have to direct capital
where to go.
Prices are not normally set according to costs. That would be a dif-
ferent kind of economic system, one directed by the Government,
perhaps.
Secondly, as to prices between this country and France, most prod-
ucts are different. You do not find the same price here as abroad. Their
income is much lower. Their personal income, I would imagine, is one-
third as much as ours, or less, and the prices are all at a much lower
level there.
Senator NELSON. Well, we have cases, of course, where the drug is
made in this country and shipped to Europe and sold cheaper in all
European countries than it is here. That is not true of the American
automobile overseas.
Dr. WhITNEY. Would you want drug companies to stop exporting?
Senator NELSON. I think you put your finger on it precisely, and we
have raised this question before: There is competition in Europe so
t;hey have to meet the competition. They are licensed exclusively in
this country and they do not have to meet any competition, so they can
charge an exorbitant, a fantastically exorbitant price.
Is that true or is it not?
Dr. WHITNEY. If you reduce the price the way you are suggesting,
to the French level-
Senator NELSON. I did not say to the French level. I just said-
Dr. WHITNEY. Something in between. There is some figure in be-
tween that would put them out of business completely. They alone d'~
something like $20 million in research a year, maybe more. Then there
is some higher price level that would cause them to continue their
work, but gradually diversify out of drugs.
Senator NELSON. When the patent runs out and everybody can
make it at that price, it will come down; will it not?
Dr. WHITNEY. I think so, and that is: one of the real risks that a
company in. this industry has to take.
Senator NELSON. Where is the risk of a company that is licensed
other than their investment in distribution and advertising? They
did not do the research.
Dr. WHITNEY. If the product is replaced, there is a risk. The clin-
ical testing they do is very great, too.
Senator NELSON. Everybody does that.
Mr. CUTLER. Smith Kline & French did it in this country. Rhone-
Poulenc did not conduct the clinical testing in this country.
Senator NELSON. You do not see anything wrong at all in charging
so much in this country, $10.80 in France, and up in Canada, selling
at $2.60 to the Government, not to the druggist?
Mr. CUTLER. I do not draw the conclusions you draw from it.
Senator NELSON. I am trying to find some explanation.
PAGENO="0184"
1734 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. WHITNEY. If they are going to `sell in France at all, they have
to meet the local price.
Senator ~ELSON. So there is no local competition here, and what
you are saying is that because there is no competition in this country,
they can charge what the traffic will bear.
Dr. WhITNEY. They are doing the same thing in France, selling in
France for what they can get in France. You h'ave not told me Smith
Kline & French sells in France?
Senator NELSON. Yes; in France. They cannot do it unless they
meet the competition.
Dr. WHITNEY. It is the same thing in France. Rhone-Poulenc has
the monopoly in France.
Senator NELSON. Has it?
Dr. WHITNEY. Has `their patent expired?
Senator NELSON. I do not know whether it has, whether it is
cross-licensed or not.
I can give you a couple of examples of drugs manufactured in this
country and sold at one-third the price in France and in Geneva after
being shipped over there. What is your explanation for that?
Dr. WHITNEY. They are meeting the local prices. Otherwise they
could not export.
Senator NELSON. And in this country, there is no competition.
Dr. WHITNEY. No, I am not admitting that. On a patent drug,
there is only one product.
Senator NELSON. Right. You are correct.
I just would like to have you honestly confess, as an honest professor,
that the reason for it is a monopoly in this country which allows a com-
pany to charge any price it pleases, and it will charge what the traffic
~will bear. And they will ship a product to a.nother country where there
`is competition which they have to meet. So, the record is clear, they
will manufacture drugs, `and charge an exorbitant price here and then
go overseas and meet the competition.
When I asked one of the manufacturers about this he said, "Well,
the standard of living is lower in Geneva; that is why the price is
lower." I said, "All right, how about Mexico City, where you charge
three times as much as you do in Geneva, Switzerland?"
He said. "I could not answer that, I would have t'o ask the comp-
troller of the company."
What you have here is a price structure that defies explanation
unless you are going to confess that they are monopoly situations,
`where the firms are going to charge whatever they can get.
Dr. WHITNEY. As a lawyer and as `a Senator, you know that the
~patent law intends to give companies 17-year monopolies and they
"do it in order to give them enough profit to induce them to find further
discoveries. You are stating what is expected. I do not think you
should call it exorbitant.
Mr. CUTLER. Mr. Chairman, on your general proposition, some of
the companies publish the proportion of their earnings that come from
their foreign business and the proportion that comes from the domestic
business. Generally speaking, their rate of return on investment on
`their foreign business is more or less the same as their rate of return
~on their domestic business. If they were by your standards charging
PAGENO="0185"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 1735
an exorbitant price at home and a competitive price abroad, that would
not be so.
Senator NELSON. That does not really explain how they will sell
cheaper in Geneva, Paris, Bonn, and Rome, and then turn around and
charge three times as much in Mexico City.
Mr. CUTLER. But they have about the same rate of return on their
foreign business as on their domestic business.
Senator NELSON. That does not explainwhat Mr. Conzen could not
explain that day. I think the fact is that there was competition there
and none in Mexico City.
Dr. WHITNEY. You expect competition when a drug is patented in
this country?
Senator NELSON. I am not talking about that at all.
Dr. WHITNEY. Oh, you are not. All right.
Senator NELSON. I am just saying I think the American public ought
to know that companies who manufacture drugs here can sell them
overseas at a third of the price they sell at to American people.
Dr. WHITNEY. But if the American public then reacts by demanding
that the price be cut by two-thirds here, that will almost destroy this
extremely valuable industry.
Senator NELSON. Well, go ahead.
Dr. WHITNEY. We are talking about generic drug prices. I will skip
a bit.
At the bottom of page 6, I say we do not have published reports on
how much generic firms are making. I say I am willing to assume that
the risk-taking companies earn higher profits than the limited-service
firms and will do so as long as, though no longer than, their research
is productive. One reason I am willing to say it is that it is the very
function of profit that my paper is about.
What does it mean, after all, to say that a price is "too high"? Con-'
sumers make this complaint when they are paying for necessities-
milk at 30 cents a quart, meat at a dollar a pound, or prescriptions at~
$4 or $5. But millions of television sets are bought at $200 or so and~
cars at 10 or 15 times as much. Buyers must think these prices are~
"reasonable" if the product is worth more to them than some other
use of their money, whether for a substitute, an unrelated article, or
an addition to saving. If a $5 prescription, or 6 of them, will keep a
patient from losing a couple of days' pay or spending a night in a
hospital, the price is reasonable. It does not cease to be so because those
whose venture capital developed the drug keep for dividends and re-
investment a 20th of the retail price, which is about average, or a 10th,
or whatever is necessary to bring this drug or a later and better one
into existence. Nor does a theatre ticket, costing what a prescription
would or sometimes more, become unreasonable if, instead of giving
one a seat at a profitless "flop," it is for a "hit" whose backers are
doubling their money in a couple of years.
Mr. GROSSMAN. Dr. Whitney, may I stop you there?
Dr. WHITNEY. Certainly.
Mr. GROSSMAN. I want to read the sentence over again:
If a $5 prescHption, or 6 of them, will keep a patient from losing a couple of
days' pay or spending a night in a hospital, the price is reasonable.
`What is the basis for your economic thinking here?
PAGENO="0186"
1736 COMPETITIVE PROBLEMS IN THE DRUG Th~DUSTRY
Dr. WHITNEY. I would not be here today if I were not able to get
these antibiotics at $4.50 a couple of days ago, which made it possible
for me to come down here.
Mr. GROSSMAN. You say you think these company prices are reason-
able. You did not have any choice. You go in and pay a couple of dol-.
lars, a.nd that is it. I just cannot believe that any serious man would
come here and say that to this committee.
Dr. WHITNEY. Try to open your mind and listen to my case. I had,
2 or 3 days ago, the question of bowing out to keep from testifying.
I would normally let my virus run its course. But in view of the pres-
ence of antibiotics, I made a deliberate choice to spend $4.50 to come
down here. What is so unreasonable about that? I made a deliberate
choice.
Mr. GROSSMAN. Frankly, I think your example is not the norm. I
`think most people do not have the choice. They just go in and get a
prescription; the doctor makes the choice for them.
You said if a $5 prescription or six of them will keep a patient from
losing a day's pay, the price is reasonable. You are just throwing this
open, right. In other words, there is no limitation as far as you are
concerned; the price is always reasonable?
Dr. WHITNEY. Oh, no.
Mr. GROSSMAN. Oh, no?
Dr. WHITNEY. If a person were given $100 worth of antibiotics-
Mr. GROSSMAN. How is he to know?
Dr. WHITNEY. If he were forced to pay $100 in order to save himself
a day's work and a day's work would pay him $20 a day, it would be
better to lose the day's work. You measure what you get.
Mr. GROSSMAN. I just do not think the buyer has that much choice.
I think you are giving him a better position in the marketplace than
he really has.
Dr. WHITNEY. Do you think the buyer is being overcharged for his
product if it keeps him out `of a one-night stay in a hospital?
Mr. GROsSMAN. I think if it is reasonably produced and a reasonable
profit could be elicited for $1 and the product costs $2, he is being
overcharged.
Dr. WHITNEY. What is your standard for a reasonable profit?
Mr. GROSSMAN. I do not think you and I agree on any social respon-
sibility the manufacturer has. I think you are more concerned with the
investor. I underst.and your argument on-
Dr. WHITNEY. I am afraid you do not understand it.
Mr. GROSSMAN. I understand your argument is the consumer is bene-
fited by the high prices because the industry has more money for re-
search `and all this.
Dr. WHITNEY. You see it.
Mr. GROSSMAN. I see what you are saying, but I think we have to bear
a lot to go as far as you do. I do not think we have to bear it. I think
~the statement here takes us way beyond anything I could reasonably
`accept.
Dr. WHITNEY. It is mainly true.
Mr. G-Ross~IAN. in other words, I am just going to set here and wait
for the industry to decide what they are going to charge.
Dr.. WHITNEY. Shall I go ahead, or shall I go on arguing about this?
PAGENO="0187"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1737
Mr. GROSSMAN. I would like also to ask you on this `other point. This
20-year business is another argument `of yours which I cannot, frankly,
follow. Twenty years are too `few, you say. I asked that earlier this
morning. I think you remember.
Dr. WHITNEY. Yes; first I skipped that this afternoon but then I
mentioned the railroads, which were considered a prime investment
for three decades.
Mr. GROSSMAN. Why do these companies not build up reserves, then?
Dr. WHITNEY. What do yon mean by building up reserves? I dis-
cussed that in terms of future dividends-
Mr. GROSSMAN. I am just wondering, in Searle's case, if Searle has
only one product, this oral contraceptive, the consumer seems to be
keeping them in a profitable business by supporting this one drug.
When you say they are struggling to find-
Dr. WHITNEY. I th'ink that `was your word, sir.
Mr. GROSSMAN. "Struggling" is `my word? I think not.
Dr. WHITNEY. Is it not?
Well, go ahead.
Mr. GROSSMAN. All I am trying to say is I do not understa'nd why the
consumer `has to pay because these companies are not diversifying.
Dr. WHITNEY. Diversifying what?
Mr. GROSSMAN. Their products. You are saying there is a risk be-
cause Searle has an oral contraceptive, and if we find something wron'g,
their whole pricing, whole profit structure is going to go tumbling
down. I am saying that since 1950, we have facts to show that probably
10 to 15 years before the oral contraceptive was developed, th'ere was
30 percent profit in 1950. You can `look down your `own chart. The
profits-they are never below 23 percent.
Mr. CUTLER. `Mr. Grossman, you could probably show an even higher
rate of return for Polaroid. Would you say that made Polaroid
excessive?
Mr. GROSSMAN. No~; I think here we go back to the social responsibil-
ity t;hat `Mr. Squibb talked about `last week.
Mr. CUTLER. Let us do th'at.
Mr. GROSSMAN. I say it is relative. I am not saying it is complete;
it is a relative thing.
Mr. `CUTLER. I understand, but your assumption seems to be that a
manufacturer should compute h'is costs,~ then he should add a reason-
able profit that he gets out of the sky, perhaps from a law or something
else, and add that on.
That `is not the way the economic system works. T'hat is `what Profes-
sor Whitney is trying to say to you.
Five years ago, in the Kefauver committee, Professor Rostow, now
the Under Secretary of State, used the example of a farmer who was
able to produce tomatoes at half the price of `his neighbors because he
was more efficient or because his land was richer, or something else.
That did not mean that he would go to the market and s~ll hi's `tomatoes
at the same profit margin that every'body else did; he would sell them
at what they would bring in the market against the competition of
other tomatoes. If you say that just shows a lack of social responsibil-
it.y, we disagree with you.
Mr. GROSSMAN. I would only say we are talking about a different
PAGENO="0188"
1738 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
market. We are talking about somebody who goes out and is told by his
doctor to buy a certain drug and he buys it.
Mr. CUTLER. Every one of these drugs competes with other drugs,
different patented drugs, other generic varieties of the same drug in
many cases. There is no drug that has a total monopoly on the market.
Everything is competing with a variety of other drugs or, as Dr. Whit-
ney said, with foregoing the use of any drug at all. There are controls
on prices created by competition. We are all in favor of having as much
competition in this industry as we can.
Mr. GROSSMAN. I want to a.sk you one last question. Does the eco-
nomic theory that you have been talking about justify the price dif-
ferentials between what the companies are charging to the hospitals
a.nd what the companies are charging to the retail druggists?
I am not talking about foreign sales. I want to know how you, as an
economist, can explain that and whether perhaps we should do some-
thing about the way the Robinson-Patman Act applies to these com-
panies.
Dr. WHITNEY. The Robinson-Patman Act, as I understand it, does
not apply to them.
Mr. GROSSMAN. That is right. Do you think it should?
Dr. WHITNEY. I do not know whether the price you are suggesting
should be higher, if that is your implication; that is what would
happen.
Mr. GROSSMAN. The prices would come up?
Dr. WHITNEY. I assume they would have to come up.
Mr. GROSSMAN. Do you think the companies are making profits on
their sales to hospitals?
Dr. WHITNEY. I think they must be making some. I assume they are
covering just the bare incrementat costs, selling at such low figures.
Mr. GROSSMAN. But you think they are making a profit, then?
Dr. WHITNEY. They may be making `a sma:ll profit covering `their
out-of-pocket costs, and perhaps a little contribution to overhead and
profit. I would not know that. But beyond making their average profit,
I would not know that. But beyond making their average `profit, if
they sold to pharmacies at the prices they sell to hospitals, t'hey would
all be out of business. You recognize that, do you not?
Mr. GROSSMAN. Can you justify the difference in prices?
Mr. CUTLER. Would you explain what you `mean by justification?
Mr. GROSSMAN. Let us talk about buyer. The buyer must think these
prices are reasonable-that is what you said. I go to a `hospital as an
outpatient and I get something for $2. I go to a druggist `and I pay $17.
Do I think `that that is reasonable?
`Mr. CUTLER. The hospital charges you $1 for an aspirin; they pay a
lot `less.
Mr. GROSSMAN. Using the example that I gave, do y'ou think that
that is reasonable? In other words, am I to interpret it-you say the
buyer. You are the `buyer and I am the buyer. Would I think I have
gotten a good deal if I went to the retail druggist and paid $17?
Mr. CUTLER. Let us come to this again, if we can, because I think
you are askking more a legal question than an economic question. If by
"reasonable" or "justifiable," you mean does it meet the Robinson-
Patman standard of differences in the cost of supplying one customer
PAGENO="0189"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1739
as compared to the cost of supplying the other, assuming both are op-
erating in the same market, the answer is probably no. Probably the
companies would have to charge substantially higher prices to hos-
pitals and to the Government, and perhaps a little bit lower prices to
pharmacies in order to average out if the Robinson-Patman Act ap-
plied to sales to hospitals and to the Government. If Congress thinks
that is what ought to be done, let the Congress do. it and the companies
will comply with that.
I assume no one here is suggesting we should raise the price to
hospitals and to the Government. The implication seems to be that
we ought to bring down the pharmacy prices to hospital and Govern-
ment level. If we did that, these figures you have before you con-
clusively demonstrate that for companies earning 18 percent, if they
reduce their prices by two-thirds on 70 percent of their business, they
would be out of business. There is no answer for that that I know of.
Senator NELSON. Go ahead.
Dr. WHITNEY. I first want to make one personal statement about
Mr. Grossman's first having said that I have shown all the way through
this interest in the investor. I am not an investor. I am not such
a quixotic person as to want to help the investor in drugs. I have
never been one of them. My first interest is in the health of the public.
My interest in this industry is as a tool to that end.
3. Although a case can be made for various special provisions for
low-income persons with heavy outlays on drugs, no case can be
made for helping them by reducing the prices to everyone. Milk
is not. sold at a cut price to all so that low-income groups can afford.
moie of this healthy food.
Senator NELSON. Just as an aside, I might say that in the school
lunch program, it is sold at a low price to everybody, for 3 cents a
glass.
Dr. WHITNEY. To make that relevant, to all schoolchildren. You
would not want to put that to the whole population, I take it?
Senator NELSON. No, just to schoolchildren. I just wanted to point
that out.
Dr. WHITNEY. How many dollars and cents would be saved by
consumers if the industry's earnings could be brought down to average
without damaging production incentives,? I have seen nothing specific
on this from industry critics.
Let us look at profit figures for 1966. The FTC-SEC reports
earnings before income taxes on stockholders' equity for all manu-
facturing as 61 percent of those for drugs. The drug profit margin-
a.lso before taxes but on sales, not equity-was 19.7 cents per dollar.
Sixty-one percent of this would mean a return on equity equal to
that of all manufacturing, but the price would be brought down only
7.7 of the 19.7 cents. Now drug manufacturers received perhaps $1.4
billion of the $3.05 billion in prescriptions dispensed by community
pharmacies. Reducing this $1.4 billion by the 7.7 percent would bring
revenues down $108 million. This is 10.8 cents in $3.05, or 12.4 cents
in a typical $3.50 prescription.
Some critics lightly assume that the pharmacist will reduce his
selling price by the same percentage that his acquisition cost falls.
Certainly under the professional fee system this does not happen.
This is not all. First, someone will have to make up the corporate
income tax payments lost and the money now being reinvested by
PAGENO="0190"
1740 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
the industry out of its earnings, unless its expansion is to cease. At
1966 ratios, the 7.3 cents remaining drop to 4 cents-that is, a 4-cent
reduction in a $3.50 prescription. This does not allow for administra-
tive costs.
All right, one might say, why not save the 1 cent, or the 4 cents~
or whatever it is? Certainly, if it costs nothing-but is it a good
gamble to imperil industry incentives for so little?
The critics seize on selected examples of drugs with the greatest
price differentials. But, no economist and perhaps no accountant thinks
that every product sold by a firm, whether department store or manu-
facturer, should carry the sa:me profit margin.
Certain trends are alarming. New chemical entities marketed per
$100 million of B. & D. expenditures, for 1959 throi~gh 19G6, were 32,.
22, 21, 11, 6, 6, 7, and 3, respectively. They may rise again to 6 or 8 as
the industry catches up with the new requirements of the Food and
Drug Administration, but more than 6 or 8 are needed. There may be
no real recovery if the profits from research are threatened.
We all know how fast medical care costs are rising; but prescrip-
tion drugs were only 9.8 percent c~f total medical care expenditures in
1966, down from 11.7 percent in 1957. In fact, drugs through their
therapeutic effects have been alleviating the impact of other costs
which are rising rapidly.
The American economy leads the world. So does the American drug
industry, as recognized by foreign medical auth&rities and as regis-
tered very practically in an annual 12 percent increase of `foreign
sales. It was not the leader until its profit opportunities became' evident
after World War II. The same economic principles account for the
success of both the economy and the industry.
As new and better drugs are discovered through profit-seeking re-
search, competitive pressures are inevitably brought against existing
drugs. None of us wants to disturb this process. More important than
the price of existing drugs, many of them declining or gradually dis-
appearing, is the health of the future, and this will depend on the
new drugs that may be, indeed must be, found.
Senator NELSON. Thank you, Professor Whitney. I appreciate your
taking the time to come here.
Dr. WHITNEY. Thank you for staying so late, Senator and members
of the staff. I appreciate that.
Senator NELSON. Thank you, Mr. Cutler.
We will now adjourn, subject to the call of the Chair.
(The complete prepared statement of Dr. Whitney follows:)
STATEMENT OF Sniox N. WHITNEY, PH. D., PROFESSOR OF EcoNoMIcs,
NEW YORK UNIVERSITY
Mr. Chairman and members of the subcommittee, I am a professor of economics
at New York University and consultant to the U.S. Economics Corporation, a
firm of business analysts retained by the Pharmaceutical Manufacturers Associ-
ation to make studies in this industry. I am appearing, therefore, on behalf of
the PMA.
I shall discuss the allegations you have heard that `branded ethical drug prices
a'nd manufacturing profits are "too high." More than one distinguished medical
witness making these charges, which are plainly economic ones, admitted lack
of training in economics. Industry Witnesses have anticipated me in part, hut I
shall speak as an academic economist who applies the established principles of
this science to the problems before you.
PAGENO="0191"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1741
Economics begins with the truths that goods are not produced without work
and progress does not occur without saving and coustructive investment. To
achieve these, experience since the end o~ feudalism, and recently even in
Communist countries, points to the desire to better oneself and one's family as
the most effective single motive. Reliance on self-interest also minimizes corn-
pulsion, conscription and penalties.
Although this incentive is clearly important to labor, people being eager to go
wherever their wages or salaries (after considering other aspects of the job)
will be highest, it is the application to capital that concerns us here. Self-interest
tells people to invest where the outlook for profits (after taking account of risks)
is best; and this in turn is normally where consumer demand is greatest. Adam
Smith spoke of "an invisible hand" by which the search for profit leads to the
general good, and also wisely said that we expect our dinners from the self-
interest, not the benevolence, of butchers and bakers.1 Our standard of living is
proof that, on the whole, the process has worked. Countries which condemn the
profit motive are glad to draw from us some of the surpluses which `this moti-
vation produces.
How does this apply to the questions before the Subcommittee? Quite simply,.
drugs are too important to the public welfare for us to withdraw or dampen
the incentives which have created and multiplied all our goods. Any university
teacher of economics denying the contrLbution of profit in attracing capital-
and `this might happen, since a reputation for originality is prized in my pro-
fession-I hope you will ask two questions: First, is he indeed deliberately re-
jecting the central principles of the science as it is now known? and second,
what alternative system does he recommend for inducing owners of capital to
put it where consumers want it, and why would this alternative work better?
Several persons interested in the present inquiry concerning the pharmaceuti-.
cal industry have granted that this industry should be allowed a normal or
average profit. But such a profit will `not attract investment where it is most
needed. Nor will it achieve the second function `of profits-stimulating cost re-
duction and efficiency. These essential jobs are done by differences, not uni-
formity, in earnings.
How has this worked in practice? From 1950, when analyses of the drug
industry were first reported separately in a leading financial publication through
1965 the rate of return on net worth for "all manufacturing" averaged 65 per-
cent of that for "drugs and medicines." 2 As a~ result, $3 of additional stock-
holders' money per dollar of net worth in 1950 went into drug manufacturing
for every $1 into all manufacturing, and about 25 cents into meat packing, the
least profitable major manufacturing industry reported. Comparing drugs with
meat packing, 314~ times the rate of earnings brought 11 or 12 times the rate
of investment.8 The free market process-consumer demand, profits, attraction
of capital-has worked in drugs, brilliantly.
1 "By preferring the support of domestic to that of foreign industry, he intends only his
own security; and by directing that industry in such a manner as its produce may be of
the greatest value, he intends only his own gain, and he Is in this, as in many other
cases, led by an invisible hand to promote an end which was no part of his intention."
The Wealth of Nations, Modern Library ed., p. 423. "It is not from the benevolence of the.
butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their
own interest." Ibid., p. 14.
2 These are based on arithmetic averages of the annual rates of return given in the
First National City Bank of New York, Monthly Economic Letter, April issues (even-
numbered years being used, since each issue covers two years). For 1966 (not included
because the latest net worth figure is for January 1, 1966) the return for all manufacturing
was 67.1 percent of that for drugs. The rate of return of "all manufacturing corporations,
except newspapers" in 1966 was 61 percent of that of "drugs" in the Federal Trade.
Commission/Securities and Exchange Commission Quarterly Financial Report for Manufac-
turing Corporations. This source was not used for two reasons: it did not separate drugs
prior to the 2d quarter of 1956; and it includes along with the larger drug companies
many smaller ones which do little or no research.
From January 1, 1950, to January 1, 1966, net worth of the drug companies included
in the First National City Bank list increased by 640 percent; that of all manufacturing
by 208 percent; and that of meat packing by 56 percent. Part of the increase was through
companies growing large enough to publish reports and be included for the first time:
drug companies increased from 19 to 39, the total of manufacturing companies from 1763
to 2279, and meat packers from 19 to 27. The additional companies are usually smaller-
thus the April 1962 Monthly Economic Letter added 7 drug companies whose average net
worth for January 1, 1960, was only $15.5 million as compared with $57.5 million for.
the 29 companies previously included. The expansion of small firms to the point where they
issue public reports, the entry of new firms, capital expansion through issue of securities,
cud reinvestment of earnings are all aspects of the inflow of capital. Reinvestment is the
principal source. Without the profits of the drug industry there would have been no money
to reinvest in expansion; had its profit outlook not continued good, earnings would prob-
ably have been used for diversification away from drugs.
PAGENO="0192"
~1742 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
If legislative or regulatory action should force drug profits to the average
level, one of two things will happen. Either new capital will avoid an area with
the continuing risks of this industry but no chance of a corresponding profit,
and managements will diversify out of ethical drugs and into products of less
importance where profits can be earned without drawing public or congressional
resentment; or human nature will be `found to have changed and people to be
employing their savings with the primary purpose of helping their fellow men.
I would expect the first.
What have `the drug industry's reinvested profits accomplished? First let me
cite two statistical series. The public demand for drugs has been met with a
rapidly rising physical volume of production. Consumer purchases of ethical
plus proprietary drugs and sundries-an increasing proportion, now over 60
percent, being ethical drugs-expanded by 5.5 percent annually from the first
postwar year, 1946, to 1966, as compared with only 3.7 percent for all other
consumer expenditures.4
The second series, research and development expenditures by the industry,
advanced from $50 million in 1951 to $416 million in 1966-faster than either
profits or net worth.5 Total R. & D. for 1957-1966 seems to have beeu greater than
~lividends paid.6 In 1965, according to `the National Science Foundation, 95 per-
cent of this R. & D. was financed privately, as against only 45 percent for all of
industry.7
Many hundreds of new drugs, as documented by earlier PMA witnesses, re-
5sulted from this profit-motivated research. Wrhat they have done for health need
not be detailed to this informed body. As an economist, I shall merely mention
±hat the monetary value of working time saved through improvement of physical
and mental health and lengthening of life is probably in the billions.
We imperil further gains if we strike at so-called "exce~" profits which, in
fact, have furnished the driving force. For a mere 6 cents in dividends per dollar
of sales of all types of products made by companies classified in this industry,8
or less than a cent and a half in dividends for each dollar of total medical care
costs,° the industry has done its essential job. This is not primarily one of basic
pharmaceutical and biological research-though it happens to lead all industries
in percentage of B. & D. going to basic research `°-but one of discovery and bring-
ing into use of new medicines. Eliminate 2 of the 6 pennies in dividends-make
them average for all manufacturing-and investor enthusiasm will decline to
average also. Does anyone really expect enough psychological satisfaction from
ensuring that this one industry earns only average returns to warrant endanger-
ing future investment in the search for cures of our dread diseases?
It is implicit in my argument that drug prices are not "too high." I shall review
briefly what seem to be the three most common reasons given by those who con-
tend that they are: the size of profits, the contrast `between some branded and
`unbranded drug prices, and inability of low-income groups with chronic illnesses
to pay.
1. To call higher than average profits a proof of excessive prices overlooks the
economic functions of profits as I have already detailed. It forgets also that
in a free market earnings tend to vary with risks. When privately financed B. & D.
Derived from US. Department of Commerce, The National Income and Product Ac-
counts of the United States, 1929-1965: A Supplement to the Survey of Current Busi-
ness, pp. 48-49. The figures are dollar estimates translated into terms of quantity by
adjusting for price changes, The percentage increases are, of course, compounded annually.
The 5.5 percent would undoubtedly be higher If the expanding distribution of drugs
through hospitals and government purchases were included.
R. & D. from annual PMA surveys. The Increase was 732 percent, compared with 627
percent for net earnings and 564 percent for net worth in the First National City Bank
sample.
6 The PMA R. & D. totals for these ten years amounted to $2,530 million. Drug industry
dividends of $2,660 million were reported in FTC-SEC ip. cit. A substantial part of these
~ilvidends, however, may be assumed to have been earned in proprietary drugs, chemicals
and other products-sales of which In 1966, according to the company reports to PMA,
were about a third of the total sales of these products plus drugs for the FTC-SEC
sample.
PMA Report, Pharmaceutical Industry Research and Development Activity, 1966-1967,
`p. 4 and National Science Foundation, Basic Research, Applied Research, and Development
in Industry, 1965, p. -20.
8 Sales in 1966 -were $7,281 million; dividends, $426 million. FTC-SEC, op. cit.
Private medical care expenditures (exclusive, therefore, of those paid by government)
-were estimated at `$31,250 million in 1966. U.S. Department of Commerce, Survey of
Current Business, July 1967, p. 23.
10 National Science ~~undatIon, op. cit., p. 78.
PAGENO="0193"
COMPETITIVE PROBLEMS IN THE DRUG ]NDUSTRY 1743
is several times that of other industries in relation to sales," and when drugs
become obsolete so fast that over half the 1966 prescriptions were written for
drugs not known in 1956, there is plainly a high built-in risk factor. Any search
for the new is more hazardous than production of the old..
According to economic theory, a short to medium term higher than average
profit is. understandable, but entry of new capital will eventually bring it `to a
long-term level which is commensurate with risk. In this industry, capital has
entered. Rising consumer demand, stimulated by the industry whenever it de-
velops a new and effective drug, and the inherent risks of research which have
prevented even more money from entering, are sustaining its earnings. Twenty
years are too few, in many cases, to deduce absence of risk from continuing
profits: look at what happened to the railroads or, for that matter, meat packing,
sugar and textiles.
2. Prices of certain drugs of one manufacturer may be higher than those of
another producer which are at least theoretically chemically equivalent. Prices
could hardly be the same when one company must cover the costs, and be rewarded
for taking the risks, of research, testing, quality control, and original distribu-
tion, and when it makes all dosage forms available where and when needed;
whereas the maker of the other drug may only `manufacture, and then,, probably
only the most profitable form. Do `the lower prices charged by generic-name manu-
facturers reflect these lower risks and costs, or are the industry critics .right in
implying that they reflect willingness to operate on lower profit margins? Until
a representative sample of such firms publishes its rate of return on sales or
investment we `shall not know the answer.12 In the meantime, I am quite willing
to assume that the risk-taking companies earn; higher profits than the limited-
service firms, and will do so as long as, though `no longer than, their research is
productive.
What does it mean, after all, to say that a price is "too high"? Consumers make
this complaint when they are paying for necessities-milk at 30 cents a quart,
meat at a dollar a pound, or prescriptions at $4 or $5. But millions of television
sets are bought at $200 or so and cars at 10' or 15 times as much. Buyers must think
these prices are "reasonable" if the product lS; worth more to them than some
other use of their money, whether for a substitute, an unrelated article, or an
addition to saving. If a $5 prescription, or 6 of them, will keep a patient from
losing a couple of days' pay or spending a night in a hospital, the price is reason-
able. It does not cease to be so because those whose venture capital developed the
drug keep for dividends and reinvestment a twentieth of the retail price which is
more average, `or a tenth, or whatever is necessary to bring this drug or a later and
better one into existence. Nor does a theatre ticket, costing what a prescription
would or sometimes more, become unreasonable if, instead of giving one a seat
at a profitless "flop," it is for a "hit" whose backers are doubling their money
in a couple of years.
3. Although a case can be made for various special provi'sions for low income
persons with heavy outlays on drugs, no case can be made for helping them by
reducing the prices to every one. Milk is not sold at a cut price to all so that low
income groups can afford more of this healthy food.
How many dollars and cents would be saved by consumers if the industry's
earnings could be brought down to average without damaging production in-
centives? I have seen nothing specific on this from industry critics-certainly
nothing in the majority repori of Senator Kefauver's Subcommittee which de-
manded drug price reductions in 1961.'~ You have heard of l'arge economies made
by hospitals `through purchases of drugs by generic name. Were all purchasers
to do the same, many research-based companies'would be put into serious straits.
1~ In 1966 PMA surveys show 10 9 percent of the domestic pharmaceutical sales dollar
of research oriented firms was allocated to the research, discovery and development of
pharmaceuticals and biologicals. The National Science Foundation shows total industry
R. & D. privately financed, as only 2.0 percent of sales. Ibid., p. 69.
12 39 drug companies in the First National City Bank sample earned $711 million
in 1966 on $3,377 million net worth (again, these are overall company figures, applying to
all products), or 21.0 percent. The FTC-SEC sample, Including these and perhaps as
many small firms, earned $787 million on $3,547 million, or 22.2 percent. The difference,
$76 million on $170 million, is 44.7 percent. Surely the small firms were not so profitable-
but, without knowing the make-up of the samples or carefully comparing the reporting
forms, we can make no affirmation.
ii Administered Prices: Drugs, Report of the Committee on the Judiciary, United States
Senate, Made by Its Subcommittee on Antitrust ana Monopoly. Pursuant to S. Res. 52,
87th Cong., 1st Sess., Report No. 448.
81-280-65-pt. 5-13
PAGENO="0194"
1744 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Let us look at profit figures for 1966. The FTC-SEC reports earnings before
income taxes on stockholders' equity for all manufacturing as 61 percent of
those for drugs. The drug profit margin-also before taxes but on sales, not
equity-was 19.7 cents per dollar. Sixty-one percent of this would mean a return
on equity equal to that of all manufacturing, but the price would be brought down
only 7.7 of the 19.7 cents. Now drug manufacturers received perhaps $1.4 billion
of the $3.05 billion in prescriptions dispensed by community pharmacies.'4 Re-
ducing this $1.4 billion by the 7.7 percent would bring revenues down $108 mil-
lion. This is 10.8 cents in $3.05, or 12.4 cents in a typical $3.50 prescription. Some
critics lightly assume that the pharmacist will reduce his selling price by the
same percentage that his acquisition cost falls. Certainly under the professional
fee system this does not happen.
This is not all. First, some one will have to make up the corporate income tax
payments lost as a result of declining profits of drug manufacturers. At present
tax rates, the 12.4 cents the public is supposedly "saving" will be reduced by 5.1
cents. Second, someone will have to replace the money now being reinvested by
the industry out of its earnings, unless its expansion is to cease. At 1966 ratios,
the 7.3 cents remaining drop to 4 cents-that is, a 4 cent reduction in a $3.50
prescription.
All right, one might say, why not save the 1 cent, or the 4 cents, or whatever
it is? Certainly, if it costs nothing-but is it a good gamble to imperil industry
incentives for so little?
The critics seize on selected examples of drugs with the greatest price differen-
tials. But, no economist and perhaps no accountant thinks that every product
sold by a firm, whether department store or manufacturer, should carry the
same profit margin. Instead of prices for each product which meet the market,
such a system would create a jungle of prices reflecting the costs of each com-
pany-if indeed costs could be determined.
Certain trends are alarming. New chemical entities marketed per $100 million
of R. & D. expenditures, for 1959 through 1966, were 32, 22, 21, 11, 6, 6, 7, and 3,
respectively.'5 They may rise again to 6 or 8 as the industry catches up with the
new requirements of the Food and Drug Administration, but more than 6 or 8 are
needed. There may be no real recovery if the profits from research are threatened.
We all know how fast medical care costs are rising; but prescription drugs
were only 9.8 percent of total medical care expenditures in 1966, down from 11.7
percent in 1957.16 In fact, drugs through their therapeutic effects have been
alleviating the impact of other costs which are rising rapidly.
The American economy leads the world. So does the American drug industry,
as recognized by foreign medical authorities and as registered very practically
in au annual 12 percent increase of foreign sales.'7 It was not the leader until its
profit opportunities became evident after World War II. The same economic prin
ciples account for the success of both the economy and the industry.
As new and bett.er drugs are discovered through profit-seeking research, com-
petitive pressures are inevitably brought against existing drugs. None of us wants
to disturb this process. More important than the price of existing drugs, many
of them declining or gradually disappearing, is the health of the future, and this
will depend on the new drugs that may be, indeed must be, found.
(The charts referred to by Senator Nelson follow:)
`5The $1.4 billion Is 46 percent of the $3.05 billion pharmacy ethical drug sales, as
estimated by the ITS. Department of Commerce.
`5Derlved from the de Haen surveys of new chemical entitles Introduced and the
PMA estimates of drug R. & D. expenditures.
16 Derived from Department of Commerce estimates of consumer expenditures for medi-
cal care and PMA estimates of expenditures for ethical drugs.
17 PMA data. The average annual gain from 1960 to 1966 was 12.4 percent.
PAGENO="0195"
DRUG PRICE COMPARISONS I
.~
Item Highest city price Lowest city price
~
O
lop price to Percent high- Percent trade
New York Defense Veterans druggist est city over name price to ~
City Supply Administra- (Red Bunk) lowest city druggist over
Agency tion 1968 price lowest insti- H
tution price H
L~i
341
357
367 o
816
230 ~.i
313
0)
340
Chlordiazepoxide Cl capsules (25 m~., 500's).. Des Moines, $45 Philadelphia, $18.50 $28.50 $21.75 $45.00 243.2
Meprobamato tablets (400 mg., 500 s) Des Moines, $31.20 Los Angeles, Philadelphia, 9. 45 $2. 30 2. 90 31, 20 330. 2
and San Jose, $9.50.
Sulfisoxazde tablets (0.5 gm., 1000's) Des Moines, $25.30 Philadelphia, $6.90 8.35 8.50 25.30 366.7
Chleramphenicol capsules (250 mg., 100's)_ Indianapolis, $22.50 Philadelphia, $3.75 6.73 4.96 5.41 ~30.60 680
Ampicillin anhydrous or trihydrate capsules Grand Rapids (Mich.), $136.75~ Los Angeles, $59.45 59.34 136.75 230.4
(250 mg., 500's).
Ampicillin oral suspension (250 mg./cc., 80/cc. Wichita and Corpus Christi, Albuquerque, $1.55 2. 70 3. 16 3. 23 4. 85 265. 8
bottle). $4.12.
Nitrofurazono soluble dressing (1 lb.) Cincinnati, Corpus Christi, Nashville, $4.08 1.50 2.30 5.00 340.0
Milwaukee, Minneapolis,
Cambridge, and Wichita,
$5.10.
Irivalent polio vaccine (10-dose package). Albuquerque, $6 Phoenix, $1.02 2.61 6.00 588.2 588 H
lsoniazid tablets (100 mg., 100's) Milwaukee, $0.48 Corpus Christi, $0.07 . 1975 .26 .143 1.14 685.7 1,629 -
Isoniazid tablets (50 mg 1000 s) Grand Rapids $2 60 Detroit Los Angeles and 79 1 30 5 18 329 1 666 ~j
Milwaukee, $0.82.
Prednisone tablets (5 mg., 1000's) Nashville, $18.00 Los Angeles, $4.35 4.58 4.52 4.10 170.00 413.8 4,146 ~
Diphenhydramine capsules (25 mg., 1000's)... Wichita and Winston-Salem, Philadelphia, $3.19 3.10 2.90 5.66 12.84 345.2 443 ~
$10.70.
Diphenhydramine capsules (50 mg., 1000's) - Des Moines, $21.27 Los Angeles, $3.00 3. 25 5 5. 20 4. 39 19. 14 709. 0 709 ~
Reserpine tablets (0.1 mg., 1000's) Grand Rapids, $19.98 Newark, $1.35 .60 1.20 23.50 3,330 3,917
Reserpinetablets(0.25 mg. 5000's) Grand Rapids, $160 Chicago, $2.09 2.95 ~4.50 05.00 188.24 7,655.5 9,007
Dextroamphetamine tablets (5 mg., l000's) - Des Moines and Newark, $22.60 Los Angeles, $0.53 - 57 1. 90 1. 00 22. 60 4, 264. 2 4, 264
Phenazopyridine HCL (0.1 gm., 1000's) Erie, Indianapolis, and Winston- Phoenix, $4.13 4.80 °11.60 32.16 48.00 1,162.2 1,162
Salem, $48.
5 $1 per 1,000.
6 $0.19 per 100. I.'4
$0.58 per 100.
1 All prices are quoted from contracts let in 1967.
2 1967.
3 $0.52 per 100.
$0.90 per 100.
PAGENO="0196"
1746 CO~PETITWE PROBLEMS IN THE DRUG INDUSTRY
PRICES OF CHLORPROMAZINE (THORAZINE) BY RHONE-POULENC'S LICENSEES 2 TO THEIR RESPECTIVE
GOVERNMENTS, 1966-67
Price of U.S. licensee
(Smith, Kline & French)
to U.S. Defense Supply
Agency
Price of Canadian
licensee (Bell-Craig)
to Canada's Department
of Veterans Affairs
Smith, Kline & French's
price to U.S. community
pharmacist
:25 mg. tablets $32.62 per thousand.
100 mg. tablets $48.22 per thousand.
$2.60 per thousand.2
$7.25 per thousand.2
$57.58 per thousand.
$92.64 per thousand.
1 Rhone-Poulenc is the discoverer and patent holder of thorazine.
2 Derived from prices for bottles of 500.
(The report "Risk and Return in American Industry-An Econo-
metric Analysis," submitted by Mr. Irving H. Plotkin follows:)
RIsK AND RETURN IN AMERICAN INDUSTRY-AN ECONOMETRIC ANALYsIs
(By Gordon B. Conrad and Irving H. Plotkin)
This report is rendered upon the condition that it is not to be
reproduced in whole or in part for advertising or other purposes
without the special permission in writing of Arthur P. Little, Inc.
I. SUMMARY
A. PUBPOSE AND SCOPE
This study was undertaken for the Pharmaceutical Manufacturers Association
in order to examine possible broad economic expinnations for the levels of profit-
ability of major pharmaceutical manufacturers. These explanations were sought
because of the continuing criticism of the pharmaceutical industry in the press
and by various members of Congress and Federal regulatory agencies directed
at the supposedly unreasonably high prices and profits of this industry. The
concept of the study was to relate levels of profitability among companies and
industries to the amount of risk they have been subjected to over an extended
period of time. In particular, our concept embraces science based industries, in-
cluding the pharmaceutical industry, and examines their usually high risks
which in part result from the high degree of technological innovativeness and
obsolescence characterizing their operations.
Our procedure involved first the development of a theoretical foundation for
describing risk and its association with levels of profitability. Testing these
theoretical constructs then reQuired the devising of quantitative measures of
the risks and uncertainties experienced by companies or groups of companies
PAGENO="0197"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1747
within specific industries. Although some of our analyses involved the stock
market investor and how he might view risk in relation to profitability, our
major concern and emphasis was on the individual company, its management, and
the risk and return to assets invested in specific economic activities. In order
to test the concept that the pharmaceutical "industry" or any other "industry"
might have a characteristic and identifiable risk environment, we grouped com-
panies by "industry" and conducted most of our analyses" on such industrial
units.
The data for these studies have been obtained from Standard and Poor's
Gompustat Tape. The basic time period covered was 1950 through 1965. The
Oompustate tape is essentially composed of audited and published financial infor-
mation analyzed and adjusted by Standard and Poor's to insure comparability.
Appendix B presents the rates of return definitions in terms of the Oompustat
tape, while. Appendix C lists the industries and number of companies used in the
analyses.
B. CONCLUSIONS
1. The theoretical measure of risk developed for this study correlates signifi-
cantly with various well recognized measures~ of profitability for the approxi-
mately 780 companies grouped into 59 Industry's sectors which have been
examined. Exhibits I and II illustrate in graphic form the statistical results
obtained (each dot representing one industry sector).
2. The risk measure we developed appears to be a valid measure of the uncer-
tainty that corporate managements ~face in deciding on the rates of return their
investments must achieve, i.e. the profitability levels' necessary for the enterprise
to survive and grow within the risk environment these managements perceive.
3. The general pattern of risk/return relationship shown in Exhibits 1 and 2
conforms with the expectations of broad'economic theory.
4. Individual industries, as grouped in this study, can be evaluated for their
relative performance with respect to risk versus return, and can be judged on
their conformity to the overall pattern of American industry.
5. The pharmaceutical industry fits welL within the overall pattern of risk!
return relationships for American industry. While displaying a high level of
profitability, the pharmaceutical industry also shows a high level of risk expecta-
tions over the period 1950-1965.
C. REPORT ORGANIZATION
The report is organized along the following lines:
In Section II we explain why questions of consumer price must be ultimately
related to rates `of return, and further why questions of adequacy and excessive-
ness of rates of return must be answered within the context of a risk environ-
ment analysis. Section III together with Appendix A develops in words and
mathematics a basic theory of risk and return. It also describes the types of
statistical problems one is likely to' encounter when attempting to validate such
a thiOTy. .
PAGENO="0198"
C)
INDUSTRY
RATE
of
RETURN
(%)
Industry Return (Net Income + Fixed Charges)/Total Capitalization
Industry Risk Average Inter-com~asy Dispersion
10
20
30 00
SO
EXHIBIT I MODEL I: INDUSTRY R1SK/RET~)RN PATTERN MEASURED AT BOOK VALUE
NOTE: REPRESENTS IS)SITION OF PHARMACEUTICAL INDUSTRY
60
70
80
IP4bUSTRY RISK
PAGENO="0199"
INDUSTRY
RATE ~
RETURN MtMt
~° ~ Risk ~° ~ ~u LBS 2W) 220 200 260 280 300 320 345 360
EXHIBIT II MODEL II: INDUSTRY RISK/RETURN PATTERN MEASURED AT MARKET VALUE INDUSTRY RISK
NOTE: (~) REPRESENTS FOSITION OF PHARMACEUTICAL INDUSTRY
4
PAGENO="0200"
1750 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Having defined our general measures of risk, Section IV defines the specific
`book" and "market" value measures of return that we investigated. In this
section we develop the concept of "Total Investable Funds." This concept facili-
tates our making balance sheet item comparisons among the diverse industries
encompassed in olir study. The results of testing of our basic models are sum-
marized in Section V. A more detailed and technical description of the statistical
analyses appears in Appendix D.
II. MOTIVATION
Economic theory has long maintained that a going concern must reward it~
Investors with a return commensurate with the risks inherent in the particular
business undertaking. Although many elaborate and technical definitions have
been offered for the term "risk," the underlying meaning is generally accepted
to be "lack of predictability." While ruling in the Hope Natural Gas Case, 1944,
the U.S. Supreme, Court stated this fact of economic life as follows:
"From the investor or company point of view it is important: that there be
enough revenue not only for operating expenses but also for capital costs of the
business. These include service on the debt and dividends on the stock. By
that standard the rate to the equity owner should be commensurate with the
returns on investment in other enterprises having corresponding risk. That re-
turn, moreover, should be sufficient to assure confidence of the financial in-
tegrity of the enterprise, so as to maintain its credit to attract capital."
In short, if an enterprise is to attract capital it must offer prospective
investor, especially equity investors, an expected return high enough to compen-
sate for any risks (lack of predictability) that might accompany this expectation.
Until recently there have been two distinct forms of corporate endeavor. The
first and most common type is the private corporation whose product prices are
set in more or less competitive markets. The second type is the natural, regu-
lated monopoly who Is given by government fiat exclusive geographical juris-
diction, and whose product prices (or rates) are set by a regulatory body
(agency). Examples of such organizations are telephone ~companies, light and
power utilities, gas pipe lines, etc.
A critical similarity between these two types of industrial organizations Is
that both raise investment capital from the general public in the national
capital markets. The question is, by what mechanism :does each type of organi-
zation balance return and risk for its investors.
The forces of supply and demand in the capital market allocate to each
private enterprise an amount of capital that balances at the margin prospective
risks and returns from the particular economic activity. It is the consumers'
market place that determines the quantities and prices of the firm's products.
In the case of regulated monopolies the regulatory agenlles seek to establish
a rate schedule that over time will guarantee a return on invested . assets of
some specified, say 6%, rate of return. In setting the price schedules for regu-
lated industries the various agencies are required~ to consider the~ effect of
these rates on the total return of the corporation.
"A public utility is entitled to such rates as:will permit it to earn' a return
on the value of the property which it employsfor~ the convenience of the public
equal to that generally being made at the same time and in the same general
part of the country on investments in other business undertakings which are
attended by corresponding risks and uncertainties; but it has no constitutional
right to profits such as are realized or anticipated in highly profitable enter-
prises or speculative ventures." 1
However, today various government `bodies are attempting a ~tnique form of
regulation. This regulation stems from the desire to control the final consumer
prices of various goods and services.2 Therefore, like the regulatory agency, the
government desires to set the market price of the product. However, unlike
the regulatory agency, the total effect of ,a change In price on the earning
abilities of the corporations is `usually not considered. Changes in price or
rate schedules have immediate effects on the ~total revenue stream available to
the corporation, and both short and long term effects on the corporation's
ability to service its debt and pay dividends to its stockholders. It would be
1 U.S. `Supreme Court, Bluefield Waterwork8 Case (1923).
2Zome have called this type of regulation "Consumerism."
PAGENO="0201"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1751
wrong to assume that one could materially affect the revenue streams of corpo-
iations without affecting the rate of return a corporation is able or expected
to earn
Although final consumer price was the question at issue in the several court
and regulatory cases mentioned above attention was directed toward rates of
return in regulated industries. Questions of "adequacy," "excessiveness," and
"equity" of final price were answered by examination of profit. Under assump-
tions of profit maximizing behavior questions of adequacy, excessiveness, and
even equity kre evaluated in terms of the rate of return and risk environment.
In studying adequacy and excessiveness of rates of return and risks, the courts
and agencies have generally found that the government is not free to change
merely the rates of return of the industries whose prices it seeks to establish.
Unless the government is willing to make certain guarantees of minimum returns
to these new "semiregulated" industries, it will leave the risk environment
unchanged, while usually lowering the rate o~f return by lower product prices.
Such action would result in a marked reduction in the flow of capital to these
corporations and, therefore, in a necessary curtailment of the normal flow of
goods and services to the public.
In summary then, any discussion of prices and exces:s'ivenes!s of prices in the
pharmaceutical industry should be focused on the underlying issues of profit-
ability of this industry, and even more importantly on the relation of this indus-
try's profit level to the risks inherent in its operation. If prices are to be
challenged or if suggestions are to be made by the Government for new price
mechanism, the industry should be in a position to deal with such matters on
grounds that are truly pertinent. What follows in this report is a theoretical and
statistical development of risk/return relationships in American industry which
can be used to place the entire issue of possible price regulation for drugs in the
perspective of:
(1) the Pharmaceutical Industry's position in our economy; and
(2) the relationship of its profit level to its risk environment
III. Paoroszn THEORY OF RISK AND RET[JRN3
Our objective in general terms has been to determine the relation~ship, if any,
between -the levels of profitability and varying degrees of risk experienced in
American industry. The first variable, rate -of profitability (or return) is rela-
tively easy to measure conceptually. As will he explained in Section IV, we have
used a number of book value and market value financial statistics to measure
it. Problems of insuring inter-company accounting comparability were minimized
by our use of the Compustat data, in whose preparation pains were taken to insure
comparability.
The concept of risk, bewever, is a more trotiblesome problem involving seman-
tics. Risk is basically -a subjective phenomenon' `and not as susceptible to precise
or direct measures What we have done therefore is to theorize that certain
objectively measurable concepts are related-tO, and to some extent describe, risk.
Our next step was, by statistical `techniques, to correlate rates of return with
these objective risk measures.4
Two types of problems can arise First of all because of some logical error or
assumption contrary to fact there may be no correspondence between the con
cept of risk and -our objective measures. Our sole technique for dealing with
this possibility has been to express as explicitiy as we can the steps in our logic
and our assumptions so that they may be critiCally examined.
* The second `type -of problem is that `we' can find `a relationship which turns out
to be spurious i e some statistical fluke This type of problem is easier to deal
with, at least conceptually. It is discussed --in `our analysis of the statistical
results below. ` -- `-- `
The basic unit of concern in our risk/return analysis must be the individual
corporation It is within the individual corporation that the balance between
Parts `of this section are based on the* Illustrations given by Paul Cootner and Daniel
Holland In' their study of Risk and Return forthe American Telephone and Telegraph Com-
pany (M I T DSR Project No 9565)
4Economlsts and -financial analysts have' long proclaimed the existence of a "risk
premium." By this phrase they'usually mean that! prospective Investors must be offered
aboye average expectations of return (premium returns) in order to induce them to in-
vest In projects having above average uncertainty (risk) Hence the higher yield on a
corporate debenture than on a government bond is believed to be caused by adding the
appropriate risk premium to the government's (riskiess) interest rate. -
PAGENO="0202"
1752 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
expected returns and expected risks is struck. However, this is not meant to
suggest that the only or best source of information about risk expectation `by
entrepreneurs is historical statistics developed on a company `by company basis.
Surely management forms its risk expectations on the `basis of experience; but
not solely on the experience of its own firm. Likewise, entrepreneurs considering
entry into an industry will assess `the general riskiness of the industry by exam-
ining the range of corporations active (or previously active) in that industry.
It is therefore, not unreasonable to seek a measure of e~vpected risks based on
historical industrial experiences.
It is our contention that returns among industry groups cannot be compared
unless one has a measure of industry risks. The logic of our model of inter-
industry risk/return comparisons follows from our beliefs about the inherent
forces which create risky or uncertain situations within individual industries
and, consequently, within companies.
For the concept of an "industry" to have any validity in risk/return com-
parisons it must be demonstrated that there exists sufficient similarities among
various companies so that they may be meaningfully divided in industry
groupings.
A listing of the various phenomenon that might contribute to inter-industry
difference in basic riskiness would certainly include the following:
(1) Differences in the ease of entry of new firms into the industry or the
ease of construction of new capacity in the industry.
(2) Differences in the income elasticity of demands for the final products
of the industry. (This would affect the response of the industry to general,
economic activity.)
(3) Differences in price flexibility in the industry.
(4) Differences in the stability of major sources of raw materials.
(5) Differences in storability and durability of products and raw materials.
(6) Differences in exposure to foreign competition.
(7) Differences in competition among existing prospective, or potential
new products.
Another major cause of differences in inter-industry risk character derives from
differences in technological and research and development bases of industries'
products. Differences in innovative processes and product obsolescence r'ates are
prime examples.
These considerations and others led us to stratify a number of companies into
various industry groups. As will be seen below, our research has indicated that
a fairly definite pattern of risk/return relationships does emerge from the indus-
try grouping based on SIC stratification.
The reasoning of our model is as follows:
Assume an investor is seeking to enter an industry, i.e., to set up an economic
concern in that industry or reinvest capital in a going concern in that industry,
in such items `as plant expansion, product development, etc. The investor being
reasonably experienced with the economic, financial and production problems
of the industry, believes that he can expect to do as well as anyone else operating
in that industry. He also knows that he is not omniscient, so that while he hopes
on occasion to be more far-sighted, imaginative or effective than his competitors,
he also knows that he is likely to be outwitted, or to outwit himself, or to
run into a number of natural calamities as often as the reverse. His view, in
that case, of the~ risks inherent in operating in the industry probably arise from
observing the impact of errors and advantages upon the rates of return of the
companies already engaged in the Industry. If the impact, on profitability, of
some above or below average `behavior is severe the risk of entry will be lrrge.
If, on the other hand, no reasonably likely error (or action of a competitor) could
push him far from the average rcturn he might assume there was low risk. Thus,
at least in theory, a concept of risk evolves `and as such is subject to statistical
testing. (Of. Cootner and Holland p. 42).
In order to perform these statstical tests it was necessary to construct `a guan-
titive measure of `the industry risk (or uncertainty). We selected `measures of
the dispersion of individual companies' rates of return about their industry's
average rate of return for a given year. An industry which is characterized by
relatively high dispersion of rates of return presents the prospective investor
with much greater uncertainty as to the return he will achieve should he invest
in that industry, than does an industry with low dispersion. Because our theory
of return is essentially a long-run theory, we averaged over the period studies
PAGENO="0203"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1753
(1950-1965) each industry's annual rate of return and dispersion. In this way
our measures should not have been unduly iiifiuenced by abnormal years.5
(It will be readily noted by economists that in measuring risk as the average
intercompany dispersion (or variance) we are, at the same time, staying with
and breaking with standard economic tradition. The word risk is generally used
by economists to represent certain aspects of the utility functions persons are
presumed to maximize in their decision making processes. Specifically, the second
moment of the expected utility function is considered the risk element 8 which
one usually tries to minimize, while trying to maximize the first moment of the
utility function. It is, therefore, natural to measure risk as some type of vari-
ance (i.e., mathematical second moment) ; in so doing we are in keeping with
tradition. However, the expected utility function concerns itself with eco ante risk.
There is no strong reason to believe that temporal variance is a good measure
of this quantity. In fact, when dealing with autocorrelated time series (as
economic series almost always seem to be) we reject the usual reliance on
individual company temporal variance.)
The inter-company dispersion of returns measures one aspect of industry
riskiness. We call it the interspacial component and view it as somewhat ana-
logous with the uncertainty of any one company's market share in a nonregulated
industry. There also exists an intertemperal component to industry riskiness.
This component is analogous with the non~-predictable element of year to year
changes in individual company or industry profitability. So defined it is very
difficult to measure because most economic times series are highly autocor-
related. We have developed some more general intertemporal dispersion measures
and have used them, as far as possible, in our analyses.
Our statistical tests on interspacial dispersion turned out to be significant and
are discussed in full below. There are, however, a number of theoretical prob-
lems which warrant discussion at this point. A few are easily disposed of. One
is the problem of industry definition. The theory depends critically upon the
idea of similarity between the companies assigned to an industry group. The
industry groupings u-c chose are as homogeneous as the SlO based Compustat
tape would allow. One may choose the precise industry composition in differing
ways and thus, because the number value of our basic measure of risk is so
critically dependent on tIle industry grouping, we believe that it was essential
to test the sensitivity of our measure to different industry groupings. It is
reassuring to report that the results were essentially unchanged when we em-
ployed a small number of quite narrow and homogeneous industries,7
A similar problem arises because of the widely different sizes of the firms
that are rightfully grouped in any industry and the fact that the firms used in
our analyses (because of the selectivity of the Compustat tapes) tended to be
the larger and more successful firms in each category. We, at this point, do not
have sufficient data to investigate the effect that this may have on our results
but have sufficient reason to believe that the inclusion of smaller firms would
strengthen the relationship we have found.
IV. DEFINITIONS OF RETURN
In studying the relation between risk and return it is, of course, necessary to
construct quantitative measures of both variables. The measure of risk has been
defined in the previous section as the average interspacial deviation of company
rates of return about the industry's rate of return. As can be seen from the
mathematical deviation presented in Appendix A, the general definition of our
risk measure does not depend at all on the specific definition of the rate of
return. However, care must be taken in defining return, for the logic upon which
the measure of risk is based maintains its economic validity if, and only if, the
return calculated is a true, overall economic rate of return.
Rates of return can be measured either at "book value" or at "market value."
Book value returns relate the yearly income flow as reported on the company
P&L statement to stock Balance Sheet items, such as total assets. Although
book value figures are subject to many imperfections, in the long run they are
the best indicator of real' economic return to invested resources. For a company
6Appendix A presents a detailed mathematical derivation of our measure of risk and
contrasts it with other measures that have been proposed.
0 C. J. H. Markowit8, Port/oflo Beiection, New York: Wiley, 1959.
May also be called cross-sectional or intercompany.
8 Throughout this paper "real" is used in its economic sense, meaning tangible or physical
rather than intangible.
PAGENO="0204"
1754 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
which employed the same level of assets in an industry for an entire year and
whose assets were financed solely by common equity, the ideal book rate of return
would be:
Net Income-Total Assets
However, such a company is quite unusual, for most American corporations
employ a seasonally changing level of assets, which are financed in varying
degrees by both debt and equity capital. Therefore, measuring total assets at any
point in time (and, consequently, at a particular season), one would find some
industries with a higher than average level of assets while others would be below
average due to seasonal factors. Further, net income for highly levered firms
would represent a much smaller portion of the total return to the invested
resources, than would it for an all equity financed company which had no fixed
charges to pay.
To overcome these objections we have developed the concept of total return to
total permanently investable funds. Total return is defined as "Net Income plus
Fixed Charges," while total permanently investable funds is "Total Assets minus
Current Liabilities." It will be immediately recognized that our asset base
measure-"Total Investable Funds" is the familiar Total Capitalization (Com-
mon and Preferred Equity plus Long Term Debt) while our total return measure
is the sum of the returns to both the equity and debt capital suppliers. We prefer
to use the more general terms for they allow us to consider financial as well as
industrial industries in the analyses.
To summarize then, in order to adjust for industry differences in peak-sea-
Sons and financing we have defined the book value rate of return as:
B- Net Income + Fixed Charges
- Total Asset - Current Liabilities
For comparisons we have also considered other commonly used but, for our
analysis, less meaningful measures of book return. They are:
B' - Net Income
- Common Equity
and:
B"- Net Income
Total Assets
B' is objectionable because it considers neither the totality of assets invested
in the enterprise nor the totallty of return. As indicated above, B" is a mislead-
ing measure when comparing industries with different degrees of leverage.
Before defining our measure of market value rate of return it is important to
emphasize one point. When we relate book return to enterpreneurial risks, we
are asking whether resources are being efficiently allocated in the real economy.
However, when we relate market returns to market risks we are concerned with
the efficiency of the capital markets as allocators of financial instruments. A
company may be experiencing monopoly returns (returns higher than justified
by risk) on its book assets (i.e., monopoly real returns) while the holder of its
equity instruments would receive a "normal" return if the monopoly profits
were capitalized when the stock was issued. Although, the relationship between
market return and risk does not directly bear on the question of efficient resource
allocation it is of interest to us. The testing of the relationship offers an addition-
PAGENO="0205"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1755
al test on the viability of our measure of risk and provides insight into the indus-
tries deemed risky by the market. If stock values are ultimately tied to real
economic phenomena, then the potential market risks should be related to real
economic risks.
Market rate of return is here viewed as the total income received by a pur-
chaser of all the securities of a company related to his purhase price, on the
assumption that he sells his holding at the end of a single year. (Other time
spans can reasonably be considered.) Therefore,
M Fixed Charges + Dividends + Change in Market Value
Initial Market Value
where
Initial Market Value = Total Market Value of all debt and equity issues at
beginning of year one,~
and
Change in Market Value = Terminal Market Value less Initial Market Value.
Fixed Charges and Dividends are those actually paid (or payable) by the
corporation which the purchaser would receive (or accrue) in the course of the
year.
For obvious reasons we could not obtain market values for debt instruments
and used Compustat book values for both debt and preferred equity. Oommon
equity is evaluated as the `total market value of the common shares outstanding.
This compromise with the ideal definition does not seem to be of great significance
because of the relative unimportance of preferred stock and the relative stability
of most corporate bonds.
V. STATISTICAL ANALYSES9
The model used to test the industry risk/return relation was of the form:
(I) Industry Return=a+b2 (Industry Risk)
The major question we sought to answer concerned the sign and statistical sig-
nificance of "b2" in the above equation. That is, if "b2" is positive and signifi-
cantly greater than zero we have demonstrated our hypothesized relationship
between industry return and our measure of risk-a relationship implying that
high returns are associated with high risks during the period studied.
Model I was fitted on both book (B) and market (M) rates of return. It is
not necessary to study the detailed statistical analyses appearing in Appendix
D to appreciate the relationship we have found. A simple graph can tell most of
the story.
Oonsidering book return and risk first, we have plotted in Figure 1 (average
book return) on the vertical axis and Risk (average inter-company variance) on
the horizontal axis. For each of the 59 industries studied we have plotted one
point identifying its 15 year average risk/return position. The fact that the pat-
tern of points form an upward sloping line demonstrates the posited relation-
ship between risk and return.
° The following description of the results of our regression analyses is intended only
to summarize the general findings. It is not a substitute for Appendix D which reports in
detail the various models tested.
PAGENO="0206"
Industry Return (Net Income + Fixed Charges)/Total Capimlization
Industry Risk = Average Inter-company Dispersion
INDUSTRY
RATE
of 20
RETURN
10 20 30 30 50 60 70 330 .90 100 110 120 130
FIGURE 1 MODEL I: ~DWTRY RISK/RETURN PATTERN MEAS~ED AT 1300K VALUE ~D~TRY R~K
NOTE: (~)REPRESENTS PoSITION OF PHARMACEUTICAL INDUSTRY
19
PAGENO="0207"
Risk Th~DT~TRY RISK
FIGURE 2 MODEL II: P'~DI~TRY RISK/RETURN PATTERN MEASURED AT MARKET VALUE
NOTE: (~) REPRESENTS POSITION OF PHARMACEUTICAL JNI)USTRY
PAGENO="0208"
1758 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Exact quantification of relationship, as well as assessment of its statistical
significance requires the full regression analysis presented in -Appendix E. From
this analysis we have drawn the regression line fitted to the points. The equation
is:
(TB) Industry Return=8.6+0.038 (Industry Risk)
The b2 (=0.038) coefficient is highly significant (F=50) at 57 degrees of free-
dom. The correlation coefficient (R) is .68 while R2 equals .46.
Figure 2 portrays the same relationship, with Return and Risk measured at
market value.
A glance at: the graph or the equation (1M) reveals much the same story:
(IM) Industry Return=14.4+.007 (Industry Risk)
In this equation b2 is again highly significant having an F test value of 82 with
57 degrees of freedom. The correlation coefficient equals .77 (out of a possible
1.00) while R2 reached 0.59-most impressive for a basically cross-sectional
analysis.
Now that the hypothesized relationship between risk and return has been
statistically valldated, we may turn to questions concerning individual industries
and their relation to the normal risk/return pattern. Given that an industry
has a higher than average rate of return, it is both meaningful and possible to
ask whether its risk is proportionately higher than average risk. For example,
in Figure 1 the pharmaceutical industry has one of the highest book rates of
return as well as one of the highest risks. Using the regression statistics we may
test whether its particular risk/return point could reasonably be generated by
the economic mechanism described by the regression line. Using a two standard
error test-i.e., greater than 95% confidence limits-we can conclude that the
point does belong to the normal pattern.15
In an attemept to examine other dimensions of risk we expanded our original
model to include temporal variances as well as the spacial variance first used to
measure risk. The yearly variance of the industry's rate of return about its
temporal average was added first:
(I') Industry Return=a+b2 (Industry Risk) +b3 (Industry Temporal
Variance)
In both equations I'B and I'~ there is some imphrovement in R2 from 0.46 to
0.51 and from .59 to .61., respectively. This improvement was at the cost of a
reduction in degrees of freedom as well as the value of the F statistic. Including
an average of the individual companies' temporal variance added little or no
independent or partial explanatory power.
Although we are primarily concerned with explaining industry rates of return,
some time was devoted to analyses of individual company returns. The basic
model used was: -
(II) Company Return=a+b3 (Company Risk) +b3 (Industry Risk)
Company risk is the standard temporal variance used by most researchers.
While it yields the expected results (b2 positive and significant) for market
rates of return, Model II yields a significafitly negative value of b2 for book
rate of return. We attribute this negative coefficient to the high degree of
auto-correlation found book value statistics. (In Section III we rejected the
temporal variance as a measure of :risk on theoretical grounds related to auto-
correlation.)
In summary, we have seen that our hypothesized measure of industry risk
has been statistically validated for both book and market measures of industry
10 Mathematically this finding is derived rom the limits of prediction formula (cf. Miller & Frei.inds,
Probability and Statistics for Engineers, p. ~235):
(a+bsro)±t~isS, .[i+1+n(xo-.~2]l/2
where t~/2=2. Setting X~ to 74 we get a y range of 19.4 to 10.2. The observed y value of 17.6 falls well within
this range.
PAGENO="0209"
COMPETITIE PROBLEMS IN THE DRUG INDUSTRY 1759
rate'~ of return Further we have found the expected upward sloping risk/return
relationship and have been able to test the éoñcurrencè of particular industries'
results with the general pattern formed by American industry. The impressive
magnitudes of both our t and F statistics as well as degrees of freedom, permit
us to assert our results with a very high level of statistical confidence; higher
in fact, than is usual in cross-sectional analysis.
APPENDIx A
MATHEMATICAL DERIVATION OF RISK/RETURN MEASURES
Notes
1. Part I of this appendix defines in a general framework the
mathematical and statistical quantities used in this study. The
notation is more general than that used in Part II which defines
the exact quantities used in this analysis.
2. For simplicity the letter "0" is used to represent any return
quantity we are measuring. For example, C may stand for Net
Income/Total Asset.
PART I-GENERAL MODEL
DEFINITIONS AND CALCUlATIONS OF RATES OF RETURNS AND VARIANCES FOR
COMPANIES AND INDUSTRIES
Each company, 1, is uniquely assigned to an industry, j. (1=1, . . ., n;
j=1, . . ., N). The value of any variable for a given company In year t, may
be written:
C~1(t), (t=l, . . ~, T)
C~1(t) is read! the value of variable C for the i" company in the j"~ industry for
year t.
The average value (mean) of variable C for company i is:
~ ~Ci~(t)
The variance of C for company i is:
T rT 12
T~J {C~,(t)]2= I ~ C~,(t)
TTIf1 ~ Lt=i
- T(T-i)
For industry j the average (or `industry') value of variable C in year t is:
~TV11(t)C~1(t)
C.1(t)= L
~ Wi1
i=i
where the W~1(t)'s are the weights assigned to each company. They may be equal
or, if assets are used as weights, unequal.
(Note that a dot "." in place of a subscript~ means that we have summed over
that subscript.)
The variance of C/1(t) is given by:
T
T ~C.,(t)]2_I ~C.1(t)
v - Lt=i
T(T-l)
21-280-68-pt. 5-14
PAGENO="0210"
1760 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
This statistic, V(C.1), is the variance of the industry average over the years
and must be carefully differentiated from V[C.1(t)J. The variance of the com-
panies about the industry average in any one year is given by:
a ~ [C~1(t)]2_[~ Cii(t)]
V[C.1(t)]= ~ n(n-1)
in the equal weight case, and by:
~
i=l -[C.1(t)]2
~ W1~(t)
V[C (t)}- ``
- ~ r ~T~1(t) 12
~ [~w~t~j
in the case of unequal weights.
The difference between V (Ci) and V[C.1 (t)] is quite significant for our
analysis of interindustry riskiness. Year to year profitability for an entire
industry may be quite stable. Yet in any one of those years, the individual com-
panies whose profitabilities are averaged to yield the industry's figure, may be
widely distributed about that mean. (That is, V[C.j (t) I may be very much
larger than V(Cj).) Using the year to year industry variance tends to hide
the important company to company variance. The risk of a change in market
share facing individual companies is not at all reflected in the time series
variance of industry sales of the detergent or automotive industries. Our pri-
mary concern in this analysis is the risk as it appears to an individual company.
Most scholars, on the other hand, have concentrated on the year to year industry
variance.
The V[O.j (t) I calculations will yield a T-element vector, V.~, each of whose
elements is one year's value of V[C.1 (t) 1. The question then arises how best to
summarize this vector (for graphic presentation and interindustry comparisons).
One method is to indicate the range of the elements. It would be preferable if
some scalar could be used rather than two numbers. Should the elements of
V~ prove to be serially uncorrelated, we could construct an estimate of the
variance of the underlying process by taking a (weighted) average of the
elements.
PART IT-SPECIFIC NOTATION
DEFINITIONS AND CALCULATIONS OF RATES OF RETURNS AND VARIANCES FOR
COMPANIES AND INDUSTRIES
The specific quantities used in our analyses are outlined below. It should be
noted that all measures are derived from the single datum point C11, which is
defined in Part I, above, as Cij (t)-the value in year t of any variable for
company i (in industry j).
COMPANY
Company value in t:
Company's average value over time: 1
Company's temporal dispersion: 2
T
V(C,1)=7.~ (C11_V~,)2
t 1=1
1 In Section V this statistic is called Company Return.
2 In Section V this statistic is called Company Risk.
PAGENO="0211"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1761
Company's temporal standard deviation:
=
Company's temporal coefficient of variation:
S/7J~
TYPICAL COMPANY (T.C.)
`T.C. value in year t:
C*t=N ~
i=1
Industry spacial dispersion about C*1 in t:
~ (C~t-C~)~
i i=1
Industry spacial standard deviation about ~ in t:
S2(C1~) =~I[V2(C1~)]
T.C. average value:
-1~;
U*=~ ~~~1C*t
t=1
Average industry spacial dispersion about C* ~:
V~(C~~) =~ ~ V2(C1~)
I t=1 i~
Average industry spacial standard deviation about C~ ~:
~(C~~) ~ ~ S2(C1~)
i £=1 i
Industry's spacial coefficient of variation (about V*):
T.C. average value:
-~
t=1
T.C. temporal dispersion:
V*(C~~)=~ ~ V(C~)
1=1 t
T.C. temporal standard deviation:
N
~ S(C~)
t 1=1 t
T.C. temporal coefficient of variation:
PAGENO="0212"
1762 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
INDUSTRY
Industry value in year t:
N
C.i=~ w~tC~~
i=l
Industry spacial dispersion about C.~ in t:
V1(C~~)=~~ (C~g_C.t)2
I t=1
Industry spacial standard deviation about C.~ in t:
S1(C~1)= i/TT1(C~)
Average industry value:
*C.=~r ~
t=1
Average industry spacial dispersion:
Vi(C~~)=~ ~V1(C~~)
i t=I i
Average industry spacial standard deviation:
~ S1(C~)
i t=1 I
Industry's spacial coefficient of variation:
~iIv.
Industry's temporal dispersion about d:
V(C. ~) =~ ~ (C. ~_V.)2
Industry's temporal standard deviation:
S(C.~) =~/TT(C.~)
t g
In Section V this statistic is called Industry Return.
4 In Section V this statistic is called Industry Risk.
5 In Section V this statistic is called Industry Temporal Variance.
PAGENO="0213"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1763
APPENDIX B
DEFINITIONS OF FINANOIAL RETURNS
The basic data source of our study was the Annual Industrial Compustat Tape
issued by Standard Statistics, Inc., a Division of Standard & Poor's Corp. For
each of the companies on the tape some or all of the quantities listed in Table
B-i are given. Their definitions are those currently used in financial analysis and
may be found in any of Standard & Poor's source books. The rates of return
used in our study have been constructed from the S & P variables and are
defined in Table B-2 in terms of the S & P number used in Table B-i.
The base period for calculating the various quantities defined in Appendix A
was 1950-1965. However, when a needed datum was not available in any par-
ticular year, that measure and all derived measures were adjusted to permit
maximum use of all available.information.
TABLE B-1.-LISTOF VARIABLES APPEARING ON COMPUSTAT ANNUAL INDUSTRIAL TAPE
S. & p. No.
Balance sheet variables
S & P. No.
Balance sheet variables
1
2
Cash and equivalent.
Accounts receivable.
17
18
Nonrecurring expenses.
Net income.
3
Inventories.
19
Preferred dividends.
4
Current assets.
20
Available for common.
5
6
7
8
9
10
11
12
13
14
15
16
Current liabilities.
Total assets.
Gross plant.
Net plant.
Long-term debt.
Preferred stock.
Common equity (book value).
21
~
22
23
24
25
26
27
~
28
29
30
~
Common dividends.
Market value and miscellaneous variables
~
Stock price, high, $1 per share.
Stock price, low, $1 per share.
Stock price, close, $1 per share.
Shares outstanding.
Dividends per share.
Adjustment factor (for changes in number of
shares outstanding).
Shares traded.
Employees.
Capital expenditures.
Income statement variables
*
Net sales.
Operating income.
Depreciation and amortization.
Fixed charges.
Income taxes.
TABLE B-2.-Definitions of variables used in stvdy in terms of
S. & P. numbers.
Symbol
Accounting definition
~
S. & P. code
definition t
B
Net income+Fixed charges
18+15
6-5tt
Total assets-Current liabilities
B'
Netlncome
Common equity
18
11
B"
.
Net income
Total assets
18
6
M
F
Ixed charges±Dividends+AMarket value ftt
15+19+20+iiMVfft
(25*24)+9+14
Market value ttt
See table B-i.
tt 6-5=9+10+11 because Total assets=Current llabilities=Total capitalization.
ttf MVg(25g*24g)+9g+lOg.
iiMVffiVg-MVg-i.
PAGENO="0214"
1764 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
APPENDIX C-INDUSTRIES USED IN THE ANALYSES
SIC No. Industry Total number
of companies
0800 Forest products 7
1000 Metals, miscellaneous 12
1031 Lead and zinc 5
1042 Gold mining
1211 Coal, bituminous 7
1311 Dii, crude producers 9
1810 Motion pictures 8
Food products 43
2000 Packaged toods 9
2010 Meat packers 7
2020 Dairy products 7
2030 Canned toods 7
2046 Corn refiners 2
2070 Confectionery
2082 Beverages, brewers 17
2085 Beverages, distillers 11
2086 Beverages, soft drinks 7
Tobacco 10
2111 Cigarette manufacturers B
2121 Cigar manufacturers 4
2200 Textile products 15
2300 Textile apparel manutacturers 11
2510 Home furnishings 9
2600 Paper 17
2650 Containers, paper 10
2700 Publishing 17
2731 Publishing books 9
2800 Chemicals 43
2830 Drugs 29
2844 Cosmetics 12
2850 Paint
Dii 25
2912 Integrated domestic 18
2913 Integrated international 7
2950 Building materials, roofing and wallboard 10
3000 Tire and robber goods 11
3141 Shoes 8
3221 Containers, metal and glass 7
3241 Building materials, cewent 10
3291 Abrasive products 7
3310 Steel 22
3331 Copper 6
3334 Aluminum
3400 Machinery, metal fabricating 13
3430 Building materials, heating, air conditioning, plumbing 14
3449 Miscellaneous metal work 7
Machinery 62
3511 Stoam generating 4
3522 Agricultural 6
3531 Construction and meterials handling 7
3533 Dii v~ell 6
3540 Machine tools 8
3550 Specialty 16
3560 Industrial 10
3569 General industrial 5
3570 Office and business equipment 14
3581 Vending machines 6
Electrical products 30
3600 Electrical and electronic leaders 6
3610 Electrical equipment 14
3622 Electrical industrial controls 3
3630 Electrical household appliances 7
3651 Radio-RV manufacturers 7
Electronic products 31
3670 Electronics 27
2679 Electronic components 4
3611 Automobile 5
3713 Autotrucks
3714 Auto parts and acceesories 17
3721 Aerospace 17
3740 Railroad equipment 11
3871 Watches 5
4210 Trucking 10
4400 Shipping
4511 Airtransport 13
4830 Radio, TV broadcasters 7
5311 Retail, department stores 16
5331 Retail, variety stores 10
5411 Retail, food chains 21
5600 Retail, apparel chains 6
5812 Eating places 6
Financial 14
6140 Finance 9
6145 Finance, small loans 5
PAGENO="0215"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1765
APPENDIX D
REGRESSION ANALYSES
Section V described two regression models:
(I) Industry Return=a+b2 (Industry Risk) +b3 (Industry Temporal Variance)
(II) Company Return=a+b2 (Company Risk) +b3 (Industry Risk).
As indicated the description was quite abbreviated and did not detail the full
models nor all the relevant statistic. In order to accomplish this, we shall use
the notation introduced in Part II of Appendix A.
Before turning to the specific models studied, let us consider a general four
variable step-wise regression model:
x, =a+b2x2+b3x3+b4x4
In this model x1 is the "dependent variable," while variables x2, x3, and x4 are tue
"independent variables." Regressing x1 on x2 alone would yield the a and b2 coeffi-
cients of the reduced model x1=a+b2x2; regressing x, on both x2 and x8 would
yield a, b23, and b3.2 of the model xl=a+b2x2+b3x3. (The a's might, of course, be
different.) In the following discussion and tables b2 refers to the coefficient of
variable 2 in a two-variable regression, b3 refers to the coefficient of variable 3 in
a two-variable regression (i.e., x1=a+bsx3), b2.3 the coefficient of variable 2 in a
three-variable regression involving x3 (i.e., xi=a+b2x2+b3xs) etc.
Reformulating and expanding the models present in Section V we have:
(I) t~. = a+ b2[Vi(C1~)]+ b3[V(C. ~)]+ b4[V*(C~~)]
i t t
tY~= a+ b2[V(C~~)]+ b3[VT(C1~) J+ b4[V(C. i)].
i t
As will be recalled from Appendix A, ~. measures the (15 year) average return
for an industry, while C~ measures the (15 year) average return for a company.
Model I was fitted with 59 observations (one per industry), while Model II
was fitted with 766 observations (one per company). Each model was fitted for
each of the four basic rates of return-B, B', B", and M (see Appendix B for
definitions).
Tables P-i and D-2 summarize almost all of the statistics for the various
regression steps. In simple regressions (y=a+bx) the F statistic serves as a test
of significance of both b and B2; in multiple regressions each coefficient has it
own F test reported a does the R2.
PAGENO="0216"
1766 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE D-l.-Regression model Ja
V.= a+ b2[Vi(C~1fl+ b3[V(C. ~)1+ b4[V*(C~~)}
i t
Definition of rate of return b
Regression statistic
M B B'
B'
a 14.4 8.6 12.2
5.6
0. 00724 0.08314 0. 00215
0. 1300
F 82.36 49.8 12. 5
25. 6
102 0.59 0.46 0.18
0.31
a 14. 6 9.9 11. 5
6. 5
53 0.00760 0. 126 0.04463
0.213
F 67.0 16.7 6.8
8.3
0.54 0.23 0.10
0.13
a 16. 1 10.7 12.4
6.4
b4 0. 00153 0. 00053 0. 00001
0. 07432
F 35.6 0.7 0.5
7.8
102 0.38 0.01 0.01
0.12
a 14.2 8. 5 11.7
5. 5
52.3 0.00502 0.071 0.00177
0. 11496
F 9.7 32.8 7.1
17.5
53.2 0.00280 0.062 0.05197
0. 10766
102 0.61 0.51 0.21
0.34
F 43. 5 29.43 7.3
14.4
S~ 3.7632 2. 1313 3. 1105
1.7869
Mc 19.0 10.8 12.5
7.2
a 14.1 8.5
5.5
52.34 0.00311 0.07149
0. 12393
F 2.3 32.8
15.9
b3.24 0.00356 0.06065
0. 13769
F 3.8 5.0
2.7
54.32 0.00047 0.0004 (0)
F 2.1 0.8
-0. 92165
0.4
0.62 0.52
0.34
F 20.4 29.4
9.6
Se 3. 7286 2.1342
1. 7966
Mc 19.0 10.8
7.2
a 59 observations.
b See appendix B.
`Decrease in significance if variable 4 allowed to enter regression.
TABLE D-2.--Regression model ~ a
t~~=a+ b2[V(C~4)I+ b3[V'i(C~1)}+ b4[V(C.~)]
t i £
Definition of rate of return b
Regression statistic --
M B B'
.
B'
a 19.3 11.2 13.5
7.5
52 0.00131 -0.00084 -0.00057
-0.01611
F 647. 1 479.6 630. 5
6.9
102 0.46 0.38 0.45
0.0009
a 16.9 8.8 12.4
5.4
53 0.00351 0.04124 0.00090
0.15534
F 61.8 68.9 0.3
86.0
102 0.07 0.08 0.0004
0.09
a 17.7 10.7 13. 1
7.0
0.00341 0.06523 -0.07799
0.11425
F 41.6 5.1 0.9
5.2
102 0.0516 0.007 0.0012
0.0067
a 16.7 8.9 13.1
5.5
bnz 0.00126 -0.00084 -0.00038
-0.02786
F 596.6 551.7 650.8
22.3
53.2 0.00083 0.69116 0.00421
0.17098
F 32.7 117.7 1.2.0
102.9
.102 0.48 0.46 0.46
0.1244
F 353.3 334.9 325.6
55.3
8, 11.4 5. 1 15. 5
3.5
M'~ 21.8 11.0 12. 5
7.3
76 observations.
b See app. B.
PAGENO="0217"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1767
APPENDIX E
SAMPLE OF DATA GENERATED ]~T1OM COMPUSTAT ANNUAL INDUSTRIAL TAPE'
2830 1500 ABBOTT LABORATORIES
2830 26000 AMERICAN HOME PRODUCTS CORP.
2830 26100 AMERICAN HOSPITAL SUPPLY CORP
2830 69400 BAXTER LABORATORIES
2830 72860 BECTON DICKINSON CO -
2830 91000 BRISTOL-MYERS COMPANY
2830 128251 CARTER WALLACE INC.
2830 202400 CUTTER LABORATORIES S
2830 313600 GILLETTE CO. S -
2830 397700 JOHNSON/JOHNSON S -
S - 2830 -406300 KENDALL CO - :
2830 428700 LILLY ELI CO - - S S
2830 464400 MCKESSON & ROBBINS, INC. (MD.)
2830 465700 MEAD JOHNSON/CO S S
2830 471000 MERCK & COMPANY - 5 -
2830 479000 MILES LABORATORIES INC -
2830 . 543200 NORWICH PHARMACAL CO.
2830 565800 PARKE, DAVIS & CO. - S
*-.-** 2830.- 579000 PFIZER UCHASCD &-COCTINCC -. . .
2830 591800 PLOUGH INC
2830 619550 R1CHARDSON-MERRELL INC.
2830 629150 RORER WM H -
2830 643300 SCHERING CORP.
2830 .648000 SEARLE G D CO . S S -
- 2830 665500 SMITH KLINE/FRENCH LABORATORIES INC - .. -
2830 693600 STERLING DRUG INC. S -
2830 704920 SYNTEX CORP. . . S S
2830 755550 UPJOHN CO . S S -. - S
2830 768700 WARNER-LAMBERT PHARMACEUTICAL CO. . . . - S . -
ABBOTT LABORATORIES-DRUGS
- . Year - mv ret
MV
Ern+int
B capit
Earnings
corn eqty
Earnings
assets
1950
1951
-8.287
42.327
-17.432.
1.814
11.131
. 19.575
15.076 .
12.806
13. 063
12.026
19.256
17.476
14.606
15. 051 -
13.852
14.021
10.894
- 9.445
. 9.777
.8.951
1952
1953
1954
1955 .
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
77777.000
.77777. 000
77777.000
.77777.000
.77777. 000
77777.000
.77777.000
77777.000
.77777.000
. 77777. 000
.77777.000
. 77777.000
77777. 000 .
77777.000
77777. 000
77777. 000
77777.000
77777.000
77777.000
77777.000 .
77777. 000
77777.000
15.169 ~.
16. 105
17.434 -
16. 188.
14. 936
. 13. 446.
11.477 -
13.118 . .
14.356
14. 887
15. 159.
9.837
.10. 585
11. 133
10. 707
10. 196
9.151
8.226
- 9.070
9.871
10. 778
11.024
No:years
5 5
1616
-
5.910 14. 509 15. 157 10.229
V/I . 529.167 9.285 3.516 1.745
S/I 23. 004 8. 047 1. 875 1. 321
(S/T)/(I)B 3.892 .210 .124 .129
C(.) 20. 523 17. 524 18. 942 12. 049
C( )B 25 860 17 998 19 701 12 259
1Any designation of 77777.000 IndIcates no data available.
PAGENO="0218"
f
CC~'~'Cl)< .833 = 1.4994 or $1.50 manufacturer's selling price.)
Brand product prices are verified in The Red Book or Blue Book at the net or
wholesale list price level. Products specified by the generic name might be filled
by a popular brand or one of a number of generic or less popular brands at a wide
price range. Therefore, the calculated manufacturers' selling price does not
always coincide with the manufacturer's actual prices since quantity discounts
or deals to the trade are not usually reflected nor does each pharmacy follow an
identical pricing policy. However, the relative positions of the products based
primarily on numbers of prescriptions and also dollar volume permit a valid com-
parative analysis of market shares and long term competitive trends of the
products discussed in this study.
The trends expressed in market shares in most cases are the same for dollar
volume and numbers of prescriptions. Differences in market share between
prescriptions and dollar volume for a product reflect:
1. Its price relative to others in the class;
2. Its potency and the size of the prescription; and
3. The length of time the product is taken
The last point may also reflect the incidence of side effects which might se-
verely limit the continuing use of any of the products. A product with fewer
side effects presumably could be given in larger doses and for longer periods
of time but actually resulting in fewer prescriptions written.
III. ANALGESICS-NON-NARCOTIO
Non-narcotic analgesic is a general classification for pain relievers, not in-
cluded in the morphine class. They are used in a wide variety of medical con-
ditions. The total category has increased approximately 400% in total number
of prescriptions written and over 700% in dollar volume between 1956 and 1965.
Non-narcotic analgesics represented 1.64% of all prescriptions written in 1956
and accounted for 4.75% in 1965. The increase in importance of this class among
all products is due to the development of synthetic types of analgesics which have
reduced the danger of habit or drug dependency developing in conjunction with
treatment. Although 16 products represented three-fourths of the total class in
1965. there were 182 products being marketed which were audited (See `Tables
1A &1B). Over 50% of the 1965 market is represented by drugs introduced since
1956.
PAGENO="0238"
CC_C F\~
0
I ~
=~
CD>
CD
3g~ 02
CD~
g.:2
~ 0
3* 0
~c) ~
C,
CI)
03
~
N_CCDOOCC_)OC_C1CC_C
* CJ000Q1
(710)0) (00 ~ 0000) -.J ~J (71
~.J(Co~N_C~)CJ1 OV.J~0)
* C
~0
C,,
0)
CD
C,
o
CD.
a0
C)
30
CD -~
- C)
C) ~
4~ 0)
C,.) -J
00
0) ~
2~ p0~4P~~J~
c)3 c~::
~ ~ H~HHHH
I I I
I -~
~ 00 ~ ~ 00 -J .-J 0000
~p5°. ~ NPN~
0000O_C0000~~CC_CN.): ~
PAGENO="0239"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1789
TABLE 2-A.-ANTIARTHRITVCS, NONSTEROI DAL
[Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 47. 5 43. 8 36. 4 41. 9 43. 3 35. 9 30. 8 33. 9 28. 9 22. 9
2 15. 0 13. 6 14. 0 11. 2 8. 2 7. 9 8. 5 6. 4 3. 9 4.
3 8.3 7.2 8.2 7.4 6.5 4.7 4.4 5.7 4.5 4.2
4 6.8 5.3 3.4 1.7 .7 .8 .9 .9 1.6 1.2
5 1.6 3.4 3.6 3.5 4.5 5.7 7.0 5.9 9.0 8.4
6 3.8 8.5 5.2 4.3 3.2 5.9 1.4 1.2 .8
7 6. 1 11. 5 15. 6 13. 9 12. 6 15. 5 15. 7 13. 5
8 .1 .5 1.8 5.5 4.0 3.1
9 9.2 14.5 14.1 11.4 10.0
10 1619
Cumulative market
share of above pro-
share of above pro-
ducts 79. 2 77. 1 80. 2 82. 4 83. 2 81. 8 86. 4 89. 3 80. 2 85. 7
All others 20. 8 20. 8 19. 8 17. 6 16. 8 18. 2 13. 6 10. 7 19. 8 14.6
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
TABLE 2-B.-ANTIARTHRITICS, NONSTEROI DAL
[Percent share of market based on dollar volume, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 52.5 46.2 36.3 41.7 41.8 34.3 28.7 32.0 25.7 19.3
2 13.0 11.7 10.9 8.9 6.4 6.0 6.6 4.8 2.9 3.0
3 7.2 6.4 7.2 6.1 5.4 4.0 3.7 4.9 4.0 3.6
4 5.3 4.2 2.8 1.3 .6 .7 .7 .7 1.4 1.4
5 3.1 6.3 6.6 6.5 8.0 9.0 11.6 10.2 14.4 13.3
6 5.4 12.2 6.8 5.9 4.3 7.9 2.0 1.3 .9
7 5.8 11.5 16.0 14.0 12. 1 15.2 14.4 11.9
8 .1 .6 2.8 8.6 6.3 5.3
9 10.3 15.4 14.9 11.8 9.9
10 19.7
Cumulative market
share of above
products 81. 1 80. 2 81. 8 82. 8 84. 2 83. 2 89. 5 93. 3 82. 2 88. 8
All others 18. 9 19. 8 18. 2 17. 2 15. 8 16. 8 10. 5 6. 7 17. 8 11. 2
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
V. ANTnoioTIcs-BR0AD AND MEDIUM SPECTRUM
Broad and medium spectrum antibiotics represent the largest product class
both in number of prescriptions written and dollar volume and are characterized
by a large number of highly competitive products. The broad range of infections
treated by antibiotics and their prophylactic use against reinfection contribute
to the size and growth of this class. It has increased between 85% and 90% both
in prescription and dollar volume over the last ten years, representing 7.3% of the
total number of prescriptions written in 1956 and 7.8% in 1965. Seven major
products used in 1956 accounted for nearly 70% of the class. By 1965, these same
products accounted for only 35% of the prescriptions written. The introduction
of other products (new agents or modifications of existing products) in the inter-
vening ten years were responsible for the dediluse. Over 40% of the 1965 market
was from products introduced since 1956. Thirteen products represented over
76% of the class in 1905 while 84 were identified in the audit (see Tables 3A & 3B).
PAGENO="0240"
1790 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 3-A-ANTIBIOTICS, BROAD AND MEDIUM SPECTRUM
Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962
1963
1964
1965
1 33.7 23.6 10.9 6.8 5.7 5. 1 4.5
4.1
3.7
3.7
2 9.7 10.6 12.9 14.7 14.5 10.0 5.6
4.9
4.9
4.0
3 9.3 5.4 4.5 3.6 5.2 7.4 8.3
8.1
7.8
7.9
4 7.4 5.1 6.3 3.0 1.1 .8 .9
.7
.5
.4
5 6.1 2.7 1.2 .6 .5 .3 .3
.2
.2
.3
6 1.8 1.5 1.9 2.1 1.4 1.6 3.3
4.8
7.6
7.2
7 1.6 5.5 5.0 5.0 4.3 4.9 5.8
5.2
5.1
5.5
8 .3 .2 .1 .1 .2 .3 .6
.8
2.7
5.9
9 14.9 22.7 22.3 14.7 11.6 10.5
9.0
9.4
8.8
10 . 1 2.0 3.7 4. 1 5.3 5. 1
5.0
5.2
4.13
11 .8 7.5 9.6 10.0 8.7
6.9
7.7
7.1
12 3.3 16.6 17. 5 17.6
15.8
15.4
15. 3
13 1.4 5. 1 7.6
Cumulative market
7.5
7.1
6.5
share of above
products 69. 9 69. 6 68. 3 72. 7 79. 3 79. 9 78. 8
All others 30. 1 30.4 13.7 27.3 20.7 20.1 21.2
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0
73. 0
27.0
77. 3
22.7
76. 6
23.4
100.0
100.0
100.13
TABLE 3-B--ANTIBIOTICS, BROAD AND MEDIUM SPECTRUM
[Percent share of market based on dollar volume 1956-651
Product 1956 1957 1958 1959 1960 1961 1962
1963
1964
1965
1 32.8 21.3 9.4 5.5 4.7 4.0 3.8 3.5 2.9 2.9
2 10.9 11.4 13.4 15.2 15.4 11.2 6.8 6.1 6.4 5.4
3 8.8 4.7 3.8 3.0 4.8 7.3 8.0 7.7 7.4 7.3
4 7.5 4.8 5.9 2.5 .8 .6 .6 .5 .3 .3
5 6.1 2.6 1.2 .6 .5 .4 .3 .3 .2 .2
6 2.1 1.7 2.1 2.2 1.5 1.6 3.5 5.0 7.8 7.4
7 1.7 5.5 5.0 4.8 4.1 4.5 5.3 4.7 4.7 5.1
8 .3 .2 .1 .1 .3 .4 .8 1.0 2.6 5.13
9 15. 9 22. 5 21. 9 14. 3 11. 5 10. 1 8. 8 8.9 8. 0
10 .2 2.9 5.1 5.4 6.7 6.8 6.7 6.8 5.4-
11 1.0 8.2 9.3 9.6 8.2 6.4 7.3 6.7
12 3. 4 16.2 16. 8 16. 6 14. 7 14. 5 15. 2
13 1.5 5.6 8.4 8.1 7.5 7.2
Cumulative market share of
above products 70. 2 68. 3 67. 3 72. 5 78. 8 80. 2 79. 2 73. 5 77. 3 76. 1
All others 29. 8 31.7 32. 7 27. 5 21. 2 19. 8 20. 8 26. 5 22. 7 23. 9
Categorytotal 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
VI. ANTIOBIOTICS-PENIOILLINS
Penicillin is synonymous with the dramatic growth in the drug industry and
frequently is regarded as the antibiotic. Penicillins represent a growth of approxi-
mately 92% in the number of prescriptions written and an increase of 108% in
dollar volume during the ten years. In 1965, they accounted for 3.15% of all
prescriptions-slightly higher than their 2.86% share in 1956. Much of the growth
can be attributed to the development of new synthetic penicillins, although some
of the old established products have continued to be prescribed more frequently
than a decade earlier. In 1965, eleven products accounted for almost 84% of the
class while 89 products were reported in the audit (see Tables 4A & 4B). Forty-
three percent of the market was represented by new products introduced since
1956.
PAGENO="0241"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1791
TABLE 4-A-ANTIBIOTICS, PENICILLINS
(Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 19.9 20.8 12.3 5.0 1.5 2.2 1.3 1.2 1.1 0.8
2 17. 1 16. 1 16. 3 12. 5 21.8 24. 5 30. 1 27. 5 24. 2 21. 2
3 14.8 11.6 11.1 8.6 6.3 5.7 5.2 3.6 2.7 1.8
4 13.2 9.9 10.4 9.2 13.5 12.1 12.0 13.9 15.0 14.9
5 8.9 13.7 10.8 7.9 2.9 3.9 4.2 4.0 2.6 1.9
6 3. 0 15. 3 19. 3 18. 9 19. 5 20. 1 21.0 23. 2 20. 8
7 1.0 3.5 2.8 2.7 2.6 2.5 4.0 5.0
8 - .9 5.0 3.7 5.5 4.7 5.1 4.4 4.6
9 .9 8.1 6.3 4.1 4.3 4.0 2.1
10 2.8 6.3
11 4.7
Cumulative market share of
above products 73. 9 75. 1 78. 1 81. 9 79.7 82. 4 84. 3 83. 1 84. 0 83. 8
All others 26.1 24.9 21.9 18.1 20.3 17.6 15.7 16.9 16.0 16.2
Category total 100. 0 100. 0 100. 0 100. 0 100.0 100. 0 100. 0 100.0 100. 0 100. 0
TABLE 4-B-ANTI BIOTICS, PENI Cl LLI NS
(Percent share of market based on dollar volume, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 24.4 24.9 14.1 5.5 3.0 2.1 1.3 1.2 1.0 0.7
2 13. 7 12. 4 12. 7 18. 2 18. 9 21. 2 26. 2 25. 1 22. 2 19. 3
3 16.9 12.4 12.6 8.6 6.5 6.8 5.2 4.3 3.1 2.4
4 9.0 6.7 6.9 6.2 8.9 8.4 8.7 9.8 9.6 9.2
5 10.8 15.7 12.7 8.5 6.4 4.1 4.6 3.9 2.7 1.9
6 4.2 18.9 23.1 21.4 21.4 21.4 22.4 24.7 20.9
7 1.3 4.2 3.3 3.2 3.0 4.0 4.6 5.4
8 1.0 5.9 4.3 6.0 5.0 5.4 4.4 4.4
9 1.1 10.4 8.6 5.4 5.3 4.5 2.7
10 5.2 10.3
11 5.1
Cumulative market
share of above
products 74. 8 76. 3 80. 2 81. 3 83. 1 81. 8 80. 9 81. 4 82. 0 81. 8
All others 25. 2 23. 7 19. 8 18. 7 16. 9 18. 2 19. 1 18. 6 18. 0 18. 2
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
\TII ANTIHISTAMINES
Antihistamines represent one of the more slowly growing therapeutic cate-
gories of the drug industry. Their performance depends on the severity of the hay
fever and cold seasons. Both the effectiveness and the side effects of the various
products vary from patient to patient. Trial of several different products may be
necessary to find the one product most effective with the fewest side effects for
the particular patient and disorder being treated. Dollar volume over the 1956-
1965 period increased 53.5% while the number of prescriptions increased 19.2%.
In 1956 this category represented 4.25% of prescriptions written while in 1965
it represented only 2.92%. Only six products accounted for over 86% of the class
although 62 products were audited in 1965 (see Table 5A & 5B).
Considered a breakthrough in the late 1940's, antihistamines reached a point
beyond which no significant new products have emerged. The leading antihista-
mine products in 1956 continue to be the leading products ten years later. They
are often switched in patients when drowsiness, the most common side effect,
occurs. The leading six products which accounted for nearly 92% of the number
of prescriptions written in 1956 accounted for 78.3% in 1965. The top seven
products accounted for 86.2% of all antihistamine prescriptions written in 1965.
81-280-OS-pt. 5-1O
PAGENO="0242"
1792 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 5-A.--ANTIHISTAMINES
[Percent share of market based on numbers of prescriptions written, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 25. 2 24. 8 21. 5 18. 2 20. 3 20. 7 20. 5 19. 5 22. 8 24. 5
2 24. 4 28. 4 27. 7 19. 8 23. 0 23. 1 23. 8 20. 6 24. 0 23. 0
3 17. 0 16. 0 16. 0 12. 0 10. 1 8. 9 9. 9 8. 4 7. 5 6. 5
4 10.3 10.5 11.0 10.7 9.7 9.5 9.5 10.5 11.0 12.3
5 9.1 12.8 10.0 7.9 7.9 6.3 5.4 5.9 5.9 5.6
6 4.8 6.5 6.6 7.0 6.3 5.9 5.6 5.0 6.6 6.4
7 .3 5.6 6.3 6.1 6.5 9.0 6.2 7.9 7.9
Cumulative market
share of above
products 91. 8 99. 3 98. 4 81. 9 83. 4 80. 9 83. 7 76. 1 85. 7 86. 2
All others 8. 2 . 7 1. 6 18. 1 16. 6 19. 1 16. 3 23. 9 14. 3 13. 8
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
TABLE 5-B.-ANTIHISTAMINES
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 20.6 19.7 16.6 14.0 15.2 15.9 15.6 14.8 17.4 18.2
2 25.7 29. 7 29. 1 20. 2 23.4 23.7 24. 1 20.9 24.6 24. 0
3 16.7 16.9 16.5 11.2 10.0 8.7 9.4 8.4 7.4 6.4
4 16. 3 15. 3 14.8 13. 6 12.3 12. 3 12. 0 13.7 13. 8 15. 1
5 7.7 10.4 7.9 6.2 6.5 5.5 4.9 5.4 5.4 5.3
6 4.8 7.0 7.4 7.7 7.0 6.5 6.1 5.2 7.0 6.7
7 .4 6.2 7.0 6.9 6.8 9.6 6.4 8.0 8.1
Cumulative market
share of above
products 91. 8 99. 4 98. 5 79. 9 81. 3 79. 4 81. 7 74. 8 83. 6 83. 8
All others 8. 2 . 6 1. 5 20. 1 18.7 20. 5 18. 3 25.2 16. 4 16. 2
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
VIII. ANTIOBESITY-AMPHETAMINES
Among the products used to control weight gain and when necessary, encour-
age weight loss are the amphetamines. Included in this product class are com-
binations with sedatives to modify the exhilaration often produced by ampheta-
mines, or tranquilizers to provide relief of tension-anxiety symptoms. The
number of prescriptions written have increased approximately 36% during the
ten year period while the dollar volume has increased approximately 64%. This
increase has been due principally to the introduction of new products through
1960. The total class represented 3.4% of all prescriptions written in 1956 de-
clining to 2.6% in 1965.
Although the eleven products shown in Tables GA & GB represent over 80%
of the market for amphetamine products the importance of this class is reflected
by the number of products which are competing in the antiobesity market. In
1965 over 80 products were being sold by brand or generic name in the ampheta-
mine class.
PAGENO="0243"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1793
TABLE 6-A.-ANTIOBESITY-AMPFIETAMI N ES
[Percent share of market based on numbers of prescriptions written, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 35. 5 35. 2 31. 6 25. 0 25. 7 20. 9 17. 9 13. 0 11. 7 12. 5
2 31. 1 29. 6 33. 9 32. 9 27. 6 27. 5 25. 1 25. 8 25. 3 21. 0
3 3.6 3.1 2.2 5.3 6.8 6.3 7.5 7.8 8.4 7.4
4 2.4 2.8 3.2 2.7 2.3 1.7 1.6 1.7 2.1 2.1
5 1.7 1.0 .7 .4 .6 2.8 2.6 2.7 3.7 5.0
6 1. 4 3. 0 5. 3 8. 4 8. 1 9. 8 10. 0 7. 4 6. 7 5. 4
7 .5 2.2 1.9 1.4 1.2 1.6 3.2 3.1
8 .2 6. 5 7. 3 10. 1 13. 5 13. 8 15. 0
9 .5 1.5 2.4 2.1 2.8 4.5
10 2.9 3.0 1.5 2.8 2.2
11 .7 1.8 2.5 3.7 3.9
Cumulative market
share of above
products 75. 7 74.7 77. 4 77. 1 80. 0 82. 8 83. 2 79. 6 84. 2 82. 1
All others 24.3 25.3 22.6 22. 9 20.0 17.2 16. 8 20. 4 15.8 17. 9
Category total 100. 0 100.0 100. 0 100.0 - 100. 0 100. 0 100. 0 100. 0 100.0 100.0
TABLE 6-B.-ANTIOBESITY-AMPHETAMINES
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 40. 5 38.7 33. 4 25. 9 24. 9 20. 2 17. 1 12.3 11. 0 11. 5
2 31. 8 30. 1 35.3 34.9 27.6 27. 6 24. 6 25. 0 25. 1 20.3
3 2.1 1.8 1.2 3.7 5.4 4.7 5.3 6.0 6.7 5.8
4 1.8 2.2 2.6 2.3 1.9 1.5 1.2 1.2 1.6 1.4
5 1.0 .8 .5 .3 .6 2.7 2.4 2.7 3.9 5.3
6 1.6 3.5 6.3 9.7 9.2 11.1 11.4 7.9 7.0 6.0
7 .5 2.1 1.7 1.3 1.1 1.3 2.6 2.6
8 .2 8.0 9. 1 13.0 17.0 17.2 18. 5
9 .8 2.0 3.0 2.5 3.3 5.3
10 2.9 3.1 1.4 2.8 2.1
11 1.0 2.4 3.2 4.7 4.9
Cumulative market
share of above
products 78. 8 77. 1 79. 8 79. 1 70. 1 84. 1 84.6 80. 5 86. 0 83.7
All others 21. 2 22. 9 20. 2 20. 9 29.9 15. 9 15. 4 19. 5 14. 0 16. 3
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
IX. ATABAXIOS
The term ataraxic is generally used to designate products which are employed
as calming or tranquffizing agents. However, these products may be used to
treat a number of different disorders so that they vary considerably in strength
and action. Our discussion is restricted to those products used to treat patients
residing at home rather than those who are hospitalized in mental institutions.
This classification of drugs represents one of the more important areas of the
drug industry and is one which has grown significantly over the past decade
in both numbers of prescriptions (93%) as well as in dollar volume (100%).
One half of the tranquilizers on the market in 1965 did not exist in 1956. In 1065
these products accounted for 7.4% of prescriptions dispensed as compared to
6.6% in 1956. Twelve products represented almost 87% of the class in 1965 of the
56 which were audited (see Tables 7A & 7B). New products introduced since
1956 account for over 50% of the market.
PAGENO="0244"
1794 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 7-A.--ATARAXICS
[Percent share of market based on numbers of prescriptions written 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 43. 0 39. 5 32. 9 32. 9 27. 4 23. 7 21. 1 17. 5 13. 6 13. 5
2 21.5 15.9 12.3 10.2 8.8 8.4 9.7 7.4 6.8 7.4
3 18.4 14.0 11.8 11.8 12.7 10.4 8.7 7.6 7.3 6.7
4 4.5 5.6 5.2 4.6 3.4 3.6 2.2 2.0 2.0 1.9
5 3.1 4.9 5.4 4.4 2.9 2.5 1.9 3.1 2.3 1.6
6 6 10.3 17.8 16.7 14.4 12.6 9.3 8.5 7.3 6.3
7 3.6 2.9 1.5 .6 .5 .2 .1 .1 .1
8 2.5 3.3 4.5 2.3 1.2 1.2 .7 .7 .7
9 5 1.6 1.1 1.4 1.4 1.3 1.7 1.9
10 3.1 7.7 6.6 7.5 7.9 7.3 5.4
11 11.4 26.0 29.4 31.4 32.1 29.7
12 2 8.0 12.9
Cumulative market
share of above
products 91. 1 96. 3 92. 1 91. 3 92. 7 96. 9 92. 6 87. 6 88. 3 86. 7
All other 8.9 3.7 7.9 8.7 7.3 3.1 7.4 12.4 11.7 13.3
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
TABLE 7-B.-ATARAXICS
[Percent share of market based on dollar volume 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 46. 3 42. 4 32. 9 32. 9 27. 0 23. 1 19. 8 16. 6 12. 2 12. 3
2 18. 8 13. 5 11. 4 9. 9 8. 8 8. 2 9. 7 7. 6 7. 1 7. 8
3 19. 7 15. 2 12. 0 11. 9 12. 5 10. 4 8. 7 7. 5 7. 0 6. 4
4 4.2 4.8 4.9 4.2 2.9 3.1 1.9 1.7 1.7 1.6
5 3.1 4.8 5.6 4.4 3.0 2.8 2.1 3.4 2.7 1.8
6 .5 10.5 17.9 16.2 12.9 10.8 7.5 6.8 5.4 4.7
7 3.8 2.6 1.5 .6 .4 .2 .1 .1 .1
8 2.7 4.3 4.8 2.4 1.1 1.3 .8 .7 .7
9 .6 2.0 1.3 1.8 1.7 1.6 2.0 2.1
10 3.1 8.3 7.0 8.1 8.9 8.5 6.1
11 12.1 27.4 30.7 32.6 32.8 30.1
12 .2 8.0 12.9
Cumulative market
share of above
products 92.6 97.7 92.2 90.9 91.8 96.1 91.7 87.8 88.2 86.6
All others 7.4 2.3 7.8 9.1 8.2 3.9 8.3 12.2 11.8 13.4
Category total 100. 0 100. 0 100. 0 103. 0 100. 0 100. 0 100. 0 100. 0 100. 0 1 00. 0
X. RAUWOLFIA-DIURETIC COMBINATIONS
Rauwolfia-diuretic combinations provide the anti-hypertensive action of ran-
wolfia plus the diuretic action (increasing the secretion of urine) of the thiazide
products. The latter is considered important in the management of hypertension
also, producing a complementary anti-hypertensive action. The reduction of
body fluid apparently is related to lessening some causal effects of bypertension~
Although rauwolfia products have been used since the early 1950's and mercurial
diuretics even longer, it was not until after the introduction of thiazide diuretics
that the rauwolfia-diuretic combination gained acceptance in medical prac-
tice. The importance of this class can be seen by the rapid rate of growth over~
the past seven years and the many new product entries into the market (see
Tables 8A & 8B). The first five rauwolfia-diuretic combinations introduced in
1959 accounted for 0.67% of all prescriptions written and by 1065 this category
represented 1.71% o2 ~1l prescriptions written. As a class, the prescriptions in-
creased by 344% from 1956 to 1965. The dollar volume increased by 334%
during the same period.
Strong interest in new combinations indicates that there is a medical need
for additional compounds to try on individual patients who w-ere unresponsive
to an earlier product or whose prognosis might be improved by a newer product.
The demand for these products clearly indicates that a need exists and virtually
PAGENO="0245"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1795
a new market has been created for products combining complementary thera'
peutic effects. Although the fixed dosages of manufactured products of this kind
limit the versatility often needed to treat individuals, they are appropriate to
treat large numbers of patients. One obvious reason for the acceptance of these
combinations is the greater assurance that the patient will take one tablet or
capsule containing several medical ingredients than he would if several tablets
or capsules had to be taken separately. If the dose of any ingredient is critical,
each can be prescribed separately and patients can be controlled on an individual
basis.
TABLE 8-A.---RAIJWOLFIA-DIURETIC, COMBI NATION
(Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963
1964
1965
1 54.2 36.0 28.9 28.5 29.1
2 20.4 19.7 11.1 8.5 7.5
3 14.1 23.4 25.6 18.0 20.5
4 5.2 10.6 6.6 5.7 2.6
5 .2 3.4 2.0 1.4 1.3
6 3.7 9.8 10.3 8.1
7 1.7 4.6 9.3 8.6
8 1.3 8.7 8.3 6.9
9 1.3 3.2 5.8
10 .7 2.3 2.6
11 .4 1.4 1.0
12
13
Cumulative markets
share of above
products 100.0 99.8 99.7 96.7 93.9
All others . 2 . 3 3. 3 6. 1
Categorytotal 100.0 100.0 100.0 100.0 100.0
22.9
4.6
15.7
3.4
1.0
9.4
12.9
10.2
5.8
3.8
2.0
1.5
.7
18.3
2.6
18.6
1.5
.7
7.5
17.6
8.2
3.5
6.5
1.6
2.6
3.5
93.9
6. 1
92.6
7.4
100.0
100.0
TABLE 8-B.-RAUWOLFIA-DIURETIC COMBI NATION
[Percent share of market based on dollar volume, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963
1964
1965
1 50. 5 34. 1 27. 9 28. 2 27.2
2 22.8 22.1 11.3 9.0 8.2
3 13. 6 23. 6 26. 5 18. 2 21.2
4 6.3 8.8 5.9 5.2 2.5
5 .1 3.3 1.7 1.4 1.2
6 4.4 10.8 11.4 8.9
7 1.7 4,2 8.1 7.9
8 1.5 9.2 7.8 6.8
9 1.5 3.4 6.5
10 7 2.2 2.6
11 .5 1.6 1.3
12
13
Cumulative market
share of above
products 100. 0 99. 6 99. 9 96. 6 94. 3
Allothers .4 .1 3.4 5.7
Category total 100. 0 100. 0 100. 0 100. 0 100. 0
22. 8
5.7
15. 7
3.2
.9
10.5
10.7
11.6
6.0
3.8
2.2
1.4
.6
19. 1
3.2
19. 5
i.~
.8
8.9
14.9
9.1
3.5
6.6
1.6
2.4
2.8
95. 0
5.0
93. 8
6.2
100. 0
100. 0
XI. CORONARY VASODILATOR5
This therapeutic category contains the drugs which are used primarily to
bring symptomatic relief of angina pectoris and essential hypertension.
The prescriptions written for these drugs during 1950-1965 increased at a
slower rate than that of the overall drug industry. The number of prescriptions
increased 72% while the equivlilent dollar volume of th~se products increased
158%. These products accounted for 1.74% of the prescriptions written both in
1956 asid 1965. Thirteen out of 08 products represented over 90% of the class in
1965 (see Tables 9A & 9B).
PAGENO="0246"
1796 COMPETITIVE PROBLEMS IN THE DRUG Th~DUSTRY
Although this product class is the subject of active research efforts, may
of the new products introduced have not made a significant impact. Patients
responding to one kind of treatment are usually not switched when new products
are introduced. Of seven products introduced between 1956 and 1965, only two
have made any impact on the market.
TABLE 9-A.-CORONARY VASODILATORS
[Percent share of market based on numbers of prescriptions written, 1956-65J
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
Cumulative market
share of above
products 64. 0 74. 9 80. 4 76. 4 80. 5 86. 1 81. 3 84. 2 86. 3 91. 1
All othero 36. 0 25. 1 19. 6 23. 6 19. 5 13. 9 18. 7 15. 8 13. 7 8. 9
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
100.0 100.0 100.0
TABLE 9-0.-CORONARY VASODILATORS
[Percent share of market based on dollar volume, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
Cumulative market
share of above
products 55. 1 70. 8 78. 4 72. 1 75. 7 79. 8 82. 4 81. 7 83. 4 90. 8
All others 44. 9 29. 2 21. 6 27. 9 24. 3 20. 2 17. 6 18. 3 16. 6 9. 2
Categorytotal 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
XII. DIABETIC THERAPY-OTHER
Diabetic products other than insulin fall into two chemical classes-sulfony-
lureas and biguanide. Each of the products shown in Tables 1OA & lOB Is a
different chemical compound except one sustained release form. None of these
products existed in 1956 and only one was available in 1957. They can lower the
blood sugar level when given orally in the treatment of selected patients with
diabetes without the need for insulin injections.
The product class has grown over 1400% in total numbers of prescriptions
written since 1957. By 1965 these products represented 1.26% of all prescrip-
tions written and 2.3% of the total dollar market. In 1965, the total product
class was represented by 8 products with the five shown accounting for over
99% of the market.
1 29. 5 39. 4 41. 8 37. 3 35. 2 36. 8 25. 6
2 19. 6 19. 4 19. 1 17. 7 22. 5 27. 5 20. 4
3 7.7 6.6 10.3 8.4 8.6 10.3 8.2
4 4.6 4.5 1.9 1.4 2.5 .9 .4
5 2.1 2.5 3.3 6.6 5.6 5.0 4.7
6 .5 2.5 2.6 2.3 1.8 1.3 1.6
7 1.0 2.1 3.0 2.5 1.7
8 .4 .6 .8 .7 2.1
9 .5 1.1 1.2
10 10.1
11 5.2
12 5.0
13
23.8 20.9 22.4
22.1 18.6 19.1
4.9 5.7 6.3
.4 1.0 .7
6.0 4.6 3.9
2.3 1.3 1.3
2.1 1.6 1.2
2.1 4.2 2.6
1.9 6.7 9.0
12.8 16.9 12.5
3.3 6.4 8.7
6.8 4.5 5.1
.5 2.1 3.5
1 32. 7 45. 0 52. 9 45. 7 44. 3 46. 1 28. 1
2 6.2 5.8 6.0 5.2 7.2 8.9 6.5
3 7.8 8.6 9.4 9.0 10.8 14.3 9.7
4 7.0 7.1 3.0 2.3 3.5 1.3 .6
5 .6 .8 1.1 2.2 1.7 1.5 1.3
6 .8 3.5 3.9 3.8 2.6 1.8 2.2
7 1.8 3.4 4.5 4.1 2.6
8 .3 .5 .6 .6 1.8
9 .5 1.2 1.3
10 18.1
11 5.2
12 5.0
13
26.4 21.8 22.1
7.0 5.4 5.3
5.1 6.0 6.2
.4 .8 .6
1.6 1.3 1.0
2.9 1.6 1.9
2.6 2.2 1.6
1.8 3.7 2.2
2.2 7.8 13.5
20.8 25.4 17.9
3.3 6.4 8.7
6.8 4.5 5.1
.8 2.9 4.7
PAGENO="0247"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1797
Although the products in this class have achieved a significant prescription
volume, they are not a replacement for insulin injections (estimated annual
sales of $22 million) where the patient's insulin requirement is high. Their
principle advantage is that where they are indicated they can be given orally
replacing the daily injection required to administer insulin.
TABLE 10-A-DIABETIC THERAPY, 0TH ER
[Percent share of market based on numbers of prescriptions written, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964
1965
1 100. 0 99. 0 82. 2 84. 0 81. 7 77. 0 77.3 75. 0
70. 2
2 1. 0 14. 5 8. 7 12. 0 15. 3 16. 6 12. 8
13. 2
3 3.3 7.3 5.9 5.0 3.5 3.8
3.4
4 .4 2.7 2.6 5.3
8.1
5 3.1
Cumulative market
4.8
share of above
products 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
All others
99. 7
.3
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
100. 0
TABLE 10-B-DIABETIC THERAPY, 0TH ER
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964
1965
1 100. 0 99. 1 85. 1 85. 7 84. 9 81. 3 79. 1 77. 6 72. 8
2 .9 12.9 9.2 10.3 12.4 16.7 11.9 12.8
3 2.0 5.1 4.3 3.3 2.2 3.0 2.5
4 5 3.0 2.0 5.1 7.6
5 2.3 3.9
Cumulative market
share of above
products 100.0 100.0 100.0 100.0 100.0 100.0 100.0 99.9 99.6
All others .1 .4
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100.0
XIII. DIURETICS
Diruetic agents are used to treat edema which may be associated with a
number of different disorders. The drugs may be used as `adjuncts in the manage-
ment of congestive heart failure, to control the severity of cardiac decompensa-
tion or to treat patients with edema associated with kidney and liver disease,
pregnancy, obesity, etc. The diuretic class `had one of the most dramatic increases
over the ten year span increasing 490% in th~ number of prescriptions written
and 560% in dollar volume. Of the total industry volume, diuretics accounted
for 1.12% of the prescriptions written in 1956 but increased to 3.68% of those
written in 19&5. Eleven products accounted for 80% of the class of 43 products
audited in 1965 (see Tables hA & hiB).
In reviewing the 1965 data it appears that diuretics that did not exist in 1956
are responsible for nearly all of the prescriptions written. The market is fairly
well fractionated with about a dozen compounds actively competing for signifi-
cant positions.
PAGENO="0248"
1798 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 11-A-DIURETICS
[Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 960 1961 1962 1963 1964 1965
1 52.6 53.5 17.0 7.8 6.6 5.6 4.7 5.1 5.0 3.7
2 13.3 12.9 3.3 1.4 .9 .4 .2 .2 .4 .3
3 6.3 2.4 .2 .1
4 5.6 3.1 .6 .3 .1 .1 .1 .1
5 . 8 72. 1 61. 8 42. 3 33. 3 32. 7 23. 7 23. 1 25. 9
6 14. 3 20.6 19. 6 17. 3 18.8 14.4 20. 0
7 7.8 11.3 7.8 6.9 4.5 3.3 4.9
8 5.4 5.0 5.4 6.3 7.1 4.1
9 1.0 3.0 4.9 5.7 6.6 5.9
10 .2 2.2 3.2 5.7 8.7 3.3
11 .1 3.5 4.1 5.0 5.1 4.7
12 .4 6.1 4.9 4.4 4.9
13 .2 2.6
Cumulative market
share of above
products 77. 8 72.7 93.2 93. 5 88. 5 80. 9 85. 6 79. 9 78. 4 - 80.3
All others 22. 2 27. 3 6. 8 6. 5 11. 5 19. 1 14. 4 20. 1 21.6 19. 7
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
TABLE 11-B-DIURETICS
[Percent share of market based on dollar volume, 1956-651
Product 1965 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 49. 6 48. 7 14. 0 7. 7 7. 3 6. 4 5. 8 6. 9 6. 7 5. 4
2 18.4 15.6 3.9 1.7 1.4 .5 .2 .3 .4 .4
3 8.2 3.1 .3 .1
4 2.4 1.2 .2 .1 .1
5 1. 1 74.5 64. 1 41.5 32.8 30. 5 22. 1 21.5 24.2
6 13.4 19.4 18.9 17.1 18.0 13.9 19.3
7 7.2 10.5 7.6 6.2 4.3 3.1 4.7
8 5.2 4.7 5.6 6.6 6.8 4.4
9 .9 3.0 4.8 5.4 5.9 5.6
10 .3 2.3 3.3 6.4 9.2 3.6
11 .1 3.2 4.0 4.4 5.1 4.5
12 .4 5.9 4.7 4.3 4.7
13 .2 2.7
Cumulative market
share of above
products 78.6 69. 7 92.9 94.3 89.6 79. 8 83. 5 79. 1 77. 1 79. 5
All others - 21. 4 30. 3 7. 1 5. 7 10. 4 20. 2 16. 5 20. 9 22. 9 20. 5
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
XIV. Hon~ioxns, CORTICOIDS
The term corticoid is short form of the name corticosteroid. The desired action
of corticoids is to favorably influence a large number of clinical conditions by the
suppression of inflammatory or immulogical processes. Unfortunately the anti-
inflammatory action of the corticosteroids is inseparable from their metabolic
effect which limits their usefulness for prolonged administration because of
rather serious side effects. Because of these side effects a great number of deriv-
atives have been prepared and are used extensively in treating patients with
various types of disorders-both acute and chronic conditions (collagen diseuses,
bronchial asthma, hay fever, allergies, dermatoses and certain inflammatory
eye diseases).
PAGENO="0249"
0
0
L~:j
0
0)
0)
C)
~0
C) ~
~
~0
3
3
00
=
0
0
(I)
C)
0
00
C)
0
00
C,)
-
~ 0)00 O~ 00 N) 00 (~C)~ (00
~
3
2
>
00
0
00
0
z
C')
C)
0 -
-4
C)
0
00
C,,
PAGENO="0250"
1800 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
XV. CoRTIcoIDs WITH ANTI-INFEOTIVES
Corticoids with anti-infectives describes a group of products which represent
the combination of the anti-inflammatory corticosteroids with various anti-
infectives in topical preparations with a maximum local effect. Because of the
broad range of competitive corticosteroids and anti-infective agents, there are
numerous possible combinations obtaining different effects (see Tables 13A. &
13B). Consequently, the many companies which have marketed preparations
make this category one of the most highly competitive in the entire field. Over the
last ten years the category has grown approximately 78% in numbers of prescrip-
tions written and 87% in dollar volume. The category has maintained its relative
position representing 1.51% of all prescriptions written in 1956 and 1.55% of all
prescriptions written in 1965. The dollar volume for this category declined
from 1.63% of total in 1956 to 1.45% of total in 1965.
It appears from the large number of products being offered in this class plus
the interest in a variety of combinations of corticosteroids and anti-infectives
that there is a need for a broad range of products to treat infections and inflam-
mation from many different causes. Twenty-five products account for over 80%
of the class and 76 products make up the other 20%. Many of these combination
products are available in specialized forms for treatment of conditions located
in the eyes, ears, and nose as well as for general topical use anywhere on the
body surface.
TABLE 13-A.-CORTICOIDS WITH ANTI-I NFECTIVES
[Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 41. 1 29.7 23.8 16.4 14. 9 12.7
2 18.2 14.7 12.0 7.4 5.7 4.9
3 8.8 4.2 3.3 1.9 1.7 1.7
4 4. 6 3. 8 3. 1 3. 0 3. 4 3. 9
5 2.8 4.3 2.7 3.4 3.3 2.9
6 2.8 3.7 4.4 6.2 6.2 8.4
7 2.6 1.9 1.3 .8 .7 .4
8 2.3 2.5 2.3 2.1 1.6 1.1
9 1.2 1.3 2.3 2.7 1.4 1.0
10 . 8 9. 5 12. 6 12. 1 11.7 10. 1
11 . 4 .4 2. 1 2.0 1.8 1. 5
12 2 6. 1 14.2 14.3 9.2 6.7
13 1. 0 1.9 2.2 3. 1 3.3
14 . 1 . 8 2. 4 2. 1 1.8
15 1 2.2 1.9 1.9
16 3. 5 2. 2 1. 4
17 1. ~ ~. ~ 6. 9
18 1. 8 7. 8 10. 7
19 1.4 1.4
20 8 1.2
21 1.4
22 .2
23 -
24
25
10.0 8.1 7.3 .7
3.3 2.3 2.0 2.0
.6 .6 .7 .5
2.4 1.9 2.9 2.5
2.5 2.5 2.2 1.7
7.9 8.2 7.8 8.7
.3 .2 .3 .3
1.1 .9 .6 .4
1.1 .5 .5 .5
9.6 11.0 10.0 9.5
1.4 1.3 1.1 1.4
5.4 4.2 3.5 2.6
3.0 2.4 2.4 2.5
1.3 1.5 1.9 1.8
1.0 1.3 1.3 1.2
.5 .5 .3 .1
8.2 8.3 9.3 11.3
13.4 11.1 13.2 11.5
.5 .4 .2 .1
1.8 1.6 1.6 2.4
7.1 6.8 6.4 5.4
.9 1.7 1.7 2.1
1.1 1.5 2.9 2.5
.8 1.8 2.7 3.2
.3 .8 1.1 1.2
86. 2 85. 5 85. 5 81. 4 83. 9 82. 3
13.8 14.5 14.5 18.6 16.1 17.7
Cumulative market
share of above
products 85. 8 83. 2 86. 9 86. 3
All others 14. 2 16. 8 13. 1 13. 7
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
100.0 100.0 100.0 100.0
PAGENO="0251"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1801
TABLE 13-B.-CORTICOIDS WITH ANTI-INFECTIVES
(Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962
1963
1964
1965
1 41.7 29.0 23.0 16.0 14.6 12.0 9.4
2 14.2 12.7 11.2 6.3 4.7 4.0 2.8
3 6.1 2.9 2.3 1.3 1.1 1.2 .4
4 4.6 3.8 3.1 3.0 3.4 3.9 2.4
5 3.4 4.7 4.3 4.3 4.0 3.6 2.9
6 2.5 3.8 4.2 6.2 6.2 7.9 7.0
7 2.6 2.1 1.3 .8 .7 .4 .2
8 2.4 2.7 2.4 2.3 1.8 1.3 1.3
9 1.4 1.5 2.9 3.6 1.7 1.1 1.3
10 .7 10.8 14.8 14.1 13.9 12.5 11.3
11 .3 .4 1.6 1.6 1.5 1.2 1.2
12 .2 6.7 13.0 13.6 8.2 5.7 4.4
13 1.3 2.1 2.7 3.6 3.7 3.1
14 .1 .8 2.4 2.1 1.8 1.3
15 .1 2.1 2.1 2.2 1.1
16 2.4 1.5 .9 .4
17 1.8 5.6 7.0 8.6
18 1.6 7.5 10.0 12.4
19 1.1 1.1 .3
20 1.0 1.5 2.1
21 1.5 7.2
22 .2 1.1
23 1.1
24 .9
25 .5
Cumulative market
share of above
products 81. 7 83. 9 88. 1 87. 0 87. 1 85. 8 85. 2
All others 18. 3 16. 1 11. 9 13. 0 12. 9 14. 2 14. 8
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
7.5
1.9
.4
1.9
2.9
7.1
.2
1.0
.6
12.6
1.0
3.5
2.6
1.5
1.6
.4
8.8
10.1
.2
1.9
7.1
2.1
1.4
1.9
1.1
6.9
1.6
.5
2.9
2.6
6.6
.3
.6
.6
11.3
.9
2.9
2.5
1.9
1.5
.2
9.7
12.2
.1
1.9
6.7
2.1
3.0
3.0
1.5
6.1
1.6
.4
2.5
1.8
7.3
.2
.5
.6
10.8
1.0
2.1
2.5
1.8
1.4
.1
12.1
10.4
.1
2.8
5.6
2.5
2.3
3.5
1.4
81. 3
18. 7
84. 2
15. 8
82. 0
18. 0
100. 0
100. 0
100. 0
XVI. ORAL MUSCLE RELAXANTS
Drugs in this group act on the central nervous system and are useful in pro-
moting relaxation of skeletal muscle spasm. They have been recommended as an
aid in the management of almost every musculo-skeletal and neuromuscular con-
dition in which painful muscle spasm occurs. Thus these agents can be used in
connection with the treatment of a number of different medical conditions.
Over the past decade the number of prescriptions for these products has in-
creased 350%, while the equivalent dollar value has increased 405%. In 1956
these products accounted for 0.56% of the total number of the prescriptions
written. In 1965 they were responsible for 1.26% indicating wider acceptance
of the use of these products in various conditions.
The top five products in 1956 were responsible for nearly 76% of all prescrip-
tions written. By 1965, one-third of the 47 products in the class accounted for over
82% of the prescriptions (see Tables 14A & 14B). The newer products introduced
the combined effect of muscle relaxants and analgesics or sedatives, or both. The
effectiveness of muscle relaxants alone was not enough in many cases for satis-
factory treatment. If muscle spasm caused pain, then pain relief was also indi-
cated. If muscle spasm could be reduced by a sedative, working in the central
nervous system, then a sedative was indicated.~ Frem these experiences, the new,
convenient forms evolved, increasing competition further.
PAGENO="0252"
1802 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Product
TABLE 14-A.-ORAL MUSCLE RELAXANTS
[Percent share of market based on numbers of prescriptions written, 1956-65)
1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
Cumulative market
share of above
products 75. 6 82.2 74. 4 76. 2 88. 3 74. 9 79. 8 84. 2 83. 1 82. 5
All others 24. 4 17. 8 25. 6 23. 8 12. 7 25. 1 20. 2 15. 8 16. 9 17. 5
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
TABLE 14-B-ORAL MUSCLE RELAXANTS
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 37.0 39.2 10.6 5.5 4.8 2.0
2 13.0 10.9 4.0 1.1 .6 .7
3 14.4 10.0 4.9 1.7 1.3 1.0
4 9.1 11.9 9.4 6.0 6.8 4.4
5 4.8 2.6 1.6 1.1 2.1 2.1
6 5.6 32.9 19.3 18.4 11.4
7 10.5 5.6 1.8 1.8
8 .5 21.8 13.0 9.1
9 10.6 15.8 14.8
10 .1 4.5 5.5
11 .1 9.6 5.5
12 6.3 9.8
13 .3 4.6
14 1.6
15
16
0.1
.7 0.7 0.5 0.5
1.6 .8 .6 .8
5.3 9.5 9.1 10.5
1.8 1.6 4.3 5.2
10.5 6.0 6.6 6.0
.4 .3 .3 .2
6.4 3.9 3.5 3.7
12.5 13.3 11.6 9.1
8.7 8.6 10.4 8.6
4.0 2.6 3.7 3.6
10.3 8.1 8.3 9.2
8.7 10.4 10.0 9.8
4.9 3.8 6.5 6.9
1.0 8.6 4.1 3.0
.7 1.3 3.2 5.7
Cumulative market
share of above
products 78. 3 80. 2 74. 4 72. 9 85. 3 74. 3 77. 6 79. 5 82. 7 82. 8
All others 21. 7 19. 8 25. 6 27. 1 14. 7 25. 7 22. 3 20. 5 17. 3 17. 2
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
XVII. PSYCHOSTIMULANTS
Drugs in this therapeutic category are used to treat depressed patients with
widely varying causes. Thus the compounds can offer different types and different
mechanisms of action. This category has shown dramatic growth although the
base was small in 1950 (0.38% of total prescriptions written in 1956 to 1.3% in
1905). The number of prescriptions written has increased 510% over the period
1956-1965, while the dollar volume of these products has increased tenfold.
The three principal products in 1956 accounted for 70% of the prescriptions
written. Two of these were withdrawn from the market in 1962. By 1965, twelve
products represented over 87% of prescriptions in a class of 36 products (see
Tables 15A & 15B). New products in this field have been viewed with guarded
optimism since the effects of psychostimulants might mask underlying problems.
As a result, each new product is tried in different kinds of patients to find where
it is most effective. The product class, therefore, has a history of product in-
troduction, initial acceptance and a decline to some lower level of use.
1 28.4 33.8 11.1 5.8 4.9 2.1
2 16.0 14.1 5.8 1.6 1.0 .9
3 15.6 12.1 6.7 2.3 2.0 1.4
4 10.9 15.2 10.4 10.0 9.8 4.9
5 4.7 2.7 2.3 1.5 2.8 2.9
6 4.3 28. 3 17. 4 17. 6 10.9
7 9.4 5.2 1.7 1.7
8 4 22.1 12.8 9.2
9 10.1 15.4 14.5
10 .1 4.0 5.2
11 .1 9.7 5.4
12 6.4 9.7
13 .2 4.8
14 1.3
15
16
.8 .8 .6 .6
2.3 1.4 1.1 1.3
8.2 14.4 14.4 15.6
2.4 2.4 5.6 7.0
10.0 5.9 6.1 5.4
.4 .3 .4 .2
6.4 3.9 3.5 3.3
12.5 13.2 10.8 8.5
7.7 7.4 8.6 7.2
4.2 2.7 3.3 3.4
10.4 8.7 8.1 8.9
8.6 10.1 9.4 9.2
4.2 3.2 4.9 5.4
1.0 8.8 4.1 2.9
.6 1.0 2.2 3.6
PAGENO="0253"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1803
TABLE 15-A.-PSYCHOSTIMULANTS
[Percent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
I 39.2 42.8 23.6 20.1 13.6 11.0 12.5 12.5 16.0 15.9
2 19.6 10.2 1.7 1.6 .4 .1
3 11.3 7.2 2.4 2.3 1.7 .8 .4 .3 .7 1.2
4 .4 11.9 27.4 8.5 2.4 .6
5 30. 1 39. 9 21. 2 19. 0 12. 8 19. 7 18. 3 12. 5
6 8.2 17.9 18.8 16.3 15.4 23.6 15.2
7 7.0 12.7 5.3 2.0 3.2 5.0 1.8
8 6.5 13.7 12.5 8.7 5.6 3.8 2.2
9 4.8 10.9 6.3 3.6 2.8 2.4 1.7
10 10.6 16.4 20. 5 22.2 29. 5
11 6.9 12.8 21.6 4.9 3.1
12 7.3
Cumulative market
share of above
products 70. 5 72. 1 71. 7 80. 1 91. 0 89. 4 83. 0 96. 8 93. 7 87. 7
All others 29. 5 27. 9 28. 3 19. 9 9. 0 10. 6 17. 0 3. 2 6. 3 12. 3
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
TABLE 15-B.-PSYCFIOSTIMULANTS
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 32.5 36.0 18.2 11.5 7.8 6.0 7.3 7.7 9.9 10.3
2 25.6 14.3 2.1 1.1 .3 .1
3 9.4 4.1 1.7 .8 .5 .3 .2 .1 .3 .5
4 .4 15.5 17.2 6.7 1.9 .6
5 30. 1 39.9 21. 2 1. 9 12. 8 19. 7 18. 3 12. 5
.6 13. 0 22. 9 23. 6 20. 8 19. 1 28. 1 18. 2
7 8.8 13.7 5.8 2.1 3.2 5.5 2.1
8 6.1 11.5 10.9 8.0 5.4 3.3 2.0
9 6.4 12.4 7.1 4.5 3.3 2.8 1.8
10 10.8 17.0 21.3 21.6 30.2
11 5.9 11.8 19.8 4.5 2.7
12
Cumulative market
share of above prod-
ucts 67. 9 69. 3 69. 3 94. 3 92. 2 90. 1 84. 5 98. 6 94. 3 87. 1
All others 32. 1 30. 1 30. 7 5. 7 7. 8 9. 9 15. 5 1. 4 5. 7 12. 9
Category total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
XVIII. SEDATIVEB-BAIIBITURATE
The barbiturates comprise an important and valuable class of central nervous
system depressants. It has been reported that they are used more to produce sleep
than for any other purpose. 1 Proper selection and rational use of the barbiturates
depend on an intimate knowledge of their pharmacology and toxicology as well as
~an acquaintance with the important variations in action and toxicity produced by
certain individual products. Barbiturates are differentiated by their duration of
action ranging from longer duration barbital and phenobarbital to shorter dura-
tion secobarbital, pentobarbital and hexobarbital. Combinations of barbiturates
are often used to provide a more balanced effect by combining those with differing
Tates of action and dissipation.
The total category has increased approximately 7% in total number of prescrip-
tions written and 22% in dollar volume between 1956 and 1965. Barbiturates rep-
resented 7% of all prescriptions written in 1956 but only 4.3% in 1965.
The lack of significant increases in demand for barbiturates may be related to
the development of non-narcotic analgesics, tranquilizers and barbiturate-anal-
gesic combinations. The importance of this class, however, is seen by the large
number of products competing. Although the five products in Tables 1GA & 1GB
account for almost three-fourths of the product class, there were 78 products on
the market in 1965.
1 Tine Pharmacological Basis of Therapeutics-Second Edition Goodman and Gilman.
PAGENO="0254"
1804 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 16-A--SEDATIVES-BARBITURATE
lPercent share of market based on numbers of prescriptions written, 1956-651
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 24. 1 24. 5 26. 4 21. 5 22. 0 22. 2 22. 9 19. 7 21. 5 20. 9
2 15. 1 15. 2 13. 4 13. 8 15. 6 16. 5 15. 5 17. 3 16. 8 14. 9
3 13.2 12.8 15.2 14.7 19.3 17.7 17.4 15.1 14.2 15.4
4 10.7 12.0 11.8 14.2 14.5 13.7 10.2 12.5 14.2 13.2
5 5.8 5.5 7.3 8.6 9.4 9.6 6.8 7.8 8.0 9.7
Cumulative market
share of above
products 68. 9 70. 0 74. 1 72. 8 80. 8 79. 7 72. 8 72. 4 74. 7 74. 1
All others 31. 1 30. 0 25. 9 27. 2 19. 2 20. 3 27. 2 27. 6 25. 3 25. 9
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
TABLE 16-B-SEDATIVES, BARBITURATE
[Percent share of market based on dollar volume, 1956-65]
Product 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
1 16.7 18.1 20.1 15.9 15.7 16.9 17.3 15.1 16.5 16.2
2 14. 7 15. 5 14. 1 13. 7 15. 8 16. 4 15. 0 16. 8 16. 4 14. 1
3 13. 2 13. 1 15. 6 14. 7 19. 5 18. 0 17. 5 15. 3 14. 1 15. 4
4 10. 6 12. 9 13. 4 16. 9 16. 8 16. 3 12. 6 15. 4 17. 3 16. 1
5 5.8 6.3 7.8 9.0 10.1 10.1 7.1 8.0 8.2 9.8
Cumulative market
share of above
products 61. 0 65. 9 71. 0 70. 2 77. 9 77. 7 69. 5 70. 6 72. 5 71. 6
All others 39. 0 34. 1 29. 0 29. 8 22. 1 22. 3 30. 5 29.4 27. 5 28. 4
Category total 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0 100. 0
XIX. SULFONAMIDES
Products in this group are used to treat a broad range of infections. Some prod-
ucts are specific for either systemic or urinary tract infections while others treat
both types. Over the ten year time span the number of prescriptions for sulfon-
amides has increased 25% while the dollar volume has increased 72%. Sulfon-
amides represented 2.85% of prescriptions written in 1965 as compared to 4.0%
in 1956.
The principal products used in 1956 continue to be major products in 1965. How-
ever, approximately 76% of prescriptions written are for 14 out of 173 products
audited in 1965 (see Tables 17A and 17B).
PAGENO="0255"
C
CD
CD
p
C
CD
I.
~0
3
3
-1
(0
C
C
0)
C)
(0 N~ 0) ()~ 0)0) - C-Na Q~
-
(0(0(000(0 ~ N) (0 N) N) C) ~J
3
2
0>
-0~ C/)
0
0
ci
ci
02
I,
PAGENO="0256"
PAGENO="0257"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
THURSDAY, JANUARY 18, 1968
U.S. SENATE,
MONOPOLY SUBCOMMITTEE OF THE
SELECT COMMITTEE ON SMALL BUSINESS,
Washington, D.C.
The subcommittee met, pursuant to notice, at 10 a.m., in room 318,
Old Senate Office Building, Senator Gaylord P. Nelson (chairman of
the subcommittee) presiding.
Present: Senators Nelson, Javits, and Scott.
Also present: Benjamin Gordon, staff economist; James H. Gross-
man, minority counsel; Susan H. Hewman, research assistant; and
William B. Cherkasky, legislative director, staff of Senator Nelson.
Senator NELSON. We will open the hearing of the Subcommittee on
Monopoly of the Small Business Committee.
Our witness this morning is Dr. Willard F. Mueller, Chief Econ-
omist and Director of the Bureau of Economics of the Federal Trade
Commission.
I have read your statement, Dr. Mueller, and I find it commendable
and lucid compared to the economic presentation given last month by
the economists sponsored by the Pharmaceutical Manufacturers Asso-
ciation, which, quite frankly, I did not understand very well.
Will you introduce your associates for the record, and, if, at anytime,
one of them wants to make a comment, will he just identify himself
so we can keep the record straight.
I appreciate very much your taking the time to come here this
morning and the obvious extensive work you have put into your
statement.
You may proceed to present it in any way you see fit.
I assume you have no objection to us interrupting you with questions
from time to time.
Dr. MUELLER. Not at all, Mr. Chairman.
Senator NELSON. And do you have a biographical statement?
Dr. MUELLER. Yes; I have a sketch which I can give to the reporter
or read into the record.
Senator NELSON. Read it into the record for us, please.
STATEMENT OF DR. WILLARD P. MUELLER, CHIEF ECONOMIST AND
DIRECTOR, BUREAU OF ECONOMICS, FEDERAL TRADE COMMIS-
SION, WASHINGTON, D.C.; ACCOMPANIED BY DR. RUSSELL C.
PARKER, ASSISTANT DIRECTOR; AND WILLIAM H. KELLY, MEM-
BER, BUREAU OP ECONOMICS
Dr. MUELLER. First, I would like to introduce my two colleagues.
To my immediate left is Dr. Russell C. Parker, assistant to the Direc-
1807
81-280-68-pt. 5-17
PAGENO="0258"
1808 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
tor of the Bureau of Economics, and next to him is Mr. William H.
Kelly of our staff.
I am Chief Economist and Director of the Bureau of Economics
of the Federal Trade Commission. Prior to my current appointment,
I was Chief Economist to the Select Committee on Small Business
of the House of Representatives. Previous to that assignment, I was
professor at the Uiiiversity of Wisconsin, from 1957 to 1961. I was
assistant professor at the University of California from 1954 to 1957.
I also have taught on a part-time `basis at American University; served
as a visiting professor at Michigan State University, and I am cur-
rently a part-time staff member of the Department of Economics,
University of Maryland.
It is a privilege and a pleasure to appear before this subcommittee.
My appearance today is in response to the request of your Chairman
that I submit testimony on the subject of profits in the drug industry,
as well as present an independent analysis of the study "Risk and
Return in American Industry-an Econometric Analysis," presented
to this committee on December 19, 1967.' The study was sponsored by
the Pharmaceutical Manufacturers Association and prepared by
Gordon R. Conrad and Irving H. Plotkin of Arthur D. Little, Inc.,
in consultation with Prof. Jesse W. Markham and Prof. P. J. Cootner.
Hereafter we shall refer to the study as the Conrad-Plot.kin study.
Before turning to the study, we shall first place in perspective the
profits of drug manufacturers by comparing them with those earned
by business enterprises in other American industries.
Figure 1 shows for 1966 the average rate of return on stockholders
investments of leading firms in 22 important American manufactur-
ing industries. Profit rates of leading drug manufacturers exceeded
those of large firms in the 21 other industries. In fact, drug industry
profits were twice as great as one-third of the remaining industries;
were 44 percent or more above those of all but four other industries;
and they exceeded even such traditionally high profit industries as
motor vehicles and computing machines.
Nor was 1966 an exceptional year. Table 1 compares over the period
1950-66, the average profits of large drug companies and large com-
panies in the 22 industries shown in figure 1. Several points are of
special interest. First, in the early years, 1950-55, average drug com-
pany profits were about equal to or somewhat above the average of
other large manufacturers. Second, beginning in 1956, however, aver-
age profit rates of drug companies were well above the average of
other large companies. Finally, since 1956 drug companies have con-
sistently ranked either first or second among all large manufacturing
industries. This indicates that during the last decade large drug
companies have occupied an especially advantaged position relative
to large companies in other American industries. Table 2 summarizes
profit data of all drug companies and all manufacturing companies
for the period of 1956-67. Over the period covered it shows essentially
the same picture as table 1. Since 1956 drug manufacturers have
failed to occupy first place in only one year.
1 The study, "Risk a.nd Return in American Industry-an Econometric Analysis," by
G. R. Conrad and I. H. Plotkin, begins at p. 1746, supra.
PAGENO="0259"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1809
This pattern of persistently high profits indicates that large drug
companies occupy a unique position in the American economy. And
they appear to have become increasingly unique since the mid-1950's.
Before turning to an analysis of the Conrad-Plotkin-Markham-
Cootner explanation of high profits in the drug industry, I shall
briefly review what appears to be a virtual consensus of opinion among
researchers in the field of industrial organization concerning the
causes of high profits of drug manufacturers.
In this connection I would like to emphasize the crucial role which
congressional hearings have played in developing the facts necessary
for scholars to study the organization and performance of the drug
industry. Prior to the Kefauver drug hearings on administered prices
in 1959-61, not a single article concerning the American pharmaceuti-
cal industry had appeared in a professional economic journal.
Senator NELSON. Are you referring to any article of any kind or
just articles in economic publications?
Dr. MUELLER. I am referring to articles on the structure and or-
ganization of the drug industry. This observation is based on the
recent study that I cited in a footnote by Professor Walker of the
University of Indiana, who surveyed the literature.1
Since the first such article appeared in 1962, there has `been a growing
volume of research literature on the subject, `all of which has drawn
heavily on the Kefauver `and subsequent congressional hearings. The
facts developed by this committee have made another enormous con-
tribution to the fund of knowledge concerning the drug industry.
I am confident that `scholars will be sifting and winnowing the facts
for years.
The preponderance of economic evidence argues that the persistently
high profits of the drug industry are the result of the absence of effec-
tive price competition in the sale of many products. Price competition
in drugs is ineffective for several reasons. Concentration in the produc-
tion of many drugs is high because of the patent privilege. And even
where there are relatively many sellers, as well as many potential
sellers-for example, in the case of unpatented drugs sold under
generic names-effective price competition often is muted by vast ad-
vertising, promotion, and other selling efforts which differentiates in
the minds of consumers the products of the largest drug manufactur-
ers selling under their own brand or trade names from those of other
manufacturers. Hence, manufacturers selling chemically identical
drugs under generic names frequently have difficulty in selling them
at any price.
Senator NELSON. Are you referring to' the retail market particularly
in this instance?
Dr. MUELLER. That is correct; the sales of druggists in the pre-
scription market.
Senator NELSON. We do find from the testimony, as I am sure you
have noted, that generic companies or generic drugs are competing
effectively in bids to nonprofit institutions and to the Federal Govern-
ment, and you will also find testimony that brand name drugs drop
substantially in their price when they bid in this competitive area.
1 The complete prepared statement submitted by ~" Dr. Mueller begins at p. 1824, infra.
PAGENO="0260"
1810 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. MUELLER. That is correct.
The evidence shows that when the buyers are well informed such as,
hopefully, the purchasers for hospitals and the U.S. Government, they
nsually purchase on a generic basis rather than a brand basis, or when
they do buy on a brand basis they purchase the brand selling at the
lowest price rather tha.n simply the most popular one or well known
one.
Senator NELSON. So, you are explaining, I take it, the difference
between the price situation in the retail market, where you will find
a brand name product sold to the pharmacist at the highest price
charged in the country, and the price situation in the institutional
market where the same company bidding its brand name product to the
Government or some nonprofit institution will offer it at a substantially
lower price; is that correct?
Dr. MUELLER. That is correct. I think every person who has studied
the record of congressional hearings where this sort of evidence has
been developed and has been explored and subjected to extensive com-
ment by informed people outside and within the drug industry, as well
as medical doctors, has come essentially to this conclusion: namely,
that there are very substantial differences between prices for products
purchased by hospitals and other large buyers and prices paid in drug-
stores for the identical product.
Senator NELSON. And I understand it to be your conclusion that in
general that is because you will very frequently find competition in
bids to the Government or bids to nonprofit institutions while. there is
much less, or in some cases not any, competition at the retail level and
the price charged by the company to the retailer and the wholesaler is
not a competitive one; is that correct?
Dr. MUELLER. And the key factor for generating the competition is
the prese.nce of an informed buyer in one case and an uninformed one
in the other, and, as a result, this broadens the market. People who have
a product to sell that is physically identical to that of other sellers have
an opportunity to bid, and, consequently, you broaden the market,
which means broadening the opportunity for competition to work.
Senator NELSON. You referred to an informed buyer versus an unin-
formed buyer. That factor is at work in purchases by nonprofit organi-
zations or government. If the purchasing agent is informed and takes
competitive bids and has the personnel to evaluate the quality of the
drugs, you have an informed buyer and you have competition and a
lower price. In the same institution, nonprofit or government, if the
buyer is not informed the price is back up high again.
Now, you are not suggesting that the question of an informed buyer
makes any difference in the retail price, are you?
In other words, is it not correct that the wholesaler or the retailer is
charged a certain price, and although quantity buying accounts for
some differential, an informed purchaser for a large drug chain is still
being charged a high price, the same price as is paid by an uninformed
purchaser for another drug chain; wouldn't that be correct?
Dr. MUELLER. That is correct. As I interpreted the evidence that had
been developed on this point, the only opportunity of getting a lower
price at this level from the ultimate consumer's standpoint, namely, the
patient. of a doctor, is if his doctor orders the drug on a generic basis.
PAGENO="0261"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1811
And then, of course, we may be able to get. the product at the retail level
at a lower price than for the branded product.
But on your specific point as to whether the branded product sells
at the same price or a higher price through drug stores in these corn-
parative bids, I agree with your interpretation of the facts.
Mr. GORDON. I think there may be another point, and that is that the
purchasing agent for the ordinary consumer is not the one who pays~
In other words, the doctor does not pay for the medicine. He does the
purchasing. He orders the drug, and the patient, the consumer, does
not have much of a choice.
Dr. MUELLER. I have heard many people speak to that point, includ-
ing doctors, who say that some doctors are more discriminating in this
respect than others. And I think increasingly, with publicity concern-
ing the alternatives available, consumers often ask their doctors:
"What is this generic business all about?" So this, I suspect, makes
the doctor a little bit more price conscious. And I think this is one of
the benefits of publicity concerning the alternatives available to con-
sumers. It is sort of a counter to the advertising which promotes brand
names so decisively in the minds of consumers today.
Senator NELSON. Go ahead.
Dr. MUELLER. The resulting wide price spread between advertised
and generic drugs often applies to unpatented as well as patented
drugs. These factors shelter the leading concerns, or any concern with
a highly differentiated drug product, from effective price competition.
This explanat ion of high drug profits is not novel. Nearly all re-
searchers who have analyzed the drug industry in detail have come
essentially to the same conclusion. Similarly, empirical studies which
have cut across many industries have identified the elements of market
structure that are primarily responsible for high noncompetitive
profits. These elements are high seller concentration, high barriers to
entry, and product differentiation.
Mr. GORDON. Dr. Mueller, for the record, will you please explain
exactly what product differentiation is?
Dr. MUELLER. I had anticipated this question. I know that Senator
Nelson was an undergraduate student in economics but now is primar-
ily a lawyer-or, I should say, primarily a statesman.
Senator NELSON. You may make that~ statement here, but no place
else.
Dr. MUELLER. I am thinking of back in Wisconsin.
There is a definition of product differentiation which appears in the
Antitrust Law and Economic Review which, presumably, is written
for lawyers. I am not sure if this is an improvement on one that I could
make for students.
That is the way they define it-they say:
Product differentiation refers to the distinguishing of substitute products from
one another by advertising and the like. Whereas buyers of homogenous products
regard the output of any particular seller as identical in all respects to that of
all other producers of that product, the seller of a "differentiated" product enjoys
a favored position over its rivals, in that the buyers consider it a superior product
and are willing to pay a "premium" price for it rather than accept the substitutes
offered by those rivals. Since new entrants must frequently accept a lower price
than established firms are able to get for a product of equal quality and cost, this
disadvantage is said to constitute a "barrier to entry," one that permits estab-
lished firms to charge a supercompetitive price without attracting new entry.
PAGENO="0262"
1812 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
And, as I said, I will later have some things to say about product
differentiation created by advertising or promotion, which in effect
creates a barrier to entry facing potential competitors. This is what
this definition is supposed to explain.
Senator NELSON. `When you say "product differentiation," are you
referring to the same chemical compounds which are differentiated by
advertising and other means?
In other words, you are talking about a drug, a basic compound,
which a company differentiates from the same drug put out by another
firm simply by the capsule it is in and the advertising. You are not
talking about two different drugs?
Dr. MUELLER. No, sir, they are chemically identical.
Senator NELSON. So, you would have two drugs that have the same
chemical composition but are differentiated in one way or another by
trade name and advertising and that sort of thing?
Dr. MUELLER. That is correct. People have an impression that there
is a difference. It is a psychological thing. You believe it is different,
and, therefore, you behave differently, even though it is physically
identical.
Senator NELSON. So, a drug such as prednisone, which is the generic
name, manufactured by a number of companies. One of them gives it
the brand name "Paracort," and one of them "Meticorten." They are
the same chemical compound, but are differentiated products in the
marketplace; is that what you are saying?
Dr. MUELLER. That is correct.
Senator NELSON. And they have the same purpose and the same ob-
jective in the treatment of the patient?
Dr. MUELLER. Yes. I am assuming that they are physically or chem-
ically identical, and yet, because of this differentiation in the minds of
the consumer you are able to have a unique demand for them which
permits you to get a higher price for one than the other.
Senator NELSON. So, you are saying that the difference, when you
use the word "differentiation," is in the minds of the consumers and
not in the chemical compOsition of the product itself?
Dr. MUELLER. That is correct. That is the way I am using it here,
as I think it applies most accurately in the drug industry. In some in-
stances, of course, there may be some physical difference in products,
and forms differentiate them further through advertising and the like.
Senator NELSON. In the method by which they are compounded or
by the coating on the tablet, and that sort of thing?
Dr. MUELLER. I was thinking of products other than drugs.
In the case of soaps, they are very similar often, but they may be
slightly different physically plus having `blue in them, and so forth,
which is similar to putting a different coating on the pill. But the basic
idea is that they are essentially identical products in an economic
sense because of the underlying physical a.nd chemical characteristics,
but that, because of this differentiation created by promotion, and so
forth, the consumer thinks they are different.
Senator NELSON. Soap would be one of the items that you refer to
as homogeneous products; is that it?
Dr. MUELLER. it is very much differentiated in the minds of con-
sumers. You pay a tremendous difference in the price of Tide, say, and
Safeway's detergent. And the only pdint I was making here, and I
PAGENO="0263"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1813
probably introduced an element of confusion by doing so, is that in
some cases there are slight physical differences as well as this psycho-
logical difference created in the mind of the consumer; whereas, in
ethical drugs most frequently the products are physically and chem-
ically identical.
Senator NELSON. Go ahead.
Dr. MUTELLER. Perhaps the most pervasive factor blocking effective
price competition in drugs is the presence Of substantial product differ-
entiation of branded drug items. A recent econometric study demon-
strates that advertising and promotion-created barriers to entry are
the single most important explanation for differences in profit rates
in American industry. The drug industry was among the industries in-
cluded in that study.
Professor Seymour Harris of Harvard University pretty well sum-
marizes the conclusions of academic scholars concerning the organiza-
tion and performance of the drug industry:
Many are concerned that at industry which comes close to being a public utility
achieves the highest profits in relation to sales and investment of any industry;
is highly concentrated in its control of the market; reveals serious monopolistic
trends; increases the ~ost to consumers by differentiating the product at a dizzy
pace, with the differentiated product usually similar to or identical with exist-
ing products; and greatly inflates the cost through record expenditures on selling.
The competition among companies to overwhelm the doctors by repetitious and
often misleading advertising, and a failure to give as much publicity to the bad
side effects as to the immediate beneficial effects, are unfortunate. Thus competi-
tion forces even highly moral firms to become less ethical in their behavior. In
the drug industry the relation of labor to total costs is minimal; and like the
soap and tobacco industries, using similar selling techniques, their relation of
labor to value added is a minimum-selling expenditures and profits are the large
items in gross receipts.
The cost of drugs is too high. I say this, though I am aware that the research
contributions of the industry are important and that the lives saved, the suffer-
ing averated, `and the acceleration of recoveries `are worth more than the $4
billion spent on drugs. But the cost could be substantially less.
Does this mean that risk plays no role in high drug profits? Not
necessarily. Although these high profits can be explained by the struc-
tural characteristics of the industry-namely, high concentration,
high entry barriers, and a high degree of product differentiation-it is
conceivable that risk also played some part. Conrad, Plotkin, Mark-
ham and Cootner testified that they believed high drug prices and
profits were due primarily to uniquely~ high risks assumed by large
drug manufacturers, and that the Conrad-Plotkin study measured the
magnitude of this risk. The relevant question, of course, is how much
of total profits can he attributed to the risk factor. Let us therefore
turn to the em~irical evidence on this subject.
Mr. GROSSMAN. Dr. Mueller, before you do that, I wonder if I could
take you back to Professor Harris' quote?
In the first line he says:
"Many are concerned that an industry which comes close to being a
public utility," et cetera.
* How do you feel about that?
Do you think the drug industry should be treated as a public utility?
Is it close to being so?
Should it `be?
Dr. MUELLER. I think what he is referring to is that one character-
istic of a public utility is that it is providing a service to the consumer
PAGENO="0264"
1814 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
which is especially unique, is required for matters of safety and public
welfare. And this, I think is the definition he is using.
Mr. GROSSMAN. When Mr. Squibb testified he said that this was a
possible end result if the drug industry did not do something on its
own, that we would eventually treat them like public utilities. Do you
think the time to do this is now?
Dr. MUELLER. I personally am among those who are very reluctant
to see more direct regulation of American business of the kind we have
in the railroads and transportation, and so on. It presents a great many
difficulties. And my personal inclination is to try to make competition
work wherever it is at all possible, and I would hope that there are
solutions to this problem short of the public utility approach.
Mr. GROsSMAN. Thank you.
Senator NELsON. Is it your opinion that if there were-whatever the
word means-effective competition at the retail level, that this would
basically resolve the problem that we are talking about here?
I realize that this involves a question of fact and all kinds of other
things.
Dr. MUELLER. If we could achieve effective competition, I think it
would resolve the problem. This gets into the question of how you
achieve it and whether it can be achieved.
Mr. GRoss~rAx. I wonder if you could explain what you mean by
high seller concentration?
Dr. MUELLER. The fact that there are very few sellers of some drugs.
Mr. GROSSMAN. Thank YOU.
Senator NELSON. I believe you address yourself to that particular
question a. little later on, with some comparisons with the auto industry
and so forth.
Dr. MUELLER. In terms of seller concentration?
Senator NELSON. Yes.
Dr. MUELLER. No, sir. I think earlier in my summary of what people
have said of the industry I made some references to concentration.
Senator NELSON. I meant lack of competition, because one or two
or three companies may be the only producers of a product.
Dr. MUELLER. That is correct..
I might ask, with your permission, that the fairly extensive foot-
notes and so on which I ha.ve in my prepared statement be incorporated
in the record. I tried to cut down on my statement by putting a good
deal of the underlying data in the footnotes and references.
Senator NELSON. The full statement, including the footnotes, will be
included in the record, and at anytime the footnotes seem to be irn-
portant to an understanding or appreciation of your ba.sic text, you
may read them or explain them extemporaneously or however you see
fit.
Dr. MUELLER. Thank you.
One way of gaining insight into the question of "risk" is to look at
what investment analysts tell investors about the drug industry. This
may seem to be a rather homespun approach to the problem. but after
all it is what investors believe about an industry that determines
investment decisions.
A perusal of studies by investment analysts indicates that they
generally advise investors that the drug industry is a rapid growth,
PAGENO="0265"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1815
high profit industry where established firms hold a strong position
relative to small companies and potential entrants. The industry is
frequently described as "depression resistant" because, as one analyst
put it, "Illness is no respector of business cycles, and Americans have
shown that they will buy the medicine they need regardless of eco-
nomic conditions." These `characteristics are considered to make drug
stocks good "defensive" investments. As a result, drug stocks sell at
relatively high price-earnings ratios, `indicating that investors are
confident of a high future payout. A Standard & Poor's analysis of
drugs summed up the factors affecting drug stocks as an investment
as follows:
Shares of drug equities have historically sold at relatively high price-earnings
ratios, owing to the industry's recession-resistant characteristics, its above-
average earnings growth rate, and its strong underlying position. Moreover, it
is difficult to enter the drug field.
Investment analysts generally emphasize that the high earnings of
drug companies make drug stocks a good buy. This is not to imply,
of course, that investment analysts view the `industry as completely
riskiess. The staff of Moody's Investors Service, after explaining a
number of reasons why drug stocks were a good investment, stated:
The drug industry cannot be risk-free. The postwar years have seen periods
of slowdown, and individual companies have' suffered temporary setbacks. The
causes have been many. Competition has led to price-cutting in popular prod-
ucts, such as penicillin, where capacity has been overexpanded. Occasionally, a
profitable new drug is found to have unsuspected and unfavorable side effects.
This, however, is less of a problem than product obsolescence or the expiration
of patents on major drugs that have been exclusive with one company. In recent
years, government regulation has been tightened at the drugmaker's expense.
Finally, the ebb and flow of respiratory diseases often causes sharp fluctuations
in drug sales.
But after enumerating the above points, the Moody's analysts
continued:
The impact of such development has caused only temporary deviations in a
growth curve that has pointed strongly upward.
Thus, while the drug industry faces uncertainties and problems,
from an investor's standpoint these "risks" apparently are no greater
than those found in many other industries. On the contrary, drugs are
considered a sound growth investment.
Investment analysts frequently make mention of the fact that
drug profits may be adversely affected by factors that will increase
price competition and thereby erode high profits. The most frequent
reference of this sort is the observation that anything threatening to
increase the use of generic drugs as opposed to brand name drugs
threatens high profits. An analysis by "Value Line" of the possible
effects of medicare on drug profits is typical of investment analysts'
views on the subject. After explaining that medicare very probably
would increase drug sales, "Value Line" concluded that drug profits
would not go up by a corresponding amount because:
Hospitals and institutions usually, wherever possible, buy generic name
drugs rather than brand names in order to reduce costs. The most profitable
business for the drug manufacturers is that which comes through drugstores,
where drugs are prescribed on a brand name basis. [Emphasis added.]
PAGENO="0266"
1816 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. May I interrupt, just for clarification of a point?
You state that the hospitals and institutions whenever possible buy
generic drugs rather than brand names in order to reduce cost. Isn't
it the actual situation that hospitals and the Government use the
formulary system when they purchase drugs. They take bids on a
generic basis, and then the contracts are made with companies that
produce only generic drugs and companies that produce both brand
names and generics. Brand name companies bid their brand, but
here is where you find the example about the effect of competition on
prices: They may very frequently win; they may very frequently
underbid the generic company in the institutional field, that is, Govern-
ment and hospitals, and therefore the trade name drug may very fre-
quently be used. But here is where you see the contrast between the
price that is charged by the brand named company in the competitive
field versus the price they charged in the retail field. In the retail mar-
ket they can hold their market price against a lower priced generic
drug, hut they can't hold it against a generic drug in a Government
purchase unless they substantially reduce the price when they bid.
Is that a correct statement?
Dr. MUELLER. That is correct. I think investment analysts view the
end effect of this process.
This observation, of course, is concerned with how medicare might
affect competition because of the increasing use of generic drugs; it
is not explaining profits associated with risks.
In sum, there is no reason to conclude, on the basis of advice being
given investors by investment analysts, that the drug industry is a
uniquely risky industry. On the contrary, the generally glowing reports
of investment analysts suggest. t.ha.t large dnig companies should have
little difficulty obtaining adequate capital should they choose to go
into the market for it. Actually, however, their profits are so large
that drug companies seldom need go to the capital market for equity
capital. And there is no reason to expect that drug companies would
have difficulty getting adequate capital even if they enjoyed profit
rates comparable to most other American industries.
But perhaps this is a too prosaic approach to the problem. Let. us,
therefore, turn to the Conrad-Plotkin-Markham- Cootner "economet-
ric" explanation of high profits in the drug industry.
My comments today concerning the Conrad-Plotkin analysis will be
limited to an evaluation of the testimony presented to this conunittee
last month. Arthur D. Little, Inc. has promised to provide us with the
underlying data used in their analysis.
Mr. GROSSMAN. Dr. Mueller, before you go on: At what point do
you think investors would not invest? How low do we have to go in
market percentages before we lose the investors?
Dr. MUELLER. In the drug industry?
Mr. GROSSMAN. Yes.
Dr. MUELLER. I am not sure just where that level is. There are com-
panies that have had fairly low earning experiences that have been able
to obtain capital funds.
I think you will recall Dr. Whitney's statement that in the early
1950's drug companies were investing at a greater rate than other
companies, by a. 3-to-i ratio, or something like that. Actually, at that
PAGENO="0267"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1817
time drug profits on the average were very similar to the average of
all American industry.
So, it is difficult to know precisely where this level is. I have little
doubt in my own mind that it is considerably below the 21 percent
level the large companies have earned in recent years, and there is
not any persuasive evidence to argue that they would have difficulty
getting capital if their profits were the same as most other American
industries.
Mr. GROSSMAN. In the same line of questioning, what do you think
is a fair return for the drug industry? We are talking in terms of
social responsibilities.
What would you say would be a fair profit?
We have had a lot of discussion about the percentage return. Is
there any specific percentage?
In other words, should we say: You can't go over 15 percent?
Isn't this all rather vague?
Dr. MUELLER. Yes, unless you are going to take a public utility
approach which I find, personally, repugnant. Until we try other
methods, I would not say that any particular percentage is appro-
priate. I would like to see the industry exposed more to the breezes of
price competition and see what reasonable rate of return would be
generated by competitive market forces, rather than to set arbitrarily
some spec~flc rate.
Senator NELSON. Is there any evidence that during the period you
refer to, 1950-55, the drug industry was receiving profits fairly com-
parable to the average in all industries. Did the industry experience
any difficulty in attracting investment capital if they needed it?
Dr. MUELLER. I just have not had an opportunity to look into that
point. I think it would be interesting to see what the experience was.
The observation I make, though, is that there are many examples of
companies that were making profits by which this 21 percent standard
would be very modest~, in fact quite mediocre profits, or close to those
earned by the rest of the American industry on the average. Yet they
have been able to generate capital on their own or borrow it and grow
effectively.
So, simply asserting that because someone is earning a high rate
of return that this must somehow demonstrate that they must earn
such a return in order to continue to grow is to just define the problem
out of existence. And I am afraid that is what Professor Whitney
might have been doing in his exchanges with you; at least that was
my impression as I read the testimony last night.
Mr. GROSSMAN. Do you give any stock to his point about the rail-
roads, how they prospered for 30 years?
Dr. MUELLER. I think he is factually correct, that they did profit
for 30 years.
Mr. GROSSMAN. Then, I assume that there is a risk, and we will have
to wait for 50 years before we can ever know?
Dr. MUELLER. I would not wait on the problems in the drug indus-
try for 50 years to find out whether or not Professor Whitney's theory
is correct.
With your permission, we will provide a brief supplemental mem-
orandum to the committee should we have any additional observations
after reviewing these data.
PAGENO="0268"
1818 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. If we thought it would `be of value, would YOU be
prepared to submit the supplemental data to the committee at a sub-
sequent date?
Dr. MUELLER. Certainly. I would be happy to.'
You will recall that Conrad and Plotkin attempted to test the hy-
pothesis that the level of an industry's profit rate is positively cor-
related with the degree of risk faced by firms withrn the industry. In
other words, the more risky an industry, the higher its average profits.
The concept of risk in investment decisionmaking theory refers to
situations where it is impossible to predict with certainty the outcome
of particular economic events. The presence of uncertainty is assumed
to affect investors' decisions. A common assumption is that investors
must be paid a "risk premium" if they have an aversion to assuming
risks. "Risk aversion" has been an underlying assumption in a number
of recent. theoretical works, particularly in the areas of portfolio selec-
tion and monetary theory. However, the assumption of "risk aver-
sion" is not a universal economic law. One need only view the crowds
at the racetrack paying for the privilege of taking a gamble to infer
that some persons regard risktaking as furnishing positive rather
than negative satisfaction. These individuals may be viewed as "risk
lovers." As a group, these risk lovers lose money at the racetrack. This
is also the case with persons gambling in commodity futures markets.
In order to explain why individuals will both purchase insurance to
guard against large losses and undertake gambles with remote possi-
bilities of achieving high returns, Friedman and Savage, have argued
that some persons regard risk-taking as furnishing positive rather
This, in a. nutshell, is what risk aversion theory is all about. But
note two important points. First, the size of the risk premium is an
empirical question. The theory tells us nothing about. the amount. of the
premium, nor even whether it is positive or negative. Second, central
to the hypothesis that is necessary to offer `a. positive premium to in-
vestors in order `to attract. adequate co~pital into a risky industry is the
idea that. risk may cause firms to incur losses, as well as to enjoy ab-
normally high profit rewards. Hence, risky industries would be char-
acterize.d by the presence of both firms with `abnormally high profits
and firms with abnormally low profits. It would be inconsistent with
risk theory if nearly all firms in an industry made very high profits
and few or none ever suffered losses.
The Conrad-Plotkin measure of risk misses this point. Risk is quan-
tified by Conrad and Plotkin by measuring the variance of individual
companies' rates of return about the industry average in a given year
and computing a simple average of these values for the 16-year period
1950-65. This measure assmues that the greater the variation in the
profit. rates of firms about the industry `average, the riskier the indus-
try. The chief conceptual shortcoming of this measure is that it does
not necessarily tell us anything about the probability `of incurring
losses. In truth, using this measure a.n industry may be defined as
risky even though all firms in it earn excessively high profits; on the
other hand, this measure may define an industry as having very low
risk even though all firms are making little or no profit. An example
will illustrate this point. By the Conrad-Plot.kin measure, the drug
I See supplemental ~taternent beginning at p. 184S. lnfra.
PAGENO="0269"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1819
industry is a high risk industry and the aluminum industry is a low
risk industry. Conrad and Plotkin's estimates of the average rate of
return, standard deviation, and variance for the two industries are as
follows, and I show that in my statement in tabular form.
The drug companies in their sample experienced an average rate
of return of 17.5 percent over the period 1950-65. The standard devia-
tion in profits around `this average was 8.6 percent. This means that the
profit ratio of roughly two-thirds of the companies in the industry fell
in `the range, 8.9 percent to 26.1 percent.
On the other hand; `the average rate of return of a group of alumi-
num companies was 7.8 percent, with `a standard deviation of 1.3 per-
cent. This means that two-thirds of the time aluminum company profits
fell in the range, 6.5 percent to 9.1 percent.
Thus, according `to Conrad and Plotkin, `the `drug industry is riskier
than the aluminum industry because `of the greater standard deviation
in the profit rates of drug manufacturers.
Just what do these facts concerning the variation in profits tell an
investor about the relative profit expecta~tions in these two industries?
They say, in effect, that there is a 2-to-i chance that profit rates
in the drug industry will fall in `a range from 8.9 percent `to 26.1 per-
cent, whereas there is a 2-to-i chance in the minimum industry
that profit rates will fall in a range from 6.5 percent to 9.1 percent.
Can anyone seriously argue that investOrs would prefer to place new
capital in the aluminum industry rath'er than in the drug industry?
The only risk that the aluminum investor is saved from is `the h'igh
probability that aluminum companies will earn less than 9 percent-
there is only 1 chance in 6 of getting more than 9 percent.
On the other hand, the risk t'he investor in the drug industry faces
is that the chances are poor that drug companies will earn a rate of
return as low as the `average return in the aluminum industry. In fact,
there is only 1 chance in 6 that they will earn a rate of return of
below 9 percent, whereas there are 4' chances out of 6 that they
will earn between 9 and 26 percent, and 1 chance in 6 that they
will earn over 26 percent. In other words, five-sixths of the time the
drug companies would be well above the `aluminum companies' average
return. Clearly, then, it is nonsense to infer from the Conrad-Plotkin
variance measure of risk `that the drug industry is riskier than the
aluminum industry in `terms of attracting new capital.
Losses, or even low profits, are practically un'heard of among large
drug companies. In this respect the `drug industry is practically
unique among important American industries. Figure 2 shows for 22
maj or industries the percent of the time the eight largest companies
fell in various profit rate categories during the period 1954-66. Large
drug companies not only earned a higher return than any other of
the major manufacturing industries shown, but none of the drug
companies ever experienced losses during the period, nor did any com-
pames experience profit rates below 5 percent. Only two other indus-
tries enjoyed this distinction, petroleum refining and cigarettes. I
might say parenthetically that I used' this 5 percent figure because
it represents `the approximate rate that someone would receive when
purchasing very `secure bonds during this period. So, in effect, this
r the upper level of very secure investn'ients. And when you receive
PAGENO="0270"
1820 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
below 5 percent, you are receiving less than you would earn in sort
of an insured, risk-free investment.
Companies in most manufacturing industries had profit rates rang-
ing between 5 percent and 15 percent. By considering this range to
`represent a sort of "norm," we can visualize the extent to which large
`drug manufacturers departed from it. Over the period 1954-66, the
eight largest drug manufacturers were in this range about 25 percent
of the time and none ever fell below it. On the other hand, 75 percent
of the time the leading drug ma.nufacturers earned profits exceeding
15 percent, and fully 17 percent of the time they had profit rates
exceeding 25 percent.
No other industry matched drugs in the frequency with which
companies had profit rates exceeding 15 percent. In only three other
industries-motor vehicles, computing machines, and aircraft-d~d
large firms have profit rates exceeding 15 percent more than 50 percent
of the time. On the other hand, all but four of the remaining 18
industries had profit rates exceeding 15 percent less `than one-fourth
of the time. Finally, some companies in all but two of the 21 industries
outside drugs earned below 5 percent at least part of the time. These
contrasting patterns cast serious doubt on the proposition that large
drug companies face a serious risk of incurring losses.
I might add parenthetically that if we take the 29 drug companies
used in the Conrad-Plotkin study, their profit performance was not
a great deal different than the eight used in that chart. The 29 com-
panies earned below 5 percent only 0.4 percent of the time. If we
had used as many companies, say 29 companies, in those other 21
industries, the red area shown on the right, these other industries-
in other words, the below 5 percent area-would be considerably wider.
Figure 3 illustrates the profit experience of leading drug companies
in each year during the 1954-66 period. Most importantly, it shows
a pattern of persistently high profits. None of these large companies
*earned below 5 percent in any year, and in only 4 years did any
`company earn between 5 percent and 10 percent; 3 of these years
~were in the beginning of the period-1954, 1955, and 1956. Since then,
in only 1962 did a company earn below 10 percent. And in the last
3 years only a small percentage earned below 15 percent.
This actual profit experience seems to fly in the face of the Conrad-
Plotkin-Markham-Cootner inference that drug manufacturing is a
uniquely risky business. The explanation, of course, is to be found
in their definition of risk. Using a different definition of risk, Dr.
Irving N. Fisher and Dr. George R Hall of the Rand Corp. concluded
4hat risk accounts for a very small portion of the high profits of
`drug companies. The findings are shown in table 3. They show that
for the period 1959-64 drug companies earned an average return
`of 18.32 percent. Fisher and Hall attributed 1.68 percent of this to
risk. They concluded that the "risk premiums" for drugs are "very
low" and that the explanation for high profits "must be sought
in factors other than risk." 1
Mr. GORDON. Mr. Plotkin testified here last month. I quote from his
Ttestimony on page 2743 of the transcript:
`The report "Risk and Corporate Rates of Return," by Dr. I. N. Fisher and Dr. G. N.
Hall, appears as Appendix II, p. 2120, infra.
PAGENO="0271"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1821
Management, it Is our contention, forms its risk expectations not solely on
the past profit history, of its own company, but also on the diversity of histories
that other companies undertaking similar ventures, to use the term again, other
companies within its industry undertake.
Would you comment on that, please?
Dr. MUELLER. Well, I think that is essentially true, that they would
look at the experience in their industry. And applying that to drugs,
they would find that if they were going to go into the industry at the
level of success of the leading concerns, that the profit experience is
very good. On the other hand, they would look at the profit experience
of the generic producer, say, who does not sell the differentiated prod-
uct, does not have a monopolistic product due to a patient, say, and
the profit experience would demonstrate a very poor prospect. So,
one must understand what it is he is looking at. And they have in
their measure tried to capture the risk factor by looking at. the (hf-
ference in the profit rates of t.he leading 29 companies. If they had
taken the top eight companies, for example, the profit variance would
be considerably less. In some other industries they have as few as five
companies. So, it turns out, that their measure is very arbitrary.
It is true that Conrad and Plotkin have found a statistically sig-
nificant relationship between their measure of risk and industry
profits. It is statistically significant in the sense that the correlations
that they came up with could not have been due simply to chance.
And in this sense it is significant statistically. This does not imply that
there is necessarily a causal relationship between the two variables
they were measuring. In truth, they have misinterpreted the causal
factors responsible for their statistical relationship.
Upon close analysis, the Conrad-Plotkin measure of risk turns out
to be a better proxy of relative market power than of risk. Their meas-
ure assumes the existence of "homogeneous" industries; that is, "in-
dustries in which all the firms product similar products, compete in
the same markets, and, in general, face the same elements of risk and
uncertainty." In fact, however, when broad industry definitions are
used, such as those in the Conrad and Plotkin study, the constituent
firms within each "industry" are frequently highly differentiated from
one another by a variety of factors. Hence, each firm in the industry
may face different risks and other factors having a bearing on profits.
This is particularly true in consumer, service, and other so-called
differentiated product industries. Because of advertising and other
factors, some firms in such industries have a pronounced and persistent
advantage over others. As a result, the most advantaged firms earn
persistently higher profits than the less advantaged firms. Such a dif-
ference between the profits of the most advantaged and least ad-
vantaged firms in an industry may provide a rough measure of the
height of the entry barriers into the industry. Economic theory pre-
dicts and empirical analysis verifies that the higher an industry's entry
barriers, the higher its profits. Hence, if intra-industry profit variance
measures the height of entry barriers, we may expect a positive statis-
tical association between industry variance and `average industry
profit rates. Thus, it is not surprising that Conrad and Plotkin find
some statistical association between intra-industry profit variance
and average industry profit rates. Unfor1~unately, they misinterpreted
the significance of their own findings.
PAGENO="0272"
1822 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The. effect of product-differentiation-created int.ra-mdustry profit
variance on the Conrad-Plotkin analysis is obvious by inspection of
specific industries with a high degree of product differentiation.
I should apologize, Senator, for this rather theoretical and, I am
sure, obscure explanation of risk.
Senator SCOTT. That is all right. We do not understand it either.
Dr. MUELLER. But I think the following illustrations explain what
I am getting at here.
Automobiles, for example, show up as a very "risky" industry in the
Conrad-Plotkin study. This is the case because of the wide disparity
in profits between the strongest, most entrenched firms and the weak-
est, marginal ones. For example, during the last 5 years automobile
companies enjoyed average profits as follows: General Motors, 21 per-
cent; Chrysler, 15 percent; Ford, 14 percent; American Motors, 6 per-
cent. During most of the 1050's while Studebaker was in the industry
it operated in the red. I think nearly all students of industria.l orga-
nization will agree with Bain that the reason for the high average
profit rates in the automobile industry is the high degree of market
concentration and the very great barriers confronting potential en-
t.rants. Thus, the persistently high average profit rates of the auto-
mobile industry are primarily due to the structure of the industry,
not its risk, as measured simply by the difference in the profit rates
of the low and high companies.
Nor are automobiles the exception. On the contrary, of the indus-
tries included by Conrad and Plotkin, eight of the nine with average
profit rates exceeding 14 percent were industries characterized by sub~
stan~t.ia.l differentiation advantage among even the largest firms, and
in each case the most advantaged firms held a substantial and persistent
profit advantage over the less advantaged firms.
The drug industry is an especially poor candidate for the explicit
assumption of the Conrad-Plotkin model that induStries must be homo-
geneous. There are great product differences among even the 2.9 drug
companies they studied. They produce varying mixes of ethical and
proprietary drugs, varying proportions of branded and generic drugs,
and then enjoy varying degrees of differentiation for their branded
drugs. All of these factors, as well as a number of others, result in
persistently higher profits for some drug companies than others. Amer-
ican Home Products, for example, not only earned average profits well
above all other drug companies, but over the entire period 1954-1966,
it had profits higher than every other firm. On the other hand, over
the same period, R.exa.li Drug had the lowest profits among the top
eight companies in all but 2 years, when it was second lowest.
Although a number of factors affect the profit differential among
drug companies, the degree of advertising-achieved-product differen-
tiat~on plays a big role. Table 4 classifies the 29 drug companies used
in the Conrad and Plotkin study by the volume of their advertising
outlays. The five companies with advertising outlays in excess of $50
million in 1966 enjoyed an average rate of return of 29.2 percent
during 1961-1965; those with advertising outlays between $10 million
and $50 million had an average rate of return of 19.7 percent; and
those spending less than $10 million earned 17.3 percent. Significantly,
all of the top five advertisers earned in excess of the average return of
those spending between $10 million and $50 million.
PAGENO="0273"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1823
The preceding reveals that much of the profit variance which Con-
rad and Plotkin found among leading drug companies is the result of
the product differentiation advantage held by some firms in the
industry.
If we are correct in believing that differences in intraindustry profit
variance actually measures differences in the degree of product dif-
ferentiation rather than risk, then Conrad and Plotkins' correlation
results may have been heavily influenced by the inclusion of highly
differentiated industries. We now turn to an examination of this
possibility.
To test the hypothesis that product differentiation caused profit
variance is largely responsible for the statistical association uncovered
by Conrad and Plotkin, we have analyzed separately their consumer
goods and producer goods manufacturing industries. Product dif-
ferentiation, of course, shows up primarily in consumer goods.
Figure 4 shows all the industries used by Conrad and PIotkin. They
lind a modest degree of correlation between intraindustry profit vaii-
ance and profits using one measure of profits rates. Using a number of
other measures they found less close relationships.
Figure 5A shows that the plotted observations of the consumer
goods industries used in the Conrad-Plotkin analysis, and figure 5B
shows the plotted observations of the producer goods industries. Among
consumer goods industries, you will dbserve, there is a quite strong
positive relationship, whereas among producer goods industries the
relationship is very weak, and is not statistically significant.
In consumer goods, 76 percent of the variation among average indus-
try profit rates is associated with the variance of intraindustry profit
rates. Additionally, the slope of : the regression line fitted to these
observations is quite steep, which means industry profit rates rise
sharply with high intraindustry profit variance.
On the other hand, when only producer goods industries are used
in the analysis, the statistical relationship is extremely weak. Only 8
percent of the variation in industry profit rates is associated with
variation in intraindustry profit variance. Moreover, the regression
line is much less steeply inclined, indicating that average industry
1)rofit rates increase very slightly with increases in intraindustry profit
variance.
These findings are extremely significant. They demonstrate that the
statistical relationship found~ *by~Conrad and Plotkin was due almost
entirely to the consumer goods industries in their sample. The fact
that no significant statistical relationship remains when only producer
goods industries are used to test their model is especially damaging to
the Conrad-Plotkin analysis. A basic assumption of their method of
measuring intraindustry risk is that the industries analyzed be homo-
geneous. Producer goods manufacturing industries are, of course, much
more homogeneous than are consumer goods industries. Hence, accord-
ing to their assumptions the "purest" relationship between "risk" and
profits should have been uncovered in the analysis of producer goods
industries. And, of course, there was none.
The close statistical relationship existing in consumer goods indus-
tries very probably results because intraindustry profit variance in
consumer industries is a rougth proxy for the height of entry barriers.
Thus, Conrad and Plotkin unwittingly have made a case for the
81-280-68-pt. 5-18
PAGENO="0274"
1824 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
inference that a substantial part of the high profits earned by drug
companies is really due to advertising- and promotion-created barriers
to entry, rather than risk. This, of course, coincides with the conclu-
sion of nearly every economist who has carefully studied the drug
industry.
Upon completing our analysis of the Conrad-Plotkin-Markham-.
Cootner explanation of risk and profits in the drug industry, I recalled
the admonition once given by the great classical economist `and `logician
John Stuart Mill. Mill cautioned economists against the pitfall of the
multiplicity of causes. We `must always be skeptical of `simple `statistical
associations among complicated economic phenomena. Professor Ken-
neth Bould'ing put it well when he said:
Some of us, perhaps, still have to learn that arithmetic is a complement to,
not a substitute for, thought, and that what my spy in IBM calls the "gigo prin-
ciple"-that is, garbage in, garbage out-is a sound approach even to the most
elegantly computerized simulation.
This more or less capsules my findings in reviewing the analysis of
drug profits and their possible association with risk. I find, to be very
brief, that the high `profit experience of the drug industry is related
`only minimally to risk and uncertainty in a casual way. On the other
hand, the high profits of the drug industry are more closely associated
with high barriers to entry of new competition. In other words, in
the classic tradition, the market power enj'oyed by drug firms has `been
achieved primarily because the leading drug companies have been able
to fence themselves off from effective competition, and in this `sheltered
position t'hey have garnered extremely high profits-profits whi~h the
economist would label as "abnormal" or "excessive," profits substan-
tially above the competitive norm.
(The complete prepared statement and supplemental statement sub-
mitted'by Dr. Mueiler follows:)
STATEMENT OF DR. WILLARD F. MUELLER, DIRECTOR, BUREAU OF Ecoxo~rics,
FEDERAL TRADE Co~rMIssIoN
Mr. Chairman and members of the Committee. It is a privilege and a pleasure
to appear before this committee. I am accompanied today by two members of the
staff of the Bureau of Economics, my assistant, Dr. Russell C. Parker, and Mr.
William H. Kelly.
My appearance today is in response to the request of your chairman `that I sub-
mit testimony on the subject of profits in the drug industry, as well as present
an independent analysis `of the study Risk and Rctnrn in American Industry-
an Econometric Analysis, presented to this committee on December 19, 1967. The
study was spoDsored by the Pharmaceutical Manufacturers Association and pre-
pared by Gordon R. Conrad and Irving H. Plotkin of Arthur D. Little, Inc., in con-
sultation with Professor Jesse W. Markham and Professor P. J. Cootner. Here-
after we shall refer to the study as the Conrad-Plotkin study.
Before turning `to the `study, we shall first place in perspective the profits of
drug Inanufacturers by comparing them with those earned by business enterprises
in other American industries.
PROFITS IN THE DRUG INDUSTRY
Figure 1 shows for 1966 the average rate of return on stockholders investments
of leading firms in 22 important American manufacturing industries.' Profit
1 This information is based on the Federal Trade Commission reports on Rates of
Return for Identical Companies in ~5elected Manufacturing Industries. The industry aver-
age is based on the 12 leading companies in each industry. The 22 industries shown in
Figure 1 are those where the 8 largest corporations had combined assets of $1 billion or
more in 1966, thereby excluding 15 smaller industries appearing in the FTC Report. None
of the excluded industries had profits as high as did the drug industry.
PAGENO="0275"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1825
FIGURE 1
RATES OF RETURN OF 12 LEAD~NG HRMS
~N SELECTED !NDUSTR~ES, 1966
DRUGS & MEDICINES
PERIODICALS
RADIO & TELEVISION EQUIPMENT
MOTOR VEH1CLES
COMPUTING MACHINES*
GLASS CONTAINERS
AIRCRAFT
NON-FERROUS METALS
MOTOR VEHICLE PARTS*
CIGARETTES*
BAKERY PRODUCTS
PETROLEUM REFINING
DAIRY PRODUCTS
RUBBER PRODUCTS
MALT LIQUORS
PULP PAPER
KNIT APPAREL
DISTILLED LIQUORS
STEEL
PLUMBING FIXTURES
CEMENT
MEAT PRODUCTS
0Excopt for Motor Vehicler and Computing Machines,
based on S firms; and Cigarettes, based on 4 firms
SOURCE: Federal Trade Commission - Rates of Return For tdenti:cslConippnies
In Selected Manufacturing Industries, 1957-1966.
15 20
NET PROFITS AS PERCENT OF STOCKHOLDERS' INVESTMENT
PAGENO="0276"
1826 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
rates of leading drug manufacturers exceeded those of large firms in the 21 other
industries. In fact, drug industry profits were twice as great as one-third of the
remaining industries; were 44 percent or more above those of all but 4 other in-
dustries; and they exceeded even such traditionally high profit industries as
motor vehicles and computing machines.
Nor was 1966 an exceptional year. Table 1 compares over the period 1950-1966,
the average profits of large drug companies and large companies in the 22 indus-
tries shown in Figure 1. Several points are of special interest. First, in the early
years, 1050-1955, average drug company profits were about equal to or somewhat
above the average of other large manufacturers. Second, beginning in 1956, how-
ever, average profit rates of drug companies were well above the average of other
large companies. Finally, since 1956 drug companies have consistently ranked
either first or second among all large manufacturing industries. This indicates
that during the last decade large drug companies have occupied an especially
advantaged position relative to large companies in other American industries.
Table 2 summarizes profit data of all drug companies and all manufacturing
companies for the period 1956-1967. Over the period covered it shows essentially
the same picture as Table 1. Since 1956 drug manufacturers have failed to oc-
cupy first place in only one year.
TABLE 1-RATES OF RETURN OF LEADING DRUG MANUFACTURERS AND ALL LEADING MANUFACTURERS, 1950-66
Ranking of leading
Year Drug industry 1 All manufacturing 2 drug companies
among all leading
manufacturing
companies
1950 19.6 17.3 6
1951 15.7 14.6 7
1952 12.7 12.7 11
1953 12.3 13.2 12
1954 12.8 12.8 8
1955 15.4 15.5 8
1956 18.2 13.8 2
1957 21.5 12.8 1
1958 20.2 9.3 1
1959 20.3 10.8 1
1960 18.4 10.3 1
1961 17.6 9.8 1
1962 17.1 10.6 2
1963 17.8 11.5 2
1964 18.9 12.3 2
1965 21.0 13.4 2
1966 21.1 13.3 1
1 Based on 8 largest companies from 1950 to 1953 and 12 largest companies from 1954 to 1966.
Based on the 8 largest companies from 1950 to 1953 and the 12 largest companies from 1954 to 1966 in each of the 22
industries shown in fig. 1 aSh the exception of the computer and motor vehicle industries in which the 8 largest firms~
were used for all years and cigarettes in which the 4 largest firms were used for all years.
Note: Rate of return after taxes as a percent of stockholders' investment.
Sources: Report of the Federal Trade Commission, "Rate of Return for Identical Companies in Selected Manufacturing
Industries, 1955-66" and Moody's Industrial Manual, 1952-54.
This pattern of persistently high profits indicates that large drug companies
occupy a unique position in the American economy. And they appear to have
become increasingly unique since the mid-1950's.
Before turning to an analysis of the Conrad-Plotkin-Markham-Cootner expla-
nation of high profits in the drug industry, I shall review briefly w-hat appears to
be a virtual concensus of opinion among researchers in the field of industrial
organization concerning the causes of high profits of drug manufacturers.
In this connection I w-ould like to emphasize the crucial role which Con-
gressiosoal hearings have played in developing the facts necessary for scholars to
study the organization and performance of the drug industry. Prior to the
Kefauver drug hearings on administered prices in 1959-1961, not a single article
concerning the American pharmaceutical industry had appeared in a professional
economic journal.2
2 In each case they ranked second to automobile manufacturers.
a Hugh Douglas Walker, "Market Power and Price Levels in the Ethical Drug Industry,"
Unpublished Ph.D. dissertation, Vanderbilt University, June 1967, pp. 2-3.
PAGENO="0277"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1827
TABLE 2--RATES OF RETURN OF DRUG
MANUFACTURERS AND ALL MANUFACTURING
INDUSTRIES,
Year
Profits after taxes as a percent of
stockholders' equity
Profit rank of the
drug industry among
all manufacturing
industries I
All drug All
manufacturers manufacturers
1956
17.6 12.3
2
1
1957
18.6 11.0
1
1958
17.7 8.6
1
1959
17.8
1
1960
1961
1962
16.8 9.2
16.7 8.8
16.8 9.8
1
1
1
1963
1964
1965
1966
16.8 10.3
18.2 11.6
20.3 13.0
20.3 13.5
1
1
2
1
1967(3quarters)
18.6 11.5
1 Rank among the 26 industries for which profits are reported separately in Quarterly Financial Reports.
Source: Federal Trade Commission and Securities and Exchange Commission, Quarterly Financial Report.
Since the first such article appeared in 1002, there has been a growing volume
of research literature on the subject, all of which has drawn heavily on Lhe
Kefauver and subsequent Congressional hearings. The facts developed by this
committee have made another enormous contribution to the fund of knowledge
concerning the drug industry. I am confident that scholars will be sifting and
winnowing the facts for years.
THE CAUSES OF HIGH PROFITS IN TIlE DRUG INDUSTRY
The preponderance of economic evidence argues that the persistently high
profits of the drug industry are the result of the absence of effective price
competition in the sale of many products. Price competition in drugs hi ineffective
for several reasons. Concentration in the production of many drugs is high
because of the patent privilege.1 And even where there are relatively many sellers,
as well as many potential sellers (for example, in the case of unpatented drugs
sold under generic names), effective prjce competition often is muted by vast
advertising, promotion, and other selling effort which differentiates in the minds
of consumers the products of the largest drug manufacturers selling under their
own brand or trade names from those of other manufacturers.° Hence, manufac-
turers selling chemically identical drugs under generic names frequently have
difficulty in selling them at any price. The resulting wide price spread between
advertised and generic drugs often applies to unpatented as well as patented
The Kefauver committee staff analyzed concentration for 51 products in the major
drug groupings: antibiotics, hormones, diabetic drugs, sulfas, vitamins, and tranquilizers.
These products represented at least two-thirds of the total value of ethical drugs in 1958.
The 115 leading drug companies controlled the production of these important products as
follows:
"In 27 of the products, or more than one-half the entire U.S. output is produced by 1
of the 15 companies. . . . In sulfa drugs, one company accounts for 100 percent of the
output in eight of the nine products. In tranquilizers the condition of monopoly prevails
in six of the seven products. In antibiotics (other than penicillin) the total output is
produced by one company In five of the nine products, and in hormones and vitamis, each,
in three out of the nine. In 8 addItional products concentration takes the form of "die-
opoly"-control by 2, while in 10 others the entire output is produced by 3 companies.
Against the typical structure of concentration in manufacturing industries, it is Indeed
remarkable that In only 6 of the 51 products are there as many as 4 producers."
Report No. 448 of the Committee on the Judiciary, U.S. Senate, made by its Subcom-
mittee on Antitrust and Monopoly, 87th Cong., First Sees., Study of .4dniinisterecl Prices
in the Drug Industry, June 27, 1961, pp. 68-69.
It well recognized that advertising and promotion effort in the drug industry is
greater than it Is In nearly all other large American industries. See, for example, William
S. Comanor and Thomas A. Wilson, "Advertising, Market Structure, and Performance:
An Empirical Analysis," Review of Economics and Statistics, November 1967, Appendix
Table 2. Of the 41 industries studied by Comanor and Wilson, all but two had lower adver-
tising-to-sales ratios than did drug manufacturers. Comanor and Wilson further point out
that advertising outlays represent less than' half of the total selling expenditures of drug
companies.
PAGENO="0278"
1828 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
drugs. These factors shelter the leading concerns, or any concern with a highly
differentiated drug product, from effective price competition.6
This explanation of high drug profits is not novel. Nearly all researchers
who have analyzed the drug industry in detail have come essentially to the
same conclusion.7 Similarly, empirical studies which cut across many industries
have identified the elements of market structure that are primarily responsible
for high noncompetitive profits.8 These elements are high seller concentration,
high barriers to entry, and product differentiation. One or more of these factors
are present in the sale of all drugs where price competition is ineffective. Per-
haps the most pervasive factor blocking effective price competition in drugs is
the presence of substantial product differentiation of branded drug items. A
recent econometric study demonstrates that advertising- and promotion-created
barriers to entry are the single most important explanation for differences in
profit rates in American industry.0 The drug industry was among the indus-
tries included in that study.
Professor Seymour Harris of Harvard University pretty well summarizes
the conclusions of academic scholars concerning the organization and per-
formance of the drug industry:
"Many are concerned that an industry which comes close to being a public
utility achieves the highest profits in relation to sales and investment of any
industry; is highly concentrated in its control of the market; reveals serious
monopolistic trends; increases the cost to consumers by differentiating the
product at a dizzy pace, with the differentiated product usually similar to or
identical with existing products; and greatly inflates the cost through record
expenditures on selling. The competition among companies to overwhelm the
doctors by repetitious and often misleading advertising, and a failure to give
as much publicity to the bad side effects as to the immediate beneficial effects,
are unfortunate. Thus competition1° forces even highly moral firms to become
less ethical in their behavior. In the drug industry the relation of labor to total
costs is minimal; and like the soap and tobacco industries, using similar sell-
ing techniques, their relation of labor to value added is a minimum-selling
expenditures and profits are the large items in gross receipts.
C In a highly important study, Professor Hugh Douglas Walker of the University of
Indiana measures the extent to which drug prices have been raised by the market power
created by patents and brand advertising. He estimates that the gross benefits of the re-
moval of both brand names and patents would amount to $617 million per year. Since
the removal of patent protection might have an adverse effect on research effort, he esti-
mates that the net benefits to society after allowing an additional research subsidy of
$192 million (the total amount financed by Industry in 1961) would be $425 million.
Professor Walker summarized his finding in a paper before the Econometrics Society
meetings, December 29, 1967. His complete analysis appears in his unpublished doctoral
dissertation, "Market Power and Price Levels in the Ethical Drug Industry," Vanderbilt
University, June 1967.
Report No. 448 of the Committee on the Judiciary, U.S. Senate, made by its Sub-
committee on Antitrust and Monopoly, 87th Cong., First Sess., t~tudy of Administered Prices
in the Drug Industry, June 27, 1961. Federal Trade Commission, Economic Report on
Antibiotics Manufacture, June 1958. Leonard G. Schifrin, "The Ethical Drug Industry:
The Case for Compulsory Patent Licensing," The Antitrust Bulletin Fall 1967, pp. 893-
915. Henry Steele, "Patent Restrictions and Price Competition in tile Ethical Drugs In-
dustry," Journal of Industrial Economics, July 1964, pp. 198-223. Henry Steele, "Mo-
nopoly and Competition in the Ethical Drugs Market," The Journal of Law d Economics,
October 1962, pp. 131-164. Seymour Harris, The Economics of American Medicine, 1964.
William S. Comanor, "Research and Competition Product Differentiation in the Pharma-
ceutical Industry in the United States," Economica, November 1964; William S. Comanor,
"Research and Technical Change in the Pharmaceutical Industry," The Review of Economics
and ~Statistics, May 1965. Frank Cacciapaglia. Jr. and Howard B. Rockmam. "The Proposed
Drug Industry Antitrust Act-Patents, Pricing, and the Public," The George Washington
Law Review, Vol. 30, June 1962, pp. 875-949. For a dissenting view see Jules Backman.
"Economics of Propietary Drugs," Annals of the New York Academy of sciences, July 14,
8 See, for example, Joe S. Bain, "Relation of Profit Rate of Industry Concentration:
American Manufacturing, 1936-40," Quarterly Journal of Economics, August 1951; Joe S.
Bain, Barriers to New Competition, Harvard University Press, 1962; L. W. Weiss, "Average
Concentration Ratios and Industrial Performance," Journal of Industrial Economics,
July 1963. Norman R. Collins and Lee Preston, "Concentration and Price Margins in Food
Manufacturing Industries," The Journal of Industrial Economics, July 1966, p. 226. A
report by the staff of the Federal Trade Commission, The structure of Food Manufacturing,
Technical Study No. 8, National Commission on Food Marketing, June 1966, pp. 202-210.
H. Michael Mann, "Seller Concentration, Barriers to Entry, and Rates of Return in Thirty
Industries, 1950-1960," Review of Economics and Statistics, August 1966, pp. 296-307.
Tjnpublished study by Norman R. Collins and Lee Preston, "Concentration and Price-Cost
Margins in Manufacturing Industries," April 1, 1966. William S. Comanor and Thomas
A. Wilson, "Advertising, Market Structure, and Market Performance," Review of Eco-
nomics and Statistics, Nov. 1967.
8 William S. Comanor and Thomas A. Wilson, ibid.
10 Professor Harris is referring to nonprice rather than price competition.
PAGENO="0279"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1829
"The cost of drugs is too high. I say this, though I am aware that the research
contributions of the industry are important and that the lives saved, the
suffering averted, and the acceleration of recoveries are worth more than the
$4 billion spent on drugs. But the cost could, be substantially less." ~`
Does this mean that risk plays no role in high drug profits? Not necessarily.
Although these high profits can be explained by the structural characteristics
of the industry-high concentration, high entry barriers, and a high degree of
product differentiation-it is conceivable that risk also played some part. Con-
rad, Plotkin, Markham and Cootner testified that they believed high drug prices
and profits were due primarily to uniquely high risks assumed by large drug
manufacturers, and that the Conrad-Plotkin study measured the magnitude of
this risk. The relevant question, of course, is how much of total profits can be
attributed to the risk factor. Let us therefore turn to the empirical evidence on
this subject.
THE RELATIONSHIP BETWEEN RISK AND HIGH DRUG PROFITS
The investment analyst's view of the drug industry
One way of gaining insight into the question of "risk" is to look at what
investment analysts tell investors about the drug industry. This may seem to
be a rather homespun approach to the problem, but after all it is what investors
believe about an industry that determines investment decisions.
A perusal of studies by investment analysts indicates that they generally ad-
vise investors that the drug industry is a rapid growth, high profit industry
where established firms hold a strong position relative to small companies and
potential entrants. The industry is freqUently described as "depression re-
sistant" because, as one analyst put it, "Illness is no respecter of business
cycles, and Americans have shown that they will buy the medicine they need
regardless of economic conditions." 12 These characteristics are considered to
make drug stocks good "defensive" investments. As a result, drug stocks sell
at relatively high price-earnings ratios, indicating that investors are confident
of a high future payout. A Standard & Poor's analysis of drugs summed up
the factors affecting drug stocks as an investment as follows:
"Shares of drug equities have historically sold at relatively high price-earn-
ings ratios, owing to the industry's recission-resistant characteristics, its above-
average earnings growth rate, and its strong underlying position. Moreover, it is
difficult to enter the drug field." ~
Investment analysts generally emphasize that the high earnings of drug com-
panies make drug stocks a good buy. This is not to imply, of course, that invest-
mnent analysts view the industry as completely riskless. The staff of Moody's
Jfl~restors Service., after explaining a number of reasons why drug stocks were
a good investment, stated:
"The drug industry cannot be risk-free. The postwar years have seen periods
of slowdown, and individual companies have suffered temporary setbacks. The
causes have been many. Competition has led to price-cutting in popular products,
such as penicillin, where capacity has been overexpanded. Occasionally, a profit-
able new drug is found to have unsuspected and unfavorable side effects. This,
however, is less of a problem than product obsolescence or the expiration of
patents on major drugs that have been exclusive with one company. In recent
years, government regulation has been tightened at the drugmakers' expense.
Finally, the ebb and flow of respiratory diseases often causes sharp fluctuations
in drug sales." 14
But after enumerating the above points, the Moody's analysts continued, "The
impact of such development has caused only temporary deviations in a growth
curve that has pointed strongly upward." 15 Thus, while the drug industry faces
uncertainties and problems, from an investor's standpoint these "risks" appar-
ently are no greater than those found in many other industries. On the contrary,
drugs are considered a sound growth investment.
11 Seymour Harris, The Economies of American Medicine, The MacMillan Company, 1964,
~ "No Cycle for Drugs," analysis prepared by the staff of Moody's Investor Service for
Dun's Review, October 1967, p. 127.
12 Standard & Poor's Industry Surveys, Drugs, Cosnietmcs-Basmc Analyses, May 4, 1967,
p. D 24.
14 "No Cycle for Drugs," analysis prepared by the staff of Moody's Investor Service for
Dun's Review, October 1967, p. 127.
15 Ibid.
PAGENO="0280"
1830 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Investment analysts frequently make mention of the fact that drug profits
may be adversely affected by factors that will increase price competition and
thereby erode high profits. The most frequent reference of this sort is the obser-
vation that anything threatening to increase the use of generic drugs as opposed
to brand name drugs threatens high profits. An analysis by Value Line of the
possible effects of `medicare" on drug profits is typical of investment analysts'
views on the subject. After explaining that "medicare" very probably would in-
crease drug sales, T7alue Line concluded that drug profits would not go up by a
corresponding amount because,
"Hospitals and institutions usually, wherever possible, buy generic name
drugs rather than brand names in order to reduce costs. The most profitable busi-
ness for the drug manufacturers is that which comes through drugstores, where
drugs are prescribed on a brand name basis." 16 [Emphasis added.]
This observation, of course, is concerned with how medicare might affect
competition because of the increasing use of generic drugs; it is not explaining
profits associated with risks. In sum, there is no reason to conclude, on the basis
of advice being given investors by investment analysts, that the drug industry is
a uniquely risky industry. On the contrary, the generally glowing reports of
investment analysts suggest that large drug companies should have little difficulty
obtaining adequate capital should they choose to go into the market for it.
Actually, however, their profits are so large that drug companies seldom need
go to the capital market for equity capital. And there is no reason to expect that
drug companies w-ould have difficulty getting adequate capital even if they en-
joyed profit rates comparable .to most other American industries.
But perhaps this is a too prosaic approach to the problem. Let us, therefore,
turn to the Courad-Plotkin-Markham-Cootner "econometric" explanation of high
profits in the drug industry.
Conrad-Piotlcin study of risk and profit rates
My comments today concerning the Conrad-Plotkin analysis will be limited to
an evaluation of the testimony presented to this Committee last month. Arthur D.
Little, Inc.. has promised to provide us w'ith the underlying data used in their
analysis. With your permission, we will provide a brief supplemental memo-
randuin to the Committee should we have any additional observations after
reviewing these data.
You will recall that Conrad and Plotkin attempted to test the hypothesis that
the level of an industry's profit rate is positively correlated with the degree of
risk faced by firms within the industry. ifl other words~, the more risky an~ indus-
try, the higher its average profits.
The concept of risk in investment decision making theory refers to situations
where it is impossible to predict with certainty the outcome of particular eco-
nomic events. The presence of uncertainty is assumed to affect investors' deci-
sions. A common assumption is that investors must be paid a "risk premium"
if they have an aversion to assuming risks. "Risk aversion" has been an under-
lying assumption in a number of recent theoretical works, particularly in the area
of portfolio selection and monetary theory.17 However, the assumption of "risk
aversion" is not a universal economic law-. One need only view the crowds at the
race track paying for the privilege of taking a gamble to infer that some persons
regard risk taking as furnishing positive rather than negative satisfaction. These
individuals may he viewed as "risk lovers." As a group, these risk lovers lose
money at the race track. This is also the case with persons gambling in cam-
modity futures markets.
In order to explain why individuals will both purchase insurance to guard
against large losses and undertake gambles with remote probabilities of achiev-
ing high returns, Friedman and Savage have argued that the same individual
may be both a risk averter and a risk lover.18
This, in a nut shell, is what risk aversion theory is all about. But note two im-
portasit points. First, the size of the risk premium is an empirical question. The
theory tells us nothing about the amdunt of the premium, nor even whether it is
1~ "Medicare: Bad for the Drug Makers ?", The Value Line Investment Survey, Edition 4,
February 12, 1965 p. 426.
17 See H. M. Markowitz, Portfolio Selection, John Wiley and Sons, New York, 1959, and
Jamcs Tobin, "LioniCity Preference as Behavior Towards Risk," Review of Economic
Studies, February 1958.
18M. Friedman and L. J. Savage, "Utility Analysis of Choices Involving Risk," Journal
of Political Economy, Vol. 56, August 1948, pp. 279-304.
PAGENO="0281"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1831
positive or negative.59 Second, central to the hypothesis that it is necessary to offer
a. positive premium to investors in order to attract adequate capital into a risky
industry is the idea that risk may cause firms to incur losses as well as to enjoy
abnormally high profit rewards. Hence, risky industries would be characterized
by the presence of both zruns with abnormally high profits and firms with ab-
normally low profits. It would be inconsistent with risk theory if nearly all finns
in an industry made very high profits and few or none ever suffered losses.
The Conrad-Plotkin measure of risk misses this point. Risk is quantified by
Conrad and Plotkin by measuring the variance of individual companies' rates of
return about the industry average in a given year and computing a simple average
of these values for the sixteen-year period 1950 to 1965.20 This measure assumes
that the greater the variation in the profit rates of firms about the industry aver-
age, the riskier the industry. The chief conceptual shortcoming of this measure is
that it does not necessarily tell us anything about the probability of incurring
losses. In truth, using this measure an industry may be defined as risky even
though all firms in it earn excessively high profits; on the other hand, this measure
may define an industry as having very low risk even though all firms are making
little or no profit. An example will illustrate this point. By the Conrad-Plotkin
measure, the drug industry is a high risk industry and the aluminum ionTustry is
a low risk industry. Conrad and Piotkin's estimates of the average rate of return,
standard deviation, and variance for the two industries are as follows:
[In percentl
Average rate of return Standard deviation 1 Risk (profit variance) 2
Drugs
Aluminum
17.5
7.8
8.6
1.3
74.2
1.6
1 The standard deviatian is defined as the positive square roit of the variance. In a normal distribution 68 percent of
the observations fall within 1 standard deviation, plus or minus, of thu average; 2 standard deviations about the average
include 95 percent of thu obsurvatines.
2 Variance measures the dispersion of obsnrvations about an average. It is computed by taking a simple average of the
squared deviations of the observatioes from the avurage.
The drug companies in their sample experienced an average rate of return
of 17.5 percent over the period 1950~1965.m The standard deviation in profits
around this average was 8.6 percent. This means that the profit ratio of roughly
two-thirds of the compasuies in the industry fell in the range, 8.9 percent to
26.1 percent.
On the other hand, the average rate of return of a group of aluminum
companies was 7.8 percent, with a standard deviation of 1.3 percent. This
means that two-thirds of the time aluminum company profits fell in the range,
6.5 percent to 9.1 percent.
Thus, according to Conrad and Plotkin, the drug indus:try is riskier than
the aluminum industry because of the greater standard deviation in the profit
rates of drug manufacturers.
Just what do these facts concerning: the variation in profits tell an investor
about the relative profit expectations in these two industries? They say, in
effect, that there is a two to one chance that profit rates in the drug industry
will fall in a range from 8.9 percent to 26.1 percent, whereas there is a two
to one chance in the aluminum industry that profit rates will fall in a range
from 6.5 percent to 9.1 percent. Can anyone seriously argue that investors
would prefer to place new capital in the aluminum industry rather than in
the drug industry? The only risk that the aluminum investor is saved from
is the high probability that aluminum companies will earn less than 9 percent-
there is only one chance in six of getting more than 9 percent. On the other
10 See, for example, Professor Bain's discussion of the effect of risk on average profit rates
of an Industry. He concludes that, "a weighted average profit rate for all firms In the
economy or In the industry (all losers as well as all winners being included) should in-
clude a true net risk return of roughly zero-and there should be no obvious risk reward
explanation of group-average excess profits." Joe S. Bain, Industrial Organization, 1959,
p. 375.
~° Gordon It. Conrad and Irving H. Plotkin, Risk and Return in American Industry, p. 12.
21 Conrad and Plotkin's computations of Industry variance and profits in 59 industries
laave been reproduced in Appendix Table 1.
PAGENO="0282"
1832 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
hand, the "risk" the investor in the drug industry faces is that the chances
are poor that drug companies will earn a rate of return ~g low as the average
return in the ~l1iffliflh1ffl lflthisfr~~ Iii fact, there is only one chance in six that
~m caril a rate of return of below 9 percent, whereas there are four
chances out of six that they will earn between 9 and 26 percent, and one
chance in six that they will earn over 26 percent. Clearly, then, it is nonsense
to infer from the Conrad-Plotkin variance measure of risk that the drug
industry is riskier than the aluminum industry in terms of attracting new
capital.
Losses, or even low profits, are practically unheard of among large drug
companies. In this respect the drug industry is practically unique among impor-
tant American industries. Figure 2 shows for 22 major industries the percent
of the time the 8 largest companies fell in various profit rate categories during
the period 1954-1966.~ Large drug companies not only earned a higher return
than any other of the major manufacturing industries shown, but none of the
drug companies ever experienced losses during the period, nor did any companies
experience profit rates below 5 percent. Only two other industries enjoyed this
distinction, petroleum refining and cigarettes.
Companies in most manufacturing industries had profit rates ranging between
5 percent and 15 percent. By considering this range to represent a sort of "norm,"
we can visualize the extent to which large drug manufacturers departed from
it. Over the period 1954-1966, the 8 largest drug manufacturers were in this
range about 25 percent of the time and none ever fell below it. On the other
hand, 75 percent of the time the leading drug manufacturers earned profits
exceeding 15 percent, and fully 17 percent of the time they had profit rates
exceeding 25 percent.
Xo other industry matched drugs in the frequency with which companies
had profit rates exceeding 15 percent. In only three other industries-motor
vehicles, computing machines and aircraft-did large firms have profit rates
exceeding 15 percent more than 50 percent of the time. On the other hand,
all but four of the remaining 18 industries had profit rates exceeding 15 percent
loss than one-fourth of the time. Finally, some companies in all but two of
the 21 industries outside drugs earned below 5 percent at least part of the
time. These contrasting patterns cast serious doubt on the proposition that
large drug companies face a serious risk of incurring losses.
Figure 3 illustrates the profit experience of leading drug companies in each
year during the 1954-1966 period. Most importantly, it shows a pattern of
persistently high profits. None of these large companies earned below 5 percent
in any year, and in only four years did any company earn between 5 percent
and 10 percent; three of these years were in the beginning of the period-1954,
1955, and 1956. Since then, only in 1962 did a company earn below 10 percent.
And in the last three years only a small percentage earned below 15 percent.23
22 This Information was obtained from the Federal Trade Commission reports on Rates of
Return for Identical Companies in Selected Manufacturing Industries. The same industries
were used as for Figure 1; namely, those in which the 8 largest companies had combined
assets of $1 billion or more in 1966. The total assets of the 8 largest firms In each industry
were grouped in 5 profit categories for each year, 1054 through 1966. The combined assets
of all firms within each category were totaled over the 13-year period and divided by the
total combined assets in all categories over the period. The result was a weighted per-
centage of the number of companies among the 8 largest firms in each industry, earning
profits in each of the 5 profit categories during the period 1954-1966.
23The profit experience of the 29 drug companies used In the Conrad-Plotkin analysis
does not differ materially from that of the eight leaders. In only two years (one company in
each case) did any of the companies eara less than 5 percent. Over the period 1954-1965,
the 29 companies fell In the various profit rate categories as follows:
Percent of
time companies
Profit Rate (Percent) : in category
25 and over 22. 5
20 to 25 13. 6
15 to 20 31. 8
10 to 15 19. 3
5 to 10 12. 7
Below 5 0. 04
PAGENO="0283"
COMPETITIVE PROBLEMS IN THE DRUG INDTJSTRY 1833
COMPUTING MACHINES
MOTOR VEHICLES
AIRCRAFT
PERIODICALS
GLASS CONTAINERS
CIGARETTES
MOTOR VEHICLE PARTS
CEMENT
BAKERY PRODUCTS
PETROLEUM REFINING
NON-FERROUS METALS
MALT LIQUORS
KNIT APPAREL
PLUMBING FIXTURES
DISTILLED LIQUORS
STEEL
DAIRY PRODUCTS
RUBBER PRODUCTS
PULP PAPER
MEAT PRODUCTS
SOURCE: Federal Trade Commission -
of Return For Identical Connpanias
In Selected Manufacturing tndustrios.
II.o~
15.2~
13.3~'
FIGURE 2
DISTRIBUTION OF RATES OF RETU:RN OF 8 LARGEST FIRMS
BY PROFIT RATE CATEGORY
22 MAJOR MANUFACTURING INDUSTRIES, 1954 - 1966
Industry :cd;~i:Ae~c:~cun Percent Distribution of Company Annual Profit Ratios
DRUGS & MEDICINES- 18.5~ ________________
15.5~
~7~4
RADIO & TELEVISION EQUIPMENT 19 1~
~!~*
//~//!!!
I
11
riIrzsrIJf,A
13.4%1'~//~~
13.3%/~,~If#%
10.9%IU~/
12.3%'W~~4
9.5~ 1~ I~-~
11.6~
10.3% -
9.2%~j
7.7%
7.O%i
6.7%1 I'
9.2%J
11.4%U
1O.8%~
10.2%~
5.6%~
80 100
0 20 40 60
KEY:U Above 25% Ota5%
~15%to25% ~Loss
5% to 15%
PAGENO="0284"
1834 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
~tv~r~
t~c~ ~ StA
///f//~
~
~Wa~~!P~
~aS
a -,
-
~t1~
r~
111
II
air
/,42á4//~/i///
lilt
ivtai~~~~
~JJj/~
~!
VAr
FIGURE 3
DISTRIBUTEON OF RATES OF RETURN OF S LARGEST FIRMS
IN THE DRUG INDUSTRY, 19541966
Veer Industry Averege Pe:ceet Distribution of Corupony Aenuol Profit Pete
Prof it Rote
1954 12.6%
1955 15.4%
1956 18.2%
1957 21.5%
1958 20.2%
1959 20.3%
1960 18.4%
1961 17.6%
1962 17.1%
1963 17.8%
1964 18.9%
1965 21.0% _______ ~pyyyy~J rp,,,jj /f/~~f
1966 21.1% ___________________________________ ________________ ~SI4~ ~
F/YVJ~~i1
~f7I~/%'
20 40 60 80
KEY: Above 25% fl 10% to 15%
SOURCE: ofRethrnFo~d:~~Cpg~ryruo. 20% to 25% 5% to 10%
teOotodedMoeotootorieo_tvdostva 15% to 20%
~avi
too
~iA/
PAGENO="0285"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1835
This actual profit experience seems to fly in the face of the Conrad-Plotkin-
Markbam-Cootner inference that drug manufacturing is a uniquely risky
business. The explanation, of course, is to be found in their definition of risk~
Using a different definition of risk, Dr. Irving N. Fisher and Dr. George R.
Hall of the Rand Corporation concluded that risk accounts for a very small
portion of the high profits of drug companies.24 The findings are shown in Table
3. They show that for the period 1959-1904 drug companies earned an average
returis of 18.32 percent. Fisher and Hall attributed 1.08 percent of this to risk.
They concluded that the "risk premiums" for drugs are "very low," and that
the explanation for high drug profits "must be sought in factors other than
risk."
It is true that Conrad and Plotkin have found a statistically significant
relationship between their measure of risk and industry profits. But they
have misinterpreted the casual factors responsible for their statistical
relationship.
TABLE 3-FISHER & HALL ESTIMATES OF AVERAGE INDUSTRY RISK PREMIUMS
[In percent]
Average observed Risk-adjusted rate Average risk
Industry group rate of return of return premium
Drugs 18.32 16.64 01.68
Aerospace 15.70 13.35 02.45
Chemicals 14.09 11.31 02.78
Petroleum 11.47 10.26 01.21
Rubber 10.96 10.21 00.75
Fond 10.72 09.15 01.57
Electrical machinery 11.96 08.57 03.39
Automotive 14.77 07.54 07.23
Office machinery 14.08 07.24 06.84
Steel 08.25 07.03 01.22
Textiles 07.89 05.94 01.95
Source: Irving & Fisher and George R. Hall, "Risk and Corporate Rate of Return," paper presented before the Econo-
metrics Society, Dec. 29, 1967.
Upon close analysis, the Conrad-Plotkin measure of risk turns out to be a better
proxy of relative market power than of risk. Their measure assumes the exist-
ence of "homogeneous" industries; that is, "industries in which all the firms
produce similar products, compete in the same markets and, in general, face the
same elements of risk and uncertainty." 20 In fact, however, when broad industry
definitions are used, such as those in the Conrad and Plotkin study, the constitu-
ent firms within each "industry" are frequently highly differentiated from one
another by a variety of factors.27 Hence, each firm in the industry may face dif-
ferent risks and other factors having a bearing on profits. This is particularly
true in consumer, service, and other so-called differentiated product industries.
Because of advertising and other faètors, some firms in such industries have
a pronounced and persistent advantage over others. As a result, the most ad-
vantaged firms earn persistently higher profits than the less advantaged firms.
Such a difference between the profits of the most advantaged and least advan-
taged firms in an industry may provide a rough measure of the height of the
entry barriers into the industry.n Economic theory predicts and empirical anal-
ysis verifies that the higher an industry's entry barriers, the higher its profits.25
24 Hall and Fisher, "Risk and Corporate Rate of Return," paper presented at the meet-
ings of the Econometrics Society, December 30, 1967. Their complete study appears in
Risk and the Aerospace Rate of Return, The Rand Corporation, Santa Monica, California,
December 1967. Hall and Fisher measure risk as the variance of the profit rates of com-
panies overtime taking into account trends in profit rates.
22 Hall and Fisher, op. cit., p. 16.
~° Fisher and Hall, Risk and the Aerospace Rate of Return, op. cit., p. 31. FIsher and
Hall conclude that not only does Conrad and Plotkin's measure of risk Involve serious
practical measurement problems, but that it also "does not fully agree with a reasonable
theoretical notion of risk." Ibid.
27 Ibid.
28 Joe S. Bain, Barriers to New Competition, 1962. The difference In the profits of the
most advantaged and least advantaged firms most accurately measures the height of
entry barriers when the least advantaged firms earn only a "normal" profit.
20 Ibid.
PAGENO="0286"
i~3b COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Hence, if intra-industry profit variance measures the height of entry barriers,
we may expect a positive statistical association between industry variance and
average industry profit rates. Thus, it is not surprising that Conrad and Plotkin
find some statistical association between intra-industry profit variance and aver-
age industry profit rates. Unfortunately, they misinterpreted the significance of
their own findings.
The effect of product-differentiation-created intra-Industry profit variance on
the Conrad-Plotkin analysis is obvious by inspection of specific industries with a
high degree of product differentiation. Automq1jjIe~, f9P 9~Ample, show up as a
very "risky" industry in the COfl~a~-Plotkin stuCy. This is the case because of the
WhO dieparity in proflts between the strongest, most entrenched firms and the
weakest, marginal ones. For example, during the last five years automobile
companies enjoyed average profits as follows: General Motors, 21 percent; Chrys-
ler, 15 percent; Ford, 14 percent; American Motors, 6 percent. During most of
the 1950's while Studebaker was in the industry it operated in the red. I think
nearly all students of industrial organization will agree with Bain ~° that the
reason for the high avcraçje profit rates in the automobile industry is the high
degree of market concentration and the very great barriers confronting potential
entrants. Thus, the persistently high average profit rates of the automobile indus-
try are primarily due to the structure of the industry not its risk.
Nor are automobiles the exception. On the contrary, of the industries included
by Conrad and Plotkin, eight of the nine w-ith average profit rates exceeding 14
percent w-ere industries characterized by ~ubstantia1 differentiation advantage
among even the largest firms, and in each case the most advantaged firms held
a substantial and persistent profit advantage over the less advantaged firms.3°
The drug industry is an especially poor candidate for the explicit assumption
of the Conrad-Plotkin model that industries must be homogeneous. There are
great product differences among even the 29 drug companies they studied. They
produce varying mixes of ethical and proprietary drugs,32 varying proportions of
branded and generic drugs, and they enjoy varying degrees of differentiation for
their branded drugs. All of these factors, as well as a number of others, result in
persistently higher profits for some drug companies than others. American Home
Products, for example, not only earned average profits well above all other drug
companies, but over the entire period 1954-1966, it had profits higher than every
other firm. On the other hand, over the same period, Rexall Drug had the lowest
profits among the top eight companies in all but twe years, when it was second
low-est.
Although a number of factors affect the profit differential among drug com-
panies, the degree of advertising-achieved product differentiation plays a big
role. Table 4 classifies the 29 drug companies used in the Conrad and Plotkin
study by the volume of their advertising outlays. The five companies with adver-
tising outlays in excess of $50 naillion in 1966 enjoyed an average rate of return
of 29.2 percent during 1961-1965; those with advertising outlays between $10
million and $50 million had an average rate or return of 19.7 percent; and those
spending less than $10 million earned 17.3 percent. Sigiilficantly, all of the top
5 advertisers earned in excess of the average return of those spending between
$10 million and $50 million.83
The preceding reveals that much of the profit variance which Conrad and
Plotkin found among leading drug companies is the result of the product differ-
entiation advantage held by some firms in the industry.
If we are correct in believing that differences in intra-industry profit variance
actually measures differences in the degree of product differentiation rather than
risk, then Conrad and Plotkins' correlation results may have been heavily in-
fluenced by the inclusion of highly differentiated industries. We now turn to an
examination of this possibility.
3° Ibid., p. 169.
31 The industries are: radio-TV broadcasters. book publishing, drugs, cosmetics, automo-
biles. radio-TV manufacturers. confectionary and soft drinks.
32 It is generally recognized that profits are more stable for highly differentiated
proprietary drugs than for ethical drugs. See, for example, James Bolog, "Forecasting
Drug Earnings." Financial Analysts Journal, July-August 1966, p. 39.
3° Some companies below the top five earn persistently high profits because they enjoy
a strong position in one or two products. Smith, Kline and French has consistently earned
high profits, in recent years averaging over 30 percent. Its two specialty items are
Thorazine and Compazine, made and sold under license from Rhone-Paulenc of France.~
Sehifrin. op. cit., p. 911.
PAGENO="0287"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1837
TABLE 4-ADVERTISING OUTLAY AND AVERAGE EARNINGS OF 29 DRUG MANUFACTURERS
Advertising expenditures in 1966
Number of
companies
Advertising
as percent
of sales
Average
earnings on
stockholder
investment 1
(percent)
Standard
deviation
(percent)
Over $50,000,000'
$10,000,000 to $50,000,000 2
Under $10,000,000
Total
5
9
15
29
24.1
11.9
(3)
29.2
19.7
17. 3
20.1
5.8
6.6
8. 3
8.6
1 Simple average of after-tan earnings for the perind 1961-65.
2 Drug companies included in the Conrad-Plotkin study with advertising expenditures of $10,000,OJ0 or more in 1966,
an reported in Advertising Age, Aug. 28, 1967, p. 36.
Not available.
Relationship between product differentiatiOn and profits
To test the hypothesis that product differentiation caused profit variance is
largely responsible for the statistical association uncovered by Conrad and
Plotkin, we have analyzed separately their consumer goods acid producer goods
manufacturing industries. Product differentiation, of course, shows up prisnarily
in consumer goods.
Figure 4 shows all the industries used by Conrad and Plotkin. They find a
modest degree of correlation between in'tra-industry profit variance and profits
using one measure of profit rates.33
Figure 5A shows the plotted observations of the consumer goods industries used
in the Conrad-Plotkin analysis, and Figure SB shows the plotted observations of
the producer goods industries34 Among consumer goods industries there is a
quite strong positive relationship, whereas among producer goods industries the
relationship is very weak, and is not statistically significant.
~ Conrad and Plotkin test their theoretical analysis against a number of measures of
rates of return and risk. They report the results of fitting eight distinct models in Appendix
Tables D-1 and D-2 of their report and refer to others in tine text. Most of the discussion
and the conclusions of tine report, however, are based primarily on the two measures which
resulted in tine l)est fitting models.
The first and most important of these is a model in which rate of return is defined as
net profit plus fixed charges over total capitalization. The risk coefficient is the average
yearly intra-industry variability of company profit ratios for the period 1950-1965. The
mechanics of computation were to calculate a variance of company profit ratios about the
industry average profit ratio in each year of the 16-year period and then compute a
simple average.
Conrad and Plotkln used the coefficient of determination (R2) to evaluate the goodness
of fit of their simple correlation-regression models. The R2 for the first model was .46.
This means that 46 pecent of the difference in industry profit levels was associated with
the intra-industry variance in profits. Corresponding R2 values for two other models using
book value rates of return were .31 (net income as a percent of total assets) and .18 (net
income as a percent of common equity).
In addition to the models using book value measures of rates of return, Conrad and
Plotkin also apply their analysis to returns to stockholders calculated as a percent of
the market value of outstanding stock. This ratio is defined as the total of dividends,
fixed charges, and the change in market value of shares of common stock during tine year
divided by the market vnlue of common shares outstanding at the beginning of the year.
The risk coefficient computed on the basis of this earnings ratio is suggested as a measure
of the market risk to present and prospective purchasers of outstanding shares of stock.
There is apparently no claim that it is an indicator of the company's ability to attract
new investment financing or that it indicates tine way in which real resoinrces in the
economy should be allocated efficiently between industries. We concur with this. The
relevant comparison in evaluating an industry's ability to attract new capital should
be expected earnings relative to cost of capital. The market value of outstanding stock
may bear little relationship to a company's earning on invested cnpital because the market
value of outstanding stock has capitalized into it expected future income resulting from
monopoly profits. Because of this, a firm enjoying the benefits of monopoly and earning
profits well in excess of the cost of capital may exhibit only a normal rate of return per
dollar of equity measured at market prices.
~ We have excluded the following industries from our analyses because none, strictly
speaking, is a manufactured product: Radio-TV broadcasters, book publishing, publishing,
trucking, eating places, department stores, apparel retailers, air transport, shipping, variety
retailers, food retailers, and financial institutions. Leading firms in all of these industries,
with the possible exception of shipping and trucking, enjoy varying degrees of product
differentiation. Therefore these industries do not meet the homogeneity assumption of the
Conrad-Plotkin model. They therefore most appropriately should be grouped with the
consumer goods industries shown in Figure 5A. When the above 12 industries are included
with the 15 consumer goods industries plotted in Figure 5A, the resulting R2 is 0.63.
PAGENO="0288"
[NDWTRY Figure 4--Ratcof return and intra-industry variance
RATE in rate of return, 59 industries, 1950-1965
of 20
RETURN
C)
C
s * Y~8.6÷,QS3X
E~p1ainec1 variance
F .46
2
o t .._._.l ` -L Z
ID 20 30 CC' 50 60 70 80 90 100 110 120 130 ~
Industry rate of retui-n = (net income + fixed charges)/total capitalization Iiitra-industry variance in rate of CI)
return
Intra-industry variance is the 16-year simple average of the variances of
company profit rates about their industry views in each of the years.
PG represents position of pharmaceutical industry.
Source: Gordon R. Conrad and Irving H. Plotkin, Risk and Return in American Industry
Figure 1, p. 1. The plotted observationsThppear in Appendix Table 1 of t~it'statement.
PAGENO="0289"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
FIGURE 5
Relationship between industry average rate of return and
intra-industry variance in rate of return
A. 15 Consumer Goods industries
10 20 2(~
Source: Appendix Table 2.
0(5
00 ItS
`Ott SU ontra-industry variance
in rate of retttrn
1839
industry mate
of return I
Industry rate
of return I
B. ~2 Prod~~er Goods Tnddutrie~
15
10
-~J------~--
- * Regression line
~ I 0~ Y~9.6t.03K
0.~
, f Iigplained variance
R~,08
81-280-68-pt. 5-19
PAGENO="0290"
1840 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In consumer goods, 76 percent of the variation among average industry profit
rates is associated with the variance of intra-industry profit rates. Additionally,
the slope of the regression line fitted to the observations is quite steep, which
means industry profit rates rise ~harp1y with high intra-industry profit variance.
On tile ~th@r liflncl, when on'y producer goods industries are used in the analy-
use statistical relationship is extremely weak. Only 8 percent of the variation
in industry profit rates is associated with variation in intra-industry profit vari-
ance. Moreover, the regression line is much less steeply inclined, indicating that
average industry profit rates increase very slightly with increases in intra-indus-
try profit variance. These findings are extremely significant. They demonstrate
that the statistical relationship found by Conrad and Plotkin was due almost
entirely to the consumer goods industries in their sample. The fact that no signifi-
cant statistical relationship remains when only producer goods inftustries are
used to test their model is especially damaging to the Conrad-Plotkin analysis. A
basic assumption of their method of measuring intra-industry risk is that the
industries analyzed be homogeneous. Producer goods manufacturing industries
are, of course, much more homogeneous than are consumer goods industries.
Hence. according to their assumptions the "purest" relationship between "risk'
and profits should have been uncovered in the analysis of producer goods
industries.
The close statistical relationship existing in consumer goods industries very
probably results because intra-industry profit variance in consumer industries is
a rough proxy for the height of entry barriers. Thus, Conrad and Plotkin unwit-
tingly have made a case for the inference that a substantial part of the high
profits earned by drug companies is really due to advertising- and promotion-
created barriers to entry, rather than risk. This, of course, coincides with the
conclusion of nearly every economist w'ho has carefully studied the drug industry.
CONCLUSION
Upon completing our analysis of the Conrad-Plotkin-Markham-000tfler expla-
nation of risk and profits in the drug industry, I recalled the admonition once
given by the great classical economist and logician, John Stuart Mill. Mill cau-
tioned economists against the pitfall of multiplicity of causes. We must always
be skeptical of simple statistical associations among complicated economic phe-
nomena. Professor Kenneth Boulding put is well when he said, "Some of us,
perhaps, still have to learn that arithmetic is a complement to, not a substitute
for, thought, and that what my spy in IBM calls the `gigo irinciple,' (that is,
garbage in, garbage out) is a sound approach even to the most elegantly com-
puterized simulation."
This more or less capsules my findings in reviewing the analysis of drug profits
and their possible association with risk. I find, to be very brief, that the high
profit experience of the drug industry is related only minimally to risk and uncer-
tainty in a causal way. On the other hand, the high profits of the drug industry
are more closely associated with high barriers to entry of new competition. in
other words, in the classic tradition, the market power enjoyed by drug firms has
been achieved primarily because the leading drug companies have been able to
fence themselves off from effective competition, and in this sheltered position
they have garnered extremely high profits-profits w-hich the economist w-ould
label as "abnormal" or "excessive," profits substantially above the competitive
norm.
~ Kenneth Boulding, "The Economics of Knowledge and the Knowledge of Economics"
American Economic Review, May 1906, p. 10.
PAGENO="0291"
Variance (risk)
116. 859
99. 085
78. 783
74. 2 13
69. 936
67.284
59. 901
58. 127
57. 631
41. 800
34. 965
34. 338
33. 259
32. 866
29. 705
28. 321
28. 217
27. 486
27. 426
26. 777
25. 412
22. 822
21. 306
20. 535
19. 873
19. 580
19. 528
18. 637
18. 599
18. 312
16. 111
15. 485
15. 476
14. 330
14. 170
13. 321
13. 320
12. 497
11. 812
9.259
9.219
8.948
8.769
8.708
8.682
8.477
8.157
8.061
7.804
7.494
7.255
6.018
5.331
5.080
5.014
4.521
3.709
2. 899
1.579
Mean (return)
18. 929
15. 477
8.797
17. 524
10. 552
18. 726
12. 231
17. 983
15. 042
14. 699
8.754
15. 214
15. 275
8.532
7. 114
12. 949
11. 950
10. 959
10. 121
11. 348
9.831
12. 949
12. 144
8.698
11. 107
9.019
9. 493
11. 852
10. 002
8. 547
10. 396
8.007
9. 199
7.878
12. 453
9.242
7.209
9.849
6.224
10. 764
11. 137
6.625
12. 177
11.666
8. 591
7.519
9. 443
10. 082
10. 819
10. 634
9.689
9.531
8. 304
9.673
8. 534
7.410
8.091
9. 546
7.778
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 1841
APPENDIX TABLE 1.-INDUSTRY VARIANCE AND RETURN BASED ON BOOK VALUE
Industry
1. Radio-TV broadcasters
2. Publishing books
3. Gold mining
4. Drugs
5.Publishing
6.Cosmetics
7. Aerospace
8.Automobilu
9. Radij-TV manufacturers
10. Confectionery
11. Building materials, heat
12. Beverages, soft drinks
13. Lead and zinc
14. Miscellaneous metalwork
15. Watches
16. Auto parts accessories
17. Oil, crude, producers
18. Electrical products
19. Machinery
20. Metals, miscellaneous
21. Beieruges, brewers
22. Electronic products
23.Chomicals
24.Shoss
25. Trucking
26. Machinery, metal fabrication
27.Copper
28. Eating places
29. Retail, department stores
30. Retail, apparel chains
31. Container, paper
32. Home furnishing
33. Food products
34. Air transport
35. Office and business equipment
36. Auto trucks
37. Beverages, distillers
38. Textile apparel manufacturers
39.Shipping
40.Tobacco
41. Building materials, rouf
42. Retail, variety stores
43. Vending machines
44. Building materials, cement
45. Blast furnaces
46. Textile products
47. Paper
48. Abrasive products
49. Retail, food chains
50.Oil
51. Forest products
52. Paint
53. Cnal, bituminous
54. Tire and rubber
55. Steel
56. Railroad equipment
57. Containers, metal and glass
58. Financial
59. Aluminum
Source: Gordon R. Conrad and Irving H. Plotkin, Risk and Return in American Industry, table F-i, p. 78.
PAGENO="0292"
1842 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
APPENDIX TABLE 2-A--CONSUMER GOODS INDUSTRIES RATES OF RETURN AND INTRA-INDUSTRY VARIANCE IN
PAT~Or RETUIST4, 1550-65'
Industry Ave
rage rate of return I
1950-65'
ntra-ifldustry variance
in rate of return 2
Drugs
Cosmetics
Automobile
Radio-TV manufacturers
17.5
18.7
18. 0
15.0
74.2
67.3
58. 1
57.6
Confectionery
Beverages, soft drinks
Watches
Beverages, brewers
Shoes
Home furnishings
Food products
Beverage, distillers
Textile apparel manufacturers
Tobacco
Tire and rubber
14. 7
15. 2
7. 1
9. 8
8.7
8.0
9.2
7.2
9. 8
10.8
9. 7
41. 8
34. 3
29. 7
25. 4
20.5
15.5
15.5
13.3
12. 5
9.3
5. 1
I Rate of return is defined as net income plus fixed charges divided by total capitalization.
2 The variance is the simple average of intra-industry variances for each year of the period, 1950-65.
Source: Gordon R. Conrad and Irving H. Plotkin, "Risk and Return in American Industry," Arthur D. Little, Inc.; appen-
dix table F-i, p. 78.
APPENDIX TABLE 2-B-PRODUCER GOODS AND MINING INDUSTRIES RATES OF RETURN AND INTRA-
INDUSTRY VARIANCE IN RATE OF RETURN, 1950-65
Industry
Average rate of return,
1950-65'
Intraindustry variance in
rate of return 2
Gold mining
Aerospace
Building materials, heat
Lead and zinc
Miscellaneous metalwork
8. 8
12.2
8. 8
15.3
8. 5
78. 8
59.9
35. 0
33.3
32. 9
Auto parts, accessories
Oil, crude, producers
Electrical products
Machinery
Metals, miscellaneous
Electronic products
Chemicals
Machinery, metal fabrication
Copper
Container, papnr
Office and business equipment
Autotrucks
Building materials, roof
Vending machines
Building materials, cement
Blast furnaces
Textile products
Paper
Abrasive products
Oil
12.9
12. 0
11.0
10.1
11.3
12. 9
12.1
9. 0
9.5
10.4
12. 5
9.2
11.1
12.2
11.7
8.6
7.5
9.4
10. 1
10.6
28. 3
28. 2
27.5
27.4
26.8
22. 8
21.3
19. 6
19.5
16. 1
14. 2
13.3
9.2
8.8
8.7
8.7
8.5
8.2
8.1
7.5
Forest products
Paint
Coal, bituminous
Steel
Railroad equipment
Containers, metal and glass
Aluminum
9.5
8.3
8.5
7.4
8. 1
7. 8
6.0
5.3
5.0
4. 5
3. 7
1. 6
I Rate of return is definod as not incornn plus fixed chargns divided by total capitalization.
2The variance is the simpla avoraga of intra-industry variances for each year of the period, 1950-65.
Source: Gordon R. Conrad and Irving H. Platkin, "Rink and Return in American Industry," Arthur D. Little, Inc.;
appundix table F-i, p. 78.
PAGENO="0293"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1843
SUPPLEMENT TO STATEMENT OF Da. WILLARD F. MUELLER, DIRECTOR,
BUREAU OF ECONOMICS, FEDERAL TRADE C01IMIssI0N
In my statement on January 18 before the Monopoly Subcommittee I raised
a number of questions concerning the theoretical and empirical credibility
of a study to Messrs. Conrad and P'lotkin of Arthur D. Little, Inc.1 Most im-
portantly, whereks their study concluded that differences in the level of profit
rate among industries are explainable by differences in the degree of risk faced
by firms within the industries, my analysis demonstrated that their statistical
findings wTere the result of the inclusion of consumer goods industries with a high
degree of product differentiation.
At that time I promised to provide a brief supplemental memorandum analyz-
ing the underlying data used by Conrad and Plotkin in their study. The results of
this further analysis give added support to the finding that the differences in
profit rates among industries result almost exclusively from the inclusion of dif-
ferentiated consumer goods industries in the sample. Such industries are char-
acterized both by high average profits and by high intra-industry variance in
profits.
EFFECT OF GROUPING SOME INDUSTRIES
Conrad and Plotkin, in choosing the sample of industries to be used in their
analysis, grouped together a number of industries which are defined separately
in both the Census of Manufacturers and in the data source from which they
computed their regression variables.1 One is immediately puzzled by the reason-
ing behind such an apparently arbitrary grouping procedure.8
This gPouping procedure is particularly questionable with regard to the "Food
Products" and "Machinery" industry groups. In the case of "Food Products" seven
4-digit SIC industries are grouped together; yet several other industries which
Census defines as food industries are excluded from the grouping and included in
their analysis as separate industries.4 Data for the seven industries were provided
separately in the banic data source used by Conrad and Plotkin, and six of the
seven had numbers of companies equal to or greater than the numbers of com-
panies which were included as separate industry observations. Therefore, to be
consistent these six should be included as separate observations in the sample
rather than combined into a single observation.
The seven industries included in the fOod group are listed in Table 1 along with
their individual intra-industry profit variances and average profit rates. As Table
1 shows, there are wide differences in both profit variances and average profit rates
among the seven food products industries. Average profit rates ranged between
5.561 and 12.481, and the variances of industry profit rates ranged between 2.190
and 32.698.
In the case of "Machinery" Conrad and Plotkin arbitrarily combined eight
4-digit SIC Industries into a single observation. Table 2 shows the variances in
profit rates and average profit rates of these eight machinery industries. These
industries are clearly distinct and should be treated as separate observations.
As is the case with food products, both the profit variance and the average profit
rate vary greatly among the various machinery industries. Intra-industry profit
variance ranged between 9.372 and 41.005 and average profit rate ranged be-
tween 8.212 and 14.552.
1 Gordon R. Conrad and Irving H. Plothln, Risk and Return in American Industry,
Arthur D. Little, Inc., May 1967.
2 Ibid., Appendix C, pp. 34-35.
Data were not submitted which could be used to determine time effect that the grouping
of certain industries had on the regression and correlation models reported by Conrad and
Plotkin. There is a suggestion, however, that the grouping may have produced higher
correlation coefficients by the fact that observations computed on an unweighted basis
had that effect. From the data submitted it was possible to compute simple or unweighted
variance coefficients. Using these, the explained variance or the R2 of tile equation with
grouped data was slightly higher than an equation with only ungrouped observations.
Conrad and Plotkin combined the seven industries shown in Table 1 as one observation,
and used confectionery, soft drinks, beer arid distilled liquor as separate observations.
For unexplained reasons Conrad and Photkin excluded the sugar industry from their
rinalysis although information was available on the industry. In addition to sugar refining,
Conrad and Plotkin excluded three other industries for which the Standard & Poor's
Compustat Tape of 1967 contained five or more companies and ten Industries which con-
tained either three or four companies. These latter ten industries are significant because
in some cases Conrad and Photkin combined industries that contained as few as two
companies into grouped observations.
PAGENO="0294"
1844 COMPETITIVE ?I~OELEMS I~ THE D1UJG INDUSTRY
TABLE 1.-INTRAINDUSTRY PROFIT VARIANCES AND AVERAGE PROFIT RATES OF 7 FOOD PRODUCTS INDUSTRIES
Industry category I
Number of
companies 2
Industry profit
rate
Intraindustry profit
variance 4
Food products combined
Packaged foods
Meatpackers
Dairy products
Cannedfoods
43
9.075
24.463
9
7
7
7
12.481
5. 561
9. 190
7.215
23.878
12. 590
2. 190
18.004
Bread and pastries
Cookies and crackers
6
5
9.458
10.134
32. 698
9.112
Corn refiners
2
9. 964
8. 632
tIndustry category as defined in the basic source used by Arthur D. Little, which was Standard & Poor's compustat
industrial tape.
2 Number of companies in the Arthur D. Little sample.
3 Net income plus fixed charges as a percent of total capitalization (total assets minus current liabilities). The ratios
shown are the 16-year (1950-65) simple averages of profit ratios for firms included industries.
4 The 16-year simple average of annual unweighted intraindustry profit variances. Weighted intraindustry profit vari-
ances could not be computed from the data submitted by Arthur D. Little.
Source: Data provided by Arthur D. Little, Inc.
TABLE 2.-INTRAINDUSTRY PROFIT VARIANCES AND AV
ERAGE PROFIT RATES OF 8 MACHINERY INDUSTRIES
Industry category 1
Number of
companies 2
Industry average
profit rate 3
Intraindustry
profit variance
Machiaery combined
Spacialty machinery
Industrial machinery
Machine tools
62
16
10
8
11. 165
11.450
12. 181
10. 906
33. 954
41.005
21. 386
21. 054
Construction and materials handling machinery
Oil well machinery
Agricultural machinery
General industrial machinery
Steam generating machinery
7
6
6
5
4
10. 580
13.096
8.212
14. 552
9.912
9.372
14.280
33. 349
38. 282
17. 478
1 Industry category as defined in the basic source used by Arthur D. Little, which was Standard & Poor's compustat
industrial tape.
3 Number of companies in the Arthur D. Little sample.
3 Net income plus fixed charges as a percent of total capitalization (total assets minus current liabilities). The ratios
shown are the 16-year (1950-65) simple averages of profit ratios for firms included in industries.
4 The 16-year simple average of annual unweighte'J intraindustry profit variances. Weighted iatraindustry profit vari-
ances could not be computed from the data submitted by Arthur D. Little.
Source: Data provided by Arthur D. Little, Inc.
The grouping procedure used by Conrad-Plotkin cannot be jiastified because
of the fewness of firms in the separate 4-digit SIC industries within these two
groups. The criterion which they used for including an industry in their analysis
as a separate observation w-as that the industry contain at least five companies.
Nine of their sample industries contained the accepted minimum of five coin-
panies. As Tables 1 and 2 show, in only one of the seven grouped food products
industries w-ere there fewer than five companies, and in only one of the eight
grouped machinery industries were there few-er than five companies.
These arbitrary groupings conflict with Conrad and Plotkin's testimony before
the Committee that they had not "massaged" the data.5 Their arbitrary groupings
also violate one of the basic assumptions of their theoretical m'od~l that the in-
dustries analyzed be homogeneous, i.e., that the companies within each industry
grouping are subject to the same demand and supply conditions.
Whatever their rationale for grouping these industries, the i~ocess of ungroup-
ing provides an additional test of the hypothesis that the positive relationship be-
tween intra-industry variance and industry average profit rate results pri-
° Plotkin testified on this point, "We have not changed one iota of their data, including-
their industry classifications as to which companies belong to which industries." Transcript,
Tuesday, December 19, 1967, p. 2745.
PAGENO="0295"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1845
manly from the inclusion in the sample of consumer goods industries character-
ized by product differentiation.6
Turning first to consumer gdods, the separation of the food products group
into its separate 4-digit SIO industries enables us to expand the number of con-
sumer goods observations from 15 to 20. TJsing data supplied by Arthur D. Little,
two regression equations were computed expressing the relationship between
intra-industry profit variance (X) and industry average profit rate (Y). The
first regression was computed using the sample of 15 consumer goods industries
included among the industries analyzed by Conrad and Plotlcin.7 The resulting
regression equation is:
Equation I Y=9.3+0.038X
with explained variance R2=0.26.8
A second regression equation was computed based on a sample of 20 indus-
tries. The increased number resulted from replacing the food products group
by the six food products industries having at least five sample companies.9 The
regression equation using the adjusted sample is:
Equation II Y=8.9+0.042X
with explained variance R2=0.30.
Increasing the size of the consumer goods sample raises from 26 to 30 percent
the variation among industry average profit rates associated with the variance
of intra-industry profits. In addition, the expansion of sample size increases the
slope of the regression line from 0.038 to 0.042, indicating a greater response
of industry profit rates to changes in intra-industry l)rOfit variance.10 These re-
sults substantiate our earlier finding that in the consiuner goods sector a strong
relationship between intra-industry profit variance and industry average
profit rates does in fact exist.
Turning now to the producer goods and mining sectors, the separation of
the machinery group into its 4-digit SIC industries enables us to expand the
size of the producer goods and mining sample in the same way as we did for
consumer goods. Using the Arthur D. Little data two regression equations
were computed-the first using essentially the same producer goods and mining
industries as were included in the Conrad-Plotkin sample, and the second
using an expanded sample adjusted by ungrouping the machinery industries
and replacing the single machinery group observation by the seven machinery
industry observations containing at least five sample companies.11
6 The process of expanding the number of observations increases the reliability of a
statistical relationship. That is, all other things equal, a larger sample size reduces the
probability that the relationship could have been due to chance. If, in addition, the
increase in the size of the sample also increases the strength of the relationship between
I)rofit variance and average profit rate, one can even be more confident that such a rela-
tionship exists. If the relationship is weakened when the sample size is increased, the
likelihood that such a relationship actually exists may be more suspect.
Observations are listed in Appendix Table 1A.
S Intra-industry variance data used In this equation were computed on an unweighted
basis. Data were not submitted by Arthur D. Little, Inc., to permit computing the addi-
tional values of the variables for Equation II on the weighted basis used by Conrad and
Plotkin for Risk and Return in American Industry. The use of weighted variables by
Conrad and Plotkin results in a higher R2 for consumer goods than the ones shown in
Equation I. (See Figure 5A of my January 18, 1968, statement before this Subcommittee.)
There is no reason to believe that the use of weighted observations should reverse any
of the relationships between Equation I and Equation II discussed below.
Observations are listed in Appendix Table lB.
10 The statistical reliability of these relationships was Increased because of both the
higher correlation coefficient and the enlarged size of the sample. By adding tile addi-
tional industry observation the statistical probability that the relationship could have
occurred by chance was reduced from 2.5 percent to about .5 percent.
11 See Appendix Tables 2A and *2B. Arthur D. Little did not furnish the underlying
data on gold mining, lead and zinc, miscellaneous metals, or forest products. Hence, these
observations were excluded from the two samples.
PAGENO="0296"
1846 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The regression equation for the unadjusted sample containing 26 observations
is:
Equation III Y=9.8+O.018X
with explained variance, R2=0.17.12
Computing a similar simple regression equation for the adjusted producer
goods and mining sample containing 32 observations, the resulting equation is:
Equation IV Y=10.2±0.016X
with explained variance, R2=0.10.13
Whereas the expansion of the sample size improved the statistical relationship
between intra-industry profit variance aiid industry average I)rOfit rate in the
consumer goods sector, exactly the opposite results occurred in the producer
goods and mining sectors. The explained variance, R2, decreased from 0.17 to
0.10, indicating that an even smaller percentage of the differences in industry
average profit rates resulted from the variance in intra-industry profits.14 More-
over, the slope of the regression line, which shows the response of industry
profit rates to changes in industry profit variances, decreased from its already
low level of 0.018 to 0.016.
In sum, the strong statistical relationship between intra-industry profit vari-
ances and industry average profit rates in the consumer goods sector proved to
be even stronger when the size of the consumer goods sample was increased,
while the already weak statistical relationship between variances in industry
profit rates and industry profits in the producer goods and mining sample became
still w-eaker when the size of the producer goods and mining sample w-as ex-
panded. This finding is significant. Together w-ith the findings I reported in may
January 18, 1968 statement to this Subcommittee, it demonstrates conclusively
that the statistical relationship reported by Conrad and Plotkin was due entirely
to the differentiated consumer goods industries in their sample. As pointed out
in my original statement, a basic assumption of their hypothesis is that the in-
dustries analyzed be homogeneous. The producer goods industries in their sample
most closely approximate this assumption. Yet, when only these industries are
analyzed, no significant statistical relationship exists between industry variance
and profit rates.
PROFIT RATES OF LARGEST COMPANIES AND OTHER COMPANIES
The underlying data furnished by Arthur D. Little provide additional facts
with which to test our hypothesis that the intraindustry variance in profit rates
is really a measure of the advantage which some firms enjoy over others by
reason of their success in differentiating their product. If we are correct that
in many consumer goods industries product differentiation creates advantaged
positions for the largest companies relative to smaller ones, then the largest
companies may be expected to enjoy persistently higher profits than other com-
panies.15 On the other hand, we generally would not expect the largest firms in
producer goods industries to enjoy a significant product differentiation advantage
over other companies in the industry. So, unless such firms enjoyed some other
advantage we would not expect the largest companies in such industries to earn
persistently higher profits than the remaining companies in the industry.
To test the above hypothesis, the Arthur D. Little sample of companies in each
industry was divided into two size classes on the basis of relative company size.
12 Values for X and Y in Equations III and IV are unweighted with respect to company
size. Unweighted variables had to be used in measuring the effects of adding the 7 addi-
tional ungrouped observations because the company size weights employed by Conrad and
Plotkin were not made available by Arthur D. Little. The effect of using unweighted values
rather than weighted values can be seen by comparing the R2 and the slope of Equation III
with comparable values for the equation shown in Figure 5B of my January 18, lOGS.
statement before this Subcommittee. This comparison shows that the relationship between
intra-indlustry profit variability and average industry profit rates is even less significant
when Conrad and Plotkin's weighted values are used.
11 Ibid
14 The statistical probability that the relationship could have occurred due to chance
increased from 2.5 to 5 percent.
15 Ideally, we should know the degree of product differentiation enjoyed by each firm
in the industry. Lacking this information, we have assumed that the largest companies
possess the most highly differentiated products. This assumption is generally valid becauce
the leading companies in consumer products industries are also by and large the leading
advertisers.
PAGENO="0297"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1847
In each industry, the four largest companies based on total assets were grouped
together to form one size class, while the remaining sample companies were
grouped in a second size class. If there were fewer than eight sample companies
in an industry, the sample was divided evenly between the two classes. The
simple average of profit rates for the period 1963 through 1965 was computed
for each size class of each industry. Table 3 summarizes the results of these
tabulations.
TABLE 3.-SUMMARY OF RELATIVE PROFIT RATES OF LEADING COMPANIES AND OTHER COMPANIES,
54 INDUSTRIES
Typeof industry'
Total
md
number of
ustries2
The number of md
ustrius in which the
earned 3-
leading companies
-
Higher profits than
other companies
The same as ether
companies
Lower profits than
other companies
Consumer goods
Producer goods
19
35
15
16
0
3
4
16
`Industries classified according to Federal Reserve Board listing for the Index of Industrial Production.
2 Arthur D. Little did not submit data for several industries which Conrad and Plotkin used in their analysis reported in
"Risk and Return in American Industry." For this reason the total number of industries is slightly fewer.
3 Net income plus fixed charges over total capitalization. This is the profit ratio used by Conrad and Plotkin.
Source: Tables 3A and 3B.
For the period 1963 to 1965, the leading companies in fifteen of the nineteen
consumer goods industries earned higher profit rates than the other companies
in the idnustry as a group. In sixteen of the thirty-five producer goods and
mining industries the group of leading companies had higher profit rates than
the group of smaller companies; in another sixteen industries the group of
smaller companies averaged higher profit rates than the group of leading
companies; and in the remaining three industries both groups averaged exactly
the same profit rates. Clearly, no great advantage accrued to producer goods or
mining companies by reason of their being leading companies in their respective
industries. The relationship between relative firm size and average profit rates
was random.
The results of this test again support the proposition that the leading com-
panies in consumer goods industries possess some unique advantages which do
not acme to the leading companies in producer goods and mining industries. It
is our hypothesis that these advantages are the result of the largest companies'
greater capacity to differentiate their products. As a result the companies which
are most successful in achieving a highly differentiated product are able to
charge higher prices and make higher profits than the less advantaged com-
panies. In contrast, the products of producer goods and mining industries are
more homogeneous. Consequently, relatively small firms'° enjoy profit rates about
equal to industry leaders.
DIFFERENCES IN PROFIT RATES AMONG CONSUMER GOODS INDUSTRIES
Economic theory posits that industries with a high degree of product differen-
tiation will experience both higher average industry profits and greater differ-
ences in profits between the leading companies and the remaining companies in
the industry than industries with a moderate to low degree of product differentia-
tion,12 We are now able to test this theory by employing the underlying data used
in the Arthur D. Little study.
In Table 4 the consumer goods in ustries listed in Appendix Table 3A have
been grouped into three categories on the basis of the total advertising expendi-
tures of the four leading advertisers in the industry in 1964. It shows that there
is a strong positive relationship between the absolute amount of an industry's
advertising and its average profit rates. In the top category, containing in-
dustries whose products were highly advertised, the average profit rates of both
`° The companies included in the Arthur D. Little sample generally included only large
and medium size companies. Very small companies were excluded from the sample since it
included only companies whose stocks were registered on principal stock exchanges.
`~ Joe S. Bain, Barriers to Neov Competition, Harvard University Press, 1956.
PAGENO="0298"
1848 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
leading companies and all other companies were substantially higher than those
in either of the other two categories. In fact, the smaller companies in the top
category earned higher average profits than the leading companies in the second
category containing industries characterized by moderate levels of advertising.
As the level of industry advertising declined, the average industry profit rates
also declined, dropping from 10.9 percent for the leading companies in the top
category to 11.0 percent for the industry leaders in the second category and to
8.5 percent for the leading companies in industries in the lowest category.
TABLE 4.-RELATIONSHIP BETWEEN ADVERTISING, PROFIT RATES AND INTRAINDUSTRY PROFIT VARIABILITY
FOR 19 CONSUMER GOODS INDUSTRIES
Advertising expenditures of 4 leading
advertisers, 1964
Number of
industries
Average of
industry
profit rates of
leading corn-
panies, 1963-65
Average of
industry
profit rates of
other companies
in 1963-65
Absolute
difference
in profit
rates
Average of
intraindustry
variance in
profit rates
1950-65
Over $150,000,000
$50,000,000 to $150,000,000
Under $50,000,000
`3
2 ~1
2 5
16. 9
11. 9
8. 5
12. 9
10. 0
6. 8
4. 0
1. 9
1. 7
89. 8
43. 1
29. 8
1 Include drugs, automobiles, and tobacco.
2 Include cosmetics, beer, soft drinks, confectionery, canned foods, liquor, tires, radio and TV, dairy products, bread
and pastries, and biscuits and crackers.
2 Include watches, shoes, home furnishings, apparel, and meat products.
Source: Tables lB and 3A. Advertising data were tabulated from Advertising Age, Jan. 3, 1966, p. 46.
Greater absolute expenditures on advertising not only lead to higher average
industry profit rates but to greater variances in intra-industry profit rates.
This occurs, as explained above, because the leading companies enjoy an ad-
vantaged position relative to those companies less capable of sustaining large
advertising expenditures. Table 4 show-s the difference in average profit rates
betw-een the class of leading companies and the class of other companies in
each advertising category. In the top advertising category the profit rates of
the leading companies averaged 4.0 percent higher than the profit rates of smaller
companies. In the second category the leading companies' profit rates exceeded
those of the smaller companies by an average of only 1.9 percent. In the third
category this difference declined to 1.7 percent. The persistently higher profits of
leading concerns resulted in greater variances in profits. As shown in Table 4,
the average of intra-industry variance in profit rates was 89.9 for the top
category. For the second category the variance dropped to 43.1, and it declined
still further to 29.8 for the lowest category.
These results again verify our earlier findings that the advertising-achieved
product differentiation advantages held by leading firms in an industry affect
both the higher dispersion of profit rates w-ithin the industry and industry
average profit rates.
CONRAD AND PLOTKIN'S UNDISCUSSED FINDING
Conrad and Plotkin did not discuss before the Subcommittee their statistically
most significant finding.'8 The regression models appearing in Appendix Table
D-2, page 40, Risk and Return in. American Industry, show that company profit-
ability is negatively related to profit variability. That is to say, the more profit-
able a company, the more stable are its profits. This finding is in direct con-
tradiction to their principal hypotheses that high profit firms earn such profits
as a reward for the "risk" associated with varying profit levels. Conrad and
Plotkin gave no explanation for their strange silence regarding this potentially
significant finding.
CONCLUSIONS
Our analysis of the underlying data for the Conrad-Plotkin study supplied
by Arthur D. Little provides further support for the hypothesis that the dif-
18 The explained variances (R2) and the simple regression coefficients for temporal vari-
ances in Conrad and Plotkin's models analyzing returns on total capitalization and net
income to common equity (Model II, page 40, .Rmsk and Return in American Industry) are
statistically more significant than those discussed in their text.
PAGENO="0299"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1849
ferences between average profit rates among industries studied resulted pri-
marily from the impact of product differentiation on industry profit rates. The
intra-industry variance in profit rates is really a rough measure of the size
of the product differentiation barrier to entry in an industry and not a measure
of industry risk. It measures the differenèe in profit rates between the most
advantaged firms and the least advantaged firms, this difference increasing
as the level of product differentiation increases. Hence, the Conrad-Plotkin
study unwittingly supports the inference that in consumer goods industries high
industry profits are caused by product differentiation-created barriers to entry.
On the other hand, careful statistical analysis reveals that their risk hypothesis
has no empirical foundation.
TABLE 1A.-RATES OF RETURN AND INTRAINDUSTRY VARIANCE IN RATE OF RETURN FOR 15 CONSUMER GOODS
INDUSTRIES, 1950-65
Average Intra-
rate of industry
Industry return variance
1950-65 1 in rate of
return 2
Average Intra-
rate of industry
Industry return variance
1950-65 1 in rate of
return2
Automobiles 11. 129 168. 889
Cosmetics 17. 852 101. 583
Drugs 17. 998 90. 827
Radio-TV manufactures 13. 721 90. 226
Watches 7.335 70. 392
Beverages, brewers 11.966 67. 558
Confectionery 13. 018 50. 358
Beverages, soft drinks 12.424 45. 950
Tire and rubber 9.914 31. 367
Shoes 9.528 26. 143
Beverages, distillers 9.057 25. 591
Food products 9.075 24. 463
Textile apparel manufactures 10. 655 23. 115
Home furnishings 7.874 16. 878
Tobacco 9.627 9. 824
~
1 Rate of return is defined as net income plus fixed charges divided by total capitalization.
2 The variance is the simple averagn of unweighted intraindustry variances for each year of the period, 1950-65.
Source: "Underlying Data for Study" by Conrad and Plotkin, `Risk and Return in American Industry."
TABLE lB-RATES OF RETURN AND INTRAINDUSTRY VARIANCE IN RATE OF RETURN FOR ADJUSTED SAMPLE
CONTAINING 20 CONSUMER GOODS INDUSTRIES, 1950-65
Average Intra-
rate of industry
Industry return variance
1950-65 1 in rate of
return 2
Average Intro-
rate of industry
Industry return variance
1950-65 1 in rate of
return 2
Automobile 11. 129 168. 889
Cosmetics 17.852 101.583
Drugs 17.998 90. 827
Radio-TV manufacturers 13. 721 90. 226
Watches 7.335 70. 392
Beverages, brewers 11.966 67. 558
Confectionery 13. 018 50. 358
Beverages, soft drinks 12.424 45.950
Food, bread and pastries 9.458 32. 698
Tire and rubber 9. 914 31. 367
Shoes 9. 528 26. 143
Beverages, distillers 9.057 25.591
Food, packaged foods 12.481 23. 878
Textile apparel manufacturers 10. 655 23. 115
Food,canned foods 7.215 18. 004
Home furnishings 7.874 16. 824
Food, meatpackers 5.561 12. 590
Tobacco 9.627 9.824
Food, cookies and crackers 10. 134 9.112
Food, dairy products 9. 190 2. 190
1 Rate of return is defined as net income plus fixed charges divided by total capitalization.
2 The variance is the simple average of unweighted intraindustry variances for each year of the period, 1950-65.
Source: "Underlying Data for Study" by Conrad and Plotkin, "Risk and Return in American Industry."
PAGENO="0300"
1850 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 2A.-RATES OF RETURN AND INTRAINDUSTRY VARIANCE IN RATE OF RETURN FOR 26 PRODUCER GOOD
AND MINING INDUSTRIES, 1950-65
Average Intra-
rate of industry
Industry return variance
1950-65 1 in rate of
return 2
Average Intra-
rate of industry
Industry return variance
1950-65' in rate of
return 2
Electrical products 11. 292 174. 584
Aerospace 12.628 113.246
Building materials-heating, air
conditioning 11. 886 71. 494
Paper 10. 753 64. 410
Oil, crude producers 12. 608 47. 690
Office and business equipment 12. 871 46.056
Miscellaneous metal evork 8.447 35. 774
Machinery, metal fabricating 9.360 34. 644
Auto parts and accessories 12. 749 34. 287
Machinery 11. 165 33. 954
Abrasive prodacts 11.814 32. 161
Containers, paper 10. 638 19. 228
Chemicals 10. 492 18. 759
Textile products 7. 843 18. 754
Copper 9.842 15.673
Auto trucks 10. 301 14. 543
Building materials, cemest 12. 558 13. 214
Steel 8.830 12. 418
Building materials, roof and walL -- 10. 609 10. 839
Point 10. 096 9.855
Vending machines 12. 493 8.406
Oil 10. 087 7. 937
Containers, metal and glass 8.066 7. 871
Coal, bituminous 8.368 7.449
Railread equipment 8.149 6.318
Aluminum 7.938 2.812
`Rate of return is defined as net income plus fixed charges divided by total capitalization.
2The variance is the simple average of unweighted intraindustry variances for each year of the period, 1950-1965.
Source: "Underlying Data for Study" by Conrad and Plotkin, "Risk and Return in American Industry."
Electrical productu 11. 292 174. 584
Aerospace 12.628 113. 246
Building materials-heating, air
conditioning 11.886 71. 494
Paper 10.753 64.410
Oil, crude producers 12. 608 47. 690
Office and business equipment 12. 871 46. 056
Machinery, specialty 11. 450 41. 005
Machinery, general industrial 14. 552 38. 282
Miscellaneous metalwork 8. 447 35. 774
Machinery, metal fabricating 9.360 34. 644
Auto parts and accessories 12. 749 34. 287
Machinery agricultural 8. 212 33. 349
Abrasiveproducts 11.814 32.161
Machinery, industrial 12. 181 21. 386
Machine toals 10. 906 21. 054
Containers, paper 10. 638 19. 228
Average Intra-
rate of industry
return variance
1950-65' in rate of
return 2
Chemicals 10. 492
Textile products 7. 843
Copper 9.842
Auto trucks 10. 301
Machinery oil viell 13. 096
Building materials, cement 12. 558
Steel 8.830
Building materials, roof and wall. 10. 609
Paint 10. 096
Machinery, construction and ma-
terialhandling 10.580
Vending machines 12. 493
Oil 10.087
Containers, metal and glass 8. 066
Coal, bituminous 8.368
Railroad equipment 8. 149
Aluminum 7. 938
TABLE 2B.-RATES OF RETURN AND INTRAINDUSTRY VARIANCE IN RATES OF RETURN FOR ADJUSTED SAMPLE
OF 32 PRODUCER GOOD AND MINING INDUSTRIES, 1950-1965
Industry
Average Intra-
rate of industry
return variance
1950-65 1 in rate of
return 2
Industry
18. 759
18. 754
15. 673
14. 543
14. 280
13. 214
12. 418
10. 839
9.855
9.372
8. 406
7.937
7.871
7. 449
8. 318
2. 812
`Rate at return is defined as net income plus fixed charges divided by total capitalization.
2 The variance is the simple average of unweighted intraindustry variances for each year of the period, 1950-65.
Source: "Underlying Data far Study" by Conrad and Plotkin, "Risk and Return in American Industry."
PAGENO="0301"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1851
TABLE 3A.-PROFIT RATES OF LEADING COMPANIES AND OTHER COMPANIES IN 19 CONSUMER GOODS INDUS-
TRIES, 1963-65
Size classes 1
Leading companies All other companies
Industry Number of Average rate Number of Average rate
companies or return companies or return
1953-65 2 1963-65 2
Cosmetics 4 22.7 7 14.3
Drugs 4 20.4 25 19.3
Automobiles 2 19. 0 3 10. 0
Confectionery 2 16.2 2 9.8
Radio-TV manufacturers 4 15.5 4 8.8
Soft drinks 3 13.4 3 19.8
Tobacco 4 11.4 6 9.5
Apparel 4 10.8 7 13.2
Dairy products 3 10.5 4 9.0
Canned foods 3 10.3 4 4.0
Malt liquors 4 9.9 13 9.5
Shoes 3 9.2 4 9.1
Biscuits and crackers 2 9. 1 3 10. 1
Home furoisfrings 4 8. 7 5 7. 8
Tires aod rubber 4 8.3 7 7.8
Distillers 4 7.8 7 11.1
Bread and pastries 3 7. 3 3 5. 7
Watches 2 6.9 3 .3
Meat products 3 6. 7 4 3. 4
I The group of leading companies consists of the 4 largest companies except when fewer than 8 companies are included
in the sample for an industry. In this case, the sample companies are divided evenly between the 2 groups with the largest
firms going into the group of leading companies. If there is an odd number of companies in industries having less than 8
sample companies, the median company is placed in the all other company category.
2 A 3-year simple average of company profit ratios.
Note: The packaged food products category was excluded from the group of consumer goods industries for this analysis
because it was too broadly defined, and hence contained a number of sample companies which did not compete with one
another.
Source: `Underlying Data for Study" by Conrad and Plotkin," Risk and Return in American Industry-An Econometric
Analysis.'
PAGENO="0302"
1852 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 3B.-AVERAGE PROFIT RATES OF LEADING COMPANIES AND ALL OTHER COMPANIES IN 35 PRODUCER GOOD
AND MINING INDUSTRIES, 1963-55
Leading companies `
All other companies
Industry
Number Average
of rate of
companies return
1963-652
Number
of
companies
Average
rate of
return
1963-652
Chemicals
4
15.7
39
10.0
Aerospace
Machinery, construction, and material handling
Office and business equipment
Machinery, general industrial
Electronic products
Electrical products
Oil
Vending machines
4
3
4
2
4
4
4
3
15.5
13.7
13.0
12.9
12.2
12.0
12.0
12.0
13
4
10
3
26
26
21
3
11.1
9.6
10.9
10.9
10.6
12.0
9.1
11.7
Miscellaneous metalwork
3
11.9
4
7.2
Machinery, oil well
3
11.7
3
11.7
Auto trucks
2
11.5
3
12.5
Oil, crude producers
Machine tools
4
4
10.7
10.6
5
4
19.7
11.4
Auto parts and accessories
Machinery, steam generating
Abrasive products
Textile products
Building materials, roof
Copper
Machinery, industrial
4
2
3
4
3
4
9.9
9.6
9.6
9.4
9.4
9.0
9.0
1
1
3
2
4
1
6
3
7
12.
3.
8.
9.
8.
10.
10.
7
9
7
6
8
7
8
Contuiners,puper
Cool, bituminous
4
3
8.
8.
7
7
7
4
10.2
7.9
Paint
2
8.
6
3
10.6
Machinery, metal fabricating
Railroad equipment
Building materials, cement
Paper
Containers, metal and glass
Machinery, agricultural
Steel
4
4
6.6
8.4
8.4
7.9
7.9
3 7.8
4 7.6
9
7
6
13
3
3
19
9.2
9.2
7.7
10.7
10.0
14.1
8.0
Machinery, specialty
Building materials, heat
Blastfurnaces
4
2
7.5
7.0
7.0
12
11
2
12.3
11.3
5.6
Aluminum
2 6.3
3
6.0
I The group of leading companies consists of the 4 largest companies except when fewer than 8 companies sre included in
the sample for an industry. In this case, the sample companies are divided evenly between the 2 groups with the largest
firms going into the group of leading companies. If there is an odd number of campanins in industries having less than 8
sample companies, the median company is placed in the eli other company category.
2 A 3-year simple average of company profit ratios.
Source: "Underlying Data for Study" by Csnrad and Plstkin, "Risk and Return in American Industry-An Econometric
Analysis."
Senator NELSON. Thank you very much, Dr. Mueller, for your very
instructive presentation.
Do you have any questions, Senator Scott?
Senator SCOTT. If your conclusions are correct, that the high profits
vf the drug industry are due primarily to the high entry barriers-
your table 1 shows three industries where the industry variance in
return based on their book value exceeds drugs-would you therefore
conclude that the profits of radio broadcasting, book publishing, and
gold mining are also high because of the high entry barrier charac-
teristic of those industries?
Dr. MUELLER. First, I do not think gold mining was among these,
but in these other industries I think the high entry barriers go a long
way toward explaining the level of their profits. But I have not studied
them individually as I have drugs.
Senator SCOTT. There is an old quotation in the publishing business,
"Making many books, there is no end."
PAGENO="0303"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1853
And there are new publishing companies in the field all the time.
I know of one within the last 6 months. Is it so difficult to enter the
publishing field?
Dr. MUELLER. I think we would have to make a distinction here.
These are the profits of leading concerns which Arthur D. Little have
included, and it is hard to enter, I would say, magazine publishing and
make a profit return comparable to Time and Look and the others in
the industry that have a very strong position. If you write a book,
you can find someone who will publish it for you. You would not very
likely end up with these industry leaders, however.
Senator SCOTT. Over the last 20 years, perhaps the easiest industry
to get in has been the i'adio broadcasting industry, as an example.
Would your parallels app'y to this? Because if they do, it is some sort
of a criticism of the Federal Communications Commission, it seems
to me.
Dr. MUELLER. I \vouldl not want. to be in the position of being up
here criticizing another regulatory agency. I would rather be criti-
cizing my own bureau, which I know best. And we all can stand some
criticism, I am sure.
But, again, I think what this measure that I am referring to here,
the product differentiation entry barrier, captures the fact that the in-
dustry leaders that have a very strong position-say the networks. I
do not have the numbers with me., but I looked at them in making this
analysis. You have somethiiig comparable to automobiles where you
have very strong firms at the top of the heap which earn persistently
high profits, and then you have a. range coming down to the weaker
ones, and then there are a large number of firms which are making
rather modest profits. I simply took the Arthur D. Little data, which
showed profits for leading firms essentially and not for all firms in
the industry. It is completely consistent with this analysis to find a
number of enterprises which have very strong positions because of
product differentiation. In drugs a number of enterprises can enter
the easy-entry part of the business, say, producing penicillin under a
generic name. Here you do not make much money because it is extreme-
ly competitive.
Senator SCOTT. You have said that the concentration of the drug in-
dustry is high because of the patent privilege. Would you suggest to
the committee an alternative to the existent pat.ent policy?
Dr. MUELLER. I am not up here today making any such suggestion
on patent policy. I might say on that point, after listening to the testi-
mony of Professor Markham and several of his colleagues last month,
in which they spoke of the short-life cycle of drugs, one might draw
the inference that 17 years is an extremely long period to enjoy the
patent privilege in this industry.
Senator Scorr. Because if they were testifying before the Patent
Committee their life cycle might gro.w longer.
Dr. MUELLER. I expect they might have.
In the case of specific drugs, such as tetracycline with which I am
somewhat familiar, the patent privilege is still valuable, even though
it has only a few years to run.
PAGENO="0304"
1854 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator Scorr. Would you recommend that Congress impose. limita-
tions on advertising in the drug industry, or for that matter in any
other area?
Dr. MUELLER. I would not make this recommendation for any area
in American industry.
I would hope there are other alternatives to this.
Senator SCOTT. I am glad you said that, because if there were to
be limitations recommended on the drug industry, it would seem to
me they would apply to other industries similarly situated, such as
motor vehicles, computer machines, and aircraft.
I do not know whether or not-I realize that the purpose of your
testimony is a. reply to other testimony, but are you inferring that
you, yourself, in your official position, would be happier if the drug
companies were making less profits than they are? Is that the point
you are making?
Dr. MUELLER. I am certainly never in the position of being happy
just because someone is making less money. Our system is built on
the profit motive, and we use profits as an allocator of resources; but
my own response on that point is that I would like to see the drug
industry exposed to more of the competitive pressures that are so
commonplace in the great bulk of American industry. They are an
exception, and I certainly would not want to be interpreted as im-
plying somehow that this is characteristic of American industry. It
is an exception. And there are very few industries like this.
Senator SCOTT. What I am thinking here is this: The Government
runs on taxes. Taxes are the result of income and profits. And the
President last night asked for more taxes, which would imply that
he would like to see more income in the country, and I am wondering
whether the desire for more income and more taxes on the one hand
is wholly consistent with the criticism of those industries which are
producing the income which provides the taxes which pay your salaries
and mine and even the President's. I just wonder whether or not the
purport of some of the hearings in Congress do not say on the one
hand "We want to stop some people from making as much income",
and immediately when they do, the people who would be hurt would
be the Government which takes a great part of it. That is not very
good economics, perhaps, but it is certainly awfully good common-
sense.
Dr. MUELLER. My only answer would be that we have a tax policy
which takes 50 percent of the profits of corporations that are earning,
say, 5 percent, or 10 percent. The average of American industry is
12 percent. And to argue that some industries should be immune to
any kind of public policy criticism just because, in addition to ex-
ploiting of consumers, they are giving half of this back to the Gov-
ernment..
So, I think that in a sort of a. commonsense way, this is not a very
good rationalization for not. doing anything about the industries which
present a public policy problem.
Senator SCOTT. Then, in 3-our judgment, is an automatic public
Policy problem only part of which industry makes a profit, a. high
profit?
PAGENO="0305"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1855
Dr. MUELLER. I think we have to make a distinction between in-
dustries and individual companies making very high profits. Some
industries earning below average industry profits, say in meat pack-
ing which happens to be a very competitive, low return industry
averaging 6 percent. on equity in many years. In meat packing there
are some companies, however, that are extremely efficient and better
managed and more aggressive, and they make high profits. Iowa beef
packers is the prime example. They are up here with the average of
drug companies, 20 percent. On the other hand, other companies make
considerably less, the ones for one reason or another just are not able
to cut it. But that is the way things are in a competitive industry.
And as long as the average return in the industry is adequate to at-
tract capital, the system works well.
So it is not tha.t some companies are making a profit, no one objects
to that., but it is the persistently high profits for the entire industry
that are an indicator of poor performance.
Senator SCOTT. I am probing to find out wh~ther there is any law
or regulation which says that a company can't make a profit, even a
high profit.
Dr. MUELLER. Certainly not..
Senator SCOTT. Therefore, why is a Government a.gency anxious to
show that it is critical of some conwa.nies making profits higher than
other companies?
Why does that become a matter of Government policy?
Dr. MUELLER. The only point that is relevant here is that high profits
are indicative of the presence of market power, and upon analysis of
the reason for the high profits we find that they are due to a certain
clog on competition.
Senator SCOTT. Now, if there is any fault in the area of market
power, is that not covered by the ahtitrust statutes?
Dr. MUELLER. I t.hink not. I think, as I explained in the beginning
of my statement, the consensus of economists is that the source of
market power is a combination of the patent law and the vast amounts
spent on promotion and advertising, which in this industry create the
so-called product differentiation : entry barrier. So that insofar as
there are solutions, it is something outside of the existing antitrust
law.
Senator SCOTT. This is the strongest country economically in the
world, and the strength of this cOuntry economically is built, in part,
upon advertising, upon promotional genius, and upon management..
Why, then, would a Governrnenta.gency come in and be critical of the
consequences of these elements which in other areas we rush up to
iraise?
Dr. MUELLER. First, since this area of a.dvertising is especially a
sensitive one for anyone in public office, I want to make clear that I
am not here speaking for five Federal Trade Commissioners who have
emphasized many times that they are not critical of advertising. I am
here as Chief Economist, the Bureau of Economics. And I am not
being critical of advertising as such either. Public policy has long been
concerned with advertising, as has the advertising profession. The
advertising profession now claims the major credit for helping to get
the Federal Trade Commission started, and I think they did play some
S1-2S0-65---pt. ~i-2O
PAGENO="0306"
1856 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
small part. The point is that we have responsibilities now to keep ad-
vertising honest. And we put up with a lot of puffing, and so on, but
there are already safeguards in this respect.
But I am here simply to try to help identify the problem in this
industry, and I am not at all in disagreement with you that advertising
can play and has played an extremely important role in many indus-
tries in the American economy. But recognizing this, I can't bind
myself to the possibility that in certain areas it creates problems.
Senator SCOTT. What kind of problems does advertising create?
Dr. MUELLER. I think it mutes the voice of competition. The fact
that it has prevented the purchase of drugs on the basis of their in-
herent qualities. It has prevented the purchase on this basis, and the
advertising has prevented, in effect, competition from working as we
ordinarily hope it would. It has gone far beyond the informational
aspects that we are most interested in.
Senator Scorr. Let us see if the corollary of your argument holds
water-I do not know whether a corollary is intended to hold water-
if you are arguing that advertising in the drug industry is so used
as to muffle competition, and to muffle the opportunity for the public
to have more information on the nature of the drugs it is `buying,
would it therefore follow that if you would advertise less the public
would know more?
That is a hard one to answer.
Senator NELSON. May I interrupt?
I have an answer for that. It is not a question of the public know-
ing more or less. The public, in fact, knows nothing, and advertise-
ments in this industry are not addressed to the public; they are
addressed to the man in the medical profession exclusively.
I have an example which I think explains the point and may `be
helpful to the Senator.
We have a case with regard to the drug prednisone where adver-
tising very clearly was not helpful to the public at all, nor to the
doctor himself, in terms of his patient's welfare. The company that
discovered prednisone and got it patented is the Schering Corp., which
gave it the trade name "Meticorten."
A large number of companies went into the marketplace with the
drug under license from Schering. The prices on prednisone, according
to the Medical Letter of June 2, 1967, ranged from 59 cents for 100
tablets to the pharmacist by one company to $17.90 a 100 to the
pharmacist by another company. The latter is the highest priced precl-
nisone on the market. The Medical Letter stated that these drugs were
of comparable therapeutic value, and advised the doctors to prescribe
the lower priced generic named drug.
Now, here is a case where tremendous advertising by one company
convinced the doctor-not the public, because the public had no notion
about it-convinced the doctor that he ought to continue `to prescribe
Meticorten even though there were 21 other drugs in the marketplace
that the Medical Letter said were of equal therapeutic value, one sell-
ing for 59 cehts a 100, one for 72 or 73 cents a 100, Deltra for $2.20
a 100, Meticorten for $17.90 a 100, and Paracort for $17.88 a 100.
So, here is a case where the advertising is to the medical profession
and not to the public and has not enlightened anybody, `but rather
PAGENO="0307"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1857
resulted in tremendous expenditures by patients in this coimtry for a
drug that was availaible on the market at a fraction of the price.
And I think the classic result of the publicity which was brought
out before this committee occurred ~when one company reduced its
price dramatically 2 months after we examined their pricing structure
and showed them the Medical Letter. They reduced their price from
$17.90 a 100 to $10.50 a 100, and the second company reduced theirs
from $17.88 a 100 to about $3.45 a 100. Now, that is a clear case
where the advertising was doing a disservice to the medical profession
and the public in terms of economics.
Senator SCOTT. Isn't Dr. Goddard currently looking into the whole
question of whether drugs are ôhemically and therapeutically
equivalent?
That has not been finally established, has it?
Senator NELSON. He is examining that, drug by `drug, as I under-
stand it.
Senator Scorr. If the investment: opportunities in the brand name
drug manufacturing industry are so fabulous, would you tell me why
the prices for the stock of these companies are not `constantly sky-
rocketing? And is not `that because of the c.onstant threat of congres-
sional investigation and threats of further regulatory acti'on?
Dr. MUELLER. You mean they keep going up, skyrocketing?
Senator Scorr. If their profits are so big, why doesn't the stock
keep going up?
Dr. MUELLER. Why does it not?
Senator SCOTT. Why does it not. skyrocket if profits are as fabulous
as you have indicated?
Dr. MUELLER. Well, first, the stock earnings ratios on drug stocks
are extremely high, they are among the highest of any American
industry, indicating that stockholders and investment analysts expect
drug stocks to keep going up-of `drug companies to continue to make
high profits in the future.
Now, as an investor, of course, you may buy into the market at
any p'oint, and you are not necessarily going to make vast amounts
on your investments in stocks because the market for drug stocks is
a very competitive one. How well an individual does in `the market
is not any indication of how profitable the industry is. Were you
referring to the variation?
Senator `Soorr. I am not familiar with these figures, but were you to
take, again, the eight leading companies, would their securities, their
common stock, on the market today be `higher or lower than it was a
year ago or the year before?
Dr. MUELLER. I misunderstood you. I thought when you said sky-
rocketing, that it was going up. But you are talking about it coming
dow-n as well, and whether this is associated with-
Senator Score. What I am talking about is the feeling that I have-
not the knowledge-that actually drug securities in the market have
not been skyrocketing or increasing in their market prices recently. I
am not sure, though.
Dr. MUELLER. They `are not, of course, entirely unique. The market
has been `a very mixed kind' of one `this year. Bu't is response to
your-
PAGENO="0308"
1858 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator SCOTT. Machinery stocks, IBM. Xerox, and others like, that
have gone up fantastically.
Dr. MUELLER. In response to your question as to whether the stock
market prices are responsive to public, polic actions in the drug inclus-
try, I would say they are. As I indicated in my discussion of the invest-
ment analysts' views of the drug industry, one thing `they discuss is
that if there is increasing use of generic drugs, profits will come down.
In other words, what they are predicting is: If your industry becomes
more competitive profits are going to come down, and drug stocks are
coming down, and this is true, I suppose, with respect to any public
policy action `that. is going to affect drug prices.
So, I would not be at all surprised if this association exists. I ran
across such comments in preparing this statement, as I looked at the
way in which investment analysts viewed the industry. This is some-
thing they talked about.
Senator SCOTT. I notice there is no chart here which would tend
to bear out or theorize-there is no chart of a steady rise in the valua-
tion of these drugs. WTe are discussing only profits.
Dr. MUELLER. As I pointed out in the beginning, the stocks over
this period have-stock price.s relative to earnings have been very
high. These ratios are far above the average of American industry.
And I did not really think it was important for my comments to show
the trends in stock prices. Over this period, however, they reflected
the profit situation very accurately.
Now, it is possible-and it has happened-that there are
variations-
Senator SCOTT. Maybe I have not made my point clear in these
hearings. Because if this industry or any other is doing anything they
ought not to be doing, we want to know about it. If there is anything'
that is statutorily wrong, we want to know about it. If corrective leg-
islation is needed, we want to know about. it.
But I get a little bit fed up, after 25 years of residence down here
in this cave of winds, with the constant plea. of Government to indus-
try to produce more, make more money, give us more taxes, and then
have other agencies of the Government beating them over the head.
all `the time they are doing it.
Now, it seems to me t.hat if the people are doing something wrong,
let us know about it. But I do not think it is wrong to make a profit in
this industry, and I have not yet heard a single witness in 25 years that
convinced me that it was.
Dr. MUELLER. Well, I certainly will say nothing here today to t.ry
to convince you that it is improper. I am not carrying a club to beat
any American industry. I have a job at the commission, and my part
of it is to try to make our system work fair and competitively; that is
all.
Se.nator SCOTT. I only wish some Government agencies would come
before some senatorial committees occasionally in praise of industry,
in praise of business, and not always appear in such an antibusiness
aspect.
Dr. MUELLER. I agree with you 100 percent., Senator. And before
the full committee-
Senator SCOTT. Because business is jobs, and if you praise business,
you are praising job-making.
PAGENO="0309"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1859
Dr. MUELLER. And before the full committee last June., with Sen-
ator Long and Senator Morse presiding, I appeared along with Mr.
Turner of the Justice Department, Professor Adams of Michigan
State University; we had a dialog with Professor Gaibraith. At
that hearing I went to considerable length to explain as accurately as I
could that our competitive system is working extremely well. People
have been selling it short. There are exceptions. But Professor Gal-
braitli has exaggerated their frequency. In that dialog, he kept
emphasizing the automobile industry. He kept telling Mr. Turner
that, "Unless you solve automobiles, you have to concede my point
that the system has had it."
Senator SCOTT. That is the affluent Mr. Gaibraith; isn't it?
Dr. MUELLER. The ver affluent. And the point, though, is that the
problem industries that are discussed most before congressional com-
mittees are the exceptions in our economy. The great bulk of the
economy works very well and, unfortunately, we do not come up here
and tell you "This industry is working perfectly," because you do not
hold hearings on industries that do not have problems. But our silence
should not be interpreted as implying that we do not think the system
is-the greater part of it-is working very effectively.
Senator SCOTT. I am glad you are always proud of it, because I
think we ought to be proud of `the system. I have seen some of `the
others. I have been in England quite recently, and when I see the mess
they have made of things it has made me more inclined to defend our
Own.
Dr. MUELLER. And the more we see of other country's problems the
more tolerant we become of the imperfections in our system.
Senator SCOTT. Thank you.
Senator NELSON. I am sure the Senator knows that Dr. Mueller
came here at the request of the committee.
Senator SCOTT. I did not mean to irnj)ly that Dr. Mueller came
up here to wage a vendetta against industry. I was speaking
broadly and generally about wanting more Government witnesses to
find occasion to praise the capitalistic system1 because it needs all the
friends it can get, especially around income tax time.
Senator NELSON. We asked him specifically to analyze the testimony
of other economists who had testified here and give his viewpoint.
And I think part of his testimony was directed specifically at the
question of factors that make it possible for prices to be held artifically
high.
Mr. GORDON. Dr. Mueller, at our hearings last month, when Dr.
Markham was here, our chairman stated to Dr. Markham on page
2775, line 18 of the transcript:
Senator NELSON. As one who majored in economics when I was in school
and listened to the arguments in those days, I just learned enough to discover
that you could find any poillt of view you wanted from any set of economists
any place in the United States. And from my discussions with my friends today,
I find that this is still true.
Dr. MARKHAM. If I may intrude :on the dialogue, Senator, I think you are
quite correct when you went to school, and that was true when I went to
school, too.
If this analysis means anything, it means I think, what precisely Professor
Cootner and the rest who have testified here have said, and that is as Pro-
fessor Plotkin has said, this takes the issue really out of the area of opinion;
it puts it in the area of quantitative fact.
PAGENO="0310"
1860 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
First of all, how new is this, really?
Is this really new?
Dr. MUELLER. Tho oconomefric approach which they used?
~\Te11, actually, all econometrics involves the application of mathe-
matical and statistical teclmiques in testing economic relationships.
Economists have been doing this for many years. The chief advantage
of this technique today is that it permits us to analyze vast volumes of
data with modern electronic computers in a short period of time, test
many relationships. Actually, the Arthur D. Little study is a very
simple one. It involves a study of only two relationships, the so-called
risk factor and profits. It does not introduce any other variables such
as market structure or competitive variables, as have some other
studies of risk.
These teclmiques have been used long before I was born; they came
into prominence in the twenties as simple and multiple regression
techniques. So they are not anything new or novel. The whole point of
it is that you can't take, as Professor Boulding pointed out-you
can't take data that are very crude and send them through a machine
and come out with something that is better than the input. In addi-
tion, you have to know what it is you are putting in.
So, simply using a sophisticated technique does not give you any
good new answers to difficult questions unless you have a good con-
ceptual framework and you have good data.
And, as we point out, when you really understand what Conrad and
Plotkin are measuring, their data show nothing at all in the so-called
producer goods industries; there is no significant statistical relation-
ship present. And, as a result, when we test their theory with, in effect,
better data than they used, we find no relationship. That is what the
scientific method is all about. As Cootner explained in his opening
statement, he had hoped this is the way science operated, the testing
of alternative hypotheses, and so on. And this is what we have tried
to do in our analysis, really test what it was they thought they had
come up with.
Mr. GROSSMAN. Dr. Mueller, two questions:
First of all, I would like to ask you just as an economist to comment
on one of the statements that Dr. Whitney made at our last meeting.
He said:
"If a $5 prescription, or six of them, will keep a patient from losing
a couple of days' pay or spending a night in the hospital, the price is
reasonable."
\~\That do you think of this "market-will-bear" type of philosophy?
Dr. MUELLER. It, in effect, permits a rationalization of anything that
you see. It is like saying: "We would all be willing to pay more for
our newspapers." Certainly we buy the Star, the Post, and the New
~fork Times, and we pay only 10 or 15 cents for these fine newspapers.
We would be willing, however, to pay more; but competition pushes
the price down; and, as a result we get more than our moneys' worth.
The economist refers to this as a consumer surplus. So, merely because
there is a great value received, a greater value received than someone
would be willing to pay, does not prove anything in terms of the effec-
tiveness of our market system, our competitive system.
Senator NELSON. As I understand that testimony given a few weeks
ago, what the economist was really saying was that if you are scuba
PAGENO="0311"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1861
diving and are under 50 feet of water and run out of oxygen, any price
you pay for it is good.
Dr. MUELLER. That is right, and you can't argue with that.
Mr. GROSSMAN. One final question, and this is more in the way of
summation-
Dr. MUELLER. If 1 may interject, I think you have devised a new
theory of value, Senator. We will call it the "scuba theory of value."
Mr. GROSSMAN. In the way of summation, when I asked you about
what you thought a fair and profitable return was, you talked in terms
of more competition rather than a specific figure. And when Senator
Scott asked you about changes in patent policy, you said: "No", per-
haps, but you had nothing specific to recommend. With regard to ad-
vertising, you had nothing specific-
Dr. MUELLER. I never answered the question, I do not think.
Mr. GROSSMAN. I wondered whether you could tell us, since we have
to rely on your expertise on this, what exactly do you recommend?
In other words, you talk about competition. And I would like to
know specifically what you think should `be done.
Dr. MUELLER. My testimony and my part in these hearings is a very
small one. I was asked primarily to analyze the significance of the
Conrad-Plotkin study. But this committee and its predecessors which
h'ave investigated the drug industry have put their fingers on the key
problem, which is the beginning' of understanding. Advertising-
crea'ted-product differentiation is certainly a. key point. And some-
thing-I am not recommending anything particularly today-but
there should be something that can' be done to erode this power. One
possibility is the greater use of generic drugs-anything that helps to
bring this development about may help; perhaps something designed
to simply give consumers, in this case doctors, more information `about
what it is they are prescribing for their patients. There are different
ways of doing this, some private and some public perhaps.
On the patent question, I am sure that you will get other witnesses
who would be more willing than I to take a particular position on that
point..
But that is the o'ther key point.
And many other countries take a.' different position on this than does
the United States.
Mr. GROSSMAN. Thank you.
Senator NELSON. Thank you very much, Dr. Mueller, for your
thoughtful presentation to the committee and the very valuable refer-
ence work you did. And if you wish to present supplemental material
in the future, as you stated earlier that you may, we will be happy to
look at it. And, as I understand yOu, you said if you thought it worth-
while, you would be happy to come back before the commit'tee and pre-
sent it personally?
Mr. MUELLER. Yes, sir.
Senator NELSON. Thank you very much.
We will adjourn until 10 o'clOck tomorrow morning.
(Whereupon, at 11 :55 a.m., the. subcommittee adjourned, to recon-
vene at 10 a.m., Friday, January 19, 1968.)
PAGENO="0312"
PAGENO="0313"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
FRIDAY, JANUARY 19, 1968
U.S. SENATE,
MONOPOLY SLTBCOMMITTEE OF THE
SELECT COMMITTEE ON SMALL BUSINESS,
Washington, D.C.
The subcommittee met, pursuant to notice, at 10 :15 a.m., in room 318,
Old Senate Office Building, Senator Gaylord P. Nelson (chairman of
the subcommittee) presiding.
Present: Senator Nelson.
Also present: Benjamin Gordon, staff economist; James H. Gross-
man, minority counsel; Susan H. Hewman, research assistant; and
William B. Cherkasky, legislative director, staff of Senator Nelson.
Senator NELSON. We will open the hearings of the Subcommittee on
Monopoly with Prof. Leonard Schifrin, head of the department of
economics at the College of Williams and Mary.
Professor, we appreciate very much your taking time to come here
thismorning. I have read your statement. It is a very well documented,
very well reasoned statement. We appreciate the time you have taken
to assemble this material.
You may present your statement in any fashion you see fit. If you
find it most effective to read it, you may. If you wish to extemporize or
add to it as you go along, that is perfectly satisfactory. I trust you will
have no objection if we occasionally interrupt you with some questions.
STATEMENT OF LEONARD G. SCHIFRIN, PH. D., DIRECTOR, DEPART-
MENT OF ECONOMICS, COLLEGE OF WILLIAM AND MARY, WIL-
LIAMSBURG, VA.
Dr. SCHIFRIN. I would like to read my statement, Senator Nelson,
because it represents a general condensation of a very vast quantity of
material.
Senator NELSON. Very well. Go ahead, Professor.
Dr. SCHIFRIN. My name is Leonard Gerald Schifrin, and I am
presently associate professor of economics and head of the department
of economics at the College of William and Mary in Williamsburg,
Va. I received my bachelor's and master's degrees from the University
of Texas, and my doctorate from the University of Michigan. I taught
at the University of Michigan and at Yale University before coming
to William and Mary in 1965. Within the field of economics my areas
of specialization are industrial organization, problems of competition
and monopoly, and governmental policy toward business. My particu-
lar research interest in recent years has been the economics of the
ethical drug industry.
1863
PAGENO="0314"
1864 COMPETITIVE PROBLEMS IN THE DRLTG INDUSTRY
I thank the members of this committee for affording me the oppor-
tunity of presenting to them some of the salient aspects of my studies
of the operation of this industry.
First, let me speak to what the economist calls industry perform-
ance-a concept of many facets, but one that essentially deals with the
effectiveness with which the industry, in its operations, serves the goals
of society. Industry performance thus deals with such matters as tech-
nological progress, the development of new products, changes in pro-
duction levels to correspond to changes in the pattern and size of con-
sumer demand, the efficiency with which development, manufacture,
and marketing are done, and the reasonableness of prices charged
consumers.
The different dimensions of performance in the ethical drug industry
fall under two headings-"product performance" and "market per-
formance." Product performance may be measured by such things as
(a) the magnitude and quality of the industry's effort to develop ne~v
and better products, (b) the tangible results of this effort, i.e., the num-
ber of discoveries and new products flowing from research and develop-
ment, a.nd (c) the human and economic impact-better health, longer
life, and greater productivity-resulting from these tangible results
of industry research and development.
In this regard, it is apparent that the drug industry is a highly re-
search conscious industry, in basic research as well as in applied re-
search and development. It has made available to the public over the
years a large number of new and better products; and these products,
together with advances in other areas of health, have had dramatic
impact on our mortality rates, our longevity, and on our general well-
being. Many dangerous illnesses have been brought under control;
much of the discomfort and even the hopelessness of illness have been
checked. The drug industry, along with other sectors of the health
industries, deserve credit for the contributions it has made in this
respect.
But the record also shows clearly that the product performance of
the industry needs qualification and contains serious flaws. A number
of authorities have demonstrated that the research performance of the
industry is exaggerated by industry officials so as to justify the very
large profits of the large drug firms (an illogical argument, by the
way). Furthermore, the H. & D. performance is quite small compared
to the promotional and advertising outlay, which generally runs four
times as large. Beyond questioning its magnitude, critics have con-
tended convincingly that much of the research is imitative, wasteful,
and aimed at patent procurement rather than progress. Further, many
of the so-called "new drugs" coming to market represent duplications
of existing drugs, combinations of drugs representing no therapeutic
improvement over their components taken separately, or new items
that are the result of molecule manipulation rather than substantive
therapeutic advancement. We have heard or seen cases of ineffective
drugs, harmful drugs, drugs without adequate warnings, dangerous
drugs coming to the market as part of the "new drug flow" for which
the industry claims credit.
Senator NELSON. May I interrupt a moment?
Dr. SCHIFRIN. Yes, sir.
PAGENO="0315"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1865
Senator NELSON. You say beyond questioning its magnitude, critics
have contended convincingly that much of the research performed by
the industry is imitative, wasteful, and aimed at patent procurement
rather than scientific progress.
Do you have any specific examples of people who have made this
criticism?
Dr. SCIIIFRIN. Yes, sir; I do. Dr. Walter Modell in his own profes-
sional articles and in testimony before several Senate committees has
made this point rather convincingly. Dr. Claude Forkner has published
in several medical journals, including the New England Journal of
Medicine, about the drug mixtures in that respect, the "shotgun
therapy," as he puts it. Dr. Dale Console, formerly with Squibb Lab-
oratories, has talked about such questionable practices of the drug
industry.
Senator NELSON. You have the articles in which they made their
criticisms?
Dr. SCHIFRIN. Yes, sir; I have these and other articles and I would
be glad to submit them for the record.:
Senator NELSON. If you would submit those to the committee coun-
we will print them in the record.
Dr. SCHIFRIN. Yes, sir.
(The material referred to follows:)
[From the New England Jounal of Medicine 259 :438-439 (Aug. 28), 1958]
MEDICAL INTELLIGENCE-DRUG MIXTIJRES 1
(By Claude E. Forkner, M.D.')
Much of the success in modern treatment of disease resides in advances in
chemistry, physiology and pharmacology. This has resulted in the isolation of
many chemical and biologic substances that have specific and sometimes powerful
effects;
For example, twenty years ago there was 1 sulfonamide available to the medi-
cal profession-namely, sulfanilaiaide. Today, according to the sixth edition of
the Modern Drug Encyclopedia, 200 different products are listed as sulfonamide
preparations. Many of these contain 1 or more of the sulfonamides in combina-
timi with other drugs, often without any indication in the name of what is actu-
ally in the drug. For example, few people would guess that Eskadiamer is a com-
bination of 2, Neotrizine a combination of 3, Ray-Tri-Mides a combination of 3 and
Terfonyl a combination of 3 sulfonamides, Powdalator is a combination of peni-
cillin G and sulfanilamide and Thizodrin is a combination of 3 drugs, 1 of which
is sulfathiazole.
I suspect that few physicians use more than 3 or 4 of the sulfonamides and
that they would like them to be marketed under standardized names rather than
under a series of arbitrary combinations of letters making meaningless and non-
existent words that add nothing but confusion, add greatly to the cost of medical
care and promote seric~1~ errors in treatment.
Who is it that wants all this nonesense on the market? Is it the doctors? No.
Is it the patients? No. Is it the drugstores? No. Is it the hospitals? No.
A couple of decades ago only a handful of antispasmodics were available, and
only 2 or 3 of these were very useful. Today over 200 differently named products
are on the market. A few examples are Alubelap, containing a mixture of alumi-
num hydroxide gel, phenobarbital and belladonna, An1esec, containing amino-
phylline, epherdrine hydrochloride and amobarbital, Atralose, containing homa-
tropine methylbromide, phenobarbital and methylcellulose, Bunesia, containing
1 Presented at the convention of the American Medical Association, New York City,
June 4, 1957.
2 Professor of clinical medicine, Cornell University Medical College; attending physician,
New York Hospital.
PAGENO="0316"
1866 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
homatropine methyibromide, butabarbital and magnesium hydroxide, and Kol-
antyl, containing 5 drugs, none of which have any relation to the name of the
product.
I have no doubt that physicians would rather use a few drugs of known com-
position sold under their own names. such as belladonna, phenobarbital, ami-
nophylline, curare and atropine, to 200 different combinations of products sold
under names that have little or no relation to the chemical compositions of the
drugs.
A few years ago one had available 3 or 4 good antihistaminics, which are as
useful today as any of the 130 marketed products, which for the most part con-
tain a variety of mixtures with which physicians cannot hope to become familiar.
There are today over 100 antacid compounds on the market for sale in drug-
stores, most of which contain combinations of drugs. One of these drugs picked
at random is called X. It contains magnesium carbonate, calcium carbonate, col-
loidal kaolin, tricalcium phosphate, sodium bicarbonate, bismuth subcarbonate,
papain and diastase. I do not believed that any good.gastroenterologist, any good
internist or any good scientist would find use for such a drug. It is like the me-
dieval prescriptions written hundreds of years ago. Such concoctions, of which
there are hundreds in drugstores, should have no place in modern therapy.
Fifteen years ago there were only 2 or 3 antibiotics. Today, a dozen or slightly
more have been shown to be of special value, and more than 270 different prep-
arations are on the market, many of which are combinations and duplications.
A few decades ago, before medicine became as precise and scientific as it is
today, doctors' prescriptions often consisted of 0, 8 or more ingredients, many
of which w-ere more or less inert. The prescriptions were written in Latin, and
the mystery of the ingredients constituted part of their virtue. Gradually, most
of this sort of unscientific and meaningless procedure was abandoned; professors
of medicine and of pharmacology taught students to use simple drugs, for pre-
cise reasons and for definite periods. Gradually. w-ith the intense developments of
chemotherapy, of antibiotic therapy, of antihistaminic therapy, of endocrinologic
therapy and of vitamin therapy there has been a mushrooming of drug manu-
facturers who are in desperate competition. The commercialistic factor has crept
into therapeutics to such an extent that physicians everywhere are confused and
misled by the literally thousands of drugs increasing in number daily. Dozens
of expensive commercial brochures, sample drugs and elegant preparations reach
one's office daily and are promptly disposed of in the wastebasket. Many of these
modern preparations are mixtures of drugs, some of which are dangerous, some
of which are useless, and most of which would be more intelligently given as
separate drugs rather than in a shotgun capsule.
There are today over 300 preparations on the American market listed as hem-
atinics designed to increase the red-cell count and hemoglobin. This, of course, is
ridiculous. It is well known to every hematologist that not more than 8 or 10
useful drugs are necessary to treat anemia. For the vast majority of cases only I
of 2 or 3 drugs is necessary or desirable.
Shotgun therapy with multiple drugs usually is unscientific, often means that
the doctor does not know what he is doing. invariably is more expensive for the
patient and not infrequently results in tragedy both for the patient and for the
doctor. An example of such shotgun therapy is the use of drugs containing vita-
mins, including B12 and fohic acid, along w-ith iron, thereby masking bleeding
from an otherwise asymptomatic neoplasm until the tumor has become incurable.
Many students believe that folic acid administered to patients with pernicious
anemia may precipitate an acute and serious exacerbation of the neurologic
symptoms.
There are about 450 vitamin preparations currently on the market. Many of
these are so-called fortified vitamins. In addition to this, milk, bread, fruit juices
amid other foods are today fortified with vitamins. Many of the preparations con-
tam a dozen or more items, including copper, iron, cobalt, iodine, phosphorus, cal-
cium, manganese, molybdenum and zinc, in addition to 8 or 10 vitamins.
Every year I am told that Americans buy over the drugstore counters about
$250000000 worth of vitamins. It is safe to say that at least $240,000,000 of this
is wasted. No reason whatever exists for the taking of vitamins by any healthy
adult American on an adequate diet. There is good reason often for correcting
PAGENO="0317"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1867
the diet of people who have faulty eating habits. The giving of vitamills in no
sense is a substitute for a faulty diet in an otherwise normal person. Vitamin Br~
is being pumped into people by the gallons all over the country by doctors who
are not aware of the fact that vitamin B12 is of no value wliatcver except in one
group of rare diseases, the macrocytic anemias.
My plea is merely the restatement of a very old and sound principle: no mcdi-
cine, and that includes vitamins, should be prescribed unless there is a good
reason for doing so. When drug mixtures are prescribed, one should remember
that one is getting into dangerous territory.
Fifty years ago there were over 300 medical schools in this country. A study
was made. More than 200 of these schools were found to be poor and were forced
to close. The American Medical Association did much to promote that improve-
inent in medical education. Today, many thousands of useless drug and vitamin
preparations exist, thousands being duplicates under misleading names. Doctors,
patients, the proprietors of legitimate drugstores, the people generally and hos-
pitals deplore this situation. Exploitation of the public by the existence of such
a situation constitutes an important item in the high cost of medical care. Who
is going to devise a remedy for the insidious disease?
[From the New York Herald Tribune, Thursday, April 14, 1960]
Ex-Dauo OFFICIAL CHARGES "QUESTIONABLE" PRACTICES
WASHINGTON, April 13 (UPI).-The former medical director of a major drug
firm accused the drug industry today of profiteering and other questionable prac-
tices and urged Congress to crack down with restrictive legislation.
Dr. A. Dale Console, of Princeton, N.J., told Senate investigators he saw little
hope of the drug companies dropping practices he said they engaged in.
Dr. Console, medical director for Squibb Laboratories until he left to re-enter
private practice, said that "unless sweeping reforms are instituted, a truly ethical
(drug) house cannot survive in the present competitive wrangle."
CHARGES LISTED
In a severe indictment of the industry, Dr. Console charged before the Senate
Anti-trust subcommittee:
That doctors and the public are subjected to a constant "barrage" of new drugs,
some of which are worthless and others which have "a greater potential for harm
than for good." He said that "since so much depends on novelty, drugs change
like women's hemlines."
That more than half of the drug companies' research e~ort is directed toward
projects that are really not worthwhile but "are pursued simply because there's
profit in it."
That the industry has high-pressure sales techniques based on the maxIm,
"If you can't convince them, confuse them."
That most medical leaders and educators "face the problem with denial, com-
placency or a sense of futility" because the industry "is unique in that it can
make exploitation appear a noble purpose."
The subcommittee also heard Dr. Frederic H. Meyers, of San Francisco, a
University of California expert on drugs, challenge the American industry's con-
tention that it leads the world in discovering new drugs.
"Far from leading in drug progress," Dr. Meyers said, "it appears that our
industry has usually followed and often after a clear `lag."
He said that much of the laboratory work by American drug firms was centered
on "exploiting and marketing" foreign discoveries. Because of `this, he belittled
the American industry's assertion that the cost of its research justifies high drug
prices in `this country.
Dr. Meyers also criticized practices used in advertising new drugs. He said
many drug ads were "at best incomplete and at worst dishonest."
"Some ads become so expensive that they approach `payola'," he said.
PAGENO="0318"
1868 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
[From Time Magazine, May 26, 1961]
Too ~L&xv DRUGS
Prescription drugs would be cheaper and more effective if manufacturers
would market fewer of them, says Dr. Walter Modell of Cornell University
Medical College, one of America's foremost drug experts. He also believes that
this is the way to bigger profits for the companies.
Writing of pharmaceutical chemists in Clinical P1~armacology and Tlicra-
pea tics, Dr. Modell asked: "Will they realize that there are too n1any drugs for
the patient, for the physician, and, surprisingly enough, for the pharmaceutical
industry?" No fewer than 150,000 preparations are now in use, of which 90%
did not exist 25 years ago, and 75% did not exist ten years ago. About 15,000
new mixtures and dosages hit the market each year, while about 12,000 die off.
These figures, says Dr. Modell, reflect the fact that new drugs are often intro-
duced not because they are better than existing drugs or because there is a real
need for them, but "to horn in on a market which has been created by someone
else's discovery." He denounces as "structural roulette" the game of making a
minor change in the molecule of a competitor's drug, to get around patent restric-
tions, and rushing the resultant analogue to market. He points to one manufac-
turer "who sells one drug entity in this country and a congener [close chemical
relative] in another country," and argues that "each is the best for the same
purpose. Since more than one drug cannot be the best for the same indication,
we simply don't have enough diseases to go around. At the moment the most
helpful contribution is the new drug to counteract the untoward effects of otl1er
new drugs; we now have several of these."
Dr. Modell recommends that manufacturers exercise self-restraint by making
and marketing only the single best drug for each purpose, and cross-license one
another to spread both risks and profits. One of the most successful of all U.S.
companies, he says, introduces the smallest number of new drugs and does the
least "molecule manipulation." The others, Dr. Modell suggests, should do the
same--to their own advantage as well as that of bewildered doctors and patients.
Senator NELSON. sow, you comment in the next statement that there
are many so-called new drugs coming to the market which represent
duplications of existing drugs, combinations, and so forth.
Do you have some specific examples of these duplicative drugs and
those resulting from molecule manipulation?
Dr. SOHIFRIN. May I answer that question this way: The usual sta-
tistics cited on this come from tile product survey put out by Paul cle
Haen. In this survey that he updates annually, he not only lists the
total number of so-called new drugs coming to the market, but a break-
down as to those that represent duplicates of drugs existing on the
market, those that represent combinations of drugs already on tile
market, and then how many of these are clearly new drugs-that is,
items containing drugs that have not appeared on tile market previ-
ously. The large majority of tile total so-called new drugs come from
the duplications and the combinations category. I would be glad to
submit the data I have on the De Haen product survey along with this
other material.
Senator NELSON. Please hand it to the reporter.
PAGENO="0319"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1869
(The material referred to follows:)
TABLE 6-PHARMACEUTICAL PRODUCTS NTRODUCED NATIONALLY
Number
of firms
New
single
chemicals
Duplicate
single
products
Cow-
pounded
products
(a) Total
new
products
(b) New
dosage
forms
Cumulative
total
(a and b)
1966
1965
1964
1963
1962
1961
1960
1959
1958
1957
1956
1955
52
65
82
89
108
111
109
106
126
127
126
124
13
23
17
18
28
41
45
63
44
51
42
31
16
23
34
43
47
33
64
49
73
88
79
90
53
73
111
152
180
191
202
203
253
261
280
282
82
119
162
213
255
265
311
315
370
400
401
403
26
22
41
52
84
106
98
104
109
96
66
96
5,618
5,510
5,369
5,166
4,901
4,562
4,191
3,782
3,333
2,884
2,388
1,921
Note: New single chemicals indicates products that are new single chemic I entit es not previously known and de-
veloped by one manufacturer.
Duplicate single products include products such as dexamethaaone or griseofulvin put out by various manufacturers.
Compounded products comprise any product having mare than one active ingredient.
New dosage form: if a product has originally been marketed in tablets and is now offered in ampules, suppositories, etc.,
the latter are considered new dosage forms.
Source: Paul de Haen, New York City.
Senator NELSON. Now, you also stated in this same paragraph that
we have heard abotit or seen cases of ineffective drugs, harmful drugs,
drugs marketed without adequate warnings, and dangerous drugs
coming to the market as part of tile "new drug flow" for which the
industry claims credit.
Do you have specific examples of such harmful drugs, those without
adequate warnings, and so forth?
Dr. SCHIFRIN. Yes, sir; I think the most harmful example of a
new drug coming to tile market was MER 29, which was withdrawn
from the market because of its harmful effect on people taking it.
Also in these hearings, the question of inadequate advertising has
come up. I believe tile Chlormycetin case is probably the most im-
portant one there. A highly promoted drug came to the market, a
potent drug, with dangerous side effects, and the information as to
the side effects was fully understood only after the problems of use
of the drug had been so serious that it was withdrawn from the market
for a year for a very careful study.
So I cite that as the most Outstanding example of a drug, a
dangerous drug, with inadequate information.
There are quite a number of other examples. The Merrell Co. was
cited for false claims with MER-29. This is an article from the New
Haven Register of December 20, 1963, that I happen to have here.
That is just one of several examples I could name.
Senator NELSON. If you would submit that last article from it, we
will print it in tile record.
(The material referred to follows:)
[From the New Haven Register, Dec. 20, 19631
DRUG FIRM INDICTED ON FALSE CLAIMS
WASHINGTON (AP).-The William S. Merrell Co. was indicted by a Federal
Grand Jury today on charges of making false statements to the government
about the anti-cholesterol drug, MER/29, and tests made to determine its safety.
PAGENO="0320"
1870 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The indictment returned here said the company and three of its research
scienti~t~ `concealed aiid coeered up, by trick and scheme, material facts
imimportant to the Food and Drug Administration." The scientists named are
Harold W. Werner, Evert F. Van Maanen and William M. King.
Mr. GROSSMAN. Could I ask you a question here?
On Page 4, you talked about the proliferation of products that
confuses rather than proves drug selection or drugs inadequately
tested or whose side effects are minimized in the race to market for
the consumer's dollar.
Are you criticizing the FDA here, as well?
Dr. SCHIFRIN. I am not criticizing-
Mr. GROSSMAN. In other words, in order for these drugs to get to
the market, they have to somehow get past the FDA, do they not?
Dr. SCHIFRIN. Let me read that part of my statement that covers
that point you raised there. That discusses the point.
Mr. GROSSMAN. Surely.
Senator NELSON. We are not above criticizing a Federal agency,
you know.
Dr. SCHIFRIN. I consider many of these criticisms valid. I offer
them as important qualifications of the industry's product perform-
ance. But many aspects of this product performance warrants praise,
and this must be recognized, criticisms and qualification notwith-
standing.
The other half of the picture is what I refer to as "market per-
formance," and deals with the efficiency with which the industry
uses society's scarce resources; that is, the extent to which economic re-
sources are used to enhance consumer well being and other economic
goals. In this regard, the important questions that must be answered
are these: Is there enough competition to place a premium on efficiency
and penalize waste? Is there enough competition to compel firms to
pursue only those activities that benefit consumers? Can firms incur
costs for activities that do not benefit consumers yet charge consumers,
in the prices they pay, for such activities? Is there. enough competition
to keep prices in realistic relation to costs, providing profits adequate
for maintaining or expanding desirable a.ctivities but not profits
derived from the exercise of monopoly power?
To generalize from the vast quantities of evidence available, I be-
lieve that the characteristics of the drug-product marketing and dis-
tribution systems are such that effective competition does not pre-
vail. As a result, firms are free to engage in many practices-most
notably in promotion and advertising but in research and development
also-that serve their own profit goals but provide no benefit to so-
ciety. A largely wasteful promotional effort costing in the hundreds
of millions of dollars per year; misdirected research; rivalry in nov-
elty, in capturing the attention of physicians; all this represents costs
of large magnitude passed on in full to the consumer, but without any
corresponding benefit-and perhaps some harm, such as a prolifera-
tion of products that confuses rather than improves drug selection or
drugs inadequately t.ested, or whose side effects are minimized, in the
race to market for the consumer's dollar. Yet these practices and the
factors contributing to them have become part of the industry-woven
deep into the design of its fabric.
PAGENO="0321"
COMPETITIVE PROBLEMS IN T~) DRUG INDUSTRY 1871
Now, if I may comment on that question, I would say that to a
considerable extent, in recent years, the Food and Drug Administra-
tion has been doing much, much better in watching out for the ad-
vertiser than trying to do something about the proliferation of drugs
that may not really contribute to drug therapy. The Kefauver-Harris
amendment in 1962 made a great contribution in this respe'ct and
I think the FDA has administered that law very capably and effec-
tively. My main criticism is not so much that the Food and Drug Ad-
ministration can cure the problems of this industry as much as that
a system prevails,, an informational system, a distributional system-
the whole process by which information about products comes onto
the market; that is, the information largely comes from the drug
firms and it is aimed at the physicians. Thus the new products and
the claims about them, the information regarding their use all come
from the same source. I believe this is a cause of many of these abuses
in product development and product promotion.
Now, as far as the FDA, there is `one-there are certain changes in
what the FDA can do that will be very helpful and if the FDA has not
done the job that I would like it to do, its failure is in not pressing
hard enough, I believe, for certain further changes that would make
it more effective.
Mr. GROSSMAN. There could not: be this proliferation of products,
certainly, without FDA approval. if the FDA did not approve it, they
just would not be marketing it.
Dr. SOHIFRIN. That is true, the FDA-as far as the proliferation of
products goes, that is true. The FDA is playing a permissive role in
this, certainly. But understand me, when I cite here the large number
of new products, I am not really criticizing the industry's performance
just because of the quantity of its new products here. I am criticizing
the new products in this respect; the industry takes credit for putting
out a tremendous number of new products and representing these
things as therapeutic advancements. Most of these new products are
not therapeutic advancements. They are new items duplicating those
already existing on the market. These may be useful for competitive
purposes, but it represents no advancement of drug therapy. Thus if
the industry is saying, "Look, one of the manifestations of our research
and development effort is this tremendous number of new products,"
we are concerned with the industry's operations and must understand
just how much of a contribution these are or are not.
Senator NELSON. If I may interrupt the minority counsel, the
Kefauver-Harris Amendment of 1962 for the first time gave the FDA
the authority to disapprove a New Drug Application unless the drug
was proved efficacious; is that correct?
Dr. SOHIFRIN. Yes.
Senator NELSON. Now, if there is a molecular manipulation that
gives you a product, which has the same therapeutic result as an
already existing drug, does the .FDA presently have the authority to
deny the company the right to market that drug?
Dr. SOHIFRIN. No, sir. Before 1962, the FDA could deny a New Drug
Application only if the product were harmful-that is, if it did not
hurt the patient, that product could be marketed. Now a product has
to be efficacious. But it does not need to be of superior efficacy to prod-
ucts already on the market to be cleared by the FDA.
81-280-68-pt. 5-21
PAGENO="0322"
1872 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. If it. is just a product differentiation, as the phrase
is used-
Dr. SCHIFRIN. Yes, sir.
Senator NELSON. And if it does duplicate a product on the market
and is efficacious, does the FDA have the authority to prevent the firm
from marketing it?
Dr. Scrn~mN. No, sir.
Mr. GROSSMAN. One further point on this:
If it is efficacious and if the FDA is agreeable to let it. go on the
market, you pointed out when you used the word "it will provide com-
pet.it.ion"-isn't that something we are trying to do here? In other
words, if it is going to be a competitive drug, does it not have some
value as far as our inquiry is concerned?
Dr. SCHIFRIN. That may be a contribution.
Mr. GROSSMAN. That is a pretty important point.
Dr. SCHIFRIN. Yes, but it may be offsetting negative contributions.
In other words, these drugs that. are duplicated may be very heavily
promoted. I think that is a waste of resources that could be used for
other purposes.
Secondly, they may be promoted under brand names, which creates
confusion and the fact that they are duplicates may be obscured in the
promotional literature. They may be represented as new therapeutic
factors when in fact, a large majority are not.
Senator NELSON. The only way in which a company can come onto
the market with a duplicate but differentiated product is if a patent
has expired on the one they are duplicating. Is that not. correct?
Dr. Scrnr~IN. I would say that is the usual way. There are rare
circumstances under which it could obtain a license, usually as the
result of some antitrust activity.
Senator ~ELSON. But that is a license from the patentholder?
Dr. ScrnFRIx. The patentholder, yes.
Senator NELSON. But usually in that case, if it is a differentiated
product, it is a product, as I understand it, of the same chemical com-
position with some insignificant difference and then is put on the mar-
ket and advertised to the medical profession as something better or
something------
Dr. SCHIFRIN. Yes, sir, the differentiation is nominal. It is not in its
chemical composition and many opinions maintain that it is not in its
therapeutic action, either. It is a nominal differentiation, and therefore
the professional thrust is to make a claim that this is a new and better
product. But it is really an identical product to many products already
on the market.
Senator NELSON. So that the reason for doing it really is to benefit
the company in a competitive situation?
Dr. SOHIFRIN. Yes, because if the product is made~ from a drug whose
patent has expired or is not patented, there are likely to be many such
items on the market and an item comes out to fill out a company's cata-
log and will get promoted. The company, hopefully, will want this
thing selected by its brand name. They will put a high price on it., very
often; but it does not give the doctor additional medicine to use for his
patients.
Senator NELSON. What I do not have clear in my own mind is
whether product differentiation by definition is another product with
PAGENO="0323"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1873
the same chemical composition, as was testified yesterday-I assume
that is correct. Is that what a product differentiation is?
Dr. SCHIFRIN. I would not define it that narrowly, as an economist.
Senator NELSON. But in any event, as to the chemistry involved, I
think the testimony was that a differentiated product was of the same
chemical composition as the one that isbeing duplicated.
Dr. SCHIFRIN. That is a common usage of the term, yes. That is the
way it has been used.
Senator NELSON. What I do not have clear is why should the com-
pany go to the bother of creating a product differentiation, whatever
that really means, when all they `would have to do is just duplicate the
product on the market, the same compound, and give it their brand
name and advertise it on the market under their brand name?
`What is the reason not following that procedure instead of going to
what is called product differe.ntiation~?
Dr. SCHIFRIN. `Well, Senator, I think there are two types of product
differentiation. One is an actual difference in the product, a physical
difference. That may be a small difference. That is what the term
"molecule manipulation" refers to, to differentiate a product, chem-
ically, and that is done very often, tO get a patent on a product that is
very much like the product of some other company that is already
patented. That is a way of entering their market.
The second way of differentiating a product is really to produce
something just like the other firms are producing but to differentiate
it by calling it some other name, a trade name differentiation.
Why would a firm do either of these? `Well, to have the physical
differentiation, they would do it for the sake of acquiring a patent
and then having an exclusive item that they could promote under a
trade name, hopefully have it prescribed heavily and have these pre-
scriptions filled, as they are at a very high price.
Perhaps a company duplicates a drug that is already on the market;
first of all if it were a small firm, in the hope that through being an
active price competitor, it could gain large sales. A large firm would
do it in hopes of making that trade name stick and thus having PIIY
sicians prescribe it by its trade name designation and those prescrip-
tions would be filled at much higher prices and thus it would be profit-
able for the large firm to attempt this nominal differentiation, if it
could make that trade name stick for this differentiated product.
Senator NELSON. So in this case, where they simply duplicate the
drug already on the market, the differentiation involved is only the
differentiation so fa.r as identification by a brand name is concerned,
is that correct?
Dr. ScmFiuN. Yes.
Senator NELSON. And the hope in that case is that by effective ad-
vertising, they will be able to get into the market and compete. Is that
it?
Dr. ScmFiiIN. Well, to get into a market and compete, but also,
here we get into the whole practice of trade names, by promoting under
a trade name, of course, they are popularizing the specific trade name
they are promoting. They are also engaging in promoting the practice
of prescribing by trade names
Now, the promotion serves both of these purposes. There may be
many, many firms producing that i~roduet. But only the larger ones
PAGENO="0324"
1874 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
that can popularize these trade names become the effective ones and
thereby become insulated against the generic firms who are in the
market and selling at a lower price. So there are very large profits to
be gained by a larger firm entering the market, even for an unpatented
product.
Senator NELSON. So then you are describing a situation where a
company, because of its distinction, can then effect a nominal differ-
entiation simply by the use of a brand name and then, by effectively
advertising its brand name to the medical profession-because that is
where they advertise-convince the majority of the profession to pre-
scribe that brand name. Once a firm is able to do that, it can artificially
set the price higher. Is that what you are saying?
Dr. SOHIFRIN. Yes, sir.
Senator NELSON. And continue to compete because there is no price
competition. There is merely promotional brand name competition in
the head of the physician who knows the name?
Dr. S0HIFRIN. That is right.
Senator NELSON. Is that what you are saying?
Dr. SCHIFRIN. Yes, sir. The people who have appeared here, many
of them, have referred to this, the competition that you are talking
about, as the competition for the ear and the eye of the physician.
Senator NELSON. To give you a hypothetical case, you could very
well have an unknown minor company that is not established by repu-
tation-that is, by name reputation. It may be a very fine company,
but it is not known widely to the medical profession. That small com-
pany could discover a new and very good drug and go into the market-
place with that drug. Even though they will get some kind of a reputa-
tion based upon this one drug, it would not be a reputation that could
match the reputation of 50 or 100 years of advertising and standing
in the medical community achieved by the big companies. So then
theoretically, you could have a case where a brand name company that
is established, could move in, once the patent has expired, and get a
position in the marketplace because of their acceptance by the medical
profession, charge an artificially high price and occupy a large per-
centage of the market even though they would be selling exactly the
same product as the small firm which developed the drug.
Dr. SCHIFRIN. Yes.
Senator NELSON. Even though the originating company may put
their product out at a much lower price and other generic companies
may come in at a still lower price. Is that what you are saying?
Dr. SOHIFRIN. Yes, and going beyond that-the example posed by
this hypothetical case, is something that could result after the patent
had expired.
But supposing a smaller company did have this development you
are talking about. The larger company would not have to wait until
the patent expired. Through molecule manipulation, they could come
up with a drug with a different chemical structure which would war-
rant a separate patent, but still be very close to the drug marketed by
the smaller company. And shortly after the development made by the
smaller firms, the larger firm could come out with its product, which
is not an improvement now, and through very heavy promotion, take
away even those initial gains that the small firm hoped to make. So the
PAGENO="0325"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1875
small firm may be jeopardized by the promotion of the large firm
during the period when the smaller firm's patent may be valid.
So it could be effective around it, but not over it.
Senator NELSON. You think that the reason the major companies
continue to emphasize very heavily the brand name rather than the
name of the company is that they can effectively compete, then, in the
retail market, where there is no price competition. They can compete
effectively against equivalent quality generic drugs that are available
at a much lower price.
Dr. SCHIFRIN. I would like to change that a little bit and I will agree
with you 100 percent. The large firm can keep the small firm from
competing with it, because the large firm, you see, which has popular-
ized trade name use, renders the small firm, which can't promote under
trade name, an ineffective competitive factor.
I think we are saying essentially the same thing.
Senator NELSON. You are aware that very frequently, the generic
drug producer does compete very effectively against a major firm in
competitive bidding situations; the city of New York, for example,
purchasing for the public hospitals and the welfare department, asks
for competitive bids and is well equipped to test the drugs-and
generic firms do compete in that situation.
Dr. SCrnEPJN. Yes, sir.
Senator NELSON. Or would it be more apt to put it the other way
around, that in that area, the brand name companies decide to com-
pete with the generic firms by reducing their prices substantially. For
example, we have cases where within the same 2- or 3-month period,
a brand name is being sold across from city hall in New York to a
community pharmacist for $8 a hundred and so forth, and then in the
same period, the same company moves in and bids to New York City
at $1.10 a hundred because they know they have to meet the competi-
tion the other brand names who are willing to compete, pius the
generic people who are in that market, too.
Dr. SOHIFRIN. Yes, sir.
And I come to this point in my statement, but to the extent the
large firms can shift the competition from price to promotional com-
petition-that is, they can shift the burden from the selection of
generic equivalents on a price basis-to have that selection done on a
trade name basis, they render the small firm very impotent in competi-
tion, you see. Perhaps this chart-~-
Senator NELSON. Just one moment before you get to that.
Does that explain in part, at least, why the highest price charged
by the manufacturer in every example I have seen, is always the price
charged to the community pharmacist? In other words, if you look at
a listing of all drugs by all companies-I have seen no exceptions-at
the start you will see that the price charged the pharmacist is the
highest price charged in America.
Then you look at bids all over :the country to the Defense Supply
Agency, the Veterans' Administration, hospitals all over America,
cities and counties-and in almOst every single instance, the price
asked by the same company for the same drug is lower, many times
one-fifth. one-tenth, one-twentieth as low as the price they charge to
the community pharmacist. And in those instances where we find a
PAGENO="0326"
1876 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
purchaser in a nonprofit institution, a municipal hospital, paying the
same price as the pharmacist does, if you will look a little further, you
will see that they did not purchase on a system of competitive bidding.
Dr. SCHIFRIN. Yes, sir.
Senator NELSON. So, the brand name identification in the mind of
the doctor, in part at least, accounts for the higher price consistently
charged `the community pharmacist. The industry is not afraid that
an unknown firm's drug would be prescribed. Is that the situation?
Dr. SCHIFRIN. Absolutely.
Mr. GORDON. May I interrupt at this point?
Dr. SCIIIFRIN. Yes, sir.
Mr. GORDON. Can I conclude that even if a small company has a
patent.. it cannot. really protect itself against the large company?
Dr. Scmrnix. That is true, because as I say, the large company can
frequently obtain a. patent on a therapeutic equivalent, of slight chem-
ical difference, you see. The difference in chemical structure would
warrant the patent application and then it would have a therapeutic
agent. here to rival that. `of the small firm and through its promotional
effort could in fact ge.t the bulk of the market. and the small firm,
although having come up with a significant development and having
acquired a patent, may not succeed very well financially in obtaining
innovational profits for t.he accomplishment.
Senator NELSON. If a. new chemical compound is discovered by a
small company and patented, can another firm actually take that com-
pound, do some molecular manipulation, and end up patenting the
slightly modified product? I have some serious doubt about that with-
out. knowing anything about it..
Dr. SCHIFRIN. Well, it seems to me that when the small firm acquires
the patent for its development, the pat.ent comes at. the price of signifi-
cant disclosure. That. disclosure, of course, can serve as `the basis for a
structural change that the large firm might accomplish to come up with
its rival therapeutic a.gent.
Secondly, I think a lot of the development, occurs simultaneously.
That. is, there is a lot, of literature in the professional journals about
progress, technology, chemical manufacture a.nd so on. It may only be
that. the small firm may beat the large firm at certain times just by a
very narrow margin in the development of its product. The large firm
is almost up to the small firm and when it sees the small firm has a
patent, it ca.n do a little sidestep and obtain a. patent. on a slightly dif-
ferent chemical substance.
Mr. GORDON. Do you have examples of that?
Dr. ScmFRIN. The only case I can think-it is a little bit different.
For example, meprobromate. We know that. it has been an e.normous
financial success for Carter and Wyeth. But me.probromate was actu-
ally discovered by Dr.. Berger in England. Rather `than initiate pro-
duction as a small manufacturer, he went with his discovery into
the employment of Carter and entered the market through a large
firm. I think this reflects the very unstable and uncertain posi-
tion that `a small firm making a development would face in the market.
That was avoided in that case by going directly to the large firm.
Senator NELSON. In some of `these developments where a number
of researchers and companies around the world are working on the
PAGENO="0327"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1877
same problem, seeking `the sathe solution, progressing at a similar
rate, is `there evidence tha~t because of the fear of a legal contest of the
patent by one versus the other, they reach some agreement th'at one will
get the patent and that `he will license the others?
Dr. SCHIFRIN. Very much so. When Patent Commissioner Ladd
appeared before the Kefauver committee, he poInted out that patent
interferences-that is, where companies contest a patent application
made `by another company-are more frequent in chemical and chem-
ical-related industries than any other industry. That is fact one.
*T'he second fact, and this came from, I think, Mr. Federico, who was
with Commissioner Ladd, is that a patent interference results in a very
careful scrutiny of a patent application. In a high proportion of `these
cases, it turns out that the result is unp'atentability. That is, a decision
that the patent should not be awarded. Large firms are researching
along these similar lines. They attempt to test the patent application,
because rewards for getting a patent. are very high in this industry.
Hence, the major firms that have pursued parallel lines of research,
often permit a patent applicant to go ahead unchallenged and to get
his patent. But the other Targe firms, for not challenging that patent,
will get licenses to produce it. In other words, they will share the patent
among three or four firms rather than have nobody get the patent. So
licensing is a result of `these agreements not to test patents by the large
finns.
Senator NELSON. When you say they share the patent., you do not
mean that three companies, more than one company, gets the patent
in their name?
Dr. SCHIFRIN. No, but as a condition for withdrawing from the in-
terference, the companies that withdraw get licensing privileges.
Senator NELSON. Is that a violation of the antitrust laws in any way?
Dr. Scrnn~IN. Well, I would say that in a considerable number of
these cases, t'he licensing agreements have been accompanied by price-
fixing agreements. We have seen that in `meprobromate, and `in tetra-
cycline. We `have seen it `several times t'hrough the industry, that the
cooperation involved in eros's-licensing is a v'ery tempting circumstance
to lead to a-if not an overt, at least a tacit price-fixing agreement.
The tetracycline is a notable example of these things.
Senator NELSON. I notice tha't there is more than one type of case. I
have seen instances w'here one company `gets the patent and licenses
two others who have been working in the same area, doing `parallel
research. T'hen there are cases where one company gets the patent and
licenses anybody wh'o wants to b,e licensed for all practical purposes.
Dr. SCHIFRIN. T'hose are rare.
Senator NELSON. There are some; are there not?
Dr. `SCHIFRIN. Yes, reserpine, I `think, is the outstanding example.
Mr. GORDON. Dr. Schifrin, if the large companies will not test the
validity of `the patents owned `by other large `companies, an'd if small
ccmpanies can't do it because of the litigation expenses, then who is
going to protect the public against the possibility of invalid patents
being used `to reap `monopoly profits?
Dr. ScrnriuN. Mr. Gordon, as of now, the answer to `that is nobody
does. The FT'C, in its tetracycline case a few years ago, under section 5
of the Federal Trade Commission Act, `d'id consider a fraudulent pa-
PAGENO="0328"
1878 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
tent application to be an iuifair method of competition. So to the extent
they are willing to do it, they can. They `have been the only ones who
have don~ thi~ and of eour~ fhe~e are infrequent examples. There is
no regular mechanism by which the true validity of patents can be
watched over and ascertained right now. One of my proposals is related
very closely `to that point.
Senator NELSON. You mean, then, that if a company has what ap-
pears to be a new product and there is doubt in the minds of other com-
panies that it `is really patentable, in order to avoid free competition,
they just do not challenge its patentability; so that it then becomes
patented and some compa.nies become licensed and then, for 17 years,
there is an artificially high price paid by the consumer because there
a patent has been granted, which, if challenged, would not have been
granted in the first place?
Dr. SornriuiN. Correct. In other words, market domination by a few
firms that exercise a monopolistic type of interest comes to exist, yes.
Senator NELSON. And you say there is no effective protection in
behalf of the public by any public agency against this kind of
occurrence?
Dr. Scrn~'Rn~. At present, there is not.
Senator NELSON. It rather shocks me that you could have a situation
where something that may be shown to be not really patentable if
challenged becomes patented and no public agency is prepared to pro-
tect the public interest.
Dr. SoirFnm-. Well, the most careful scrutiny of a patent's validity
occurs in the court, not at the Patent Office. Thus, things for which
patents are issued by the Patent Office may, if contested, be found to
be invalid. The patent may be found to be invalid if contested. Thus
the burden is put on usually some small firm. And the small firm, as
Mr. Gordon correctly pointed out, usually can't bear the expense of
litigation.
Senator NELSON. And you are advocating that some mechanism to
protect the public interest in this kind of instance be established?
Dr. SOHIFRIN. Yes, sir; I think an advisory panel to the Patent
Office, reflecting the highest degree of expertise in the medical field
should be established, and should give these patents the very careful
scrutiny that would never arise otherwise.
Senator NELSON. Are you satisfied that there is no mechanism under
the present law by which any Federal agency could, on its own motion,
challenge the patentability of a product?
I do not mean challenge a conspiracy to monopolize or any such
thing, but challenge just the patentability of a product?
FDA, Federal Trade Commission, anybody?
Dr. SOHIFRIN. I am not positive as to the interpretation that these
agencies give their own enabling legislation, but I will say this: If
any such organ exists, it certainly is not used rigorously, if at all.
Senator NELSON. Do you think that your proposal is feasible as a
practical matter? That is, would it really put into the hands of public
representatives adequate tools to protect against this specific
circumstance?
* Dr. SOHIFRIN. Yes, sir; becanse I believe that much of this exper-
tise, and from time to time, several related important functions, have
PAGENO="0329"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1879
been filled by people who are in private industry-I should say mainly
in medicine. Many of our people who are outstanding physicians, who
teach in medical schools, certainly are competent to assist in this
matter.
Senator NELSON. You certainly would not give this kind of group
final authority; would you?
Dr. SOHIFRIN. No; but I would hope, sir, that their expertise would
be recognized by the Patent Office in determining patentability; that
these people serve as a rather influential body of experts.
Senator NELSON. And what is the next step in the event that this
independent body that you would establish made a recommendation
which they felt very strongly about but was not accepted by the Patent
Office. Would you provide a mechanism then, which would allow the
Government to go into court to challenge the patent?
Dr. SOnIFRIN. That is one alternative. I would consider that as
feasible; yes. There is another alternative that is coordinated with
that one.
I know we have wrestled with the problem of patent standards very
often in our industry, and since the drug industry occupies a rather
unique position in American industry, I would not be opposed to a
unique patent law applying to drug patents, with clearly different and
higher standards than are now provided by our general patent law.
I think the drug industry warrants separate treatment.
Senator NELSON. If this is, as I believe, a serious problem, and it
may very well be, would you not be reluctant to place the final arbi-
trary authority in a commission or~ committee such as you recommend,
or even in the Patent Office itself? In other words, would you not
think that if you had a commission and they made a recommenda-
tion that an item was not patentable and the Patent Office did not
agree with that conclusion, that the legal arm of the Government,
the Justice Department, ought to be able to move to test it? Or if the
Patent Office did agree and the patentholder did not, they still ought
to have a right to go to court?
In other words, should you not have as the final resort for both the
Government and the company or companies involved the right to go
to court?
All you are trying to suggest here is that the public interest be
protected by establishing some agency with the authority to challenge
the patentability of drugs and with the ultimate authority, in behalf
of the public, to recommend to the Justice Department that the patent
issuance be challenged in the courts?
Dr. S0nIFRIN. Ultimately; yes.
Senator NELSON. Go ahead.
Dr. SomriuN. I have mentioned about the practices and the factors
that have become a part of the industry's operation. Now let me de-
scribe how this design came into being, this product and market per-
formance that I have reviewed now.
In the late 1940's and early 1950's the major drug firms became
alarmed by vigorous price competition in the sale of drugs, most
notably penicillin and streptomycin. They sought to insulate them
selves from further price competition for these and other products
through the device known as the "specialty" item-that is, finished
PAGENO="0330"
1880 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
products sold under unique trade names and, to the maximum extent
possible. made from drugs for which patents had been acquired. These
two devices-the trade name and the patent-have., to an astounding
degree, eliminated competitive conditions from the markets in which
drug products vie for sales. The vast majority of all new drugs have
been covered by patents, sometimes of questionable validity.
Patents monopolize the sale of products for a single seller or, in
some cases, for it and a limited number of licensees. Even for products
sold by many firms, the popularization of the use of trade names in
prescription writing has led to the dominance in those markets by the
few major companies in that product line who can promote through
massive advertising outlays the trade names of their specialities.
Monopoly or oligopoly has thus become the usual situation-in the
antibiotics, the hormones, the mental drugs-and down the list of areas
of drug therapy. The competition among firms in the market is no
longer one in which producers of comparable items seek customers
through more attractive prices, but one in which a single seller often
exists alone, or, if lie shares the market with a few rivals, that rivalry
is in advertising, claims, and trade-name repetition-certainly not in
price.
This is a large, costly, and wasteful competition. It is excessive, con-
fusing, and largely ignored, but it does serve its purpose. It. popularizes
particular trade names and strengthens the use of trade names in gen-
eral, thereby rendering price competition ineffectual. Such advertising
a.dds little if anything to drug therapy, yet is a large cost, easily shift-
able to the consumer.
To some extent the peculiar dependency of the physician on the drug
firm for information places some of this activity above ordinary adver-
tising-~but even this "promotional" literature often has contained mis-
leading, insufficient, exaggerated, or otherwise inadequate information.
The same features that e.ncourage such wastes and enable drug
firms to charge their costs fully to consumers also condition the profit
margins contained in drug prices. These margins provide for rates
of return to the industry as a whole and to all the large drug pro-
ducers far higher than for the economy as a whole, year-after-year.
Furthermore, these extraordinary profit levels are remarkably stable.
Year after year, the ethical drug industry is right up at the top of
the industry profit ranking. Year after year the after-tax rates of
return on stockholders' equity for large drug firms exceeds those of
other firms; year after year the ranking in "Fortune" show a. dispro-
portionate frequency of drug firms among the most profitable large
firms in the economy; year after year the FTC-SEC profit data in
manufacturing reveals the outstanding performance of drugs. The
high profitability reflects the absence of competition. The stability of
profits demonstrates the absence of risk to investors. If risk were to~
exist, one would expect to see the high gains of some firms accom-
panied by occasional losses-to themselves or to others-but such evi-
dence of risk is virtually nonexistent. Unable to exploit the "risk"
justification for its profit levels, spokesmen for the industry have
tried to rationalize the great profits with the "research" argument-
that high profits are necessary to finance the industry's research effort.
PAGENO="0331"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1881
This is an illogical theme, however, since profits exist only after all
costs, including research, have been covered. A firm that breaks even,
or earns only a normal profit, is financing its research just as fully
and capably as is the firm earning profits far aibove normal.
As that argument folds, the industry turns to the "growth industry"
argument-that is, that the drug industry is a so-called growth indus-
try and that high profits are needed to finance that growth. In
response to that theme I contend that reinvesting your earnings instead
of taking them as dividends is one thing; but exacting from consumers
a double profit and more, to sustain and even increase dividends, on
the one hand, and simultaneously to finance capital expansion and
stockholders' equity on the other hand, is uneconomic, unjustifiable
exploitation. The consumer who purchases his prescription is thus pay-
ing in that price for these things-for activities necessary to bring
that product to him; for activities that provided no benefit to him and
hence were unnecessary; an attractive dividend to the stockholders, and
some part. of a new machine, a new plant, a piece of property that
will belong to the stockholder. The dramatic expansion of the industry
in the past 15 to 20 years has been almost entirely, if not totally,
financed out of profits-yet the industry throughout this period has
maintained a payout rate., dividends per invested dollar or per share of
stock, that compares favorably with other industries. That, to me, is
something like having your cake, eating it, and seeing it grow bigger
all at the same time.
Thus the market performance of the ethical drug industry, meas-
ured against the criterion that "prices paid by consumers reasonably
reflect. the costs of efficiently providing useful activities" is seriously
deficient. The prices paid by most consumers of drug products are
excessive for two major reasons-they are inflated by wasteful cost
elements, and they are further inflated by the excessive profits they
provide.
The question that now arises is this: What features of drug markets
render consumers so exploitable?
First, there is the peculiar importance of the product, more so than
almost any other commodity; then there is the "prescription relation-
ship," in which someone other than the consumer actually decides
what will be bought-someone who may be unaware of the availability
of alternative products, unaware of their relative prices, or indoctri-
nated in the practices of prescribing high-priced trade name specialties.
Second, given this vulnerability of consumers, to exploitation, is
monopolization-the basis of the power to exploit the consumer. This
monopolization is both result and. cause of the wasteful competition in
development and especially in advertising, and is a prime determinant
of excessive profits. all of which the consumer bears. The large drug
firms, as I have indicated, strive to create monopoly through patents
and trade names. But what is it that permits them to succeed so
impressively in that endeavor?
To answer that question we must distinguish two separate facets
of the industry's operation-on the one hand there is the development
and manufacture of drugs, i.e.,,: the active chemical substances that
go into pharmaceutical preparations. On the other hand there is the
PAGENO="0332"
1882 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
manufacture of the finished dosage-form products or preparations
containing those drugs. As breakthrough discoveries have opened up
the fields of antibiotics, cortical steroids, mental drugs, antihistamines,
vitamins and nutrients, antidiabetic drugs, and others, a pattern of
specialization has emerged. Because of differences in their back-
grounds, in their interests, in their early activities, and also because
of their inability to pursue many diverse research lines simultaneously,
the larger firms have focused their efforts and resources on one, or
perhaps a few, areas of exploration. In most of these areas, perhaps
with the exception only of mental drugs, developments after the
breakthrough discoveries have been of a "stepladder" nature. Those
firms first on the ladder have kept ahead of others; entry on the
ladder is impossible for smaller firms and extremely difficult even for
large firms once the early entrants have gained momentum and ac-
quired patents, thereby protecting all the previous rungs as well as
the ones they are now perched on, from being used by potential rivals.
Thus, scarcity of opportunities for entering into the manufacturing
of drugs-the chemical substances-has closed entry into the man-
ufacture of ~nished products, except in those few cases where the bulk
drugs are available to all fabricators of preparations containing them.
In most drug product markets the only sellers are those large firms
who hold the patents on the drug ingredients. In a few cases other large
firms are licensed to use the patents; and the many small firms in the
industry are thus limited to the manufacture and sale of products con-
taining unpatentable drugs, drugs whose patents have expired, or
drugs which are available because licensing by the patent holder is
required as a consequence of antitrust guilt. But even in these few sit-
uations, the small firms are not equal competitors of the large ones-
because the popularization of trade names (possible only for large
firms) in prescriptions renders the generically designated items of
the small firms as inconsequential competitive threats.
As a result of this specialization, patent acquisition, and the use
of trade names, there have developed two different groups of drug-
product markets, which I have depicted in this diagram.
Across the top, that says, I believe, "Products Made from Drugs
that are," and then the lefthand column "Patented," in the righthand
column, "Tlnpatentecl."
Since most drugs, most commercial items, a.re patented, I have that
column to the left larger than the one to the right.
Now, reading down the column, the columns then show the products
made from drugs that are patented on the left and on the right, the
products made from drugs that are unpatented. Reading across, first
we have the private prescription market. This is the market in which,
you know, the family physician prescribes a medicine and the prescrip-
tion is filled at the drug store.
Down below, I have the institutional market. Since approximately
70 percent of drugs-maybe 60 percent now-are sold in the private
prescription market, I have made the first row across larger than
the row below it.
PAGENO="0333"
PRO]XJCTS
PATENTED
CONTAINING
DRUGS URPATENTED DRUGS
A
A
where trade names
.
A
-
A
are used
where generic
names are used
.
B
Dr. SOrnFRIN. Now, what does this show? Those areas on that dia-
gram that are marked with "A" shows where monopoly power exists.
If a product-that is, a preparation-is made from a patented drug,
there will be very ~few sellers, perhaps only one, of that drug. Regard-
less of whether that drug then is purchased in the private prescription
market or in the institutional market, there cannot be price competi-
tion because of an unavailability of competitive alternatives.
Thus, all the items in the first column will be "A," the absence of
competition.
In the second column, we have drug products containing unpatented
drugs. There will be many sellers here. But in the private prescription
market, where selection is done by trade name, only those few large
firms that can promote trade names are effective sellers. That is, they
dominate the market. So they can continue to charge the high prices
that accompany trade name products and thus, even though these
products contain unpatented drugs, there is an effective monopoly or
oligopoly nonetheless.
Now, for products containing unpatented drugs sold to institutional
purchasers, I have both "A" and "B" there. If the institutional pur-
chaser for some reason or other continues to do his prescribing, his
purchasing, by trade name terms, then the small firms are still cut
out of the market.
Thus, only in the institutional m~trket, only for drug products made
from unpatented drugs, and only where the purchaser-that is, the
institution-considers these items to be generic equivalents and goes
on a price basis-only in that small area, "B," is there a price competi-
tion in the market for ethical drug:products. That represents approxi-
mately 5 percent of all drug sales in this country. In other words, all
the areas in "A" represent 95 percent of dollar volume of drug sales.
The competitive sphere, "B" is 5 percent, approximately.
The largest two dozen firms account for the approximately 95 per-
cent of total industry sales represented by markets labeled "A"; the
remaining hundreds of smaller firms share the 5 percent of industry
sales accounted for in market "B."
The past history of the industry can be described as largely one
of creating and expanding "A" relative to "B." The continuing goal
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1883
(The diagram referred to follows:)
PRIVATE
PRESCRIPTION
MARKET
1INSTITIJTIONAL
~4ARKET"
PAGENO="0334"
1884 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
of the large flrms has been the enhancement of their individual and
collective profitability by preserving and expanding those sectors
oi tue market where effective competition does not prevail. This o~oa.l
is sought in these ways-(1) by mnakin~ sure that the new drugs ~at
replace older ones, particularly unpatented ones, are protected by
patents-even though their patentability might not stand close
examination.
Mr. GORDON. You stated before, Dr. Schifrin, that there is a sort of
gentleman's agreement among the large firms in the drug industry
that., "you won't question my patent and I will not quesHon your
patent." Is that a fair summary of your position?
Dr. SCHIFRIN. That does not summarize my entire position.
Mr. GORDON. I mean in the existing situation.
Dr. SCHIFRIN. That is part of the existing situation, Mr. Gordon.
The second way in which the large firms try to enhance situation
"A" is by avoiding patent interference suits, which are fairly frequent
in the chemical technology industries because of parallel research by
the specializing firms, through often elaborate cross-licensing agree-
mnents. These agreements, over and above the price fixing that not
infrequently has accompanied them, are enacted to avoid the pos-
sibility that a party losing an interference suit may challenge the
patentability of the discovery.
Mr. GRoss~IAN. Dr. Schifrin, could I ask you to he specific about
this? You stay that these agreements were made "over and above price
fixing that not infrequently has accompanied them."
That is a strong allegation. Could you be specific about this fre-
quent price fixing?
Dr. SCIIIruIN. Well, in recent years, two of the most successful drugs
that have been developed, conmiercially successful and quite important
with therapeutic significance, have been meprobrornate, the tran-
quilizer, and tetracycline, the broad spectrum antibiotic. In the case
of meprobromate, Carter had the patent. Wyeth had the detail men,
et cetera. Carter entered into a licensing agreement with `Wyeth for
them to also be able to sell meprobromate and to promote it very ex-
tensively. Thus for many years the only two meprobrornates on the
market were Equanil and Miltown. Price fixing in these two com-
panies was found by the courts in recent years and compulsory licens-
ing of meprobromate resulted from that case.
The second of these two examples I use is tetracycline. Just recently,
of course, the Federal court found that there was a Sherman Act
violation as a result of price fixing. The Federal Trade Coimnission
case that I cited earlier tied the price-fixing agreement very closely
to the cross-licensing agreement that brought the five sellers into the
market.
Mr. GRoss~rAN. I am aware of those cases. You used the words "not
infrequently." That is why I wondered if you had other examples and
whether the Justice Department has prosecuted either civilly or
criminally on price fixing frequently.
Dr. SCHIFRIN. I have a list-at one time, I did compile a list of such
cases. They go back to 1040 and I would be glad to provide that in-
formation at a later time. It is just a matter of finding the appropriate
footnotes.
PAGENO="0335"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1885
But I think-
Mr. GROSSMAN. I-low do you define frequently?
Dr. SOHIFRIN. I would define frequently in that-in this way: have
there been found to be enough price-fixing agreements in the industry
so that the reasonable observer would examine carefully licensing
agreements for instances of price fixing?
Have there been enough violations of the Sherman Act in order to
warrant the assumption that there is certainly a definite possibility
that price fixing accompanies cross licensing.
Mr. GROSSMAN. You will submit that for the record at a later date?
1)r. SCHIFRIN. Yes.'
Mr. GROSSMAN. One other question at this time: I think Mr. Squibb,
when he testified, talked about the industry's desire to get into the
teaching institutions and therefore to cut their prices to go into these
hospitals, to make sure that the young doctors see their product or
become familiar with them.
Do you think that the consumer, who I think more frequently buys
the "A" products there, is paying for this activity by the industry?
In other words, he is paying a lot more because the industry is cut-
ting these prices in the institutionalareas?
Dr. SOHIFRIN. No. I do not believe that the prices to the consumers
would be lower than they are if the drug firms did not engage in that
activity; no. I believe that the consumer pricing is based, really, on-
in fact, I have definite information on what consumer pricing is
based on.
I have correspondence from people in drug firms going back quite
a number of years and this you mention was never determinant as a
factor in setting prices.
Mr. GROSSMAN. Do you think the drug companies make high profits
on those sales to institutions at very low levels?
Dr. SOHIFRIN. Yes; I do.
You are familiar with marginal cost pricing. One reason for my
belief that the consumer bears the expense of the research and devel-
opment, bears the expense of the promotion, bears the expense of the
large profit. The actual costs of manufacture of most drugs is very,
very small. I am sure the price charged to virtually any person covers
at least the direct cost of production. The firms are not losing money
on those sales to hospitals.
Mr. GROSSMAN. Thank you.
Dr. SOHIFRIN. Careful scrutiny of patentability is a threat to drug
firms because there is a sizeable chance that no patent may actually
be deserved; without patent protection, many firms can manufacture
or obtain the drug, produce preparations containing it, and sell it
with subsequent price competition a possibility. In any case, the mar-
* ket will be shared by more sellers than otherwise.
(3) Even where patent protection has not been garnered, trade
names accomplish a nominal differentiation largely accepted by phy-
1 Dr. Schifrin subsequently stated, "In addition to the meprobromate and t~ r; cy~line
cases, other price fixing cases have been: U.S. v. Eli Lilly and Co., USDC (lie) 1941,
(insulin) ; U.S. v. Schering Corp., at al., IJSDC (NJ) Civil Action No. 1919, 1)141 (hor-
mone products) ; U.S. v. Alba Pharmaceutical Co., et al., USD0 (SDNY) 1941 (imports)
U.S. v. Eli Lilly and Co., et al., USD0 (NJ) Cr. 173-58, 1959 (Salk polio vaccine)."
PAGENO="0336"
1886 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
sicians, which effectively eliminates any possible competitive threat
posed by the generically labeled output of the small firms. Trade
names supplement patent monopoly where it exists and substitute for
it where it does not.
Thus the large drug flrms have u~ed two devices, patents and trade
names, to eliminate virtually all tendencies toward price competition.
They have led to a different sort of competition-one in which the
consumer comes up a loser. Not forgetting the contribution to better
health and longer life made by the industry, there has emerged in it
a new competition-one that rewards molecule manipulation, ques-
tionable patent tactics, excessive promotional claims, and oftentimes
a product inadequately tested or cautioned. Such abuses are part of
the industry's record, and have generated an increasing surveillance
and regulation by the Food and Drug Administration.
Inevitably we come to the matter of public policy in regard to
this situation. Specifically, the question as I see it is how to im-
prove the market performance of the industry while not impair-
ing and hopefully even improving its product performance, as I
have used those terms. Stated in perhaps a more meaningful way, the
question is: How can public policy restore effective competition to the
manufacture and sale of ethical drug preparations and thereby make
their prices more reasonable, while preserving sufficient incentives for
the discovery and development of new and better products?
First, there is the matter of standards of drug patentability. Higher
standards than those now prevailing are necessary to halt the routine
issuance of patents whose validity is not substantiable in court. Higher
standards of patentability will continue to reward true accomplish-
ment and even induce more of it by affording it more protection than
is now possible; patents for insignificant or substantial coattail
developments or modifications would be eliminated. Such a change
would greatly limit the financial gains available from molecule manip-
ulation, but increase the gains from significant discovery, thus re-
directing research and development funds away from imitative into
innovational channels.
This is the context in which I suggested this commission of experts
to assist the Patent Office. I believe they could provide a good deal of
influence on this higher standard of patentability for drug patents.
Several years ago the Federal Trade Commission found that the
tetracycline monopoly was built on patents obtained with "unclean
hands and bad faith." Both the ability to acquire patents in such a
manner and the economic motivation to do so must give way. My first
specific recommendation, then, is that a special group, representing
knowledgeable legal and medical expertise, serve as constultants to
the Commissioner of Patents in reviewing and determining drug patent
applications.
My second proposal also deals with patents, but is further reaching
in its impact. This recommendation focuses on the duration and scope
of drug patents, especially pertinent in view of the monopolizing effect
of such patents. It is my view that in duration as well as scope, drug
patents provide excessive protection from competition, to the detriment
of consumers. Accordingly, I offer two alternative plans for making
PAGENO="0337"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 1887
drug patent protection relate more closely to the realities of this indus-
try. These two plans are alternative policies for the compulsory licens-
ing of drug patents.
The first alternative is to require compulsory licensing of all drug
patents after some specified period, such as 3 years.
Three years of exclusive patent prOtection is a reasonable period in
an industry characterized by rapid product turnover and a high rate
of obsolescence. Numerous studies have shown that the greatest portion
of sales of any product is likely to occur in the first few years after
its introduction. Company price policies explicitly take into account
these considerations, and most companies, if not all, estimate very con-
servatively the anticipated market life of their proctucts, usually tak-
ing 3 years as the period to recoup outlays and earn a profit.
If licensing were required after the first 3 years of a product's life,
that is, after its estimated life expectanc,y for pricing purposes were
ended, there could occur entry by other firms into that product market
and, hopefully, competition in price among the rivals. Of course, bene-
ficial results to consumers would be possible only for those products
with a therapeutic or commercial life longer than 3 years, `but despite
the swift product turnover in this industry, data on product sales in-
dicate that the majority of sales in any recent year represents those
of products on the market longer than 3 years.
For them, the patent holder would continue earning entrepreneurial
profits, though perhaps at a lower rate than before, on his own fin-
ished-product sales and those of licensees, and consumers possibly
could now purchase their prescriptions at lower prices.
`The impact of such a policy on research and development does not
seem unfavorable; it provides a time long enough, from the companies
own viewpoint, to earn profits justifying the innovational effort. It
might even promote greater research and development by inducing
even more rapid product turnover. In many cases, a realistic period of
exclusive use and compulsory licensing at a fair royalty rate afterward
seem unlikely to deter research and innovation.
My second alternative patent-licensing policy focuses on scope, as
contrasted with the first alternative and its focus on duration.
This view of drug patents raises the question as to the justification
for any period of exclusive patent use. The contention that none is
necessary is based on the fact that a drug patent gives its owner a
monopolistic position in either of two markets, that of bulk sales, that
is, of drug substance itself-or that of dosage form products.
To the extent that he takes his profits in the sale of bulk or in roy-
alties from licenses for its manufacture, he must share the market for
finished products; if he retains his monopoly in the latter-the fin-
ished product market-he cannot reap profits from bulk sales or licens-
ing, but of course, can earn substantial profits as the sole seller of
finished products.
Compulsory sales of the bulk drug or licensing of its manufacture
merely specifies that the patent holder must reap his gains in the bulk
market rather than the final-product market; it does not take away
the opportunity to earn a profit jusifying the effort behind the dis-
covery. And entry into the preparations market would occur at the
81-280-68-pt. 5-22
PAGENO="0338"
1888 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
time when the item is introduced, rather than 3 years later, and corn-
petition can possibly come into play that much earlier.
Perhaps in this plan the royalty might have to be higher than in
the earlier approach, but not by so much as to prevent licensing from
being an effective check against exorbitantly high bulk prices.
Patentees would earn reasonable and adequate profits by charging
attractively high bulk prices or by imposing the most favorable
royalty rates permitted. But the cost of bulk ingredients is usually
only a small fraction of the total cost of production, and high bulk
costs or royalties would be more than offset by the economies in re-
source use, particularly reductions in the vast and largely wasteful
advertising effort, and the more reasonable profit margins that price
competition would bring about.
Compulsory licensing, because it permits rival firms to enter into
the market, thus is a necessary condition if price competition is to be
restored to the industry. But it is not a sufficient condition for that
competition to arise, it opens the door to the entry of additional firms
into markets closed by patents but it does not make them effective
competitors of the dominant one or few.
I come then to my last major recommendation-the prohibition of
trade name designations. Trade names, as you well know, are those
unique company names for its products-simple, catchy, and easily
remembered-Syncillin, Achromycin, Tetracyn, Pen Vee, Miltown,
Ledercillin, Orinase, and on and on and on throughout the catalogs
of the large drug houses. Such names are totally unnecessary in every
respect. If differentiation of drug products is necessary, and I am not
fully convinced that it is, let it be done not through a proliferation of
new names that are intended to displace generic terms for the product,
but in the same way as differentiation is made in virtually every in-
dustry, by the use of the manufacturer's name. Thus, the names "Car-
ter: meprobromate" and ~WTyeth: meprobromate" tell us much more
than do the words "Miltown" and "Equanil," while preserving com-
pany differentiation.
(These statements of mine parallel very closely the testimony of
Dr. Garb on June 20.)
The use of brand names that combine the company and generic
comparability becomes clear and un~bscured, contrary to the purpose
and effect of trade names.
The elimination of trade names will go far in establishing the facts
of generic similarity to physicians. Those doctors who want to select
the speciality of a particular firm can continue to do so by using the
brand name; but those who feel, as many do, that generic equivalents
are therapeutic equivalents can thus prescribe by generic name alone,
or by the brand name of a reputable seller whose product bears a com-
petitive price tag.
These three proposals of mine-a special drug patent board, some
form of compulsory licensing, and the elimination of trade names-
can, together, go very far in restoring opportunities for competitive
entry into markets, in restoring price competition in the place of waste-
ful and often harmful promotional competition, and in bringing about
reasonable prices to consumers, while preserving the incentives for the
research and development effort behind the industry's generally com-
mendable product performance.
PAGENO="0339"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1889
The opportunities for extraordinary profits still exist, but only for
innovation or discovery of therapeutic merit-not through monopoli-
zation of markets and exploitation of consumers who have no alterna-
tive but to pay prevailing prices.
All this seems to me to represent a considerably more acceptable situ-
ation than now prevails in the ethical drug industry.
Lest the tone of my proposals, or the spirit in which I offer them
be misunderstood, I would like to offer a sort of epilog to my prepared
statement.
I do not view Government as something necessarily separate from
the citizenry of this country, nor do I view it as a monolithic force seek-
ing to create for itself an increasing power over the individual persons
or business firms in our economy. I firmly believe that, as a Nation, we
have maintained a prima facie case for individualism, for free enter-
prise, and for the free market economy.
But our economy, as other facets of our society, requires a system of
checks and balances to prevent the generation of private economic
power and its subsequent abuse. Our general commitment to competi-
tion is based on the importance of the checks and balances that compe-
tition imposes on the pursuit of economic self-interest, thus channeling
the operation of the economy in a direction that serves the interests
of society as a whole.
But the absence of effective competition in most sectors of the drug
industry must be recognized, and its implications in terms of that
industry's operations have been the basis and the main concern of this
committee's hearing.
The large firms in this industry have acquired an economic power
incompatible with our conception of a free market economy; oppor-
tunity is severely limited for the many small firms in the industry,
and the purchasers and consumers of ethical drug products are denied
the benefits of competition in the manufacture and sale of those
products.
Even if we opposed only the abuse of such power, rather than its
existence, the ethical drug industry has not restrained itself in any
significant way. The profit record of the industry, unjustifiable by
any of the accepted standards Of economic performance, attests to
this lack of restraint. Wastes in promotion and other aspects of
development, manufacture, and sale add to the costs imposed on
society because of the absence of effective competition.
Thus, in the absence of restraints on individual power through
market competition or self-imposed restraints, it falls to Government
to induce an improved total performance from this industry. This,
as I see it, is a proper and necessary role for Government in the
economic field, and a role not at all incompatible with our commitment
to a free enterprise economy.
I believe we have come a long way in our understanding of the
factors contributing to both the positive and negative aspects of this
industry's operation and performance-and that we can devise ap-
propriate changes in the context of its operation that will greatly
enhance its total performance, while preserving the incentives necessary
for its existence and growth.
At this point, Mr. Chairman, I would like to request that a recent
article of mine, "The Ethical Drug Industry: The Case for Com-
PAGENO="0340"
1890 CO~ETITWE PROBLEMS LN TI~ DRUG INDUSTRY
pul~ory Liceneing," the Antitrust Bulletin, fall 1967, be included in
the record of these hearings in the appropriate place, for its relevance
to the proposals I have made.
I thank you for honoring me with the opportunity to make these
remarks to you and for your kind attention in my presentation of
them to you.
Senator ~ELSON. The article on the "Case for Compulsory Li-
censing" will be printed in the record at this point.
(The material referred to follows:)
[From the Antitrust Bulletin, fall 1967]
THE ETHICAL IJRTJG INDUSTRY: THE CASE FOR COMPULSORY PATENT LICENSING
(By Leonard G. Schifrin~)
Introduction
On December 7, 1959, the Subcommittee on Antitrust and Monopoly of the
Senate Committee on the Judiciary, then popularly known as the Kefauver Corn-
mittee, shifted the focus of its investigation of administered prices in the Ameri-
can economy to `the ethical drug industry.1 In the Spring of 1967, the monopoly
subcommittee of the Senate Select Committee on Small Business began hearings
on ethical drug prices, particularly on the often large price differential `between
finished products sold under company-assigned brand or trade-names and finished
products of the same generic designation sold under the chemical or generic
name,2 an issue originally raised by the Kefauver Committee. In the seven-and-a-
half years between the first and the most recent Senate `hearings, `at least nine
other series of hearings dealing with this industry have been conducted, and at
least three additional committee reports `or studies have been submitted to Con-
gress. `Some legislation has resulted from this extensive examination, but the only
action of real substance, the Drug Amendments of 1962, deals mainly with ques-
tions of drug safety and perhaps owes its passage as much to the Thalidomide
tragedies in Germany and other European countries as to the economic and medi-
cal issues raised in these many hearings.
Despite the lack of legislative accomplishments, the time and attention spent
in scrutinizing this important industry have, for the most part, been productive-
The industry grew to maturity without drawing attention to its practices and
performance, perhaps because of its close relation to the medical industry which
traditionally does not publicize its economic activities or perhaps because of its
continuing high profitability. The investigations, however, have revealed flaws in'
its `operation, specifically its wasteful use of resources in promotion and research,
the dubious contribution of some of its output, and the uneconomic relationship
b~tween the costs and prices of its products. In the years in which the industry
has been so frequently studied, public concern regarding the health services and
products available to consumer-patients has grown: Medicare has become part of
our Social Security law; support for the construction of health facilities has-
multiplied significantly. These activities are, of course, only part of our growing
concern for ever more numerous facets of the quality of human life. To the extent
*Head, Department of Economics, College of William and Mary.
1 It Is symbolic of the character of the ethical drug industry that the term "ethical drug"
itself has become obsolete. Drugs, technically, are the active ingredients which go into
dosage-form products or pharmaceutical preparations, rather than the products or prepa-
rations themselves. Now, however, the large majority of all preparations are pre-
fabricated, I.e., already In dosage form when sold to pharmacies and doctors. Hence, the-
modern ethical drug industry Includes firms primarily engaged In the fabrication, finish-
ing, or sale of drug products or preparations in finished dosage forms such as pills, capsules,
tablets, etc. Although the industry would be more accurately described as the ethical-drug.
products or preparations industry, common usage still retains its now-dated designation.
2 All ethical drugs have generic, I.e., common or chemical, names. In order to distinguish
their items from rival products with the same generic designation, some firms (mainly
large, prominent ones) employ trade-name or brand-name designations as well. A trade
name Is an original, trademarked, name assigned by a firm to its own item, such as Lederle's-
use of the trade name Achromycln and Pfizer's use of the trade name Tetracyn for the
tetracycline capsules each produces. Brand names, which combine the product generic
name and the name of the producer or seller, are used less often than trade names. Examples
of brand names are "Cortisone: McKesson and Robbins" and "Armour Thyroid."
PAGENO="0341"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1891
that Oongression'al attention has fostered and promoted this central goal of public
policy, the examination of the ethical drug industry has been a worthwhile en-
deavor on its part.
To the economist concerned with the relationships among market structure9
industry behavior, and economic performance, important findings have `been added
by these hearings and reports to his catalogue of knowledge. This knowledge
serves him well, for it strengthens his understanding of the forces contributing
to the desirable `and undesirable aspects of market performance. Conclusions al-
ready suggested by other industry studies have gained additional substantiation;
other tentative conclusions have been amended, qualified, or rejected in the face
of new knowledge gained from these invnstigations. Most important, we have
eome closer to an understanding of the determinants of market performance and
the manifestations of their influence in this industry.
This paper has as its frame of reference `this economic approach. Its goals
are two-fold: first, to evaluate the performance of the ethical drug industry on
the basis of the material presented in the various hearings and other sources,
and, second, to suggest how substantial improvement in this performance can be
achieved through amendment of the patent laws applying to drugs.3 The thesis
presented is that compulsory patent-licensing is essential for the needed improve-
ment in the "market" performance of the industry (where market performance
refers to those aspects of the industry's activities that determine the cost and
profit elements c'overed by the prices consumers pay for finished products). It
will also be argued that such a patent policy will not impair, and may even
improve, the industry's product performance (where product performance refers
to those aspects of the industry's activities that contribute to better health
through the flow of new products to consumers as a result of research and
development expenditures).
I. THE SALIENT STRUCTURAL AND BEHAVIOR FEATURES OF THE INDUSTRY
`The ethical drug industry can trace its lineage back many decades and even
centuries to the fields of chemistry, pharmacy, and medicine. Nonetheless, in its
present form it is a young industry, arising out of our `break in World War I
with Germany, on whose sophisticated and knowledgeable chemicals industry
we were then dependent for our drugs. Building on the foundation afforded by
wartime successes in drug synthesis and manufacture, medical science and drug
therapy began a co-operative effort tha't soon led to the discovery of insulin in
1921 and its commercial manufacture the next year. There followed discoveries
and development of sex hormones; vitamins, first from natural sources `and later
by synthetic processes; barbiturates; germicides; intravenous anesthetics; im-
proved forms of suiphanilamide; the commercial manufac'ture of penicillin; and
then advances in the n-hole area of antibiotics. In the post World War II period,
successes came more and more quickly: steroid hormones; tranquilizers, anti-
depressants, and other mental drugs; oral antidiabetic drugs; polio and measles
vaccines; oral contraceptives; and a host of other new types of drugs.
The development of new products has been the main source of the industry's
growth. The rapid expansion of the indust.ry's output, to a current domestic level
of $3 billion per year at the manufacturers' level, is one measure of the increasing
success that the industry has had in its research and development activities and
the applicability of its discoveries to an, expanding range of illnesses and injuries.
But of equal significance to its rate of growth of output are the characteristics
which the industry has acquired as it has matured.
A. The emphasis on "specialties"
In the late 1940's and early 1950's the ethical drug industry faced serious
problems. The discoveries of penicillin and streptomycin caught the attention
of many firms. Penicillin, a so-called "product of nature," was unpatentable; the
streptomycin patents were held by Rutgers University and freely licensed. As
a result, markets for these products were easy to enter. The government encour-
aged the expansion of facilities, and new production methods greatly increased
yields. The combined result of these factors was a large overcapacity in the pro-
It is imnortant to note that drugs, which, technically, are the active chemical substances
In drug products, are patentable, as are the processes In drug and drug-preparation manu-
facture. The finished products are notpatentable per se, although the brand or tradenames
under which they are sold are copyrighted.
PAGENO="0342"
1892 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
duction of products containing these drugs, drastic price competition in their sale,
and, frequently, sizeable losses.
The situation was chaotic from the producers' point of view. With many firms
manufacturing items of generic equivalence and selling them under generic labels,
the only form which competition could take was in price. The major firms soon
realized that shifting the basis of competition to some form other than price and
reducing the extent of competitive rivalry were the keys to increased profitability.
Thus each of the large firms turned its attention to the production of "special-
ties," i.e., differentiated and hopefully, exclusive items. The emphasis on special-
ties had two manifestations: the increased reliance on obtaining patents covering
new drug discoveries, and a shift to the use of trade-names for the items, both old
and new, in company catalogues. Both of these pursuits have been impressively
successful. Few of the many drugs discovered and developed since 1950 have not
been protected by patents awarded to private firms; and reliance on trade-names
in prescribing drug products has become the rule rather than exception, particu-
larly in the dominant private-prescription market, in which private physicians
prescribe medicines for patients, to be purchased at drug stores.
As a consequence of the shift to specialties, the three dozen largest firms in the
industry, which represent about five percent of the total number of firms but
account for over 95 percent of all research activity and selling outlays, have
come to dominate the industry. These few- firms account for virtually all the sig-
nificant research advances coming from within the industry; have acquired pat-
ents on the large majority of new- drugs; and have successfully induced physicians
to make product choices in trade-name terms. The growth of patents and of trade-
name use have effectively changed the nature of competition in most of the in~
dustry's markets from frequent and substantial price competition to competition
in product development and promotion and advertising. Price competition is
quite a rare phenomenon in all but a few limited instances. Presently, only in-
stitutional purchasers-mainly hospitals and government agencies-employ ge-
neric designations to any substantial extent. "Specialties" have accomplished
their primary purpose.
B. The industry's research effoft
The high rates of new-product introduction and rapid obsolescence that charac-
terize most drug markets give testimony to the size and effectiveness of the
industry's research effort. Even in those cases u-here discoveries were made in
universities (e.g., Salk vaccine) or government-sponsored research (e.g., penicil-
lin), it usually has been the developmental work done by drug firms which made
the product commercially available. This is not surprising. The drug industry is
the most research-conscious of all non-defense industries, with more company-
financed research and development in relation to sales than any other industry.
In 1964, the all-industry totals for research and development expenditures as
a percent of sales was 4.4 percent; for drugs and medicines, 4.7 percent. For
company-financed research and development, moreover, the all-industry total
was 1.9 percent, as compared to 4.5 percent for drugs and medicine.4 For the pe-
riod 1956 to 1964, the average annual increase in research and development cx-
penditures for the drugs and medicines industry w-as 13 percent, compared to an
economy-wide increase of slightly less than 10 percent in total industry expendi-
tures and between six and seven percent in total company-financed expenditures.5
Furthermore, while it is true that most research and development expendi-
tures in the drug industry are for applied research and product development, the
drug industry devotes a greater proportion of its research budget to basic re-
search than does the economy as a whole or the industrial sector.
For the drug industry, then, research and development expenditures have
shown marked increases in recent years, compare quite favorably with expendi-
tures in other industries in relation to sales. and are devoted to basic research.
not just product development. The research and development record of the drug
industry is commendable.
National Science Foundation. Basic R~search. Applied Resenreh, and Development in
Industry, 1964, WashIngton, D.C.: U.S. Government Printing Office, 1966, p. 62. There
are several industries that have greater research and development expenditures relative
to sales than the drug industry. These industries, however. are in scientific. military,
and engineering fields and receive the bulk of their funds from the rovirnm~nt.
~ Ibid., p. 62.
PAGENO="0343"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1893
C. Product introductwn
One prominent result of the extensive research effort of the industry is the
market impact of the large volume of new items introduced annually. In any
given year, most drug sales are likely to be accounted for by items introduced in
the preceding five or six years. However, many products have a longer commer-
cial life, and any decline in their share of total sales may be due as much to the
acceleration of total sales through the introduction of new products as it is to
their actual displacement by newer items. Nonetheless, market success is not an
inevitable result of research and discovery or of new product development, and no
more than a short life is assured to those products that are well received
initially.
D. Promotion. and advertising
Just as a large part of the industry's research effort reflects its efforts to
develop patentable specialties, much of its outlay for promotion and advertising
reflects its efforts to establish trade-names as a basis for prescription writing.
In the ease of research, no doubt the conquering of illness and legitimate profit
considerations are complementary factors. In the case of promotion and adver-
tising, the important need for conveying technical information to physicians
so that they may employ drug products with maximum effectiveness is a reason
which complements profit considerations.
A number of factors in medical practice have rendered physicians increasingly
dependent on producers as their major: source of information. Among these
factors are the proliferation of products containing new and older drugs; increas-
ing specialization; increasing demands on doctors' time; the time-lag in the
publication of journal articles; and the fact that most of the pre-introduction
testing of new products and the continued testing of older products is done by
or under the supervision of the drug firms themselves. All of these factors have
contributed to the growth of the promotional element of selling costs.
The responsiveness of physicians to the use of trade-names is both result and
cause of a large-scale advertising effort.: As a consequence, advertising outlays
have grown at a pace at least equal to promotion. Expenditures for advertising
and promotion now account for roughly 25 percent of the sales dollar and
one-third of total costs of production of large firms. Selling outlay ranks second
only to the rather inclusive category "cost-of-goods-sold" as a cost component
and is about four times greater than the research effort of which the industry
is so proud. The advertising element alone is approximately twice that of
research and development for the typical large firm. Few other industries compare
closely with this in advertising effort relative to sales.
E. Market concentration
Of the 700 firms in the industry, the twenty largest account for more than 90
percent of total sales, and another dozen or 15 account for half of the remainder.
The high costs of research and developiñent and of effective promotion of trade-
names have set the large firms off frOm their many smaller rivals. The new
products come almost entirely from these few large firms; in the markets for
older products, the popularization of trade names has rendered the firms selling
generically-designated items ineffective as competitors. Only in the production
of products containing freely-available bulk drug ingredients (which are dimin-
ishing in relative economic importance) and in sales to those institutional
buyers who purchase by generic designation can the many small firms participate.
Iii all other market situations, the large firms dominate the picture.
This uneven division of shares of total industry sales between large and small
firms is but one dimension, of concentration. Concentration is an even more mean-
ingful concept in individual product areas than it is for the industry as a whole,.
for it is this aspect of structure that conditions price policies and the nature and
extent of market competition. Among the large firms there has emerged a pattern
of specialization that tends to break them into smaller, rather exclusive groups,
each group sharing a product area such as antibiotics, or steroid hormones with
little fear of entry even by other large firms.
° Promotion and advertising are interrelated activities, but separable in concept, func-
tion, nnd perhaps also In magnitude. Promotion, in essence, Is the conveying of technical
information about drug products that makes possible their use in therapy. Advertising is
mainly directed at establishing and reminding physicians of the trade-names of company
specialties. Estimates within the industry indicate that total selling outlays are divided
roughly evenly between the two categories.
PAGENO="0344"
1894 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Thus, as patented specialties occupied a growing prominence in drug markets,
and as the high costs and cumulative nature of research effort induced specializa-
tion by drug firms, drug markets increasingly became oligopolized by the few
large firms specializing in each of the different areas of research and production.
While the exact order of firms in a market may change, the positions of leader-
ship are effectively preserved for the large firms specializing in that area. Con-
centration thus tends to be both high in degree and stability. Even in the case of
those older (and a small number of newer) products not protected by exclusive
patents, large firms, with their trade-named specialties and their successful pro-
motion campaigns, have come to dominate sales.
As an example, in antibiotics-the largest of the various ethical drug classes-
there are but ten prominent sellers and even they are unevenly participative
within the area. Only six of them are important manufacturers of broad-spectrum
antibiotics, and only five (including three of the broad-spectrum oligopolists)
engage in the manufacture of the leading medium-spectrum products. There are
many firmS producing products w-hich contain the older, unpatented penicillins,
but an estimated 70 percent of all such sales are accounted for by Lilly, Wyeth,
Abbott, and Squibb, four of the ten major firms. Comparable concentration
is to be found for other aiitibiotics, including streptomycin and dihydrostrepto-
myciri, penicillin-dihydrostreptomycin combinations, and penicillin-sulfa combi-
nations. A similar pattern of high concentration can be found in the manufacture
of hormone-drug products, mainly the cortical steroids, where seven firms' domi-
nate the market.
The same high degree of concentration is to be found in virtually every other
product area in the `industry. In the most recent study of market concentration
in the drug industry, the four largest firms in each of 13 major product areas
accounted for between 60 and 80 percent of sales, sometimes substantially more
and rarely less.
The main characteristics of the ethical drug industry-its emphasis on spe-
cialties, its large research and selling effort, the growing patent protection
afforded its developments-have changed the structure and nature of rivalry
very much since the penicillin era. But high concentration and a changing nature
of competition are one thing, and undesirable market performance and the need
for corrective policy changes are another. Judgments based only on industry
structure and the over-all nature of rivalry may be unwise; thus, it is to an
examination of the various dimensions of the performance that is conditioned
by these factors that we must now turn.
II. INDUSTRY PERFORMANCE
There are many dimensions and facets to the performance of industries and
the markets in w-hich they function. Perhaps it is best to consider those of the
ethical drug industry and its markets as being of two main types. These group-
ings, as noted earlier, may be called product and market performance criteria,
respectively.
A. Prodvct performance
On the product side, the main concerns are the efforts to develop new and
better drugs, the increasing availability of these drugs, and their impact on
life and health.
The large outlays by the industry for research and development have already
been discussed. In both absolute and relative terms the industry is justly proud
of its emphasis on basic and applied research. While certain qualifications con-
cerning both the magnitude and quality of this effort may be in order, on the
whole the industry's research input is impressive.
The results of this activity have been substantial. In the sixteen year period
1948-03, the total new products introduced amounted to 5,386.~ Most of these new
products were duplicate items of products already on the market, new dosage
forms of previously known drugs, or new compounds. 618, however, contained
chemical entities not previously known (as did an indeterminate number of the
new compounds) .~ All of these, including the duplicate single products, may
~ Study of Drug Purchase Problems and Policies, U.S. Department of Health, Education,
and Welfare; Washington, D.C., U.S. Government Printing Office, 1966, p. 11.
8 Ibid.
PAGENO="0345"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1895
represent contributions to improved therapy. All increase the range of choice
for physicians; many represent new alternatives and some represent break-
through discoveries into areas where no effective products had been available
previously.
The benefits of the new drugs cannot easily be measured, since most drugs
provide symptomatic relief rather than cures. While the reduced hospitalization
and shortened durations of incapacity may be quantifiable, within limits, the
great reductions in suffering can only be recognized in general. We do know the
preventive and curative drugs have greatly reduced the incidence of certain
illnesses. Influenza, tuberculosis, pneumonia, and syphilis no longer are the
severe killers they once were. The childhoOd diseases of measles, meningococcal
infections, whooping cough, gastritis, duodenitics, colitis, and enteritis have
been brought under control. Poliomyelitis is no longer the great crippler it had
been. Steroid hormones have greatly reduced the pain and crippling effects of in-
flammatory diseases, particularly arthritis. Tranquilizers and other drugs have
done mu~h to reduce the seriousness and the hopelessness of mental illness.
Indeed success over illness has become an expectation by the public, and this
expectation has been considerably met by the industry. These contributions by
the drug firms in the form of new products warrant the industry's pride in its
accomplishments and its recognition by society. Added to this record is the rapid
expansion in the industry's productive capacity and output, which has made
both new and older products available in increasing quantity.
Yet, while the achievements in research and development effort, new product
introduction, and expansion in output have been impressive, certain criticisms
of both the magnitude and quality of these activities must be noted.
It is apparent that the industry exaggerates its research and development
effort, perhaps to convey an impression of extreme risk, or competitiveness, or
exceptional enterprise, with the purpose of justifying high profits. The industry,
when speaking of its research and development activities to stockholders or to
the public, defines them very broadly, being far more inclusive than the Internal
Revenue Service or National Science Foundation. For example, industry spokes-
men appearing before the Kefauver Committee put the 1959 research and de-
velopment outlay at $198 million; the NSF figure for that year was $154 mil-
lion, or 22 percent lower.9
Not only does the industry over-estimate the quantity of this activity, but
many critics, including a large number within the medical profession, have
questioned its nature and direction. There are serious allegations that much
research activity is not related to product improvement but is imitative in
nature, so (as to generate specialties that are not really needed, or is directed
toward the acquisition of patent protection. Furthermore, it is contended that
the great profit potential awaiting new products induces their introduction before
there is sufficient knowledge of their limitations and dangers.
These allegations have been at least partially substantiated. It has been shown
above that duplicates, new dosage forms, and mixtures represent the large ma-
jority of the products being introduced, and the therapeutic advances they repre-
sent may not be very great. Many of the new single chemicals, it has been claimed,
represent no significant progress in drug therapy. It would seem, then, that only
a few "new products" represent real progress. Further, if research outlays in
certain ways are excessive, in others they may be inadequate. When the Food
and Drug Administration raised the standards for testing of new drugs as a
response to the Thalidomide experience, the number of new drug applications
and of approved new products declined sharply.
Certainly the behavioral patterns and motives that are alleged to lie behind
these criticisms of the industry are compatible with the structural and other
features of the industry noted earlier. These are important criticisms of the
industry's product performance and must be taken seriously.
It thus seems likely that the industry's product performance is not as great
as claimed, nor is it as great as it might be with a re-direction of its research
and development effort. Despite these qualifications, the available data support
the conclusion that the industry has contributed greatly to people's lives and
indirectly to the economy. For these contributions the industry deserves great
credit, whatever the flaws that exist in its structure and operation.
g Research and Development in American Industry, 1962, Washington, D.C.: U.S. Gov-
ernment Printing Office, 1063, p. 9.
PAGENO="0346"
1896 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
B. 1[arket performance
The other side of performance is what we have called market performance.
Though it is related in varying degrees and manner to product performance, such
a distinction is a useful one for this study. Market performance concerns pri-
manly the efficiency with which the industry uses the economic resources avail-
able to it In the development, production, and marketing of its output, and the
relationship between prices charged and costs incurred in making goods avail-
able to consumers. The criteria of good market performance are quite traditional
in economic analysis :-that (1) the costs incurred by sellers be closely related to
activities from which consumers derive benefit, be they expenditures needed to
develop new and better products, direct and indirect cost of manufacture, or
marketing outlays that lead to better product selection or expansions in demand
that justify larger-scale, but lower-cost methods of production; and that (2) the
prices consumers pay bear a reasonable relation to these costs; that is, prices
provide profit margins that are sufficient to reward enterprising firms for ac-
tivities that serve consumers, but not so large as to represent gains from behavior
not in the consumers' or society's interest.
These criteria are now used to evaluate the drug industry's performance. First,
in regard to the cost elements it has been indicated above that research and de-
velopment and selling outlays are large, although quite unequal. Further, criti-
cism that the research outlay is, at least in part, wasteful has been noted. While
the extent of such waste from misdirection of outlay may not be closely determin-
able, there appears to be logical support for the contention that the goal of ob-
taming patentable specialties may conflict with that of achieving maximum thera-
peutic gain from research effort, particularly since the commercial rewards for
`new" products are sizeable. Protection of an oligopolistic position by imposing
patent barriers around existing key products and processes also seems to influence
the direction of research effort to some extent. Again, the question as to fre-
quency and extent of wasteful research effort is incapable of precise answer; yet
the large financial rewards awaiting new products, coupled w-ith the large pro-
motional outlays preceding, accompanying, and following their appearance, means
that significant therapeutic progress need not be a necessary condition for large
profitability.
Second, as noted above, selling expenses in the industry are about four times
as large a cost factor as is R & D. The promotion part of selling and the advertis-
ing element are roughly eq)ial and each is twice the R & P component. Promotion
and advertising, both of which may be important marketing functions, nonethe-
less include very costly outlays for activities that may provide the consumer
no benefit, directly or indirectly, or may actually do him a disservice. There is
considerable opinion that promotional excesses are closely related to the appear-
ance of new products that are of limited merit but which might become profitable
through large-scale promotions.
The costs of promotion are, perhaps to a large extent, necessary under present
institutional arrangements. Physicians have, for reasons already noted, be-
come dependent on drug firms for their information on new products. Thus much
promotional effort, parti~ularly that providing information on new and old prod-
ucts. on the incidence and nature of side effects, and on other technical matters,
is necessary. But advertising, whose function is to popularize and remind physi-
cians of trade-named specialties, serves merely to raise the costs that ultimately
become part of prices while not improving the drug selection process. Also, the
quality of information conveyed has been questioned, largely by the physicians
at whom it is directed. Incomplete information, excessive claims, and non-report-
ing of side-effects are frequent among their criticisms. Thus, advertising may
even be detrimental to sound drug therapy.
That there is excessive, wasteful, and mis-directed expenditure for research
and promotion in the industry, and that such waste is of sizeable proportions has
been reported numerous times in the hearings involving the industry. Prices con-
ditioned in large part by these costs are, therefore, necessarily higher than they
otherwise would be.
Drug prices are uneconomically high for another reason. They provide drug
firms with profits that are consistently well above those of most other industries.
There is little justification of the profit levels that prevail. With rapid turn-
~over of products and successful research results uncertain, the industry has been
described as one of high risk, in which fortunate firms can be expected to earn
PAGENO="0347"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1897
substantial returns. Yet the evidence that high risk exists would require low
profits for the unfortunate firms, and suchevidence is not apparent.
Table 1 shows data for profits in the drug or pharmaceutical `° com-
pared to broader aggregative groupings for the past decade. The data show quite
clearly the contithiing and substantial above-average profitability of drug manu-
facture and sale.
TABLE 1.-ANNUAL RATES OF RETURN AFTER TAXES ON STOCKHOLDERS' EQUITY, DRUG INDUSTRY AND ALL MAN-
UFACTURING CORPORATIONS (EXCEPT NEWSPAPERS), 1957-61; PHARMACEUTICALS AND ALL INDUSTRY,
1962-66
(In percenti
~:
~
Drug industry
All manufacturing
corporations (except
newspapers)
1957
18.6
10.9
1958
17.7
8.6
1959
`17. 8
10.4
1960
16.8
9.2
1961
16.7
8.8
Pharmaceuticals
All industry
1962
14.4
8.9
1963
1964
1965
14.7
16.3
18.0
9.1
10.5
11.8
1966
20.3
(1)
1 Not available.
Sources: 1957-61 Federal Trade Commission-Securities Exchange Commission Quarterly Financial Report; 1962-65
`Fortune Annual Directory of 500 Largest Industrial Corporations; 1966, Federal Trade Commission, cf. New York Times,
June 4. 1967, p. 14F.
Within the drug industry, company profits for the recent past snot only nre high
on the whole, but fairly stable. For example, in the decade 1952-61, for 18 of the
`largest drug firms for which data have `been obtained, the average after-tax rate
of return to net worth was 18.5 percent: Only four firms earned below 10 percent,
a reasonable estimate of the all-industry average, as based on Table 1, at any
tiiaie in this period. Of these four, Mead-Johnson fared below 10 percent (9.3 per-
cent) only in 1952; Merck had annual profits of 9.6, 8.9, and 9.6 percent in the first
three years and averaged 15 percent in ,the last seven; Bristol-Myers had annual
profits of 7.1, 6.8, and 9.3 percent in each of the first three years and averaged
16 percent in the last seven. Olin-Mathieson is the fourth. It entered the industry
`with its acquisition of Squibb in 1955, and, in the seven years within this period
in which it participated in drug manufacture, had profits below 10 percent in four.
But Olin-Mathieson is a large conglomerate firm, and its drug operations are a
small part of its total activities. Merck,'Mead-Johnson, and Bristol-Myers are also
conglomerates, though in lesser degree, owing to their sizeable activities in bulk
chemicals, dietary products, and proprietary drugs, respectively, and in each
case the drug-operations of these conglomerates `have been substantially more
profitable than other operations.11
These and other data show that profits in the drug industry as a whole, and
for the large majority of its leading firms, rank substantially and consistently
above those of industry in general. Examples of unstable or occasionally `below-
average company profits are difficult,' to find. Thus evidence in support of the
existence of a high risk element facing the large firm is lacking.
Another defense of the industry's high profits is that these profits are needed to
finance the industry's large research effort. But, as we have seen, the research
in The data for the "drug industry" and "pharmaceuticals" Include both ethical
and proprietary (non-prescription) products. Since ethicals account for 70 percent or
more of total pharmaceutical sales, however, the Inclusion of proprietaries does not
seriously distort the accuracy of the data.
15 Subcommittee on Antitrust and Monopoly, Senate Committee on the Judiciary, Admin-
istered Prices, Drugs, Report, Washington, D.C.: U.S. Government Printing Office, 1961,
p. tLI.
PAGENO="0348"
1898 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
effort is not as large as is often suggested. Nor is the argument a logically valid
one. Profits represent the difference between revenues and full costs, including
research. To add to costs an additional element for future research (which, of
course, will also be paid for by future sales) is to charge consumers twice for
the research component.
Another point suggested in regard to profits is that the profit record of the
industry may reflect the continuing success in new product development by all
major firms. But many of these new products, including a few that have had
the greatest impact on profits, are produced under license and do not reflect
the firm's own research. American Home Products and Smith Kline and French
consistently have been among the most profitable firms not only in the drug
industry but in all industry as a whole. For each, one or two ethical items
dominates the product mix. American Home Products' (Wyeth) leading ethical
specialty is Equanil, manufactured and sold on license from Carter. SKF's
leading specialties are Thorazine and Compazine, made and sold on license from
Rhone-Poulenc of France. Most major firms sell a number of specialty items
containing bulk drugs purchased from patent holders or manufactured on license
from them. Furthermore, superior profitability is not limited to new product
success. A number of firms, e.g., Abbott, Parke, Davis, and Richardson-Merrell,
have contributed few significant developments in recent years, yet their profit
rates have consistently been well above that of most firms in the economy. It is
noteworthy that Parke, Davis' profit rate was substantially above the all-
industry average in those years when its one major development and leading
sales item, Chioromycetin, was temporarily but almost virtually off the market
because of the appearance of drastic side effects.
Certainly an alternative thesis that is equally plausible is that high profits
are due to meager competition in most of the industry's markets, including the
entire private-prescription market and most product areas of the institutional
market. The w-eakiiess of competition is largely attributable to the fewness of
competitors which results from the patents on most new discoveries, from limited
licensing, and from financial and technical obstacles to entry. Patents and limited
licensing are, of course, in accord with public policy. And many of the criticisms
regarding wasteful cost elements, excessive and inaccurate promotion, and
unjustified profitability have been directed toward many industries. But the
operation of the ethical drug markets warrants special concern and special
public policy measures for a number of reasons.
First, there is the fact that health is at stake. New products which confuse
drug selection, and "information" and claims that mislead, may impair drug
therapy and harm the public.
Second, the peculiarly exploitable nature of the patient-consumer, which is
conditioned by his need for drug products and the fact that physicians rather
than consumers make product selection, renders the reasonableness of drug
prices particularly important. The greatest harm befalls those who are unable,
because of prices higher than necessary, to afford all the medication they need.
III. PUBLIC POLICY
The essential question is how to reduce the wasteful outlays by firms and the
excessive profit margins and pass these reductions to consumers without impair-
ing new product development. The answer lies in inducing those changes in the
structure and operation of drug markets that will generate price competition,
while retaining the availability of rewards sufficient to promote the discovery
and development of new and better products. A number of co-ordinate policy
changes are necessary.
First, free entry into finished-product markets must be encouraged for en-
trants who can become effective competitors of the established oligopolists.
A substantial increase in sellers is likely to lead to more vigorous competition,
particularly in price, and the elimination of wasteful and uneconomic prac-
tices. For entrants to become effective competitors, the barriers posed by patents
on drug ingredients must be lowered and the use of trade-names in prescription
writing must be prohibited. Stricter quality controls by the FDA and its require-
ment of more prominent mention of generic names in labelling and advertising
have perhaps already laid the groundwork for an increased use of generic ter-
minology in prescription-writing; recent publicity regarding the large differen-
tials usually found between generically labelled items and trade-named special-
PAGENO="0349"
COMPETITIVE PROBLEMS IN, THE DRUG INDUSTRY 1899
ties hopefully will spur further efforts in that direction. Better and more
accessible information from professional rather than industry sources may
reduce the impact of advertising on physicians, thus reducing the magnitude of
its use. Higher FDA standards for new drugs will reduce the appearance of new
items of little need, thus eliminating mis-directed research and development
expenditures and the promotional outiays and abuses that sometimes accom-
pany them. These and other changes need to be effected for an improved industry
performance. The net result of these changes can be the selection of items by
physicians on the basis of price rather than company identification.
The pertinent question remaining is how to reduce the patent barriers to
entry without dangerously weakening the incentives for research and develop-
ment. The economic philosophy behind Our patent system is that monopoly
grants of a temporary nature and of limited scope, while perhaps creating
imperfections that temporarily mitigate price competition, serve as induce-
ments for rivalry in research and development which provide long run benefits
to consumers that more than offset the temporary mitigation of market price
competition.
Antitrust decisions have defined the limits of patent protection by con-
demning tying restrictions and other devices which owners of patents valid in
one market have used to reduce competition in other markets. Flagrant abuses
of patent rights, such as practiced by the United Shoe Machinery Company,
have led to imposition of compulsory licensing on the offending firm. While
compulsory licensing has been imposed to date only as a punitive measure on
flagrant violators of the antitrust laws, the special importance of the drug
industry to society's well-being and the critical flaws in its structure and
behavior warrant the adoption of special drug-patent policies that include
compulsory licensing as a general condition. This policy is a necessary condi-
tion for improved industry performance in terms of costs, profits, and also in
the critical matter of prices. Together with other policy changes already adopted
or proposed, that induce physicians to prescribe generically, compisory licensing
may also be a sufficient condition for improving substantially the market
performance of the industry.
`It is doubtful that the continuation and advancement of drug research would
be impaired by such modification of drug patents. The industry is characterized
by rapid product turnover and obsolescence; studies have shown that the greatest
portion of sales of any product is likely to occur in the first few years after its
introduction. Company price policies explicitly include these considerations, and
most, if not all companies, estimate quite conservatively the market life of their
products, taking three years as the average period to recoup outlays and earn a
profit. If licensing were required after the first three years of a product's life, the
period taken as the estimated life expectancy, entry in'to the market and the
price competition it would create could contribute to lower prices and profit-
ability after the patent holder has earned a profit justifying his innovation. Of
course, the beneficial results would be confined to those products with a com-
mercial life greater than three years. Although product turnover is swift, data
on product sales indicate that the majority of sales in any recent year represents
those of products that had been on the market longer than three years. For them,
the patent holder would continue earning profits, though at a lower rate, on his
own finished-product sales and those of licensees, and consumers would be able
to buy at lower prices. In regard to the impact of the proposed policy on research
and development activity, it might generate even greater effort by inducing
swifter turnover. In any case, a realistic period of exclusive patent use and a fair
royalty rate afterward seem unlikely to deter research and innovation.
Another view of the role of drug patents raises a question as to the need for
any exclusive use period. The contention that none is necessary is based on the
fact that a drug patent gives its owner a monopolistic position in either of two
markets, that of bulk sales or that of dosage-form products. To the extent that
he sells a drug in bulk or licenses is manufacture, he shares the market for
finished products; if he retains his monopoly in the latter, he cannot reap profits
from bulk sales or licensing. Compelling bulk sales or licensing would require
the patent holder to rely on the lulk market for his innovational profits, and
would promote price-competition in the finished product market at the time when
the item is introduced, `rather than after a few years as in the above scheme. In
this plan, the royalty rate would have to be higher than in the previous' scheme,
PAGENO="0350"
1900 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
but not by so much as to prevent licensing from being an effective check against
exorbitantly high bulk prices. Patentees would earn sufficient profits by charging
high bulk prices or imposing the most favorable royalty rates permitted. But the
cost of bulk ingredients is only a small fraction in total cost of production, and
high bulk costs or royalties would be more than offset by the economies in
resource use and by the smaller profit and selling elements in drug prices that
price competition would bring about.
Either of these proposals would lead to lower prices for consumers while main-
taining the profit incentive for drug discovery and development. Of `the two,
perhaps the first is more feasible politically, but that is another consideration,.
for examination at another time.
Senator NELSON. I want to thank you very much for a most thought-
fill and valuable contribution to these hearings.
We appreciate this very much.
Do you have any questions, Mr. Gordon?
Mr. GORDON. I have just one.
The drug industry very frequently states that high profits are neces-
sary to finance its large research expenditures.
What do you think of that argument?
Dr. SCIIIFRIN. Well, I think it is entirely fallacious, deliberately so.
Profits, of course, are the residual between revenues and costs. Thus
you have profits only when all your costs are met. Research is part of
your cost. Thus, the high profit exists after the research outlay has
already been accounted for.
A tw~ist of this is to say tht high profits are necessary to finance
future research. But, of course, in future prices, there is an element
that covers the research going on. To justify a high price because it is
necessary for future research is in fact to charge the consumer twice
for the research. He is paying for the present research and future re-
search. The future consumer will also pay for present and future
research, and on and on.
It is an argument that does not justify the consumer paying double
in the cost.
Mr. GROSSMAN. We talked about proliferation of products before.
You stated-that at least that brought in competition. But I think
the point you discused with Senator Nelson before was very important
and I wonder if your solutions really cover it. That is the failure of
the small firms really to be able to compete due to the promotional
problem.
Dr. SCHIFRIN. Sir, do you want me to comment on how my pro-
posals-
Mr. GRoss~rAN. Yes; can we ever meet this?
Dr. SCHIFRIN. Yes. You see what keeps the small firm from being
effective in the market now are two barriers. The patent may keep
him out of the market. If there is no patent barrier, the emphasis on
trade names keeps him out because he can't promote his trade name.
As I say, compulsory licensing is necessary, but not sufficient. They
cannot be an effective competitor as long as prescribing is done by
trade names. Thus I think trade names have no basis for existence.
If brand names are used, they will clarify the matter as to generic
equivalency. Once that is established, the small firm with a generic
product will have a better chance of getting its items selected, cer-
tainly once its low prices become recognized.
PAGENO="0351"
COMPETITIVE PROBLEMS IN IHE DRUG INDUSTRY 1901
Mr. GROSSMAN. Do you think it will also help if the larger firms
will cut down on its advertising?
Dr. SCHIFRIN. Definitely, because you will not be advertising a
trade name trying to make it seem something different from other'
products with generic equivalency.
Mr. GROSSMAN. So a lot of wasted effort is cut down?
Dr. SCrnFRIN. Right.
If you could eliminate the waste, it would reflect on the consumers
in two ways: They would either pay less for what they get or get
more for what they pay.
Senator NELSON. I thank you very much.
We will take a 5-minute recess and then hear Dr. Steele.
(Recess.)
Senator NELSON. Our next witness is Dr. Henry Steele, professor
of economicS at the University of Houston, Houston, Tex., and an
economist with very distinguished credentials.
We appreciate very much your coming to testify today, Dr. Steele..
You have presented to the committee a. very detailed and well-prepared
analysis of testimony given by the eëonomists represent.ing the Phar-
maceutical Manufacturers Association before this committee as well
as additional material. You may present this material in anyway you
see fit. All of your statement, as well as your crit.ique of the presenta-
tions of the economists for the PMA will be printed in full in the
record.
You may present your material in any way you see fit.
STATEMENT OF HENRY B. STEELE, PH. B., ASSOCIATE PROFESSOR
OF ECONOMICS, UNIVERSITY OF HOUSTON, HOUSTON, TEX.
T)r. STEELE. Thank you, Senator Nelson.
I would like to read the first half dozen pages or so from my major
statement and then turn to the supplemental statement commenting
on the PMA presentations, if I may.
I greatly appreciate the privilege of being invited to make this state-
mont before this subcommittee, and it is my hope that t.he information
which I am able to present will be of some use t.o you in your delibera-
tions regarding the vitally important economic problems arising out
of the market context within which the drug industry operates.
I am an academic economist with major research interests in indus-
trial organization and the regulation of industry, and have done much
work in the area of medical economics a.nd drug industry regulation.
I received my Ph. D. degree in industrial economics from MIT in
1957, and since then have been engaged in teaching and research, as
well as in consulting for private firms, U.S. Government agencies, and
foreign governments. At present I am an associate professor of eco-
nomic at the University of Houston. My research in the drug industry
has continued over the last 7 years, and I have written three articles
on drug industry economics and regulation and coauthored two arti-
cles on the supply and distribution of physicians' services, all of which
have appeared in professional economics journals.
In March 1965, I presented a liaper on drug industry regulation
before the University of Illinois Medical School at the invitation of
PAGENO="0352"
1902 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. Harry Dowling, the chairman of the ck~uncil of Drugs of the
American Medical Association; and in February 1967, I presented a
comprehensive program for the reform of Canadian drug laws and
regulations before the Special Committee on Drug Costs and Prices of
the Canadian House of Commons.
Senator NELSON. May I interrupt just a moment?
Do you have with you the three articles on drug industry economics
and regulation and the two articles on the supply and distribution of
physicians' services?
Dr. S~riiiIAE. I have the three articles on drug industry regulation
with me.
Senator NELSON. Would you submit them for the record?
Dr. STEELE. Yes; and I can submit the other two articles as well.
Senator NELSON. Would you also submit the other two articles, and
they will be printed at the conclusion of your testimony.1
Dr. S~u~I~E. Thaiik you. I continue.
In making the Canadian presentation, I represented the government
of the Province of Alberta, and of the 14 recommendations which I
submitted, 11 were incorporated in the final report of the committee
to the Canadian House of Commons. All of my research on the drug
industry in the United States has been entirely self-financed, and in
presenting this statement to the Subcommittee on Monopoly I wish
to make it clear that I represent no one but myself.
I. INTRODUCTION
What can an economist contribute to hearings on drug industry
problems?
It is curious that in all the hearings held in the United States, as
well as those in England and in Canada, the original demand for the
hearings has come about because of the conviction that prices are "too
high," but very much of the hearings have been occupied by investi-
gations into the safety and efficacy of drugs, and medical and phar-
macological considerations have quite generally been predominant
over economic issues.
Yet, for every person wtho is moved to voice a complaint over poor
drug quality, there must be a hundred who complain about high prices.
But the emphasis on drug safety and efficacy inevitably redounds to
the advantage of the major firms, who would much rather fight on
the battleground of relative quality than of relative prices. Their argu-
ment `has two parts:
First, drug prices are related to costs, particularly quality control
costs. Second, drug quality and hence safety and efficacy is related to
these same costs. Hence it is asserted that one caunot divorce questions
of cost and price from questions of safety and efficacy; therefore, since
problems of drug quality, being matters of life or death, are obviously
more crucial than drug price prdblems of mere dollars and cents, steps
to assure high quality should take precedence over economic reforms.
Economic reforms are correspondingly delayed. (This was the net
effect of the otherwise admirable Kefauver-Harris Act `of 1962.)
1 The five articles submitted by Dr. Steele begin at p. 1950, Infra.
PAGENO="0353"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1903
This argument is faulty in several respects:
First, drug prices are not related to drug costs, but instead to de-
mand and ability to pay.
Second, while drug quality obviously depends upon care exercised
in manufanture, the cost of quality control has been shown to be a very
small part of the total costs, and the difference in cost between a mini-
mal satisfactory program and a deluxe program would not begin to
account for the difference in prices `between the generic drug and its
brand name equivalent.
But to `return to the question: What can an economist contribute to
drug law reform hearings?
If the data were made available, he could analyze cost-price condi-
tions within the individual drug firms, and the pattern of mterfirm
price and product competition, and arrive at an informed judgment
regarding the status of competition in the industry. But such data
have not been made available, even to economists retained to defend
the industry.'
In view of the absence of the data in the analysis of which the econ-
omist has a comparative advantage, what constructive role can he
play? Primarily that of coordinating and synthesizing the economic
aspects of the data which is in the record, and evaluating the economic
relevance or credibility of certain of the arguments advanced by the
drug interests.
It is noteworthy that drug spokesmen produce arguments in their
defense which either stress or ignore similarities or differences between
drugs and other industries to suit their convenience. Thus, in the PMA
studies presented last month, one study treated the drug industry just
like any other industry in relating the variance of the earnings of
member firms in an industry (rather arbitrarily called "riskiness") to
the average rate of earnings in that industry, while the other anaJyzed
product competition in drugs in a vacuum as it were, without introduc-
ing comparative data from any other industry. But both the similarities
and the differences of the drug industry should be analyzed and allowed
for before making any comparative study of drug prices, costs, and
profits in relation to those of other industries.
1 As Professor Markham stated before this Subcommittee on December 19, 1967, in
response to just such a question, "you are just not going to get those data, and I do
not think-I would be less than honest if I said I would try to get them, implying that I
could get them for you." (Transcript, volume 23, p. 2805.) Markham apparently referred
not only to the confidential status given the information, but also questioned whether or
not drug firms bothered to make all the cost allocations Involved. Although it Is to be
admitted that many of the calculations can be made only on the basis of arbitrary assump-
tions, one would expect that well-managed firms would find it prudent to undertake such
analyses for their own information. In fact, Dr. M. A. Phillips, in his Sainsbury Committee
memorandum to the British Ministry of Health stated that the drug Industry was no
different from other organic chemicals industries in observing the customary precautions
of making detailed cost studies prior to engaging in producing projects. These studies
include the costs of research and development and of promotion. Dr. Phillips' statement Is
unusually authoritative in that he is a drug industry conslutant who has made many
economic evaluation studies for drug firms. Phillips complains that "It has been found
very difficult to obtain figures for the cost of research and development and of promotion
and advertising, although this must be known to those who have to spend this money
in these ways . . ." and explains that even with the approximations his organization has
to use in estimating these costs, he is satisfied that the accuracy of the estimates for
these items Is within 25 percent. See Competitive Problems in the Drug Industry, Part I,
pp. 54-55, of the Hearings before this Subcommittee on the present matter. (It might be
observed that only if there is a very large gap between cash flow and expenditures is a
company actually likely to indulge in som4 carelessness or negligence in the relating
of total costs to individual items sold.)
81-280--68-pt. 5-23
PAGENO="0354"
1904 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
The major similarity between drugs and other industries is that the
firms are privately owned and are managed in the interests of maximiz-
ing profits and the value of the shareholder's investment.
Provided that there is effective competition among firms, the profit-
maximization goal is an excellent coordinating mechanism since it
motivates managements to produce a. given output at minimum total
cost and hence stimulates efficiency.
At the same time, price competition among firms will keep price
levels from exceeding for any great period of time the equilibrium
levels necessary to elicit from producers that level of production which
consumers demand at that price. This is basic to the classical economic
doctrine of Adam Smith (who was mentioned more than once with
approval by PMA witnesses last month) that the force of price corn-
petition taking effect through the market, acts as the famous "invisible
hand" which by allocating resources efficiently makes each economic
agent serve the general welfare even though he is only interested in
furthering his own private fortunes.
But unless effective competition prevails, private and public welfare
in the market are not consonant. In the drug industry, the invisible
hand is invisible chiefly because it is so deeply buried in the consumer's
pocket. And it is the extreme vulnerability of the drug buyer to eco-
nomic exploitation which makes the drug industry (both in economic
and public policy terms) a unique market which cannot be compared
directly with any other. This peculiar vulnerability of the drug buyer
to exploitation is related to several major characteristics of the drug
market which prevent price competition from acting as a safeguard
and which also tends to make an economically unregulated drug in-
dustry productive of much misallocation of resources in its attempt to
maximize the profits of the individual firms.
I do not intend to advocate punitive regulation of the drug indus-
try. But the industry at present enjoys the benefits of what amounts
to public regulation in its favor, through the availability of the patent
privilege, trademark and copyright protection, and the laws support-
ing prescribing by brand name, to name only the major advantages.
This stacks the cards heavily in favor of the industry and against the
drug buyer. I am in favor of corrective legislation to redress the bal-
ance and increase the chances of the patient's getting fair value for his
prescription dollar.
At the same time, it must be stressed that to be critical of the drug in-
dustry is by no means to be critical of private enterprise as such. Most
industries are routinely accorded exemption from special economic
regulation because they naturally tend to function tolerably competi-
tively in a free market environment.
But if the drug industry is permitted to retain its present special
position of institutionalized protection in the economy, it will continue
to display elements of both monopoly and rivalry. Spokesmen for the
industry habitually refer to the intense degree of competition among
firms. Unfortunately, however. the "competition" referred to is of the
type which raises costs instead of reducing prices. This category of
activity is generally referred to by economists as "rivalry" rather than
"competition" since the latter term is usually reserved for the eco-
nomically beneficial activity specifically of price competition. Ob-
PAGENO="0355"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1905
viously, the industry which is subject to intense price cornpeti~ion at
all times is the exception rather than the rule in today's economy.
Nevertheless, the perennial threat, and occasional outbreak, of price
competition does much to keep the price policies of the typical industry
on the side of moderation. My research and consulting experience in
the field of industrial organization during the last 10 or 12 years has
been such as to convince me that the great majority of product inarket~
in the United States are more or less workably competitive, but the
specific legal and marketing arrangements which the drug industry
enjoys are such as to make it virtually a foreign body in an otherwise
workably competitive economy.
Then if I may I would like to turn to my supplemental statement,
which focuses on the presentations of the PMA witnesses last month.
Senator NELSON. We will print in full the statement that you just
read from in the record at this point and we will then start your sup-
pleinentary statement.
Dr. Smi~. Thank you.
(The complete prepared statement of Dr. Steele follows:)
STATEMENT OF HENRY B. STEELE, Pu. D., ASSOCIATE PROFESSOR OF ECONOMICS,
UNIVERSITY OF HOUSTON
I greatly appreciate the privilege of being invited to make this statement before
this subcommittee, and it is my hope that the information which I am able to
present will be of some use to you in your deliberations regarding the vitally im-
portant economic problems arising out of the market context within which the
drug industry operates.
I am an academic economist with major research interests in industrial
organization and the regulation of industry, and have done much work in the
area of medical economics and drug industry regulation. I received my Ph. D.
degree in industrial economics from MIT in 1957, and since then have been
engaged in teaching and research, as well as in consulting for private firms,
United States government agencies, and foreign governments. At present I
am an associate professor of economics at the University of Houston. My
research in the drug industry has continued over the last seven years, and I
have written three articles on drug industry economics and regulation and
two articles on the supply and distribution of physicians' services, all of which
have appeared in professional economics journals. In March 1965 I presented
a paper on drug industry regulation before the University of Illinois Medical
School at the invitation of Dr.0 Harry Dowling, then chairman of the Council
of Drugs of the American Medical Association, and in February 1967, I
presented a comprehensive program for the reform of Canadian drug laws and
regulations before the Special Committee on Drug Costs and Prices of the
Canadian House of Commons. In making the Canadian presentation. I repre-
sented the government of the Province of Alberta, and of the fourteen recom-
mendations which I submitted, eleven were incorporated in the Final Report
of the Committee to the Canadian House of Commons. All of my research of the
drug industry in the United States has been entirely self-financed, and in pre-
senting this statement to the Subcommittee on Monopoly, I wish to make it
clear that I represent no one but myself. 0
I. INTRODUCTION
What can an economist contribute tO hearings on drug industry problems?
It is curious that in all the hearings held in the United States, as well as
those in England and in Canada, the original demand for the hearings has come
about because of the conviction that prices are "too high", but very much of
the hearings have been occupied by investigations into the safety and efficacy
of drugs, and medical and pharmacological considerations have quite generally
been predominant over economic issues. Yet, for every person who is moved
to voice a complaint over poor drug quality, there must be a hundred who
PAGENO="0356"
t906 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
complain about high prices. But emphasis on drug safety and efficacy inevitably
redounds to the advantage of the major firms, who would much rather fight on
the battleground of relative quality than of relative prices. Their argument
has two parts: First, drug prices are related to costs, particularly quality control
costs. Second, drug quality and hence safety and efficacy is related to these same
costs. Hence it is asserted that one cannot divorce questions of cost and price
from questions of safety and efficacy; therefore, since problems of drug quality,
being matters of life or death, are obviously more crucial than drug price prob-
lems of mere dollars and cents, steps to assure high quality should take prec-
edence over economic reforms. Economic reforms are correspondingly delayed.
(This was the net effect of the otherwise admirable Kefauver-Harris Act of
1962). The argument is faulty in several respects. First, drug prices are not re-
lated to drug costs, but instead to demand and ability to pay. Second, while drug
quality obviously depends upon care exercised in manufacture, the cost of
quality control has been shown to be a very small part of total costs, and the
difference in cost between a minimal satisfactory program and a "de luxe" pro-
gram would not begin to account for the difference in prices between the generic
drug and its brand name equivalent.
But to return to the question: what can an economist contribute to drug law
reform hearings? If the data were made available, he could analyze cost-price
conditions within the individual drug firms, and the pattern of inter-firm price
and product competition, and arrive at an informed judgment regarding the
status of competition in the industry. But such data have not been made avail-
able, even to economists retained to defend the industry.1
In view of the absence of the data in the analysis of which the economist
has a comparative advantage, what constructive role can he play? Primarily
that of coordinating and synthesizing the economic aspects of the data which
is in the record, and evaluating the economic relevance or credibility of certain
of the arguments advanced by the drug interests. But since much of the evi-
dence and many of the arguments transcend the realm of economic analysis as
such, an economist is vulnerable to objections that he is exceeding the limits
of his professional competence.
Certainly the economist is not alone in this. During the drug industry investi-
gations in the English-speaking countries, testifying physicians have been criti-
cized for not being economists, economists have been challenged for not being
physicians or pharmacologists, medical educators have been chided for not being
doctors in full-time private practice, etc. But until the ideal witness appears,2
someone who is less than fully qualified has to stick his neck out and attempt to
put the entire picture together. There are reasons why an economist who special-
izes in the area of industrial organization and regulation is not the least qualified
of all specialists to make such a presumptuous attempt. First and foremost,
1 As Professor Markham stated before this Subcommittee on December 19, 1967, In
response to just such a question, "you are just not going to get those data, and I do not
think-I would be less than honest if I said I would try to get them, implying that I could
get them for you." (transcript, volume 23, p. 2805). Markham apparently referred not only
to the confidential status given the information, but also questioned whether or not drug
firms bothered to make all the cost allocations involved. Although it is to be admitted that
many of the calculations can be made only on the basis of arbitrary assumptions, one would
expect that well-managed firms would find it prudent to undertake such analyses for
their own information. In fact, Dr. M. A. Phillips, in his Sainsbury Committee memorandum
to the British Ministry of Health stated that the drug industry was no different from
other organic chemicals industries in observing the customary precautions of making
detailed cost studies prior to engaging in producing projects. These studies Include the
costs of research and development and of promotion. Dr. Phillips' statement is unusually
authoritative in that he is a drug industry consultant who has made many economic
evaluation studies for drug firms. Phillips complains that "It has been found very dif-
ficult to obtain figures for the cost of research and development and of promotion and
advertising, although this must be known to those who have to spend this money in these
ways . . ." and explains that even with the approximations his organization has to
use in estimating these costs, he is satisfied that the accuracy of the estimates for these
items is within 25 per cent. See Competitive Problems in the Drug Industry, Part I pp.
54-55, of the Hearings before this Subcommittee on the present matter. (It might be
observed that only if there is a very large gap between cash flow and expenditures is a
company actually likely to indulge in some carelessness or negligence in the relating of
total costs to individual items sold.)
2 The ideal witness would probably be a practicing family physician who has also in the
past been a pharmacist, a pharmacologist, a medical educator, a hospital administrator a
drug industry executive, a compiler of official therapeutic compendia, a medical public
welfare program administrator, a high official variously with the AMA, the PMA and the
FDA, a patent attorney, a research scientist, an economist, and a member of Congress.
PAGENO="0357"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1907
firms in the drug industry are economic enterprises. The economist George
Stigler has defined business as a collection of devices for surmounting barriers
to profits, and the splendid profitability records of the drug firms indicate that
their business-oriented managements have very efficiently achieved this goal.
Secondly, the anomalous price structures and price-cost relationships in the
drug industry are so striking as to be obvious to the layman,3 while other evi-
dences of delinquent performance in the, discharge of the drug industry's
responsibilities are not exactly so subtle or equivocable as to be apparent only
to those with highly specialized backgrounds in professional disciplines.4
It is noteworthy that drug spokesmen produce arguments in their defense
which either stress or ignore similarities or differences between drugs and other
industries to suit their convenience. Thus, in the PMA studies presented last
month, one study treated the drug industry just like any other industry in
relating the variance of the earnings of member firms in an industry (rather
arbitrarily called "riskiness") to the average rate of earnings in that industry,
while the other analyzed product competition in drugs in a vacuum as it were,
without introducing comparative data from any other industry. But both the
similarities and the differences of the drug industry should be analyzed and
allowed for before making any comparative study of drug prices, costs, and
profits in relation to those of other industries.
The major similarity between drugs and other industries is that the firms are
privately owned and are managed in the interests of maximizing profits and the
value of the shareholder's investment. Provided that there is effective competi-
tion among firms, the profit-maximization goal is an excellent coordinating
mechanism since it motivates managements to produce a given output at mini-
mum total cost and hence stimulates efficiency. At the same time, price com-
petition among firms will keep price levels from exceeding for any great period
of time the equilibrium levels necessary to elicit from producers that level of
production which consumers demand at that price. This is basic to the classical
economic doctrine of Adam Smith (who was mentioned more than once with
approval by PMA witnesses last month) that the force of price competition taking
effect through the market, acts as the famous "invisible hand" which by allocating
resources efficiently makes each economic agent serve the general welfare even
though he is only interested in furthering his own private fortunes. But unless
effective competition prevails, private and: public welfare in the market are not
consonant. In the drug industry, the invisible hand is invisible chiefly because
it is so deeply buried in the consumer's pocket. And it is the extreme vulner-
ability of the drug buyer to economic exploitation which makes the drug
industry (both in economic and public policy terms) a unique market which
cannot be compared directly with any other. This peculiar vulnerability of the
drug buyer to exploitation is related to several major characteristics of the drug
market which prevent price competition from acting as a safeguard and which
also tends to make an economically unregulated drug industry productive of much
misallocation of resources in its attempt to maximize the profits of individual
firms. I do not intend to advocate punitive regulation of the drug industry. But
the industry at present enjoys the benefits of what amounts to public regulation
in its favor, through the availability of the patent privilege, trademark and
copyright protection, and the laws supporting prescribing by brand name, to
name only the major advantages. This stacks the cards heavily in favor of the
industry and against the drug buyer. I am in favor of corrective legislation to
redress the balance and increase the chances of the patient's getting fair value
for his prescription dollar.
At the same time, it must be stressed that to be critical of the drug industry
is `by no means to be critical of private enterprise as such. Most industries are
routinely accorded exemption from special economic regulation because they
naturally tend to function tolerably comni~etitive1y in a free market environment.
As Senator Nelson observed, In Inquiring of Professor Cootner bow a price differential
of $32.62 versus $2.60 might be justified between the United States and Canada, "I didn't
think I was asking a question that required expertise." (Transcript of these hearings,
rolume 23, p. 2706.)
Drug Industry spokesmen have patronized Morton Mintz, author of The Therapeutic
Wightmare, referring slightingly to his having become an expert on the drug Industry
during the Kefauver hearings. He did not become an expert on the Industry, but upon the
abuses practiced by the industry-as did anyone else who read the entirety of the Kefauver
Hearings carefully.
PAGENO="0358"
1908 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
`But if the drug industry is permitted to retain its present special position of
institutionalized protection in the economy, it will continue to display elements of
`both monopoly and rivalry. Spokesmen for the industry `habitually refer to the
intense degree of competition among firms. Unfortunately, however, the "com-
petition" referred to is of the type which raises costs instead of reducing prices.
This category of activity is generally referred to by economists as "rivalry"
rather than "competition" since the latter term is usually reserved for the eco-
nomically beneficial activity specifically of price competition. Obviously, the
industry which is subject to intense price competition at all times is the exception
rather than the rule in today's economy. Nevertheless, the perennial threat, and
occasional outbreak, of price competition does much to keep the price policies of
the typical industry on the side of moderation. My research and consulting
experience in the field of industrial organization during the last ten or twelve
years has been such as to convince me that the great majority of product markets
in the United States are more or less workably competitive, but the special legal
and marketing arrangements which the drug industry enjoys are such as to make
it virtually a foreign `body in an otherwise workable competitive economy.
iA,s I have elsewhere stated, the market characteristics of the drug industry bias
it in the direction of inefficient and non-competitive performance in- five major
respects.5
(1) Essential to the effective operation of a free market is the ability of the
buyer to choose among suppliers on the `basis of an adequate knowledge of the
price and quality of the alternative products which they may provide him. But
in ethical drugs, the buyer `has no practical means of gaining access to knowledge
of the range of price and quality alternatives in the market; indeed, his purchas-
ing agent, the prescribing physician, is constantly oversupplied with biased infor-
mation and even misinformation which facilitates confusion and ignorance of
prices.
-(2) The price-conscious -buyer should be able to identify the lowest-priced seller
and purchase from him without `artificial impediments. Instead, the possessor of
a newly-written prescription is unable to buy any -but the specified drug, regard-
less of price. The willingness of the price-conscious physician to prescribe lower-
priced -drugs may be compromised if he `h-as been exposed to repeated attempts
to disparage low priced drugs on the part of representatives of brand na-me drugs
who contend that -low price means low quality. And even if a generic prescription
is written, the `buyer -has no power to compel the dispenser to sell him a reason-
ably priced generic drug instead of substituting a less reasonably priced brand
name equivalent.
(3) There must be freedom of entry into the industry by new firms, such that
high profits being made by existing firms will attract new competitors who will,
by engaging in price competition, drive profits down to competitive levels. But
freedom of entry in drugs is greatly lessened by the existence of the patent
privilege, the trademark device, and the necessity for newcomers to match the
enormous advertising outlays of existing rivals.
(4) There should be an adequately large number of competitive sellers offer-
ing buyers genuine alternatives in terms of product price and quality; none of
the sellers should be so large that he overshadows the magnitude of his com-
petitors and poses a potential threat should they incur his displeasure. In drugs,
restricted entry limits the number of sellers, and while there are few if any
genuine product monopolies, the size of the major firms is certainly appreciably
greater than that of their smaller generic-name competitors.
(5) A market is not workably competitive unless all firms act independently-
there must be no overt or tacit collusion, no passive acquiescence in prior deci-
sions arrived at by others and established by mutual consent . . . (in the drug
industry there are) two circumstances which act to hamper independence of
action. First, there is the practice of price leadership and the pricing of new
medications at exactly the same levels charged for existing substitute drugs.
Second, there is the fertile field of patents. While an individual patent confers
a inonpoly, the scope of the monopoly privilege is limited. But in an industry with
complex technology, the efficient production of a drug may require the use of
processes controlled by rival patent-holders. The negotiation of the resulting
Submission of the Government of the Province of Alberta to the Special Committee on
Drug Costs, and Prices of the Canadian House of Commons, Minutes of Proceedings and
Evidence, No. 33, pp. 2427-2429, Feb. 14, 1967.
PAGENO="0359"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 1909
cross-licensing agreements requires the mutual compromise of patent monopoly
positions, and may well stimulate such meetings of the minds as will lead to the
development of a greater sense of community of interest in policies regarding
prices, production, and participation in world markets.
In connection with this last point, it should be emphasized that the United
States drug industry is not unacquainted with outright collusion, in December
1967, Pfizer, Lederle, and Bristol Myers were convicted by the Federal District
Court of New York Oity of conspiracy in restraint of trade, conspiracy to monop-
olize, and monopolization in the tetracycline market. (Squibb and Upjohn were
named co-conspirators but not defendants.)
A further respect in which the drug market can be distinguished from other
markets is in the involuntary nature of the purchase. While an individual ordi-
narily ought to pay no more than the competitively determined full supply price
for a product, he should also pay no less than this price since in order to supply
his wants, the economy has had to allocate scarce resources which could have
been used alternatively in the production of other products. But there is a differ-
euce between paying the full cost of financing an activity deliberately engaged
in, in contrast to one forced by accident or misfortune upon the buyer through
no fault of his own. This is to suggest that the financing of drug purchases, like
other aspects of health care, has an element of insurance against risk in it. Such
insurance arrangements could conceivably be either public or private. A prudent
man of sufficient income might participate in a voluntary health insurance pro-
gram including drug costs. But the required income to make participation attrac-
tive does not depend upon the "full competitive cost" of drug supply, but instead
upon actual prices charged in highly non-competitive markets. Unless drug prices
can be made reasonable, the possible expenses of drug therapy under a comprehen-
sive private health insurance program might be so great that enormous premiums
would be required. Under these circumstances, the expenses of drug therapy
w-ould not constitute an insurable risk for practical purposes. This is all the more
applicable to public health insurance and welfare programs, of the medicare vari-
ety or otherwise. Truly comprehensive drug coverage under such plans might
allow sellers of patented high-price drugs to levy a publicly-underwritten tribute
on the sick and afflicted and divert a not-negligible portion of tax revenues and
the national income into the hands of the pharmaceuticals industry. The only
difference, fortunately, is that public authorities are in a position to exert more
effective efforts to discipline high drug prices than are private insurance corn-
panies.
The above consideraions should be kept in mind when assessing the effects of
the great variety of drug industry activities and expenditures on the price of
drugs, when such costs are generally borne in full by persons involuntarily
afflicted, whose earning power and ability to pay may be greatly reduced by the
very circumstances which make medication imperative.
The absence of workable competition among sellers is compounded by the
barriers which consumers face in obtaining information regarding drug prices and
quality. If neither a seller's customers nor his rivals can force him to compete,
what limits are placed on his ability to, exploit his customers? Essentially only
two-self-restraint and public constraint.
Self-restraint is ordinarily an impediment in the management of a business
enterprise, and under competitive market conditions would detract from efficiency.
During the Kefauver hearings several witnesses referred to their impressions
that there bad been within memory some decrease in the degree of self-restraint
in marketing among drug firms. Actually, it is to be doubted that self-restraint in
itself ever posed much of a barrier to high profits. Upjohn, for example, made
over 30 per cent after taxes on its net worth in each of the deep depression wears
1930-1935.° But this does not necessarily contradict the observations that self-
restraint was still more prevalent among drug firms prior to the second world
war. In the post-war era, however, it became obvious to all that the profit possi-
bilities inherent in the "Miracle Drugs" era of the industry's history were simply
too vast not to be fully and intensively exploited. Although it did not prove pos-
sible for new small firms to enter the market and become genuine factors to be
reckoned with in the industry, larger firms found it possible to diversify by
merging with existing drug houses, and producers of bulk chemicals and fine
chemicals found it profitable to integrate forward into drug making and market-
6 on Administered Prices, Part 20, p. 11082.
PAGENO="0360"
1910 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ing. Some firms which had previously specialized in patent medicines began to
produce prescription or "ethical" drugs. For this reason and others, the change
in the composition of the industry was accompanied by an increased emphasis on
sales promotion. Several physicans and medical educators testified during the
Kefauver hearings that as the quantity of promotional matter increased greatly,
there was a concomitant decrease in its average quality. It was argued by some
that the industry should be saved from itself, it being alleged that if only one
or two indifferently scrupulous firms saturated the physicians with marketing
appeals of dubious merit, all the others had to imitate their tactics in self-defense.
These observers tempered criticism with courtesy by omitting to single out the
firms involved, but I have been amused by a phrase which I keep running across
in my own contacts with people in the medical community who are interested
in drug reform: "Where there's smoke, there's Pfizer."
It seems safe to conclude, in any event, that the intensified rivalry which
transformed drug industry marketing practices in the 1950's in all likelihood
eliminated the factor of industry self-restraint from the category of feasible
solutions to drug industry problems.
Public constraint has so far been somewhat asymmetrical with regard to safe-
guarding the drug buyer. While the acts of 1900, 1938, and 1962 have progressive-
ly improved the quality of drugs in terms of increased assurance of their safety
and efficacy, legislative concern for the economic health of the patient has not
been nearly as marked. This is unfortunate since a sick man's economic health
may often be more delicate than his physical condition. Economic convalescence
can be much more prolonged than physical recuperation. And tw-o or three acute
illnesses in succession, requiring drugs and hospitalization, may put a person of
limited means into the ranks of the chronically medically indigent. Even the
legislation regarding drug quality was obtained only with great difficulty. There
seems to be two reasons for this. First, the lobbying and other governmental
relations activities of the drug industry, made effective not only through the
individual firms and their trade associations, but also through their allies in
organized medicine and some areas of pharmacy, and probably through the
"army" of some 15,000 detailmen. Second, public apathy tempered with the
w-idespread uncritical attitude that the makers of miracle drugs must be miracle
workers who are without fault. Even so, these obstacles to legislation have not
rendered the industry completely reform-proof.
The Elixir Sulfanilamide tragedy of 1937 prompted the passage of the
reforms embodied in the Food, Drug, and Cosmetic Act of 1938. And although
the industry has not been backward in its efforts to cultivate the public
image of the private drug firm laboratory as having a complete monopoly
over the creation of "miracle drugs," it has been more neglectful in publicizing
the fact that, by the same token, miracle drugs can produce "miracle diseases."
Dr. Walter Modell of Cornell University Medical School has observed that
some 40 or more new diseases have been recognized and attributed to the em-
ployment of modern drug therapy.7 But in 1961 the point was made more dra-
n1atically through the creation of a sizeable "miracle generation" under the
auspices of thalidomide. Again, the reaction of the public to another tragic
blunder by the drug industry was to pass a reform law, the drug amendments
act of 1962, to reduce the possibility of the recurrence of such a calamity
and to further safeguard the public by insuring more adequately the safety and
efficacy of drugs. But the danger is that public interest rapidly subsides after
each calamity, while the industry itself is continuously in search of expedients
which may be employed to minimize the impact of regulatory reforrns.s
The net result is that while the drug buyer who obtains a drug of substandard
quality (such that his health is impaired for this reason alone) is probably a
very rare individual, the drug buyer who obtains his purchase for a genuinely
reasonable price is probably a still rarer individual. Reform laws are now needed
to institute price competition in the drug industry. One certainly hopes that such
a law can be passed without the necessity for a fresh tragedy to mobilize public
opinion. Still, it is hard to imagine a sufficiently striking tragedy in purely
economic terms, short of some revelation that a large number of low-income indi-
viduals with no access to welfare case status had died of malnutrition during
Drug Industry Antitrust Act Hearings, Part 1, p. 318.
8 See, for example, Morton Mintz, The Therapeutic Nightmare, pp. 222-220.
PAGENO="0361"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1911
some period of time because purchase of medications at inflated prices had
usurped the entirety of their budgets.°
But one must first inquire into what sort of reforms might most appropriately
be embodied in legislation to erhance the likelihood of price competition and
lower prices in the drug industry. To do this requires a study of supply and de-
mand in the drug industry-of drug costs in relationship to prices. What func-
tions does the industry perform? Of what benefit are these functions? How ef-
ficiently does it perform them? Could any of these functions be more effectively
performed under Other auspices? An attempt is made to answer these and other
questions in the next section.
II. AN ECONOMIC ANALYSIS OF SUPPLY AND DEMAND FOR PRE5CRIPTION DRUG5
No economic analysis of the operations of a privately-owned industry is
properly oriented unless primary attention is given to the operation of the profit
motive in influencing the allocation of economic resources within the market.
On this point, drug firm managements, investors, trade associations, and industry
critics should find themselves in agreement. But while price competition presup-
poses mutually opposed interests on the part of separate competing firms striving
to increase their share of the market, it also requires a pervading fundamental
understanding that the institutons of the market-the price mechanism, for
example, and the contractual arrangements of purchase and sale-shall be ad-
hered to in a cooperative manner, such that the "rules of the game" exclude
so-called "unfair" competitive tactics like the disparagement of a competitor's
goods, outright dishonesty in financial dealings, and the like.
One has not said much when he simply states that the relationship of supply
and demand in a particular market determines the price charged, and the quan-
tity bought and sold. One must also investigate the factors determining the sup-
ply and demand functions. Unfortunately, many of the determinants of these
relationahips are rather subtle and tend to elude quantification. Generally, speak-
ing, supply is influenced by the costs of production, promotion, and distribution;
by the arrangements in the terms of which production costs are allowed to influ-
ence pricing and production; by the channels of distribution employed; and by
the legal provisions affecting the cost and availability of imports, the techniques
of promotion, the ability of new products and new firms to enter the market,
and the strategies available to sellers to temper competitive pressures. Demand is
influenced by the severity of the patient's need for the drug, his ability to pay
(either privately or through eligibility for public aid or reimbursement), and the
degree to which sales promotion efforts succeed in capturing the attention of the
physician in such a way as to influence him to prescribe a certain brand of drug.
A. An analysis of the factors influencing the supply of prescription drugs
1. Costs incurred by drug manufacturers
The major costs incurred by drug manufacturers may be categorized as follows:
(1) basic research; (2) applied research; (3) product development; (4) manu-
facturing the active ingredient; (5) preparation of finished dosage forms; (6)
sales promotion outlays. Each cost categOry may be best discussed by contrasting
activities actually undertaken with those which would be appropriately in an
effectively competitive drug industry.
Basic research-The drug industry has a distinct weakness for arguments to
the effect that high drug prices are due chiefly to high research costs, that most of
the research is basic in nature, and that high profits are necessary to finance these
large research budgets. The argument is beginning to seem a bit anachronistic
in view of the repeated demonstrations that the research budget is only a very
small part of the sales dollar, being lil the vicinity of 6 per cent. But there are
further weaknesses in the argument. It is obvious that while high prices might
be necessary to finance large research outlays, high profits are not. As has often
been pointed out, profits are a residual after deducting all costs, including re-
search costs. High profits indicate simply that a large part of the sales dollar
w-as not accounted for by research costs-or any other costs. Furthermore, the
majority of what is included under the category of "research" is something
° B. D. Bransome of the Arthritis and Rheumatism Foundation, while fortunately hav-
ing no mortalities to report, did state that some individuals with severe inflammatory
diseases would go without food in order to be able to afford to buy drugs. Hearings on
Administered Prices, Part 14, pp. 7992-7993.
PAGENO="0362"
1912 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
quite different than fundamental research. (It appears that within the last few
years, some drug industry spokesmen are becoming more willing to admit that
the share of their research budget devoted to truly basic research is a minor
one; but there is no sign of any retreat from the position that the type of research
which they do engage in is of immense benefit.)
It is not surprising that drug firms devote relatively little time to strictly
fundamental research since this sort of activity is inherently risky in that the
firm incurring the costs may not be in a position to capture all the benefits. Basic
research, in the sense of activities which extend the frontier of knowledge, is
potentially the most productive form of research, but also the riskiest. Risky to
whom? Here in a narrow sense the answer is clear-to the firm which invests
in this activity, since failure to obtain results will mean the loss of investment.
But to whom is basic research productive? It depends upon the ultimate implica-
tions of what you find, which are quite unpredictable. It may help the firm doing
the research, but it may help its competitors even more. Or it may help firms in
different industries. Or it may simply accrue to the benefit of humanity generally.
To engage in basic research, a drug firm must satisfy itself that this use of
the company's funds is more productive than an alternative use in applied re-
search, marketing, or the like. Such a decision is contingent upon obtaining satis-
factory answers to the following questions: 10
(1) What is the risk of failure of a given project? (Fundamental research
is always highly risky; this is its very nature.)
(2) If the project is successful, will the findings ever be commercially
applicable? (Many findings are too generalized to be immediately applicable,
although they may lead to subsidiary findings which will have commercial
applications.)
(3) Will the resulting findings ever lead to patentable discoveries? (Often
the findings are not patentable in the form in which they are ebtained, and
unless they can be embodied in some sort of marketable item, may never
be profitable even though patentable.)
(4) Will the time horizon between initiating the research project and its
fruition in the sales of commercially marketable products be sufficiently
short that the discounted rate of return on the investment will justify the
outlay? (If initial risk is high and the research and development period is
long, the net return over time may be lower than can be earned through
such alternative uses of the funds as marketing, even though the project is
eventualiy successful.)
(5) Will the "gestation period" of product development for patentable
discoveries `be short enough that patent protection will be commercially
profitable?
(6) Will the discoveries prove to be of equal or greater benefit to the
rivals of the firm? (With the well-developed state of the art of molecular
manipulation, this is a great risk and may discourage basic research to some
extent.)
(7) Will the discoveries prove to be of greater application in industries
outside the pharmaceuticals field? (This is a distinct possibility since phar-
maceuticals is a very specialized field. It might be more likely that a major
firm in the chemical industry could undertake basic research that might lead
to either a chemical or pharmaceutical discovery, and in the latter event it
might be possible to diversify into drugs. The reverse is less likely.)
(8) Will the discoveries pose the threat of obsolescene to presently profit-
able products?
These ~onstitute a large number of difficult hurdles to clear, and so it is not
surprising that drug firms should do little truly basic research. Actually, funda-
mental research is ultimately a philanthropic activity in the sense that it always
has some potential for benefiting society generally. Hence it is appropriate that
most basic research be carried out by endowed foundations and universities as
w-ell as publicly financed agencies specialized for this purpose.
It may be concluded that under efficiently competitive conditions, private drug
firn1s would not do much basic research because greater profits would be more
likely if the money w-ere invested in other functions. Instead, the industry would
rely upon the results of basic research made publicly available by more broadly-
10 See Submission of the Government of the Province of Alberta, op. cit., p. 2435.
PAGENO="0363"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1913
financed agencies. And it appears that in fact the drug firms do not engage exten-
sively in truly basic research. But this does not necessarily imply that drug in-
dustry research activities give rise to no distortion in the spectrum of basic re-
search efforts. It has been said that too little basic research in areas relating
to health and therapy is done by non-industry organizations, at least partly be-
cause the ability of the industry to pay high salaries (in turn dependent upon
the high profitability of drugs under present market conditions) diverts too large
a portion of the very small pool of qualified investigators and technicians away
from public employment in basic research and toward private employment in
applied research and product development, testing, and application-all of which
are lower-priority uses' for their very scarce skills. The remedy would seem to
lie not in increasing the amount of basic research done by private firms, but in
taking steps `to reduce their ability to drain off the best scientific personnel for
work in less productive employments than they are capable of pursuing.
Applied Research--The economic rationale of applied Tesearc'h is quite
straightforward in any industry: to serve as one means of implementing a profita-
ble marke'ting operation. The direction and emphasis of this research in the field
of drug therapy is influenced by (1) the' nature of the patent system: (2) the
impact of the patent system on the organization of the industry; and (3) the
effect of industry activities on research outside the industry. The mere existence
of the patent privilege for drugs biases research toward patentable inventions
and away from areas where no patents can be obtained. This discriminates
against basic research and stimulates applied research. It also discriminates
among different channels of applied research. From the medical point of view,
research is unbalanced due to an unduly intense emphasis on chemotherapy,
while the complementary fields of nutrition, public health, biochemistry, and
preventive medicine are underemphasized. Antibiotics provide the most sharply
focussed example. Concern has been widely expressed that antibiotic therapy may
ultimately prove to be a blind alley due to overuse and the development of
resistant strains of micro-organisms. It would seem wiser to spend less effort on
activities which tend to make micro-organisms increasingly resistant to control,
and more effort on attempts to make man naturally more resistant to micro-
organisms.
Furthermore, by biasing efforts toward applied research, the patent system
will reduce the scope of basic research findings which can be applied, and ulti-
mately will depress the productivity of' applied research. There has been much
discussion in recent years of the "increasing cost" of drug research per new
discovery. (This was true even before the FDA began to implement more strin-
gent controls over new drugs.) But to speak of increasing costs is simply to
refer indirectly to the decreasing productivity of efforts. Again, antibiotics offers
a good case in point. Applied research here was productive for a good many years,
in large measure because fundamental research in this field had already elu-
cidated much of the mechanism of infection by micro-organisms. Bacteriology
was already an established field of study. But the same is not true in the other
major areas of drug research, such as tranquilizers and oral antidiabetic agents.
Here there has been a less prolific output of various useful drugs and less enthu-
siasm among independent authorities regarding the extent to which the later
drugs are advances over the earlier drugs, and perhaps even over related drugs
which antedated the "miracle drug" era.
Another bias of some interest is related to the fact by protecting new prod-
ucts, either as such or through exclusive process privileges, the patent law
biases applied research in the direction of concocting new products rather than
fully investigating the properties of known compounds. As Prof. George Wright
of the University of Toronto has contended, it seems to indicate rather overly
one-sided emphasis that new drugs coming from drug houses are almost invari-
ably novel concoctions and therefore patentable, while the reservoir of some two
million already known compounds hasP only been pharmacologically investiga'ted
to a very modest degree.1'
To the extent that patent reforms can reduce biases of this type, resources
will be allocated more efficiently throughout the drug research sector of the
~` Professor Wright advocates more screening of known compounds rather than an
exclusive emphasis on the concocting of new ones, on the assumption that "much is yet
unknown about the association between chemical structure and pharmacological action,"
observing that the screening approach originally brought the sulfa drugs into existence.
Canadian Henrings, op. cit., no. 8, p. 540.
PAGENO="0364"
1914 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
economy. Basic research will increase relative to applied research, so that there
will be a greater variety of fundamental research findings on the basis of which
upplied research can be conducted. The potential which resides in the areas of
nutrition, biochemistry, public health, preventive medicine, and other areas will
be more nearly capable of full achievement. The bias in favor of contriving new
compounds instead of systematically conducting an empirical study of existing
compounds will be reduced. And the amount of exclusively imitative research,
and of other types of research of secondary importance aimed at finding a
patentable vehicle for a "blitz" sales promotion campaign, will be reduced or
even eliminated.
This latter phenomenon of duplicative activities is a major element in drug
research today, as conditioned by its patent orientation. Basically, there is an
over-intensive exploitation of those approaches known in the past to have yielded
profitable drugs. Since the number of known approaches is limited, it is within
the capacities of major firms to explore several of them, and since all firms are
conscious of the commercial advantage of being able rapidly to duplicate the
successful new drugs which their rivals may find, the research programs of large
drug firms tend to duplicate, at least in part, the programs of their major rivals.
(This was attested to not only by many witnesses at drug industry hearings, but
also by the near-simultaneous discovery of several drugs by two or more firms.)
This constitutes a compounded misallocation of resources: not only are scarce
talents diverted from basic to applied research, but wasteful duplication of effort
on precisely the same applied research projects seems to be common.
Much of the criticism of the "molecular manipulation" approach can be most
appropriately directed at this phase of the industry's operations. The ideal
manipulated molecule is one which is pharmacologically identical with the profit-
able product of a rival, but is legally distinct in the sense that a patent may be
obtained. However, it is the latter criterion which is crucial, not the former, and
the typical me-too version of an existing drug is of dubious superiority, if not
absolutely inferior, to the original drug which it is intended to supplant. The
most impressive testimony regarding the prevalence of misdirected research in
the major drug houses came during the Kefauver hearings from two physicians
who formerly served as medical directors for major firms. Dr. A. Dale Console,
formerly with Squibb, when asked whether there was much drug research which
produces nothing worthwhile and is not intended to, replied:
"I think the majority of it is in that category . . . and I should point out
that with many of these products, it is clear while they are on the drawing board
that they promise no utility; they promise sales. It is not a question of pursuing
them because something may come of it . . . it is pursued simply because there
is profit in it." 12
He also reported that imitative research could crowd out productive work:
"When a `crash program' comes along in which some product is being pushed
in order to get it out before a competitor gets it out, it is not unusual for a
worthwhile research program to be postponed so that the people can be taken
off it to be put on the `crash program'. Very frequenty some of these programs
are never picked up again. So I think that good research is actually hampered
by this type of thing." 12
Dr. Haskell J. Weinstein, formerly w-ith the Roeririg division of Pfizer, tie-
nounced industry managements for wasting the time of their research personnel:
"Their talents should not be expended on patent-bypassing chemical manipula-
tions, on ridiculous mixtures of drugs, or inconsequential additives to established
drugs. Since the number of well-trained capable scientists is severely limited,
their potential should not be wasted. The long-term benefits of the appropriate
utilization of the abilities of these skilled individuals would be immeasurably
greater." 13
This illustrates some of the subsidiary distortions in applied research resulting
from the patent incentive: not only modified molecules, but the development of
often irrational combinations of existing drugs which lack flexibility and com-
pound the problems of dosage and toxicity, and the devising of additives which
12 Hearings on Administered Prices, op. cit., part 18, p. 10379.
12 Ibid., part iS, p. 10254.
PAGENO="0365"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1915
may constitute questionable and/or necessary flourishes in the interest of in-
creasing a drug's absorption rate, guarding against side effects, etc.'4
Product development-The category of product development includes many
activities, such as experimental and clinical testing, determination of appropriate
dosages and dosage forms, obtaining FDA approval for marketing new drugs,
constructing pilot plant facilities, etc. Subsequent to initial marketing there
would be product application work relating to long-run evaluation of the total
effects of a drug, improvements in dosages, revisions of brochures, and related
activities.
In an efficiently competitive drug industry, profit prospects from marketing
new drugs would be moderate, and the temptation for extreme haste in product
development would be correspondingly slight. It is very important that this
temptation be minor, since proper testing and evaluation of new drugs is an
important and time-consuming task. And, as Dr. C. D. Leake of Ohio State
Univeriity has observed, "There is no shortcut from chemical laboratory to
clinic, except one that passes too close to the Morgue." 15 But before certain
needed reforms were legislated in 1962, many firms yielded to the temptation
to rush new drugs thru the development phase and on to the market as soon as
possible, limiting experimental and clinical work to the minimum acceptable
levels under the old legislation, harassing FDA staff members into approving
inadequate applications,'6 and even skipping such seemingly essential product
development stages as pilot plant operatio~." Furthermore, with an inflated num-
ber of drugs being clinically investigated in the expectation of reasonably rapid
FDA approval, the available time of the most highly qualified investigators was
soon completely employed, and recourse to less trained, less capable, and in some
instances less scrupulous individuals was necessary. But drug evaluation by
unqualified investigators can be worse than useless.'8
Since the passage of the 1962 legislation, there has been much improvement
in this area. More stringent requirements for approval of new drug applications
have been imposed; the number of new drugs being evaluated has apparently
declined, making possible an increase in the average quality of evaluations, and
the morale and effectiveness of the FDA has greatly improved. This is apparently
one area in which drug safety reforms may have been successful in eliminating
certain economic wastes as well as improving drug quality. But it should be
noted that any reduction in total drug development outlays would be likely to
result from a reduction in the number of new drugs under investigation; the
average cost of investigation per drug is likely to increase, and this increase
is certainly in the best interest of public health.
Manufacturing of the active ingredient-In an efficiently competitive drug
industry, each stage in the production process would be carried out at minimum
14 Dr. Harry F. Dowling of the University of Illinois Medical School cited an excellent
example involving both molecular manipulation and the use of Inconsequential additives.
Lilly discovered erythromycin in 1952, and in 1953 Pfizer retaliated with a molecular
shadow, carbomycin, which proved less effective in human disease than in the test tube,
and was finally withdrawn from the market in 1960. Pfizer tried again in 1956 with an-
other chemical echo of the erytliromycin, eleandomycin, and in 1957 modified its own
modification, called it triacetyloleandomycin, and advertised it widely as a major break-
through in that the same oral dose as eleandomycin produced somewhat higher concentra-
tions of the drug in the bloodstream. Lilly responded in 1958 by modifying its original
erythroinycin and marketing it in the form of its propronyl salt, claiming a higher blood
concentration rate than could be achieved with triacetyloleandomycin. None of the four
later drugs had any real advantage over the original discovery, since slightly higher doses
of the original drug would have been as effective as the later variants. Ibid., part 24
pp. 14167-14168.
15 Ibid., part 15, p. 10418.
1~ See testimony of Dr. Barbara Moulton, Ibid., part 22, Pp. 12025-12032.
"Lederle bypassed the Pilot-Plant Stage with its Triamcinolone. See Fortune May
1960. p. 276.
18 Dr. Maxwell Finland, Harvard University Medical School, cited an instance u-here
a clinical investigator had reported successful treatment of 100 cases of staphylococcal
pneumonia u-ithout a mortality. Since the usual mortality rate among the patients con-
cerned is 50 percent, the drug would appear to be miraculous. But upon investigation
Dr. Finland concluded that not a single case of staphylococcal pneumonia had been
present, and inferred that the investigator was incompetent to diagnose the presence of
the true disease from the laboratory cultures. He concluded pointedly: "This is the sort
of thing that I say is dangerous because another doctor who knows how to make a
diagnosis of staphylococcal pneumonia will use that drug to the peril of his patient."
Hearings on Administered Prices, op. cit., part 24, pp. 450-451.
PAGENO="0366"
1916 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
cost, and without incurring the cost `burdens of excess capacity, which one
suspects are characteristic of industries like drugs which shun price competi-
tion and rely on product differentiation.19 Although very large investments in
plant and equipment are common in the heavy chemicals industry, in the
pharmaceuticals field this is not the rule. The investment in facilities required for
the production of the active ingredients varies considerably from drug to drug.
For those active ingredients which can most efficiently be produced by truly
mass-production techniques, production by makers of fine chemicals or even
bulk chemicals would be indicated. But for many drugs, the required investment
is relatively modest in comparison with the supply of funds avalilable in
capital markets. Relatively small firms can efficiently produce the active ingredi-
ents for such drugs. Mass production methods are not appropriate for many drugs
since the physically minute quantities used in dosage forms require only a small
total annual volume of output. But it still might be more efficient for a small
firm to buy the basic ingredient under contract from a larger firm. If the market
were competitive, the relative economies of scale in produnction versus those
in dosage form preparation and in distribution should govern the extent of the
functions assumed by different producers at different stages in the industry.
If patents posed no real barrier to entry into drug selling, chemical raw mate-
rials and intermediates could be made by bulk chemical companies, the active
ingredient could be made by fine chemicals producers, and the finished dosage
forms could be tableted and packaged by drug makers to be distributed through
various channels. Without barriers to entry, the comparative costs of each
stage in the industry would determine the allocation of functions among differ-
ent firms. As far as financial requirements are concerned, there is no reason
why a large number of relatively small firms might not compete effectively in
the drug market.
But under present market conditions there are a number of factors which dis-
tort the division of labor among firms and introduce other criteria than compara-
tive costs as determinants of The Degree of Specialization within and Among
Firms. A relatively small drug maker might find a new drug, patent it, and under-
take production of the active ingredient on his own premises, despite the circum-
stance that his costs might be high due to his inexperience, inappropriate facili-
ties, and overall lack of adaption of his operation to the requirements of fine
chemicals manufacture. Production would still be undertaken at higher costs,
however, if the patent holder wished to prevent the "know-how" which is ordi-
narily not disclosed in the patent, from being acquired by another firm. The inef-
ficiencies inherent in this arrangement could be partially overcome by the merger
of the small firm with a larger producer of bulk or fine chemicals, but this would
increase the market power of the formerly small firm relative to its rivals.
Forward integration by merger is also stimulated by marketing practices. The
intense sales promotion of drugs under brand names particularly thru nationwide
advertising and detailing, is a practice which creates economies of large scale
marketing even though none may exist in production. This is unusual; ordinarily
it is economics of large-scale production which prevent the successful operation
of a large number of small firms, but in drugs it appears that while efficient pro-
duction might occur at a very small level of output, the exploitation of modern
marketing techniques in the drug market context can be taken full advantage of
only by a very large firm.
Where the factor of "know-how" is not important, drug firms may contract out
the production of the active ingredient to specialists. In such cases, the ratio of
the price of the bulk drug to the market value of the substance when embodied
in final dosage forms and sold to distributors is well worth noting. Ratios of the
order of magnitude of one hundred to one are not unknown.2° This is very simply
explained. There is price competition among the firms which make the active in-
gredient, but none in the sale of the finished product. If competition were to be
19 By way of example, when Bristol was producing about one-third of national tetracyc-
line output, it still had 80% excess capacity in this drug. Drug Industry Antitrust Act
Hearings. purt 4, p. 2056.
20 During the recent Canadian Hearings, Empire, a small generic firm, estimated that
it could manufacture the drug diazepam for $68 per kilogram and a kilogram of diazepam
embodied in dosage forms is worth about $20,000-a ratio of 125 to 1. (This makes diaze-
pam worth about 16 times as much as gold.) See Alberta Government Submission, op. cit.,
p. 34.
PAGENO="0367"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1917
introduced at the finished product level, drug prices would decline until they were
more in line with production costs for the active ingredient.
Preparation of jinished dosage forms.-The simple technology of the prepara-
tion of most finished dosage forms, the low operating costs of these processes,
and the modest capital requirement for such facilities, renders this stage of
the industry ideally suited for workably competitive market performance. The
processes involved for most dosage forms are technologically routine and ele-
mentary, tabletting and bottling being particularly trivial operations technically.
(After all, every pharmacist is taught-and taught well-to do such compound-
ing operations on his own premises. It is both amusing and dismaying to ob-
serve industry attempts to convince the general public that there is some magic
in the preparation of even the simplest dosage forms, which is by implication
a secret known only to the major brand name firms.) It has been shown by
evidence presented at drug hearings both in the United States and Canada
that for the typical drug, "factory costs" (producing the active ingredient and
making and packaging the dosage forms) are a minor part of the wholesale
price. There is no purely economic reason why numerous small firms could not
contract out the manufacture of the active ingredient and then tablet and
package the finished dosage forms on the basis of a quite moderate total invest-
ment. Brisk price competition between many small sellers of drugs might develop
if production costs were the only barrier to entry. And this in fact is the
prevailing mode of market behavior for those small firms which produce generic
name drugs for which tight patent control could not be achieved over their
manufacture and/or the sale of their bulk powder. These firms can either
produce or buy the bulk powder at the lOw prices which result from competition
among bulk suppliers. These firms then tablet, package and sell the drugs at
low prices representing their low costs of production. But in most markets
these generic drugs compete with their presumed brand name equivalents, and
it is likely that the true production costs of the brand name sellers are even
lower than those of the generic firms. But does this mean that the large firms
choose to undersell the small generic houses? By no means, they charge prices
up to ten or more times as high. But does this not mean that they are not able
to make any sales at these high prices? Again, by no means. They outsell the
lower-priced drugs ten or twenty to one. To an economist who has been trained
to expect that quantity sold is inversely related to prices charged, this is a
dumbfounding situation. The answer of course, as is well known, is related to
sales promotion tactics.
$ales promotion ontlays.-Any spokesman for the domestic drug industry will
tell you that its outstanding accomplishments have been in the area of research.
But any well-informed expert on marketing is much more likely to tell you
that the drug industry's real expertise lies in the area of sales promotion. And
indeed the relationship of the marketing budget to the research budget suggests
as much. But the myth seems to persist in the general mind that research
budgets exceed advertising budgets, despite repeated demonstrations that the
latter is several times as high, as can be verified by even a superficial examination
of the financial statements of any large drug firm. It seems to me that public
education can sometimes be furthered more effectively, therefore, by witty
ancedotes or epigrams than by mere statistics. For example, in my experience,
Dr. A. D. Console seems to have done more than anyone else to expose to
salutary ridicule the dubious nature of much drug industry "research" in his
memorable statement during the Kefauver hearings; "They stress that there
are many failures for each successful drug. This is true since it is the very
essence of research. The problem arises out of the fact that they market so
many of their failures." ~` The same device can be used to put drug prices
research costs, and sales promotion outlays in perspective. During the recent
Canadian drug hearings, one participant observed: "for every dollar the dru°
industry spends on research, they spend four dollars telling you about it, and
charge you ten dollars more for listening." If anything, this witty statement
errs on the side of moderation. Research costs apparently amount to something
less than one-fifteenth of sales price, and somewhat less than one-fourth of
sales promotion outlays.
~` Hearings on Administered Prices, op. cit., part 18, pp. 10372-10373.
PAGENO="0368"
1918 COMPETITIVE PROBLEMS IN THE DRuG INDTJSTRY
In an efficiently competitive drug industry, sales promotion activities would
be adapted so as to reflect the structure of information needs and to take ad-
vantage of the ability of the relevant individuals to absorb information, in order
to minimize the cost of providing the necessary data. Since ethical drugs are
ethical in that they cannot be bought over the counter, advertising to final con-
sumers would seem unnecessary. Instead, advertising need be directed only to
prescribing physicians. But doctors would seem to constitute a specialized market
for several reasons. First, they are highly trained professional men. Second, they
are extremely busy. Third, they are unquestionably prosperous. A rational ad-
vertiser would adapt his sales strategies accordingly. Since the physician is quite
intelligent and well-trained, he can respond adequately to modicum of purely
informative advertising material and need not be bombarded with masses of in-
sistently persuasive promotional appeals. (If a word to the w-ise is sufficient, the
drug firms are guilty of a colossal insult to the physician's intelligence.) Since
doctors are busy, the time they can devote to the study of drug information of
all sorts is limited. And since doctors are men of some financial means, and profit
from the availability of good drugs, they should be expected to pay the costs of
being supplied with adequate drug information. The sales promotion outlays of
the drug industry, like all their other expenditures, are paid for entirely by the
patient. Even if doctors find that detail men are convenient sources of informa-
tion, the patient is still subsidizing the doctor by providing him with a number
of tutors, and biased tutors at that. It is debatable whether this sort of subsidy
is justified. However, the amount of the subsidy, rather than the nature of it,
is the chief point. If the doctor were to be required to bear the costs of obtaining
his own drug information, he could still pass the costs on to his patients through
higher fees. But the costs of alternative means of being supplied with informa-
tion-subscriptions to official compendia and their periodical supplements, or to
independent newsletters such as The Medical Letter, would be a very small frac-
tion of the amount spent at present on sales promotion by drug firms.
Obviously, the necessary information on drugs must somehow be supplied. It
may be supplied by the companies, by independent evaluating organizations, or
by government. If supplied by government it can either be made available upon
subscription or can be distributed free of charge. By whatever medium it is com-
municated, it shOuld achieve four objectives: (1) insure adequate flow of accu-
rate and unbiased information; (2) minimize the volume of redundant communi-
cation; (3) make informative communications more concise; (4) eliminate all
misinformation.
The provision of this information by the companies themselves has been
sharply criticized, chiefly by the recipients of these promotional attentions, on
numberless occasions. This is not the place to deal in detail with these criti-
cisms, but the chief complaints may be recapitulated in terms of failure to achieve
the goals of a satisfactory communication system: (1) constant interference of
commercial bias; (2) excessive communication, such that the volume of indiffer-
ent information and just plain "noise" minimizes the likelihood of the detection
of the occasional communication of genuine value; (3) emphasis on persuasion
and suggestion rather than upon providing genuine information; (4) redun-
dancy of communication as a result of the mutually offsetting nature of sales pro-
motion rivalry among firms not competing on a price basis; (5) the presence of
a certain amount of outright misinformation, chiefly in regard to inadequate dis-
closure of side-effects or contraindications.
Evaluation by independent sources, or by public bodies, should eliminate com-
mercial bias and minimize the temptations to indulge in excessive or overly per-
suasive communication. Intentional misinformation should also be eliminated,
although the fallibility of human agency will render any drug information sys-
tern less than perfect. It is, however, doubtful that the first goal of an informa-
tion system-insuring an adeqvate flow of information-will be achieved by any
system which leaves the aquisition of the information source up to the discre-
tion of the physician. Proof of this is supplied by the small fraction of the
medical profession subscribing to the Medical Letter: only about 15 percent,
which is to be deplored. (It is of course possible that if the institution of price
competition in the drug industry lowered profit margins and eliminated the de-
tail-man, more physicians would subscribe.) In England the Prescriber's Journal
is distributed free to all doctors, and a similar step is being considered in
Canada. We in the United States should be hesitant to extend such a subsidy
PAGENO="0369"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1919
to the members of a well-paid profession, but this may still be the least costly
way of insuring at least the availability to all doctors of adequate and un-
biased drug information. And while an element of subsidy to the doctor would
remain, the subsidy would now be financed by the public as a whole, rather than
that smaller and to some extent disadvantaged segment of it which in a given
period of time finds it necessary to purchase drugs in varying amounts.
If adequate drug information were to be made available from unbiased sources,
by how much might drug industry sales promotion costs decline? Unless steps
were also taken to institute price competition in drugs, expenditures would
probably not drop at all, and might even increase, each firm trying not only to
offset advertising efforts of other firms, but also the now more widely-dis-
tributed data from unbiased sources. To explain why this outcome is likely, the
crucial role of sales promotion in the drug industry must be further explored
at this point.
The key role of sales promotion lies in its ability to substitute for productive
research on the one hand, and for genuine price competition on the other. If
a firm's research department discovers a highly effective new drug of un-
mistakable value, the drug tends to advertise itself. Physicians themselves
rapidly spread the news. On the other hand, as Dr. Walter Modell of Cornell
University Medical School has stated, "The more a drug has to be peddled,
the more one begins to wonder why." 22 Dr. Console, formerly of Squibb, was
more explicit-about the relationship between unproductive research and the
advertising budget: "Advertising is called upon to make successes of research
failures." 23 Since it appears that almost any drug will sell, at least for a while,
if promoted intensely enough, it does appear that the good offices of the sales
promotion department may compensate greatly for the indifferent fortunes of
the researchers-but at quite a cost, in terms of premature and usually unneces-
sary obsolescence of existing products. Even these costs, however, are con-
verted by drug spokesmen into rhetorical~ capital: they are construed as meas-
uring the risks of product obsolescence, allegedly due to the rapid development
of superior medications, and these "risks" are supposed to constitute a justifica-
tion for the high profit rates characteristic of the industry. The "risks" involved
are not wholly illusory, but are not, as claimed, inherent in the research process.
Instead, they arise from the way in which drugs are developed and promoted.
The high profits cannot be justified by the high risks because the height of the
profits induces expensive and disruptive product competition which manifests
itself in such "risks". But if one were tO create price competition and reduce
profits, this mode of product competition would be too expensive to support, and
the "risks" of product obsolescence would be diminished in proportion to the
decline in profits.
Even more important is the degree to, which sales promotion can substitute
for price competition. This is true in most industries, but drugs are unique in
that such efforts can almost completely suppress price competition and further-
more can seriously discredit the price competitor. Again, the sheltered institu-
tional circumstances of drug marketing should take the credit. If a company can
monopolize the eye and ear of the prescribing physician, it monopolizes the most
important drug market, given the existence of laws supporting brand-name pre-
scribing. But monopolizing the physician's attention becomes increasingly ex-
pensive when more than one firm tries this strategy. It soon becomes so expensive
that only a limited number of firms have the resources to continue-hence, as
mentioned before, economies of large scale marketing act as barriers to entry
where production costs in themselves would not be prohibitive. Smaller firms
cannot ordinarily bring their products to the attention of the physician, even
though they may be selling at perhaps 90% less than the brand name drugs.
And there is yet another way in which the sales promotion techniques of the
industry put the small firm at a prohibitive disadvantage. Although the small
low-price firm cannot ordinarily make its presence in the market directly known
to the physician, the doctor may eventually become aware of its existence and
perhaps wonder how the midget firm can undersell the giant by a ten-to-one
ratio. The detailman is ready to supply the answers. It appears likely, on the
basis of the testimony of many doctors at public hearings, that a major reason
22 Drug Industry Antitrust Act Hearings, Part I, p. 32P.
28 Hearings on Administered Prices, op. cit., part 18, p. 10372.
81-280-68-pt. 5-24
PAGENO="0370"
1920 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
why drug firms spend so much on detailmen is because of their peculiar ability
to disparage, with relative impunity, quality of the products of the low-priced
generic firms. The conscientious physician is naturally vulnerable to having his
confidence in low-priced drugs undermined by such disparagement since he is
personally not in a position to evaluate the relative quality of drugs. Whenever
a buyer lacks full information on the nature of the product at the time of pur-
chase selling efforts take on some of the coloration of a "confidence game" where
the buyer is induced to take the seller's word for it that his product is better
than his rivals', it is suggested that to be unknown is to be suspect in the drug
industry, and that it is common knowledge that most generic drugs are sub-
potent, adulterated, etc., and may even be made by counterfeiters. Thus the
i;hysician may be induced to equate low price with low quality and hence shun
generic prescribing. Hence drugs priced 90% below the market may account for
less than 10% of all sales. For these reasons there is no other industry in exist-
ence where the disparagement of the quality of lower priced products can so
completely substitute for price competition.
Is there any substance to this disparagement? There should be very little
reason to suppose that low price need be associated with low quality. If any-
thing, doctors should be more hesitant to prescribe by brand name, since counter-
feiters (who make more on $100 bills than on nickels and dimes) naturally
specialize in the high-priced brand name drugs. As far as economic motivations
to save costs by cutting corners are concerned, these should be minimal in drugs.
The strength of the positive motivation-cost reductions-is relatively small,
since neither the costs of quality control nor of the active ingredient itself are
particularly large components of total cost. For a major firm, quality control
costs seem to range from about one to three percent of the sales dollar, and al-
though the figure would probably be higher for a small firm, it should still not
be a controlling factor in costs. And while official compendia specify a certain
range of allowable variation for drug potency, the typical range is only about
90 to 110 percent. But since the cost of the active ingredient in a given drug is
usually only a minor part of total cost, the cost saved by orienting the production
process to produce an average content of 90 percent of stated label potency would
save at most only 10 percent of this cost. Furthermore, it would inevitably mean
the production of a number of substandard drugs and would expose the firm to
punitive actions by the FDA. This brings up the negative aspect the deterrents to
substandard performance. Both generic drugs and their brand-name equivalents
must meet official standards specified in drug compendia. Experts have testified
that there is no therapy gain to be achieved by producing to purity standards
"exceeding official" "minimum" standards. The products of all producers are
held to the same inspection standards, and a small firm will be even more strongly
motivated than a large firm to conform to the regulations since the impact
of a given fine will be much more disruptive to its finances.
Brand name firms have alleged that there is no therapeutic equivalency even
among drugs which sati~fy the requirements of official compendia. I sympathize
with such witnesses as Dr. Solomon Garb, who professes to find this sort of
argument both elusive and baffling.24 I respect Dr. Garb's opinion and share his
suspicion that the differences are trivial and that they cannot be meaningfully
specified. It is very hard to follow drug industry arguments which suggest that
because no two drugs, or capsules, are absolutely identical, that one should buy
brand names and shun generic names. The emphasis on the unique nature of
each pill is reminiscent of the philosophical doctrine of nomanilism, which im-
plies that no generalizations are possible since everything is in a unique category
by itself. I submit that drug firms are more pragmatic than nominalistic in their
serious moments.
The most authoritative testimony of this point would appear to have been
given by Dr. Lloyd Miller, Director of Revision of the United States Pharma-
copeial Convention, before this Subcommittee, in stating that "there are not
24 Dr. Garb expressed himself as follows: "It seems to me that if any group of drug
manufacturers wish to use the argument that their brand name drugs are better because
of certain differences, and that the doctor knows what these differences are, they should
show how the doctor finds out these differences . . . I think the differences are trivial,
but my point is I do not know that they are trivial, because I cannot find out why they
are. I have never been able to find out what the difference is between one brand of the
drug and another brand of the drug." Competitive Problems in the Drug Industry, op. cit,
Part 2, p. 545.
PAGENO="0371"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1921
more than a dozen examples where the difficulty (questionable clinical equiva-
lency of drugs meeting DSP standards) has been discovered, and it is not gen-
erally true even for all of them." My own evaluation of the "generic equiva-
lency" or more precisely therapeutic equivalency dispute is that the brand
name firms have greatly overstated the significance of any valid or partially
valid elements in their position, and that the argument on the whole is with-
out merit. The verdict of lack of merit seems inescapable since there should be
no reason to assume that the therapeutic equivalency of two brand name prepara-
tions is any greater than that of a brand and a generic name preparation. As
Dr. Martin Cherkasky recently testified before this Subcommittee, "I must tell
you that I feel quite insecure at the same time about the performance of some
of the major drug companies in this country." 26 The intensity of the dispute
is all the more puzzling since it appears to take for granted that pharmacology
is much more of an exact science in practice than in fact it isP
It would seem that the best argument which the major drug firms have in
support of their charges that unsafe drugs are on the market is the contention
that FDA inspection is not adequate. This is a perennial charge, and one which in
recent years has become much less plausible because of long-needed increases in
FDA staffs and budgets. But if any doubt remains as to the adequacy of FDA
inspection, it should be resolved by providing FDA with still more funds to the
full extent which may prove necessary. Perhaps. Still, it is the direct respon-
sibility of the FDA to safeguard the public from poor drugs, in a much more
immediate sense than this responsibility is shared by USP. Batch certification
of all drugs might be one solution. Dr. Solomon Garb's proposal that all drug
manufacture be subject to continuous federal inspection is another solution.
Dr. Miller charges that the former proposal would be unnecessarily costly; Dr.
Garb reports that drug makers contend that the costs of his plan would be
astronomical (although, as he points out, there is private inspection today at less
than astronomical cost, and it would seem quite feasible simply to substitute pub-
lic for private inspection with avoidance of excessive duplication of efforts.) An
economist would wish to make one point. Regardless of the methods adopted, the
physician must be made confident that all drugs on the market are safe to use.
Relative costs of the methods employed to bring about this confidence are im-
portant, but it is very unlikely that the costs involved would fail to justify the
resulting benefits: complete eliniination of substandard drugs and the opening
up of much of the prescription market to price competition between low-priced
generic drugs and currently high-priced brand-name equivalents. The cost of the
industry's current system of "insuring" drug quality by advertising intensively to
promote brand-name drugs and discourage generic prescribing does tend to in-
crease the sales of the former drugs and reduce the sales of the latter. But
beyond that, it perpetuates a great barrier to price competition and places an
~ Ibid., Part 2, p. 514.
~ Ibid., Part 2, p. 668.
~ A layman can perhaps be forgiven for Introducing Into the record the Idea that with
few exceptions (but among them some admittedly very important ones) in the present
state of our knowledge, drugs tend to be relatively blunt Instruments when employed In
human therapy, so that the question of "therapeutic equivalence" when generalized to
Include all drugs frequently becomes subsidiary to the question as to whether or not the
drug can be depended upon, In a particular circumstance, to do its job at all. Placing some
emphasis upon this may put the issue of therapeutic equivalence In a more adequate per-
spective. The significance of this Issue would become much greater If one could always
assume that If each drug Is optimally manufactured and administered, It will accomplish
a predictable, effective, and exclusively beneficial therapeutic result upon each adminis-
tration. But where the effect of the drug Is unpredictable, Imperfect, and tempered by
side-effects, therapeutic equivalence as such is a much less paramount consideration, and
the Importance of other factors, such as biological variations among human recipients,
incidence of side effects, and the inherent uncertainty regarding the mechanism of action
of the active Ingredient Itself, become more significant. The notion that many drugs are
blunt instruments is not propagated by the Industry, but It Is hard to avoid drawing such
a conclusion, If for no other reason than the Inability, to date, of molecular engineers to
eliminate side effects. Candid doctors have submitted that these effects should not' be
rerarded as incidental disadvantages, but are an integral part of the total action of the
drug, and should rather be referred to as "concomitant effects." Just as burning down
the house would be a "broad spectrum" recipe for roasting a pig, so also must one regard
antibiotics which Indiscriminately kill both harmful and beneficial organisms within
their range of activity as being relatively blunt Instruments. One is reminded of Buck-
lngham's characterization of his parody of the dramatic hero, Drawcansir, in the play
The Rehearsal. Drawcansir's battle cry was:
Let petty kings the names of factions know;
When e'er I fight, I slay both friend and foe.
PAGENO="0372"
1922 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
unjustifiable price premium on the brand name drug. This system amounts to
insuring drug quality by what is tantamount to private taxation, and levied
through the charging of prices far above production costs. But this does not
eliminate the production of substandard drugs, since FDA studies show that a
disturbingly high percentage of both brand and generic name drugs do not meet
requirements. It should be cheaper to institute universal FDA inspection, of
plants and products. While the cost to the government might increase (unless
as in the case of meat inspection, the companies were to pay the salaries of the
inspectors) the resulting increase in price competition which should be expected,
once arguments that low price means low quality become manifestly specious
to all, should reduce the expenditures of drug buyers greatly.~8
2. Influence of Economic Aspects of Drug Distribution on the S'uppiy of
Drugs
There is one complaint voiced by the drug manufacturers with which I have
some sympathy. While about 50 per cent of the retail price of drugs is accounted
for by distribution costs and markups, rather more than 50 per cent of the
criticism of high drug prices has been centered upon the manufacturers. The
proximate cause of the seeming disproportion in attention given the drug makers
would appear to be the great visibility of their profit achievements-outstand-
ingly high average profit levels over the last 15 or 20 years-in contrast not only
with the absence of any data to show that the druggists have been similarly
successful in feathering their nests, but even a general impression to the con-
trary. The number of retail pharmacies seems to have declined during the period
of greatest drug firm success, and many pharmacists who have failed to make
what seemed to them a reasonable living in their own profession have reportedly
been hired by the drug industry, where they can obtain more remunerative-if
perhaps less productive-employment.
But low profit margins in themselves do not necessarily attest to effective coin-
petition, since not even a complete monopolist is assured of high profits unless
he can operate efficiently. On the other hand, not all inefficiencies are necessarily
traceable to the mismanagement of the individual firm. Some inefficiencies may
be thrust upon the distribution level from without; others may be unintended
consequences of policies fostered within group itself. But before looking into
these questions, it is first useful to distinguish between wholesale and retail
distribution.
The wholesaling function seems to be the most efficiently performed stage in
the industry, chiefly because the wholesaler operates in the most competitive
market. Drug manufacturers have their markets protected by patents, trade-
marks, sales promotion outlays, and the relatively small number and large aver-
age size of the major firm. Druggists also enjoy rather protected markets be-
cause of the practice of brand-name prescribing, antisubstitution laws, and other
regulations which put the consumer at a disadvantage, plus the advantages
associated with being a closed profession regulated by semi-autonomous profes-
sional associations which have at least the potential for limiting the number
of qualified practitioners and hence influencing the rate of entry into pharmacy.
The wholesaler, however, has no comparably strong bargaining position. There
are many w-holesalers, mostly very small, and no appreciable barriers to new
entry. Furthermore, if drug makers can perform their own wholesaling func-
tions more efficiently than the independent distributor, they will integrate for-
ward and sell directly to retailers. And if retailers can do better for themselves
~ I am aware of no recent estimates of the cost of making FDA inspection fully adequate.
However, an order of magnitude approximation of the relation of probable costs to re-
quired cost reductions can be made from data presented at the Kefauver Hearings. In
1959, FDA Commissioner Larrick stated that it would take a budget increase of $3,418,000
to permit adequate drug inspection. Profits of the 22 major drug firms were $5G2 million
in 1958. A decline in drug prices sufficient to cut drug firm revenues by $7,121,000 would
of course cut before-tax profits by the same $7,121,000. With a tax rate of 52% in effect
at that time, this reduction in pre-tax profits would have cut tax receipts by $3,705,000.
The net gain can be roughly measured as the $7,121 000 saved by drug buyers, minus the
$3,705,000 in reduced tax receipts, or the required ~3,418,000. Thus If adequate drug in-
spection could create confidence in lower-priced drugs to the extent that the resulting
competition would lower major drug firm prices by enough to cut total profits by as little
as 1.27 percent before taxes, the savings realized would pay for the expanded enforce-
ment budgets. If total profits are about 20% of gross revenues, the necessary percentage
price cut would be as little as ii of the 1.27 percent, or i4 %. See Hearings of Adminis-
tered Price, op. cit., Part 22, p. 12132.
PAGENO="0373"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1923
by forming their own wholesale supply agency, on a cooperative or other basis,
they caa integrate backward and buy directly from the factory. Hence the in-
dependent wholesaler must meet the test of a competitive market, and provide
efficient, reasonably-priced services, or find himself by-passed. If all stages of the
drug industry were similarly competitive, there would probably be considerably
fewer complaints about drug prices.
But at the retail level the pressures of competition do not work as beneficially,
and more inefficiencies and impediments to the proper allocation of resources are
present. Some inefficiencies are probably traceable, at least in part, to certain of
the tactics of drug makers designed to maximize profits at the manufacturer's
level: the proliferation of branded products, combinations, dosage forms, etc.
which increase the druggists inventory costs; the liberal distribution of free
samples to physicians, which probably reduces average retailer drug turnover
rates; the economically more complex question of the "discriminatorily" low
prices made to large-volume non-profit buyers, which again probably reduces drug
turnover rates; and possibly certain aspects of policy on returns of unsold or
outdated drugs. But certain of these marketing policies are not without costs
to the drug companies as well as the druggists, and it is probably unwarranted to
impute any primary hostility on the part of the drug makers toward their re-
tailers. But certain inefficiencies have also been forced upon druggists by the ac-
tions of their own spokesmen and trade associations. The National Association of
Retail Druggists, for example, was certainly the prime mover in facilitating
adoption of the so-called "fair-trade" laws by state after state in the l93O~s.29
And yet although these laws prevented or greatly limited price competition for
trademarked drugs, by enhancing the unit profit margins on these items, more
dealers were induced to sell them, and the resulting increase in the number of
sellers reduced turnover and earnings on these products. And druggists supported
the passage of the Robinson-Patman Act, which prevents them from taking ad-
vantage of any possible cost savings available through obtaining supplies on com-
petitive bid, and in other ways prevents~ the lower-cost distributor from benefit-
tiug commensurately from the potential economies in his operations.
From this, one might conclude that not all drug price problems originate at
the manufacturer's level. The druggist's markup on the average prescription item
is no doubt higher than it might be, but~ then his unit costs are also higher than
they might be for many reasons, including those outlined above.
Clearly, the druggist has his problems. But the drug buyer has his problems,
too. These include: (1) inability to purchase a low-price generic drug if he has
been given a prescription for its brand name equivalent; (2) inability to shop
around for the lowest available price on a prescription, regardless of its manner
of specification, if the medication is needed quickly for treatment of an acute
condition, or if the prescription holder is otherwise suffering marked distress
pending the securing of his medication; (3) ignorance of the content of the
prescription, in many cases, which can simply mean inability to decipher the pre-
scriber's jargon, or lack of knowledge of the brand and/or generic name of the
drug-either of which may give rise to collateral inabilities, such as (a) inability
to determine whether or not a generic prescription was actually filled with a
brand name equivalent, and (b) inability to determine whether a generic pre-
scription actually filled as written was dispensed at the lowest generic price; (4)
buyer ignorance or docility such that he does not even realize that the prescrip-
tion form is his own property and does not have to be surrendered to the first
pharmacist to whom it is presented-who may be the one whose name is on the
prescription pad; (5) the frequently poor prospects for reasonable prices present
even for the unusual buyer who does shop around for a low price, due to the
tradition of hostility among most druggists toward price competition, and the
way in which this tradition is fostered and buttressed by the inhospitable atti-
tudes of pharmacy agencies toward price competition and the advertising of
prescription prices, by the state "fair trade" laws, by the Robinson-Patman Act,
and by still other influences.
How can greater efficiency be obtained in the retail distribution of drugs? The
characteristics of an efficiently competitive retail drug market can be broadly
outlined in a few sentences. All sellers should act quite independently with re-
spect to pricing policies; no formal or informal arrangements which would facil-
20 See, the example, ClaIr Wilcox, Public Policies Toward Business, Third EditIon, 1066.
pp. 707-710.
PAGENO="0374"
1924 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
itate uniformity of action on prices should exist. Prescriptions should be written
so as to facilitate the ability of drug buyers to stimulate price competition
among pharmacists. There should be no barriers to the dissemination among
buyers of information on the prescription drug prices of individual pharmacies.
Buyers should be free to seek out the lowest-price seller, both for the original
dispensing of a prescription and for refills. Entry into the retailing of drugs
should be free from any artificial barriers, legal or otherwise.
Under present market arrangements, there is no real incentive for the druggist
to stock low-priced generic drugs. First, there is little demand for them, physi-
cians' prescribing habits having been influenced as they have by industry efforts.
Second, if the usual two-third markup is added to invoice cost, the unit profit to
the pharmacist is proportionately smaller for the lower priced drug. Third, the
same logic applies to the wholesaler, so that even if a druggist wishes to stock
generic drugs, he may find it hard to obtain them.
It is obvious that the substitution of the "professional fee" approach in the
place of the uniform percentage markup would make the dispensing of generic
drugs relatively more attractive to druggists. But the application of compensa-
torily higher percentage markups to the lower-invoice-cost drugs would accom-
plish the same purpose. Optimal economic efficiency in the dispensing of drugs
would require that relative markups on individual items be determined by price
competition among sellers. The markup should be at the minimum rate above
cost which is consistent with the retailer's cost of distribution, including a com-
petitively-determined rate of return on an appropriate level of investment. If
genuine competition exists, the method by which the markup is arrived at will
be less important than the amount of the markup, since competition will re-
quire that this amount be substantially equal among competing sellers. The
notion of adopting a uniform professional fee for any and all prescriptions has
drawbacks. It lacks the necessary flexibility in the pricing of services which must
exist if price competition is to prevail. And the level of the fee is very important.
While I doubt if the size of the fee will be set at too low a level, setting it too
high will not insure druggist prosperity. Instead, the high unit profit margin on
each prescription will induce new entry into the industry. Many pharmacists
now among the ranks of the detailmen w-ill be encouraged to return to pharmacy.
As the number of sellers increases, average turnover declines to the point where
a balance is achieved between high unit profits and low turnover, and further
entry is finally discouraged because of low total profits. In comparing this situa-
tion with the low prices and high turnover which would prevail under price
competition, it is apparent that competition is to be preferred since prices are
lower and excess capacity and investment in underutilized resources is mini-
mized, while the profits on investment should be about the same in either case.
A few words should be devoted to contrasting druggist retailing of drugs with
other drug dispensing media. One can readily understand the unhappiness of
retail druggists who pay the full dealer list price when they read about the much
lower prices obtained by hospitals and government agencies in response to com-
petitive bids. Drug firms have tried to account for such price differences by ref-
erences to economies of large scale selling, and to promotionally low prices for
the sake of introducing their products to hospital physicians. But the price dif-
ferences are clearly too great to be accounted for merely as the equivalent of
quantity discounts. And the "promotionally low prices" argument can he dis-
missed as a rationalization since it is not characteristic of major drug firms to
be so negligent of sales promotion possibilities that the doctor would be likely
to overlook a drug if he did not have it on hand in a hospital. The basic reason
for the price differences is simply the fact that price competition can often he
kindled between brand and generic name drugs and even among major pro-
ducers of brand name drugs by means of the competitive bid approach. It has
been contended that sales to druggists at high prices "subsidize" the lower price
sales to hospitals and public agencies. If this is construed to imply that the
latter sales are actually made at a loss, it is no doubt an error. From all evi-
dences, drug production costs are very low. And a firm can always add to its
total profits by sel1in~ goods at special low prices, provided these prices are
above the out-of-pocket costs incurred on the sale, and further provided that
these transactions do not affect the prices received on other sales. To the extent
that firms have excess capacity, they will be more intensely motivated thus to
increase their rate of output and spread the overhead costs of total productive
PAGENO="0375"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1925
capacity over a larger volume of production. Differently considered, the seller
is simply charging as high a price as he can in two markets which can be sep-
arated from each other. Separation exists since hospitals do not resell to drug-
gists at higher prices. By doing so, the seller is engaging in price discrimination
(in an economic if not legal sense) since he can get much higher prices in one
market than in the other because of the different type of demand and supply
situations in the two markets. The supply situation differs not because the cost
of production differs, but because sellers can be induced to compete in price in
one market and not in the other. The difference in the demand situations de-
serves more attention.
The druggists' demand for drugs is derived from the demand of the individual
patient, which is of course very insensitive with regard to price because of the
reasons outlined above. Retail druggists can afford to pay high prices for drugs
since they can charge even higher prices to their customers without discourag-
ing sales. The demand on the part of hospitals and other public agencies is dif-
ferently constituted. These are non-profit agencies which operate within a gen-
eral budget, and while they do not concern themselves directly with selling in-
dividual items in accordance with a market-oriented schedule of charges, they
are concerned with lowering total operating costs and staying within budgets.
Their purchasing agents may also take pride in the successful exercise of pro-
fessional skills through economical buying. But although an interest in eco-
nomical purchasing is a necessary condition for obtaining low prices, it is not
sufficient unless rival sellers can actually be compelled to compete in terms of
price. Other things being equal, retail druggists might have less interest in
economical purchasing because they could more readily pass on their high costs
to their paying customers. But hospitals with charity patients and a whole host
of other exotic financial problems may be less able to forego possible savings
from economy in purchasing. Furthermore, hospitals may manufacture some of
their own drug needs if they cannot obtain reasonable bids, and this is another
factor which makes their demand for drugs less insensitive to price. With
demand more responsive to price, hospitals and public agencies can obtain lower
prices for many of their requirements, particularly where absence of patent
monopoly makes it possible for generic firms to compete. But where patent mono-
poly obtrudes, there may be no avenue through which prices may be lowered
for any buyer.
B. Factors influencing the demand for prescription drugs
It is chiefly the nature of demand for prescription drugs which makes the drug
industry an inappropriate vehicle for the unregulated exercise of market power
by sellers. Demand is so insensitive to prices charged that there is little ex-
aggeration in stating that prices have no relation whatsoever to costs. This is
contrary to the economics of almost all other industries, where price is broadly
determined by the relationship of demand and supply, and where supply is at
least directly conditioned or influenced by costs of production. In a purely com-
petitive industry the relationship is conceptually precise: market price is deter-
mined by the relationship between the supply price of a good, defined as the
cost of production of a given rate of output (including in costs the competitive
rate of return on necessary investment) and the demand price which the market
is willing to pay in order to buy a given volume of output. While supply price
depends upon production costs, demand price depends upon the consumer's need
for the product and on his income level. Wealthy and needy buyers will con-
stitute the highest-price or most price-insensitive segment of demand, while
buyers with low incomes and/or slight interest in the product will be potential
buyers only at very low prices. But since those who are willing to pay a very
high price for the product will also buy it at all lower prices, a reduction in
price will increase sales, and while it attracts new buyers into the market, it
also benefits buyers by cutting their cOsts of purchasing so that in a very real
sense they obtain a "free" increment of "use value" over and above the price
which they pay. Since price discrimination is impossible in pure competition, all
buyers pay the same price, regardless of their incomes or their relative need
for the good.
Under monopolistic market circumstances, even assuming that the structure
of demand and the costs of production are the same as in competition, the sellers
are in a better position to restrict output, limit sales only to those who are most
willing and most able to pay high prices, and in this sense charge "what the
PAGENO="0376"
1926 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
traffic will bear." Now as long as the act of purchase is voluntary, no one will
literally pay more than what the product is worth to him. In both monopolistic
and competitive markets, the market price is equal to the evaluation placed on
the good by the least interested buyer who actually does make a purchase. But
in pure competition, with no way of mutually restricting output, total produc-
tion will generally be large and prices low, while under monopolistic circum-
stances, output will be reduced in the interest of enhancing prices. This means
that the chances of the typical buyer obtaining a substantial free increment
of "use value" above price paid is much less under monopoly. But the major
difference between monopoly and competition is in the relationship between price
and cost of production. In pure competition the two are identical. But if monop-
oly power is present, price is also a function of the elasticity of demand and
will exceed production cost in direct proportion to the insensitivity of demand
to price.
Therefore when spokesmen for the drug interests argue that the consumer
is paying no more than the drug is worth to him, the obvious answer is: of course
not!! Unless force is used, no one can be induced to pay more for anything than
it is worth to him, no matter what sort of fleecing or price-gouging scheme, may
be employed. It is not surprising that a drug may be worth more to a sick person
than its cost of production, but this does not justify charging more, and in a
reasonably competitive market, prices would be much closer to cost than to
need-value.50
The above is a general treatment of the contrast between the relationship of
demand to prices in monopoly and in competition. In drugs, the argument applies
with even more force because of the extreme insensitivity of demand to price,
and the way in which this final consumer demand is mediated through the
physician. How should the needs of the sick be reflected in the market demand
for drugs? Ideally, the total potential market for a drug or group of related drugs
is measured by the total need for medication on the part of the individuals
afflicted by all the various disorders which are capable of being treated best by
the drug or group of drugs. Economically, the total effective amounts demanded
at the level of market price may fall short of total physical need in the case of
those with low incomes and no access to public care. But effective market demand
may also exceed ideal tOtal physical need to the extent that individuals not
suffering from those conditions for which the drug or drugs are of use may
nevertheless be treated with them. For any given drug it then follows that the
actual market is comprised by the total effective demand for medication on the
part of all individuals who can be induced to consult physicians, and who are
afflicted, or can make it seem convincing that they are afflicted, by those disorders
for which doctors may be inclined or persuaded to prescribe the particular
drug. The challenge to drug marketers then consists primarily in nersuading
physicians, hut also to some extent in spreading the good word to the general
public that Brand X can cure symptom Y.
And drug marketers doubtless earn their salaries. Ch~nres in the effective
demand (ie., prescriptions written and purchased as w-ritten) for individual
drugs are brought about by the familiar techniques of direct mail advertising,
journal advertising, the dividend of free samples, the financing of symposia,
the rental of exhibit space at conventions, and above all the insistence of the
ever-present detailman. Although advertising cannot yet manipulate the total
incidence of genuine disease, it can readily shift effective demand from one
drug to another. And advertising can in a sense actually create demand, even
for drugs~ Articles planted in newspapers or magazines may mention the name
of a drug alleged to be useful in treating certain conditions, and may thus bring
to the attention of more people who suffer (or imagine they suffer) from such
~ Examples where representatives of drug firm interests have defended high drug prices
as being no more than what the drug is "worth" to the buyer are legion. The most recent
exposure of this Subcommittee to this arrument is embodied in page 7 of Professor Simon
Whitney's written statement in behalf of PMA, where he states: "If a $5 prescription, or
6 of them, will keep a patient from losing a couple of days pay or spending a night
in a hospital. the price is reasonable." This can scarcely be taken seriously by an econo-
mist unless the price is also commensurate with the competitive supply cost of the drugs.
But no one who has presented this argument has so far been able to outdo Austin Smith.
who mused publicly during the Kefauver Hearings: "I wonder If any member of this
Subcommittee knows how much it costs to die? . . . death costs about $900 . .
Hence Smith has proved conclusively that any price less than $900 for a handful of
pills (a2 in this case) is a bargain since it must be worth at least $900 to the patient to
avoid the expense of demise. Hearings on Administered Prices, Part 19, p. 10615.
PAGENO="0377"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1927
conditions the fact that specific drug therapy might be purchased. Two of the
greatest drawbacks of sales promotion in drugs stem from these characteristics
of demand. First, doctors may be oversold on a drug which is then overpre-
scribed, often for minor conditions where it can do no good and may cause
mischief. Although antibiotics are usually cited in this context, other drugs
may also be overused and abuse is compounded when it is administered for
chronic, rather than acute, conditions.31 A second drawback, associated with the
first, is that patients themselves often insist on unnecessary drug medication.
III. APPROPRIATE DIRECTIONS FOR PUBLIC POLICY
The difficulties involved in reducing prescription drug prices to reasonable
levels can scarcely be exaggerated. There are four major parties to the typical
prescription drug transaction: the drug maker, the physician, the druggist, and
the patient. The only party with any direct interest in reducing drug prices is
the patient, and he has by far the least bargaining power. In fact, at the time
the transaction is made, his interest in low prices may usually be quite sub-
ordinate to his concern over his health. The major interests of the other parties
in drug prices lies in different directions.
The drug corporation, whether large or small, has to maximize profits to keep
the stockholders happy. The doctor's chief professional interest is in healing,
and if he can be made to believe that quality of one drug is superior to that
of another, he will be inclined to prescribe it regardless of price. His insensitivity
to price is naturally increased by the fact that he does not pay for the medica-
tion he prescribes. But the doctor himself is also a business man and may not
be entirely unconcerned with maximizing his own net income. It may hell) his
reputation if he is always among the first to prescribe all the new drugs, and
it may increase his prestige if he prescribes the higher priced drugs. And as an
independent businessman in an age of "organization men," he may even admire
the buccaneering tactics of the more flamboyant drug firms. Beyond that, he may
well own stocks in one or more drug firms.32 But whether he likes this or not,
he is vitally affected as a practicing physician by the policies of the AMA, which
since has in recent years received over half its revenues from the drug com-
panies. Hence under present institutional circumstances, the average doctor
has little direct interest in prescribin gthe lower price drugs and is contained
within a professional environment which may discourage such tendencies as
he may develop in that direction.
At the drug retailing level, druggists, like other dealers, resent price coinpeti-
tion because although it is a good servant to the consumer, it is a harsh master
to the producer. Druggists, however, are somewhat unusual among retailers in
that they have been more active and more successful than the others in securing
the passage of laws aimed at limiting price competition and protecting the inter-
ests of the existing group of competitors at the expense of the vigor of competi-
tion. This fact, in conjunction with the closed profession aspect of pharmacy
suggests a relatively poor prognosis for the rapid development of price compe-
tition at the druggist level. Even so, the awareness that it is desirable to restrict
competition, and even the presence of institutional arrangements which might
be used to implement this awareness, do not necessarily combine to produce the
hoped-for prosperity of the profession. As long as entry is reasonably possible,
and unit profit margins high, low turnover and excess capacity are likely to de-
velop and cancel out the advantages of high price levels.
A satisfactory solution to the problem of high drug prices must await the adop-
tion of a series of related reforms which will alter marketing and prescribing
~ Professor Mark Nickerson of the University of Manitoba Medical School reported
that the sales of adrenal steroids in the United States and Canada in 1960 was about
$250,000,000, and commented: ". . . personally I feel that I am being very liberal when I
say that fifty million of that was needed." Report Concerning the Manufacture, Distribu-
tion, and Sale of Drugs, Restrictive Trade Practices Commission, Department of Justice,
Ottawa, 1963.
~ Dr. Calvin Kunin of the University of Virginia School of Medicine submitted to this
Subcommittee for inclusion in the record an article reporting on a survey among medical
students, interns, and residents at the University of Virginia Medical Center. Six out of 73
owned stock in one or more drug firms, and 58 of the 67 non-stockholders stated their
belief that such stocks were good investments. As these students and residents continue
to pursue their careers and increase their affluence, it is likely that many of the favorably
disposed non-stockholders will buy stocks. Competitive Problems in the Drug Industry
Part 2, p. i34.
PAGENO="0378"
1928 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
practices and bring about price competition at both the manufacturing and retail
levels. The elements of such a solution would include:
(1) Abolition of drug product patents and compulsory licensing of drug
process patents at reasonable royalty rates.
(2) Outlawing of brand names for drugs, with the requirement that
drugs be identified and hence advertised and sold only by use of the generic
name(s) of the active ingredient(s) in conjunction with the company name
of the seller.
(3)Provision of FDA with sufficient authority, staff, and funds to permit it
to carry out a drug inspection program adequate not only to prevent the
sale of substandard drugs but also the plausible insinuation of the possibility
that substandard drugs might be on the market.
(4) Elimination of unnecessary barriers to entry of new drug firms into
the industry. If a drug has been cleared for marketing as the result of
adequate data compiled by one applicant, the same drug should be approved
for marketing by any firm capable of producing the identical drug. Unneces-
sarily burdensome requirements by way of conducting studies which merely
duplicate existing studies should not be imposed. In this regard, the sug-
gestion of FDA Commissioner Goddard before this Subcommittee that such
drug data submitted to FDA be made a part of the public record is an
excellent one.
(5) Provision of the medical profession with more accurate, systematic,
and objective drug data. If the price competition injected into the industry as
a result of reforms succeeds in reducing profit margins and eliminating the
detailman and if the medical profession does not then respond by subscribing
adequately to independent newsletters, the provision of a publicly sponsored
newsletter, similar to the Prescribers Journal in England may become
necessary.
(6) Exertion of every feasible effort to infuse more price competition into
drug retailing. Serious consideration should be given, at all relevant levels
of government, to the liberalizing of the requirements for operating drug-
stores and dispensing prescriptions, so that the further development of lower-
priced outlets such as discount pharmacies and drug mail order houses can
be stimulated.
(7) Recognition of the possibility that even the above reforms may not be
sufficient to reduce drug prices. If after a reasonable period of time, prices
have not declined sufficiently, consideration should be given to such addi-
tional reforms as (a) compulsory licensing of imports of patented drugs: (b)
complete abolition of drug patents; and (c) price control or public utility
regulation. The interrelationships of these recommendations may briefly be
summarized.
As has been ephasized by other witnesses, the absolute nature of the drug
patent privilege in this country is paralleled only in Panama and Belgium. All
other countries with drug patent laws provide either for the denial of drug
product patents, for compulsory licensing under certain circumstances, or for
both. The abolition of product patents and the making available of compulsory
licenses on patented drug production processes will increase the number of firms,
both large and small, making and selling each type of drug. This will stimulate
price competition, particularly since the small firms will naturally be selling at
low prices in order to counter the initial advantage of the highly advertised
brands. But the limitation of promotion to generic name plus drug company name
w-ill reduce the relative appeal of the major firm's drugs in the market, and this.
coupled with the cancellation of disparagment efforts by adequate FDA inspec-
tion. will make sales promotion efforts less differentially profitable. And since
production costs are low, price competition between large and small firms will
greatly reduce unit profit margins and in time will reduce the ability of major
firms to engage in sales promotion contests among themselves.
By such means. drug prices might in time be very substantially reduced. Drug
firm spokesmen claim that even if all profits were eliminated, prices would not
be cut by more than 15 or 20 per cent.33 But this overlooks the amounts spent on
~ Dr. Harold Burrows, for exn male. testified before this Committee to this effect:
"If Parke. Davis. for our 1966 year. had reduced our prices by 20.5 percent. we would not
have made any money . . . This is the maximum marrin that we are talking about
Compel it~re Problems in the Drug Industry, op. cit., Part 2, p. 812.
PAGENO="0379"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1929
1)1omotion, the costs of excess capacity, and the like. In an efficiently competitive
drug industry, sales outlays could probably be cut by something like 90 per cent,
and excess capacity costs would be substantially eliminated. It is premature to
speculate in regard to the possible magnitude of price reductions, but bringing
prices more in line with production costs would reduce them by more than half.
The major firms would have a certain amount of time in which to adjust to
changed circumstances, owing chiefly to the backlog of "goodwill" built up in
their behalf by intensive selling efforts. Although it has been argued many times
that the physician must be free to write brand name prescriptions because he
only trusts one particular company's product, the use of the brand name rather
obscures the identification of both the company and the drug. The use of generic
name plus the company name will still allow the doctor to specify the maker of
the drug. It is likely that much of the appeal of brand name prescribing lies in
its convenience; if it is made less convenient to the doctor by requiring that he
specify the company as well as the drug, it is probable that more purely generic
prescribing will result. If these factors are operative, plus an increased reliance
on purely generic drugs due to confidence in the adequacy of FDA inspection, then
in time the great initial advantage of the major firms in terms of their "ethical"
image will be dissipated and they will tend to lose their favored position. It is
not certain that the major firms' physical sales volunie will decline, although the
unit profit on such sales will certainly fall. But the impact should be sufficiently
gradual to allow major firms to diversify out of drugs and into other areas, such
as luxury goods, where the marketing methods which the managements have per-
fected at such cost can be applied in ways less mischievous to society. As to the
impact on research, there will always be a place in the industry for the firm
which engaged in basic, and in the long: run, truly productive research. Unfortu-
nately, there seem to be fewer of these, firms in the industry than one has been
led to believe. It is hard to attribute credit properly because to do so requires the
judicious deflation of the barrage of outrageous assertion surrounding each firm's
own public estimation of its research accomplishments, rather like trying to find
out which Hollywood spectacular really is the most Super-Colossal. One firm,
Merck, does rather stand out, if only because of the credentials of the Nobel
Prize winning "character witnesses" (the phrase is Dr. Louis Lasagna's) it has
been able to summon in its defense.
While a firm which is interested in truly fundamental research may not earn
extremely high returns on the funds it invests in such research, in the long run
it is probably more likely to survive. Without doubt, the greatest single obstacle
to the securing of drug patent reforms has been the argument that drug research
would suffer. The issue should be faced head on. Would a reduction in the expendi-
tures which the firms classify under the heading of research necessarily be detri-
mental to public health? This depends upon the types of research outlays which
are reduced, and whether or not any possible decline in productive private firm
drug research might be offset by increases in productive drug research undertaken
under other auspices. After drug law reforms, the level of drug firm research
expenditures may be reduced except in those firms where research is permitted
to be pursued in large part for its ownsake. But it is in the environment created
within such firms that research is likely to be most beneficial in the long run.
On the other hand, research of the "copyshop" type is likely to dwindle, but this
is a gain to the extent that such research typically produced less of genuine
social value than it consumes in terms of the alternative uses of the human
resources employed, even-or perhaps a particularly-in such an operation as
Pfizer, where molecular manipulation reportedly attained the status of a true
art.34
Even if total drug industry research spending does decline, professional person-
nel resources will probably be shifted into non-profit channels. It can be argued
that a ma3or diversion of pharmaceutical research endeavor from private firms to
public, university and foundation channels will in due time result in equally
major gains. Private firms appear to carry on relatively little fundamental re-
search, and more of this is needed at present to make applied research eventually
more productive. Non-profit research will also mean less waste of very scarce
human resources in imitative and duplicative programs, and in marketing-
oriented activities masguerading as research.
Silberman, "Drugs: The Pace~ Is Getting Gurious," Fortune, May 1960, pp.
PAGENO="0380"
1930 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The possibility remains, however, that drug manufacturers' prices might be
reduced by competition and yet these reductions might not be passed on very
efficiently by retailers. To increase the likelihood of competitive performance
on the part of druggists, the appropriate levels of government should review
the requirements for entry into drug retailing and compare them with the re-
quirements for satisfactory performance of drug merchandising services under
current conditions. Clearly, the pharmacist's function has changed from that
of actively compounding prescriptions to those of passive merchandising; many
of his skills have become obsolete as the result of the revolution in the drug
industry. Those interested in drug industry reform should join in urging the
maximum practicable liberalization of the traditional restrictions limiting entry
into drug retailing. This liberalization should be such as to constitute recogni-
tion that the traditional pharmacists' distinctive functions are being altered
away from professional competence in compounding and toward skillfulness
in merchandising. Such recognition would be likely to stimulate new entry by
those not traditionally opposed to price competition. In many lines of retail
trade, dealers were inertia-bound and distribution methods unprogressive until
price competition developed from such sources as chain stores, supermarkets,
and mail order houses.
In my own state of Texas, where we have never been burdened by resale price
maintenance laws, discount pharmacies and the drug departments of large dis-
count houses are doing a thriving business, and have not only given the price-
conscious drug buyer an alternative source of supply of both brand and generic
drugs, but have exerted downward pressure on the margins of traditional drug-
gists. Similar competition is rendered much more difficult in those states in
which fair trade laws are enforceable, but one of the advantages which drug
nomenclature reform should possess is in eliminating trademark names for
drugs and thus making them ineligible for the protection against price com-
petition which the fair-trade laws now allow. Some modifications of the other
laws limiting price competition, such as certain provisions of the Robinson-
Patman Act, should be also accomplished so as to allow druggists at least
the possibility of soliciting competitive bids from suppliers.
All of these reforms represent movements in the direction of creating a market
framework within which a freely competitive privately-owned industry can
efficiently operate. If these reforms do not prove sufficient to bring about the
desired result, two further measures would then become relevant. First, the
patent privilege for drugs might be further modified by allowing the importation
of patented drugs from abroad if the dealer could more cheaply purchase them
abroad than produce them domestically. This would fall short of patent abolition
in the sense that the U.S. patent holder would still be able to collect a reason-
able ad valorem royalty on sales by the importer. (Naturally, the quality of the
imported drugs would have to be acceptable, and the importer might perhaps
be required to pay for the costs of FDA inspection at the port of entry.) If this
did not prove sufficient drug patents could be completely abolished. And only if
drug price levels still prove impervious to reduction after all these reforms
should such measures as price control or comprehensive public utility regulation
of the drug industry be imposed. These latter remedies are likely to be less effi-
cient in operation because of the absence of a competitive market criterion for
prices of drugs under price control, and the general unsuitability of the drug
industry as the subject of regulation of the conventional public utility variety.
Hence these expedients should be regarded as last resorts, to be used only after
every effort to inject price competition has been exerted.
Dr. STEELE. Prof. Paul Cootner, professor of finance at MIT, made
a statement in which he discussed risk and rate of return in very gen-
eral terms and made little reference to the particular economic situa-
tion of the drug industry.
I. STATEMENT OF PROF. PAUL COOTNER
(A) In his presentation, Professor Cootner makes each of these
statements:
1. First, he admitted quite candidly, and I quote:
Now, I do not appear here as an expert on the drug industry, either with re-
gard to its pricing policy or the riskiness of its investments (p. 1).
PAGENO="0381"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1931
2. And again I quote:
* * * Neither Schumpeter nor I, nor indeed any responsible economist, will
argue that industrial abuses should not be corrected, when found (p. 4).
3. Finally, what I think is a very significant statement on page 9:
* * If one were to decide, by administrative action, to reduce the average
rate of return in a risky, competitive industry without at the same time reducing
the risks, we would find an immediate impact on that industry's investment
policy (p. 9).
(B) In general, I think that Professor Cootner's paper is a sound
pedagogical exercise which in commonsense terms conveys some of the
major subjective factors which influence an investor's frame of mind
in appraising investment prospects.
In fact, I find myself in agreement with virtually every statement
he makes. But I particularly agree with the three statements quoted
above. Statement No. 3 is sound economic theory. Statement No. 2 is
also true, in fact a truism if one defines responsibility in an economist
in terms of sensitivity to industrial abuses. And if one agrees with
statement No. 2, surely one must agree with statement No. 1, since
Professor Cootner neither suggests that he is aware of drug industry
abuses nor proposes corrections for them.
(C) The consequences of Professor Cootner's admitted lack of ex-
pertise on drug industry economics is that his paper, although educa-
tional in a general sense, is misleading because it suggests that the
industry is like any other industry in that high returns are likely to be
associated with high risk, that the drug industry's aggressiveness in
anticipating demand and "promptness in accepting innovation and
change" (p. 10) is socially benefièial, and that if industry risks are
reduced it will lead to a reduction in the net social productivity pre-
sumably due to these risky investments.
But things don't seem to work this way in the drug industry. Con-
sequently, the net effect of Cootner's paper is misleading because he
says all the favorable things about the productivity of risk-taking in
industry generally, without elucidating any of the drawbacks of the
policies which result in high profits and hence in allegedly high risks
in drugs.
I would like to emphasize that: 1. Statement No. 3 does not really
refer to drugs since it specifies "a risky, competitive industry" while
drugs are a profitable and rivalrous industry, not too much troubled
by true price competition. Among price-competitive industries, one
can expect the average profit levels of firms showing positive profits
in risky industries to exceed those shown by similar firms in safer
industries. But once we drop the assumption of price competition,
there is no such clear-cut relationship.
A pure monopolist of an absolute necessity could make enormous
profits in perpetuity and face no risks. But in the drug market there
are elements of both monopoly and rivalry. Patents confer monopoly
power with respect to a certain product and extremely inelastic
demand allows enormous unit profit margins. But these generous
margins will attract new entrants who will find it profitable to spend
vast sums in imitating the patented product legally. Once a rival com-
pound is concocted, how can the new drug take sales from the old?
Price competition is one route but a very costly one, and unless
there are a large number of rivals, it is not likely to break out. Instead,
PAGENO="0382"
1932 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
sales promotion outlays will be the vehicle of the rivalry, and the origi-
nally enormous profit margins are whittled down progressively by the
necessity of engaging in increasingly extravagant sales promotion
campaigns to counter those of one's rivals.
Senator NELSON. Is the ordinary consumer of drugs benefited in any
way by this rivalry?
Dr. STEELE. No, I would say the benefit is essentially negative in that
rivalry functions in the drug market as a substitute for price competi-
tion on the one hand and genuine research advances on the other.
When Dr. Console, former medical director of the Squibb Co., testified
before the Kefauver hearings, he said the drug industry is an un-
usually safe industry. Risks are low because if the research department
fails to make an advance, the advertising department can substitute its
expertise and make the drug appear to be an advance. So in this sense,
this rivalry prevents price competition by the ways which Professor
Schifrin has just described. And it also confuses the issue regarding
drug information and the substantive advantages, if any, made by
new drugs.
Senator NELSON. Thank you.
Dr. STEELE. In order to make the greatest profits per drug it is
usually necessary to be first in the market, otherwise the advertising
cost of wresting the market away from the first (and also heavily
advertised) drug is disproportionately great. Hence the motivation
to devise new drugs. But at the same time, the new drugs found by
others must be rapidly copied, so that the costs of research, both
primary and imitative, come to mount up. And the fact that every-
one is trying to copy and/or improve everyone else's drugs leads to
an overly rapid rate of product obsolescence and an artificially in-
duced "risk" of short commercial life for the average product.
Since many doctors have testified that there is generally no net
surplus of advantages over disadvantages for the manipulated mole-
cules, rapid changes in market shares betray motion but not progress.
Thus sales promotion and product competition substitute for price
competition, and unit profit margins decline not through price reduc-
tions but through cost increases. It might even be contended that this
route to profit erosion roughly equates risks and rates of return, such
that even though the risks are self-created by the seller's own choice
of rivalry tactics, they are real risks and only enough will be invested
in sales promotion and patent bypassing research to keep the rate of
return on investment from dropping below the minimum satisfactory
rate relative to the artificial risks built into the market.
One might agree with this analysis and still contend that it would
be socially beneficial to alter the market structure so that price com-
petition would be forced upon the sellers and lowered profit margins
would become insufficient to support constant devising of new drugs
and their rapid copying. Hence the "risks" would decline in direct
proportion to the decline in rate of return.
But it is also possible to dissent from this analysis. First, inherent
commercial risks in the drug industry are probably lower than in most
industries because of the depression-proof character of the industry.
Drugs are needed in fair times and foul, and a sick man will buy as
many drugs as lie needs regardless of his income, right up to the limit
of his ability to ffnance current purchases. Second, drug firms may
have come to realize in recent years that the market is practically satu-
PAGENO="0383"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1933
rated with advertising appeals, such that although it is necessary for
each firm to maintain current sales promotion levels in order to offset
the efforts of others, incremental advertising expenditures will have
no real impact; therefore, profit rates remain above risk-determined
levels because additional spending on sales promotion, and possibly
even on production competition, simply doesn't lead to even the pros-
pect of further differential gains. In such a case, the only remaining
tactic to bring profit rates down to equilibrium levels consistent with
risk (regardless of the nature of the risk, whether inherent or artifi-
cial) is price competition, and drug industry repugnance to this force,
plus the fact that the number of sellers in each market; that is, effective
competitors is still too small to compel price competition, generally re-
suits in a sort of high-profits truce between companies which refuse to
engage in the price competition which would effectively reduce profit
levels to equilibrium relationships in comparison with true risk. (How-
ever, Mr. George Squibb's testimony with regard to the frequency of
special deals to distributors indicates that although list price com-
petition may be unknown, there may be a substantial amount of hidden
competition in terms of such special inducements to dealers. And I sus-
pect that excess capacity in production processes is probably a major
reason for this.)
Mr. GORDON. Are you aware that one of the risks attributed to
the drug industry is the possibility that the industry may become
competitive?
Dr. STEELE. I think this is true. I think Mr. Squibb also alluded to
this risk in his statement.
Senator NELSON. Well, in economic terms, is competition considered
a risk?
Dr. STEELE. Well, I think it is really a pleasant risk. Competition
would not be a risk in itself. In comparing two equally competitive in-
dustries, one would expect that the~rate of return would be greater in
the inherently more risky industry.~ But if an industry, let us say, had
been monopolized, and there is a likelihood that for one reason or
another, competition will break out, then investors will see this as a
risk in the sense of the institutional frame of risk or something like
that, but it won't be something that is inherently in the market struc-
ture of ordinarily competing sellers.
(2) Professor Cootner does formally recognize the possibility that
the firms might react to profit reduction measures by reducing the risk-
iness of the ventures in which they engage. But the relevance to the
drug industry is limited since the example mentioned assumes that the
risk taking activities hypotheticali curtailed are socially productive,
and do not consider the possibility that these risk taking activities are
also creating.
(3) Cootner makes two statements which are not quite compatible.
On page 8 he states: "This basic conclusion is that as risk rises so does
the required rate of return." But the "required" rate is a subjective
expectation; it is basically an ex ante phenomenon. However, on page
4, Cootner states that "one should not be surprised to find large
average profits in risky enterprises * * ~ which is not quite the
same thing. One would expect to find large average profits for com-
panies making profits, but if true risk is significant, also large average
PAGENO="0384"
1934 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
losses for companies failing to break even. I would say petroleum
exploration would be a good example of this. If you take the com-
panies which only show profits and average their returns, the indus-
try looks better than if you take companies which incur losses. In
such an industry, if you take their returns and add them in, it is not
clear that the average ex post rate of return on all investment in a
risky industry should always necessarily be higher than in a safe
industry where losses are very rare. This is, of course, one of the
maj or weaknesses of statistical measurement of the risk rate of return
hypothesis: the data are usually available only for the largest firms in
an industry, which because of their very size face relatively less risk
than small firms (since riskiness is basically a property of individual
investment projects, and not of the firm itself, which is collection of
many projects) the small ventures which fail, or do not get included
in the blue ribbon listing of firms, may so regularly make losses that
total industry rate of return is significantly depressed.
II. STATEMENT OF PROF. SIMON WHITNEY
(A) While I am in general agreement with almost all of Professor
Cootner's statements, I find myself in strong disagreement with the
majority of Professor Whitney's. Time does not permit the criticism
of more than a limited number of these statements.
(1) Whitney's statement as a whole is based on the implicit assump-
tion that if private drug firms don't do drug research, it just won't
be done. But great sums of public money are now spent on research
in health and medicine. Even if drug industry reforms do reduce
private drug research, there is reason to believe that the researchers
could be more productively employed by universities, foundations,
and government agencies in doing much the same type of research.
I would like to comment now on the statement of Prof. Simon
Whitney and preface my comment by saying that I am in general
agreement with almost all of Professor Cootner's statements. The
only thing wrong with them is they are rather irrelevant. But I find
myself in strong disagreement-
Mr. GORDON. You say you are in general agreement. Yesterday,
Dr. Mueller of the FTC, stated:
Thus Conrad and Plotkin unwittingly have made a case for the inference
that a substantial part of the high profits earned by drug companies are really
due to advertising-and-promotion-created barriers to entry, rather than risk.
This, of course, coincides with the conclusion of nearly every economist who
has studied the drug industry.
Dr. STEELE. I would agree with that. I would regard myself as one
of those who agree with the statement.
But the Cootner statement is different from the Conrad and Plotkin
statement. The Cootner statement says, really, very little about the
drug industry. just discusses the problems of risk and so on.
Mr. GoRDoN. I see. He doesn't discuss the drug industry at all.
Dr. STEELE. Not really.
Whitney's statement as a whole-
Senator NELSON. Doctor, has anyone done a study to find out how
much research has been done by Government in the health field-by
PAGENO="0385"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1935
NIH and other agencies-that is really directly related to the drug
field?
WTe know what the total research budgets are. We have received
some material from time to time indicating expenditures by Federal
agencies of hundreds of thousands or millions of dollars, part of which
was spent in research on a particular drug or testing of a drug, but
the amount spent on the drug alone was not clearly isolated so you
could specifically identify it.
Are you aware of any studies, reports, by anyone who has attempted
to identify the amount of money that goes into research that you would
say is related rather directly to the drug industry?
Or to put it another way, research of the kind that the drug indus-
try does?
I am not talking about molecular manipulation or anything like
that.
Are there any studies of that kind?
Dr. STEELE. Unfortunately, I do not know of any. All the ones I
know of share the defect you have just described. They refer to pro-
ject research. Only a small proportion of the project may be devoted
to evaluating the particular drug.
Senator NELSON. We intend at some stage to have testimony from
the various Government agencies on this but I though perhaps some-
body had done a study and made some breakout of the funds spent in
this area. You are not aware or any such study?
Dr. STEELE. No, I am afraid I am not.
On page 3, Professor Whitney states that:
"* * * $3 of additional stockholders money per dollar of net worth in 1950
went into drug manufacturing for every $1 in all manufacturing * *
Stated in this manner, it sounds as if new stock issues were sold to
raise the three additional dollars.
Actually, as Whitney explains in a footnote (without appearing to
appreciate the significance of this factor), most of the increase came
about from reinvestment of earnings. Professor Whitney is trying to
have the best of two worlds: the Adam Smith world of atomistic com-
petition, a.nd the modern corporate finance world of noncompetitively
high prices and profits, and discretionary plowbacks of retained earn-
ings by management. On pages 2 and 3 Whitney implies that the first
function of profit is the allocation of investment, which is true in the
classical sense. But the classical theory of consumer sovereignty ap-
plied also to investors; in principle, all profits in a corporation, under
this theory, should be paid out to shareholders, and the decision as to
whether to reinvest in the same enterprise or in others should be made
anew upon the receipt of each dividend. This would insure a more im-
partial appraisal of alternative uses of dividends; whether for rein-
vestment or for consumption spending, than does the modern system
of having managements, rather than owners, decide upon the reinvest-
ment of about half of total earnings. (The obvious tax disadvantages
of higher dividends to high-income stockholders and underwriting
costs for new securities have been instrumental in transferring discre-
tion over total profits from stockholders to managers, but this does
not alter the basic principle involved.)
81-280-68-pt. 5-25
PAGENO="0386"
1936 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
And with price competition, relative profit margins in different
industries would reflect the scarcity of supply relative to demand and
hence indicate the relative need for expansion of investment and pro-
ductive capacity.
But if monopoly power exists, output may be restricted and prices
and profits can be at high levels although the monopolist may actually
have excess capacity and no need to expand his facilities.
Furthermore, discretion over prices may result in prices which are
too high relative to costs and in resource misallocation relative to the
outcome which would have prevailed in a purely competitive market.
If excessive profits, made by overcharging buyers, are plowed back
into the industry by bypassing the investor's discretionary power
over all profits, the investor benefits from what is tantamount. to a
capital levy on the consumer. The drug buyer thus contributes much
of the capital-the great majority of additions to capital investment-
on the basis of which the drug stockholders now expect high earnings
because of the "risk" to which "their" capital is subjected. Further-
more, Whitney takes it for granted that this increase in capital value
resulted in at least commensurate social gains; on page 4 he identifies
these gains with increases in drug sales and with research and develop-
inent expenditures.
But in neither instance is it necessarily the case that the true value
of drugs or drug research is measured by dollars spent. The individu-
als most qualified to judge these matters are physicians and medical
educators, and their judgments as recorded in public hearings on drugs
have not been such as to encourage those who wish to equate dollars
spent and value received in drugs.
(3) "Many hundreds of new drugs, as documented by earlier PMA
witnesses, resulted from this profit-motivated research" (p. 4). This
is misleading if the reader naively interprets this to mean genuinely
new chemical entit~es. Non-PMA witnesses supply different "docu-
mentation." Dr. Martin Cherkasky has previously stated before this
committee that the industry's claim in the early 1960's to have PI~-
duceci over 400 new drugs required more than 90 percent deflation.
He said:
On examination, only 29 of those were really new contributions. The rest of
them were qimmicks, new dosages, new combinations that really hadn't much
value ("Competitive Problems in the Drug Industry," pt. 2, p. GiG).
And during fiscal 1967, FDA Commissioner Goddard stated that of
the 83 New Drug Applications approved, 62 were for "what have been
called `me too's' or molecular manipulation." (Ibid., pt. 2, pp. 757-
758.)
(4) On page 6, Whitney suggests that the lower prices charged by
generic name firms reflect the absence of research, quality control, and
original distribution costs on their part. This sort of approach is a
favorite with pharmaceutical manufacturers associations. The Cana-
dmn P~iAC made similar charges against Canadian generic name
firms during their drug hearings for 1966-67, and, not unsurprisingly,
it developed that small generic firms also incurred costs for quality con-
trol and research.
The chief cost savings of the generic firm is in the area of sales
promotion, which was somehow overlooked by Whitney-unless
PAGENO="0387"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1937
original distribution" is taken to be a very inclusive euphemism
indeed.
In any event, the major firms are simply voicing their indignation
about having to be faced with price competition even if it is only
effective in the minute interstices of their total market.
I would refer again to Professor Schifrin's diagram, the roughly 5
percent of total sales by dollar volume, which would be a relatively
higher percentage of the physical volume of drugs sold because of
lower average cost.
What these firms do not realize is that any argument against price
competition is an argument for price, control. It is a long-established
principle of public policy that where competition will not work,
regulation must be substituted. "Do the lower prices charged by gen-
eric name manufacturers reflect willingness to operate on lower profit
margins?" (p. 6). By no means. Price competition, such as prevails
among generic firms, compels produëers to be efficient. and to accept
the low-profit margins determined by competition. But the major
companies would ordinarily have little direct experience wit.h this.
(5) "What does it mean, after all, to say that. a. price is `too high' ?"
(p. 7). Any trained economist would answer that it is too high rela-
tive to production cost, since this is the standard of efficient pricing
in a competitive market. But Whitney invokes a sort of value-of-
service standard in connection with prescription drug prices, implying
that if the price does not exceed what it is worth to the buyer, then
it is reasonable. But this is just the monopolistic practice of charging
what the traffic will bear; the full value to the consumer is the absolute
minimum price which can be exacted from him, and the major benefit
of price competition to the consumer is that it allows him to obtain
the goods at a price related to production cost as well as to demand,
such that the typical buyer pays less for the goods than its actual
maximum value to him, and thus enjoys some measure of what
economists call "consumer surplus."
Under monopoly pricing, the average value of this surplus will be
reduced, and under systematic price discrimination-the exercise of
which requires considerable monopoly power-it may disappear'
entirely.
(6) "You have heard of large economies made by hospitals through
purchases of drugs by generic name. Were all purchasers to do the
same, many research-based companies would be put into serious.
straits" (p. 8). The implication here is that these companies would
cease to do research. But since they spend only about 6 percent of'
their sales dollar on research and 25 percent on sales promotion, it
would seem that much greater scope for economies lie in the market-
ing budget.
But even if research outlays were cut, much of this research could
probably be more efficiently accomplished under nonprofit auspices,.
as mentioned,under point 1. And a more equitable distribution of the
cost of drug research might also be accomplished if more of it were
publicly financed. Although drug spokesmen have defended the price
of drugs by reference to research costs on innumerable occasions, I
have never once heard of them raising the question of the propriety
PAGENO="0388"
1938 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
of having these costs financed entirely by the sick. And this is a ques-
tion which deserves much consideration.
Senator NELSON. You suggest that the more equitable distribution
of the cost of drug research might also be accomplished if more were
cofinanced. Would you elaborate on that? How and in what way?
Dr. STEELE. Well, the argument is that drug research, particularly
fundamental research in drugs is ultimately a philanthropic activity
which benefits everybody. You might say in a narrow sense, if a
person is sick, he benefits and benefits exclusively from taking a drug
which he has prescribed for him and which he pays for. However.
the narrow view is overly narrow, in that availability of the drug
protects the health of everyone. To the extent that individuals who
have diseases prevent the disease from continuing and affecting others,
the society as a whole will gain. The availability of good medicatioi~
increases the health standards, the health potential, of the economy as
a whole whether or not the medication has to be used.
It is a potential benefit.
If the sick benefit entirely, then the people who by virtue of the
sick's having obtained medication do not themselves become affected.
So they benefit in a sense almost unjustifiably at the expense of those
who have had to pay for medication and because of their sickness,
have been in perhaps a worse position than others to pay for the medi-
cation and hence for research.
Senator NELSON. What you are saying is that since it is a benefit to
the whole public, the individual and the public who never becomes ill,
it is in the nature of an insurance benefit. You may own insurance,
never have an accident, never become ill, but it is a protection if you
do.
Is that what you are saying?
Dr. STEELE. Yes, this is it.
Senator NELSON. How would you finance more of this research?
In what way? WTould you contract with companies to do it?
Dr. STEELE. You could do that provided that the benefits to society
were not disproportionately appropriated by the companies involved.
I think that primarily, basic research is an activity which is not done
too efficiently or effectively by profit-oriented firms. Obviously, you
cannot tell what is going to happen when you start on a basic research
program. You want to increase your knowledge. You may end up bene-
fitting your competitor rather than yourself.
Research may be a fruitless activity for years, and the loss a com-
pany may incur on this basic research may deter it from doing further
research. So I think research either under public foundations, uni-
versities and so on, or public financing of private research, with ap-
propriate restrictions on patent monopoly could be carried out more
effectively.
Senator NELSON. Thank you.
Mr. GORDON. Is not this the principle of roadbuilding, of building
highways? The Government pays for highways throughout the coun-
try and in States as well. You do not expect the users of the roads to
pay for them. Do we not consider that all of the society benefits by it
through the opening of communication and so forth? Is that not the
same idea?
PAGENO="0389"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1939
Dr. STEELE. There is a very good analogy. But I think that in the
case of drugs, the argument is even stronger, because a person who
buys a car does so voluntarily. He engages in the use of roads in order
to further his own gains. But a person who buys a drug is not doing
something voluntarily in the sense that he has told society to use its
resources specifically to cure his disease. Instead, he is reacting to a
situation which has been forced upon him against his will.
What Senator Nelson said, using the insurance in case of a risk,
rather than using a voluntary activity.
The next point is a rather technical one. Professor Whitney is in
error when he multiplies profits per dollar of sales in drugs by the
ratio of profits per dollar of equity capital in drugs to profits per dol-
lar of equity capital in all manufacturing and claims that this brings
profits per dollar of equity capital in drugs down to a level equal to
that earned in all manufacturing.
The fallacy is that he is assuming a capital turnover ratio of unity
for the drug industry. Profits on equity capital are equal to the prod-
uct of two terms: Profits per dollor of sales and sales per dollar of
capital during some time period suèh as a year. In manufacturing,
the product of these two ratios is said by Whitney to be only 61 per-
cent as high as for drugs. But if this ratio is applied to one of the two
terms, rather than the product of both of them, in the drug industry,
the implicit assumption is that the value of the other term is unity.
But Whitney cites the FTC-SEC survey as showing that the drug
industry earned 19.7 percent on sales but 22.2 percent on investment,
hence the implicit capital turnover ratio is about 11/s. Hence prices
could be reduced by 8.8 cents in the dollar, instead of 7.7.
(8) "Someone will have to make up the corporate income tax pay-
ments lost as a result of declining profits of drug manufacturers"
(p. 9). In the first place, the Government would save 8.8 percent on
all its drug purchases at the manufacturer's level; this would go a
good way toward offsetting the lower income tax receipts from drug
firms. And to the extent that the Government indirectly finances other
drug purchases, the cost savings would be still greater. But the argu-
ment is a peculiar one. There is the suggestion that one should not
lower excessive profits because the profit maker is liable for Federal
income taxes on the full amount of the profits. But the same is true
for incomes from fraud, embezzlement, extortion, and other illegal
activities, and yet the possible adverse effect on income tax liabilities
does not deter policymakers from trying to eliminate such activities.
The next point I think is certainly i~nportant. Whitney states:
Certain trends are alarming. New chemical entities marketed per $100 million
of R. & D. expenditures, for 1959 through 1966, were 32, 22, 21, 11, 6, 6, 7 and 3,
respectively, * There may be no real recovery if the profits from research
are threatened (pp. 9-10).
This is not alarming if it simply means that tighter FDA regula-
tions have reduced the total flow of ~mew drugs by eliminating the in-
efficacious ones, which FDA did not have authority to do until 1962.
But the fact that the decline wa~ steady even before 1962 probably
indicates, in addition, that diminishing returns to the molecular engi-
neering techniques of applied research are being increasingly felt as
that method is more and more intensively and extensively applied to
PAGENO="0390"
1940 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
a fixed universe of discovery possibilities. What is needed is money
spent on fundamental research, to make the major breakthroughs for
the applied researchers in the drug firm subsequently to exploit. Since
profits rose throughout the period of declining discoveries, there is no
reason for Whitney to suggest a linkage between recovery in discovery
rate and the lifting of the threatening atmosphere regarding profits
from research. Actually, profits have little to do with it since the drug
firms do not find it attractive to invest enough funds in the fundamen-
tal researches which constitute the only path to eventually reestablish-
ing the productivity of applied research.
(10) On page 10, Whitney notes that drugs as a percent of total
medical care costs have been declining. This is due not to declining
drug costs but chiefly to the phenomenal increase in hospital costs. And
most of this increase is owing to the inclusion of hospital charges in
health insurance programs without adequately disciplining hospital
costs and charges.
Health insurance programs increase effective demand for hospital
services without at the same time taking any steps to increase their
supply. The natural result is an increase in charges. And if drug cover-
age increases under health insurance plans, including medicare, the
same thing is likely to happen in drugs. So the fact that drug costs
have increased less rapidly than hospital costs does not mean that drug
pricing has been characterized by self-restraint but simply that the
coverage of drug costs under health insurance programs has been less
comprehensive than their coverage of hospital costs. As pressures build
up to increase drug coverage, steps must be taken to discipline drug
prices and increase the supplies available at competitive prices.
Senator NELSON. Have some of the declines in drug prices been due
to the expiration of patents?
Dr. STEELE. Undoubtedly, this is true.
Senator NELSON. That is, on a particular drug, as the patent expires,
there is a tendency for the price to consumers to go down?
Dr. STEELE. This is true, yes, especially in Professor Schifrin's mar-
ket "B." Some of the major drug firms may sell at lower prices. The
patent holder himself may not reduce his own price, or may reduce it
only more slowly.
Next I would like to make some comments on the Arthur D. Little
report.1
III. COMMENTS ON THE A. D. LITTLE REPORT. "RISK AND RETURN IN AMER-
ICAN INDUSTRY"
A. Criticisms of the basic orientation of the study
(1) The study is designed to test the hypothesis that there is a posi-
tive relationship between risk and rate of return. But the definition of
risk simply as average intercompany variance within an industry is
rather arbitrary, even though this has frequently been used as a meas-
ure of risk. "Average Intraindustry Variance and Return in American
Industry" would have been a more accurate title for this study.
I am indebted to my colleague Prof. James Willis for suggestions and comments In
regard to this presentation.
PAGENO="0391"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1941
(2) It is misleading to include the drug industry in a comparative
study where a large number of industries are analyzed on the basis of
measured average rates of return and variance, and conclusions are
drawn as to the relative risks encountered by the groups of industries
as a collection of observations. This is because the markets of the drug
industry are so protected from price competition by patents, trade-
marks, extremely inelastic demand, brand name prescribing, the
mediation of effective demand through a financially irresponsible pur-
chasing agent, etc., that there is no other industry in which the seller
has a comparably great power to rise superior to price competition.
The drug industry is truly in a category by itself in this respect.
Ideally, one should hold constant the degree of competition among in-
dustries in making a study of this sort. The relationship between risk
and rate of return should be most clear cut if all industries studied
were purely competitive. The value of the study is reduced to the ex-
tent that industries far removed from price competition are included.
After all, a pure monopolist could conceivably earn enormous returns
with no risk.
(3) It is difficult to understand why the investor is conceived of as
measuring risk in terms of variance alone, without regard to the aver-
age rate of return in the industry. It seems likely that an investor will
regard two industries as having different risks, if one has an average
rate of return of 25 percent and the other has an average return of 5
percent, even though each has a variance of 50, for example.
B. Specific criticisms
(1)~ On page 11, the authors state: "It is within the individual cor-
poration that the balance between expected returns and expected risks
is struck." This is contrary to the approach taken in ecomonic theory
and in financial analysis, where the basic unit is the individual invest-
n-ient project, not the firm as a whole. The firm is a collection of pro-
jects, some of which may be very risky while some are not.
(2) On page 14, the authors admit that they were unable to allow
for possible biases resulting from the fact that their data source limited
them to the larger and more successful firms in each industry. This
might be the source of considerable difficulty. It is generally known
that larger firms make higher profits than smaller firms, both as a
general rule and within the typical industry. Hence if an industry
composed of 50 medium sized firms is compared with one made up of
25 large firms and 25 small firms, the variance and hence the riskiness
of the latter is likely to be greater, although it is debatable if this is in
fact the case. And limitation of data to the larger firms tend to under-
estimate the risks faced in a given industry. The fact that the firm is a
collection of investment projects provides an additional reason why the
larger firm may be able to make a higher rate of return than the small
firm. Not only is it able to take advantage of economies of large-scale
production, transportation, distribution, advertising, and finance, but
it can also adopt a broader scale risk diversification program which
makes it less vulnerable to the possibly unfavorable outcomes of in-
clividual projects. Hence the industries containing the largest firms
may actually face the smallest risks, and yet their rate of return on
investment will be relatively higher.
PAGENO="0392"
1942 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(3) Another reason why it is unfortunate that data could not be
obtained on the smaller and less successful firms is that which is given
as comment C-3 on Professor Cootner's statement.
(4) The industry rate of return is defined simply as the average
rate over the entire period. But if there has been a significant upward
or downward trend in an industry during the period, investors would
react to the trend as well as the average. For example, the average
might be 15 percent in two industries, but if one had declined from 20
to 10 percent, while the other had increased from 10 to 20 percent dur-
ing the period, investors would be likely to regard the former indus-
try as a much riskier investment even if both average rates of return
and variances were the same.
(5) On page 16 the author states:
A company may be receiving monopoly returns (returns higher than justified
by risk) on its book assets (i.e., monopoly real returns) while the holder of its
equity instruments would receive a "normal" return if the monopoly profits were
capitalized when the stock was issued.
In model No. I, we note that the drug industry's rate of return is
significantly higher than the value given by the regression line for its
variance. The computed value is about 14.75 percent, while the actual
value is 17.52 percent, or about 20-percent higher.
Since this is the book value regression, one might conclude that some
degree of monopoly returns are present, since the returns made are
higher than those which would be justified by the risk versus rate of
return relationship embodied in the regression line. (Here, a statement
from page 5 of Professor Cootner's statement comes to mind: "If
abuses are found, one must take care to eliminate only excessive, not
not necessary returns on investment * * ~." Hence one might conclude
tentatively that about a 20-percent reduction in drug industry returns
would still leave returns commensurate with the risks as measured by
Conrad and Plotkin.)
In other words, as far as the investor, the stockholder is concerned,
the rate of return he is receiving is just about the average on the basis
of Conrad's-that is, the A.D. Little study's-regression line.
And at a glance at figure II, shows that the market value basis is
slightly lower than the value indicated by the regression line. 1-lence
one might interpret the two regressions as showing the presence of sub~
stantial monopoly returns on book value consistent with only competi-
tive returns to stockholders due to the competitive nature of the capital
market in distinction to the monopolistic nature of the market in
which prescription drugs are bought and sold.
(6) The statement from page 16 quoted just above is not entirely
consistent with the listing on page 11 of "the various phenomenon
that might contribute to interindustry differences in basic riskiness."
On page 16, monopoly returns are defined to consist of those which
are higher than justified by risk. But on page 11, certain of the ele-
ments contributing to monopoly power-or conversely, to its absence-
are listed as factors contributing to basic riskiness. These include, for
example, differences in ease of entry. in elasticity of demand. in i'~~
flexibility, in exposure to foreign competition. and finally, differences
in competition among existing, prospective, or potei~tial new products.
it would appear that there is serious confusion between the monopoly
PAGENO="0393"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1943
elements which contribute to basic risk and those identical elements
which, by resulting in monopoly returns, provide earnings greater
than those justified by risk. The confusion can only be resolved by
assuming that all industries being studied are purely competitive. If
this assumption is contrary to fact, then it is not sufficient simply to
speak vaguely about factors influencing risk, and about monopoly re-
turns, as if they were two different things. Instead one must devise a
better theory to relate rate of return to risk and monopoly power more
explicitly and satisfactorily.
TV. COMMENTS ON THE STATEMENT OF GORDON CONRAD, AND ON THE A. D.
LITTLE STUDY, "TRENDS IN MARKET SHARE FOR ETHICAL PHARMACEUTI-
CAL PRODUGTS'~
A. General comments
(1) On page 1 of his statement, Conrad states that the Little study
shows "a significant degree of interproduct competition." But we do
not know how significant the data are for drugs until we have data
from other industries with which to compare them. Risk, being subjec-
tive, is a relative matter and until it ~ shown that other industries have
less interproduct competition, the Little study will remain in-
conclusive.
(2) Even more important, this so-called interproduct "competition"
cannot be beneficial to the consumer unless it results either in price
competition or in genuine improvements in the quality of the prod-
ucts. Otherwise we have change, and perhaps wasteful rivalry, but no
progress. Yet on the very first page of the Little report we read: "This
report does not explain the reasons, for competitive changes over the
time period since this would require revealing product names and
company strategy." This effectively prevents the study from making
any real contribution to answering the real question: is the eco-
nomic performance of the drug industry beneficial to the economy and
the consumer? 1
(3) Conrad's statement concludes:
These results illustrate one aspect of the potentially high risks facing pharma-
ceutical manufacturers, that of the genuine uncertainties as to the length of time
any one product can be expected to contribute to the company's profits.
1 This same point was debated during Kefauver's hearings on the drug Industry anti-
trust bill in 1961. At that time Professor Markham placed great emphasis on the amount
of turnover or change in the rank order of market shares by products In a particular
therapy category. But to assess the degree of workability of competition evidenced by
such turnover, one should determine how it was brought about: by price competition?
by product improvement? or by less beneficial means? But Markham seemed to believe
that turnover was a good thing for its own sake, and at least at the time of his appearance
had not analyzed its causes. When asked just what was the value to the consumer of turn-
over if there were no price competition, he responded: "~ * * I would still prefer, even
if the prices are the same, and this I know nothing about, that the firms that are trying
to serve my needs as a consumer feel that somehow or other they, through product
innovation, or by whatever means-the development of new products, new processes,
new drugs-are getting my consumer's outlay In terms of competitive activity" (pp.
2105-2106). During questioning, Markham conceded that he had not examined the facts
as to whether or not any drug firm had experienced a change in relative sales rank
because of price competition (p. 2096). Markham agreed that price competition is of
paramount importance to the consumer, but concluded his contribution to the hearings
with this statement: "I have not made any careful study of the workability of com-
ietition in the ethical drug industry. I was examining primarily these particular issues
that seemed to be important" (p. 2111). This suggests that to Markham the issue of
workability of competition was not important-but since he Is known as one of the
foremost students of the problem of workability of competition, the statement remains
an anomaly.
PAGENO="0394"
1944 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Again, it should be emphasized that this is a risk to which the in-
dustry intensively contributes, and in fact fosters, by its policy of
imitative research and product development programs, sales promo-
tion strategies, and the like. And very little of this is likely to be
socially productive.
B. Specific comments
(1) The summary section of the report bristles with unsupported
"plugs" for the drug industry. On page 4 it is stated that "the devel-
opment of effective products with fewer and less severe side effects"
is one of the most outstanding features of the market. This is a more
sanguine opinion than was stated by many doctors during the Kefauver
hearings, particularly those whose statements are contained in part
18 of those published hearings. L. Meyler's book, "Side Effects of
Drugs," is highly educational in regard to the general failure of
later modified drugs to have side effects much different from the earlier
drugs.
(2) The summary section also contends that combination products
are improvements. Again, there is much medical opinion to the con-
trary. The U.S. Pharmacopeia does not list combination drugs; the
National Formulary has only a few. Dr. Maxwell Finland stated
during the Kefauver hearings that combination drugs lacked flexi-
bility and compounded the problems of dosage and toxicity.
The summary refers to "combination products in which the in-
gredients provide synergistic effects." Dr. Finland referred to syner-
gism claims as "incorrect and misleading," observing that such activity
"is a highly specialized property related to individual strains of bac-
teria and is recognizable only after special tests. Thus, so-called syner-
gistic drug combinations can only be tailormade to an individual
strain of bacterium after such tests" (p. 13928).
(3) It is also contended that there are "obvious economies" in pro-
ducing and distributing combination products, but no evidence of
this is given. But even if economies are realized~ of what value are
they to the consumer? If drug prices are based on the "value" of the
medication-that is, what the market will bear-then costs are ir-
relevant, and cost savings simply contribute to larger profit margins.
V. COMMENTS ON THE STATEMENT OF PROFESSOR FIRESTONE
A. Geneiai comments
(1) Much of this paper is similar in style and approach to that of
Professor Cootner: It is a straightforward pedagogical exercise in
which some of the rudiments of index number measurement methods
and problems are discussed.
(2) But it differs from Cootner's statement, and resembles the
"Trends in Market Share" study in that Firestone has a. tendency to
compliment the drug industry for alleged advances which have been
regarded somewhat more skeptically by medical men. For example,
on page 7 and again on page 20, he refers with approval to sustained-
release medications despite the fact that Dr. B. IV. Burack previously
stated before this subcommittee that such a preparation "remains un-
predictable at best" (part 1, p. 330). And on page 21 he speaks with
approval of combination drugs, the criticism with regard to which
PAGENO="0395"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1945
have already been mentioned in connection with the comments on the
"Trends in Market Share" study.
(3) On page 23, Firestone develops an argument akin to Whitney's
that the price of drugs should be judged in relationship to their "value"
to the consumer rather than their production cost. Firestone's presen-
tation, although less provocating phrased, is open to the same crit-
icism that was made of Whitney~that use of the "value" criterion
betrays a monopolistic pricing strategy.
B. Specific comments
(1) On page 8 it is argued that because there are 110 drug prod-
ucts in the wholesale price index, this "probably results in the drug
category being more representative than most other categories of like
weight" since the inclusion of only 21 products would be required to
give the drugs category adequate coverage in terms of the relative "im-
portance" of this category in the WPI compared with the total num-
ber of products included in the WPI. But it should have been made
clear that the number of products included does not necessarily guar-
antee the répresentativeness of the data reported unless the reported
products are themselves representative of the category in question.
(2) On page 11 it is stated that the PMA wholesale price index
is limited to brand-name drugs only, which in itself makes it less than
perfectly representative of all drugs. Firestone should have esti-
mated how much bias might be introduced into the PMA index by ex-
cluding generics, and should have indicated how much of the indus-
try's sales are in generic form.
(3) On page 16 it is claimed that patented drugs declined in price
by 24.8 percent between 1949 and 1966, while nonpatented drugs had
increased by 1.1 percent.
While it is not surprising to discover that prices in the competitive
segment of the industry's market presumably increased, and while it is
not likely that any segment of the industry could have experienced
substantial price declines during an inflationary period unless initially
uncompetitively high prices prevailed, the contrast between the price
movements would seem to require some additional explanation.
There are two questions.
First, what sort of criterion was used to distinguish patented and
unpatented drugs at various points in time? And, second, why did the
price index for patented drugs seem to drop so appreciably?
Neither question is answered by: Firestone, although one suspects
that to answer the first question may largely answer the second. A de-
tailed explanation of the criteria used to distinguish patented and
unpatented drugs should have been provided.
Senator NELSON. When they use the term "patented drugs" here, are
they referring to drugs that were patented but subsequently the patent
ran out?
Dr. STEELE. That would be true, of any drug which was patented.
Senator NELSON. It is one thingto say patented drug prices declined
while the patent was still in effect and another to say they declined, if
they declined, after the patent expired, is it not?
Dr. STEELE. Quite true.
Senator NELSON. Then if during a certain period of time, as you sug-
gest, when costs are rising, the nonpatented drug prices rose, you
might very well expect that, as in most products.
PAGENO="0396"
1946 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. STEELE. Yes.
Senator NELSON. Would it not indicate that if during the same pe-
riod, patented drug prices declined, they had set an arbitrarily high
price in the first place, since drug industry costs are increasing, too?
Dr. STEELE. Yes; I would agree they must have set the price arbi-
trarily high so there is only one direction in which it could change.
Firestone does refer to large drops in hormones and antibiotics
prices, but one must not forget that a major antibiotics price cut was
announced in 1960 just before the Kefauver hearings were to investi-
gate the antibiotics market. Since this was the first price change in
almost 10 years, the motivation for that price cut at that particular
time is somewhat problematic. The significance of the reported price
trends for nonpatented drugs, however, may not be very great. Since
the PMA index is restricted to brand name drugs, it is doubtful if it
is very representative of nonpatented drug prices, since much of the
sales of these clrugs-in ph3~sical if not in dollar terms-is probably
under the generic name.
(4) On page 19, Firestone states: "An index cannot tell us whether
prices are too high or too low." Quite true. This should be kept in
mind when appraising the significance of reported declines in price
indexes for brand name drugs.
(5) On page 14, Firestone refers to "initial exploratory prices" for
drugs during the introductory period of their use. This is an interest-
ing phrase and its implications should have been enlarged upon since
it suggests market testing by a seller possessed with discretion over
prices and an interest in experimenting with prices to determine what
the traffic will bear.
(6) Although Firestone states that the 17 items included in the BLS
Consumer Price Index for all prescriptions since 1967 "should be rea-
sonably reliable to measure price movements for all prescriptions";
(p. 18) it is apparent from the study by Agnes Brewster, "Examina-
tion of Anomalies in Prescription Drug Prices and Utilization," that
neither the drugs included nor the method of pricing them is such as
to permit the BLS index to make any claim of being representative.
Of the 14 drugs in the 1964-66 index, 11 could be purchased generi-
callv. Only 12 of the 200 most frequently prescribed drugs in 1965 were
not sold under trade names, and these 12 drugs constituted only 6.2
percent of all prescriptions. Yet six of the generic drugs included in
the BLS index were members of the group of the 12 most frequently
prescribed generic drugs.
It is likely that these six drugs accounted for only 3 or 4 percent
of all prescriptions. Yet they constituted over 40 percent of the num-
ber of drugs in the 1964-66 index. More important, brand-name drugs
account for 90 percent or more of all drug sales, but only about 57
percent of the number of drugs could be purchased under brand names.
Hence generic drugs were overly intensively represented, and brand-
name drugs underrepresented.
The method of pricing the drugs is even more unsatisfactory. All
drugs were priced generically, even though over 90 percent of the
prescriptions are written by brand name, and probably half or more
of the remainder are dispensed by brand name.
PAGENO="0397"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1947
For only three of the 14 drugs would the generic and brand-name
prices be identical, since these three drugs were sold exclusively by
patent monopolists.
(7) The entire issue of the relative, movements of drug prices and
average prescription charges merits more unraveling than Firestone
has attempted. The price of a prescription is based on two factors:
The cost of the drug to the druggist, and the druggist's monetary
makeup. The cost to the druggist varies with the identity of the
drug, the dosage form of the drug, the unit cost of the drug, and the
number of units embodied in the prescription.
The druggist's markup is a monetary sum added to his cost, a]-
though it may be determined eithe r as a percentage markup added
to cost, or as a. service charge whk h does not vary directly with the
drug cost itself. In either case, the amount of the markup will be
influenced by the druggist's costs and the state of competition in his
marketing area.
(a) As Firestone suggests, one possible explanation for the decline
in drug prices as measured by certain indexes and the increase in
average prescription charges is that an increase in the distributor's
cost has more than offset the decrease in the price of drugs to the
distributor.
(b) Elsewhere Firestone makes a point of expressing his convic-
tion that pronounced variations in prescription prices for the same
item cannot be due to "the misdeeds of the manufacturer" (p. 22).
One may infer that this conviction includes a concern to show that
increases in prescription charges are similarly not to be attributed to
misdeeds at the drugmaking level. ("What has been most serious in
the misuse of average prescription prices is the use of these prices
for measuring what has happened to manufacturers' prices," Fire-
stone states on page 21.)
(c) However, certain drug firm practices might increase prescrip-
tion charges even though drug prices were declining. In terms of
the above analysis of the factors determining prescription charges,
the following avenues might be exploited to increase prescription
charges despite declining prices for each dosage form of each in-
dividual drug.
These points are possibly quite significant.
First, with regard to the nature of the drug: Sales promotion may
succeed in changing prescribing habits so ~s to increase both the pre-
scribing of drugs in general, of the more expensive drugs in particu-
lar, and even of the more expensive dosage forms or modes of each
drug. Drug industry critics have claimed that overselling the doctors
means overprescribing and overmedication.
Last year, I think Dr. Frederick Wolff testified before the subcom-
mittee that in his opinion, which he thought was shared by most of
his colleagues, something like 60 percent of all prescription drugs pre-
scribed were unnecessary. Thus doctors may be persuaded to prescribe
more expensive dosage modes, such as sustained-release and combina-
tion forms, instead of the simpler mode. They may also be induced to
prescribe the drug in situations where they would previously have
recommended the use of a proprietary preparation, or simply no medi-.
cation at all.
PAGENO="0398"
1948 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
(2) With respect to the identity of the drug: Doctors may be in-
duced to use newer products as rapidly as they are marketed, and these
drugs may be marketed at successively higher prices. If this is the case,
then steady declines in the price levels of individual drugs may be
quite consistent with increases in average prescription charges since
new drugs are being prescribed at higher prices in the place of the
older drugs the prices of which are declining.
Senator NELSON. So if the price index being used is composed of
older drugs and this process is in effect, it does help explain at any
rate the average prescription cost increases, is that correct?
Dr. STEELE. Yes; and this could be a real factor in the BLS index.
Firestone's index is the chain index type, so possibly some of this may
be allowed for. But not immediately.
Senator NELSON. So you would have to keep updating the drugs in
the index in order to have a really accurate reflection of a large pro-
portion of the prescriptions in the country, right?
Dr. STEELE. That is true, but there are also some unavoidable biases
in the construction of chain indexes-well, in any indexes. An index
has a number of problems which cannot be allowed for satisfactorily.
I will say a chain index is better than an index which does not ever
change its weights and allow for such biases, but even a chain index
can't allow for all of them.
Senator NELSON. But if you use an index to look at drug prices and
the number of dollars spent for drugs, and the index includes only
old drugs no longer being widely prescribed, that does not give you a
very informative view of what is happening to drug prices; does it?
Dr. STEELE. If as you say the drugs now included in the drug index
are no longer being used, this is quite true.
With respect to the quantity of the drug prescribed: the greater the
number of chronic conditions which doctors can be persuaded to treat
with continuous drug therapy, the greater the average prescription
quantity is likely to be, to the extent that such patients buy less fre-
quently but in larger individual sizes, to take advantage of any econo-
mies of the "large economy bottle" variety. (Presumably this trend
could continue until long-term drug therapy was available for all
chronic conditions-and possibly even beyond that, to the extent that
the progressive superannuation of the population as a whole increased
the incidence of chronic disorders.)
However, the emphasis should be placed upon the possible difference
between the need for such therapy, and the degree to which drug firms
are successful in creating the impression of the need for such therapy.
As Dr. Frederick Wolff, of the George Washington School of Medi-
cine, has stated before this subcommittee:
It is known that with appetite suppressants, the patients become tolerant to
them after some eight weeks and they have no effect. This is not generally recog-
nized by physicians, so they are being prescribed indefinitely, and occasionally
with very harmful side effects (Part 3, p. 836).
d. Each of the above avenues would tend to increase total revenues
from drug sales as a conscious end in itself. But the net effect of the
various means used to gain this end: a steady stream of "new" drugs;
an ever-increasing number of combination drugs, and a medically aim-
less proliferation of brand names-all these would act as an uninten-
PAGENO="0399"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1949
tional "misdeed" on the part of manufacturers in that they would in-
crease the druggist's inventory and other costs of a related nature, and
give him an incentive to charge higher prices to cover these increased
costs.
VI. COMMENTARY UPON THE STATEMENT OF G. R. CONRAD
AND J. W. MAEKHAM
The argument that "somewhat peculiar" riskiness attaches to the
drug companies because of their failure adequately to discharge their
responsibilities for conducting every stage in the product develop-
ment process in sufficient depth to eliminate all hazards of "unantici-
pated side effects," of deficient quality control procedures, of addiction
potentials, and of toxicity or inefficacy in use, is in itself somewhat
peculiar.
If drug development were undertaken from the perspective of the
best interests of the public health in the long run, such evaluation
might easily take 20 years before a suitably conscientious manage-
ment could satisfy itself that the drug merited general commercial use.
During such a time period, many related drugs might be evaluated
concurrently. If during this period a new drug were discovered which
proved clearly superior to other drugs being tested, none of the inferior
drugs would ever be marketed. Hence the risk of "the development of
a competing product superior to" another already on the market should
scarcely exist if drug development were as thorough as it should be.
(As far as present market risks per se are concerned, however, the
word "superior" can be replaced simply with "newer".)
In describing the operation of risk in drugs, the author states: "The
types of collapse we refer to do not offer hope, in most cases, of subse-
quent recovery of the product's market position" (p. 3). But in his
oral presentation, Markham illustrated his somewhat peculiar risk No.
2, "the discovery of unanticipated side effects," by reference to Parke,
Davis' brand of chioramphenicol, the so-called Chloromycetin. This
was an unfortunate choice. When, as early as 1950, this drug proved
that its lethal potential extended to the infected, as well as the infect-
ing, organism, some apprehension regarding its use developed, and it
was even taken off the market for a 2-month period during 1952. When
reinstated, it was only on condition that strong warnings regarding its
use be placed on the label and the package insert. For a considerable
period of time, chloramphenicol sales were greatly reduced. But
Parke, Davis marketing strategies rose superior to FDA precautions.
The firm's detail men were given instructions which included memoriz-
ing clever and misleading sales spiels and gambits-see the report on
the Kefauver administered prices in drugs hearings, pages 192-198-
and before too long chioramphenicol sales had risen to the point where
it was the most lucrative single brand name drug.
It would appear that the period of temporary shrinkage in sales,
even though limited in duration, had benefited the drug in the long
run. Micro-organisms had less exposure to it than to other antibiotics,
and fewer strains resistant to chlorampherncol developed in the early
1950's. In this instance, the somewhat peculiar risk turned out to be a
w-indfall in disguise.
Senator NELSON. Thank you very much, Dr. Steele.
PAGENO="0400"
1950 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Do you have any questions, Mr. Gordon?
Mr. GORDON. No.
Senator NELSON. Your critique and your original statement were
very well done and a very fine contribution to these hearings. They
will be very valuable in the hearing record. The committee appreciates
very much your taking the time to come up here today to testify.
Thank you very much.
Dr. STEELE. Thank you, Senator Nelson.
(The supplemental information submitted by Dr. Steele follows:)
[From the Journal of Law & Economics, Vol. V, October 1962]
MONOPOLY AND Co,IPETITI0N IN THE ETHICAL DRUGS MARKET 1
(By Henry Steele, Rice LTniversity)
Great concern has recently been expressed in regard to the ethical drugs
industry as an object of public policy.la At first glance such concern seems dis-
ijroportionate in view of the relatively small fraction of the consumer's dollar
which is devoted to drug purchases. In 19~9 total sales by manufacturers to
dealers of all pharmaceuticals were estimated at ~2.25 billion. Of this total,
$1.7 billion consisted of ethical drugs, which are available only upon a doctors
prescription, and S560 million of proprietary drugs and medications which can
be purchased over the counter without any prescription.2 The conventional mark-
up imposed by the druggest (66% per cent of invoice costs) would increase the
price to the drug consumer of ethical pharmaceuticals by $1.13 billion, making
the total cost of ethical drugs to the final consumer $2.83 billion. This figure
is only .9 of one per cent of the 19~9 total personal consumption expenditures
of $313.8 billion.
Concern with the policies of ethical drug makers is, however, not misdirect ed.
since the industry occupies a position of importance in the economy out of
proportion to its size. There are several reasons why it is very important to
analyze and evaluate the performance of the ethical drugs industry. First, if
competition is desirable and monopoly undesirable as a matter of principle,
industry size is immaterial, and there is no reason why smaller industries should
be permitted to misallocate resources even if, for example, increases in their
prices do not have a dramatic effect on aggregate price indices. Second, for the
policy maker, there are compelling reasons why the quality of the performance
of the ethical drugs industry should be given greater relative w-eight in his
social welfare function than its dollar volume of sales might seem to justify.
During a given interval of time, the incidence of disease is far from uniform.
and those w-ho pay the largest drug bills are, because of their disgbilities, likely
to have incomes substantially below the average. The industry shares with the
medical profession much of the responsibility for the maintenance of health,
and health is naturally a value in itself as well as an important determinant
of the productivity of human resources in the economy. The existence of ding
regulation and inspection by the Food and Drug Administration testifies to
the social premium placed upon adequate performance in this industry.
Third, the industry operates under certain unique conditions which provide an
interesting case study for analysis by the economist. These conditions include
(1) the separation of the authority to prescribe from the responsibility to pay,
w-hich in inherent in the status of the prescribing physician as an independent
purchasing agent for his patient, and which minimizes the influence of prices on
the volume of prescription sales; (2) the use of research to effect often minor
product changes which can be marketed with surprising success as major thera-
peutic advances; (3) the extreme degree to which product differentiation may be
The author is indebted to Professors E. 0. Edwards, M. A. Adelman, G. V. Rimlinger.
and J. H. Auten for valuable criticisnis of earlier drafts.
in See particularly, Report on Administered Prices of Drugs, Subcommittee on Antitrust
and Monopoly of the Senate Comm. on the Judiciary, S. Rep. No. 448, 87th Cong., 1st
Sess. (1961). (Hereinafter cited as "Subcomm. Report.")
2 Data obtained from a survey made by Arthur D. Little, Inc., and published in Silber-
man. Drugs: The Pace Is Getting Furious, Fortune, May 1960, p. 139.
PAGENO="0401"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1951
accomplished by different firms merely assigning different brand names to iden-
tical substances arid then advertising such brand names intensively; and (4) the
presence of a very high degree of concentration in the production of most of the
important ethical drugs, in the absence of any apparent economies of large-scale
production.
An investigation of the production and marketing of ethical drugs compels the
conclusion that the industry may easily be considered delinquent in regard to its
impact on the allocation of economic resources. This paper will attempt to
demonstrate that such misallocation of resources has been made possible by the
existence and abuse of the patent privilege for ethical drug products and proc-
esses and by a number of measures which the industry has taken to foster and
exploit the remarkable degree of imperfection of market information which the
structure of the ethical drugs market pernuits. Such measures can, however, be
understood adequately only in the context of actual drug marketing practices,
which will be described below.
I. THE ECONOMIC FRAMEWORK OF THE ETHICAL DRUGS INDUSTRY
Broadly speaking, the economic framew~ork within which the ethical drugs in-
dustry operates is conditioned by the nature of demand, the structure of costs,
the existence of the patent privilege for drug products and processes, the eligi-
bility of branded drug products for protection under the trademark laws, and
the existence of the Food and Drug Administration's drug product inspection
facilities as an external means of insuring quality in drug products.
The nature of demand is such as to distinguish the prescription drugs market
from virtually all other markets.3 Sole purchasing authority, as well as the neces-
sary initiative, lies with the prescribing physician, who orders the specific drug
for which the patient must pay. Chemically identical drugs may be selling under
different names at widely varying prices, but the physician has no direct motiva-
tion to prescribe the lowest-priced brand or even to become aware of prices at all.
(Indirectly he may reason that his fee for services may more readily be capable
of collection if drug prices are low, but this would seem to be a distinctly second-
ary consideration.) Ideally, the total potential market for a drug or for a group
of related drugs consists of the total need for medication (either for cure or for
the relief of symptoms) on the part of all individuals afflicted by the various
disease entities or clinical syndromes which are capable of treatment by the drug
or group of drugs. Economically, total effective demand at given drug prices
may fall short of total physical need in the case of individuals with very low
incomes and no access to welfare case status, and it may exceed total physical
need to the extent that individuals not suffering from those disorders for which
the drugs are of use may nevertheless be treated with them. For any given drug,
therefore, the actual relevant market is àomprised of the total effective demand
for medication on the part of all individuals who can be induced to consult physi-
cians arid who are afflicted by those disorders for which physicians may be in-
clined or persuaded to prescribe the particular drug.
Either for single drugs, or for a group of related drugs as a whole, the demand
curve is likely to be extremely inelastic. It has been reported that individuals
with severe inflammatory diseases and lOw incomes sometimes do without food
in order to buy drugs.4 Shifts in the demand curves for individual products are
brought about by direct-mail advertising, medical journal advertising, and by
the insistence of itinerant sailesmen or "detailmen," employed by the major drug
firms. Advertising cannot manipulate the total incidence of diseases (although
to a limited extent, news articles in newspapers and magazines may mention that
a given drug can treat a given condition, and thus make more people who suffer,
or imagine that they suffer, from a given condition, aware that drug treatment
is available) ; but it can shift the existing effective demand from one product to
another, and such shifts can perhaps be visualized as parallel rightward or lef t-
Exception might be taken to this statement by college professors, who may be said to
"prescribe" textbooks for their students. Here, however, the student has alternatives not
open to drug buyers. such as the existence of a second-hand textbook market and the
possibility of textbook sharing by two or more students.
Testimony of E. P. Bransome of the Arthritis and Rheumatism Foundation. Hearings
on Administered Prices before the Subcommittee on Antitrust and Monopoly of the Senate
Comm. on the Judiciary, 86th Cong., 1st Sess., pt. 14, at 7992-93 (1960). (Hereinafter
cited as "Hearings on Administered Prices.")
81-280-68-pt. 5-26
PAGENO="0402"
1952 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ward movements of nearly vertical demand curves. The absence of price-con-
sciousness on the part of the physician, the inability of the patient to purchase
any but the specified drug, and the marked inelasticity of the individual patient's
demand curve, all are conducive to the possibility of charging a price which is
extremely high relative to marginal cost.
The cost structure of the industry is such that there seem to be few, if any,
notable economies of scale in the production process. In this regard, a sharp dis-
tinction must be drawn between the production of "heavy" or "bulk" chemicals
which are commonly produced in enormous quantities by continuous-flow proc-
esses (sulfuric acid would be a good example) and the production of "fine"
chemicals, such as pharmaceuticals. Batch methods, which allow far less scope
for economies of scale, are the rule, especially where fermentation is the key
phase in the production process, as is the case in the production of antibiotics
(except chioramphenicol) and synthetic corticosteroid hormones.5 These two cat-
egories of ethical drugs are the largest in terms of total sales volume. Fermenta-
tion vessels are "amazingly standard." 6 Large antibiotics makers usually employ
10 to 15 such vessels.7 Changes in output levels for such products may be accom-
plished by the employment of a differing number of individually identical ferm-
entation vats, a circumstance conducive to constant returns to scale. A second
reason for the absence of important economies of large-scale production is the
frequent absence of truly large-scale production. The physical volume of output
of the active ingredients in drugs is typically very small, total national output
being perhaps in the neighborhood of a ton pear year.8 Actual cost data are kept
secret by the drug firms. The inconclusiveness of some of the economic aspects of
the 1059-1960 Senate hearings on the ethical drugs industry is due to the "trade
secret" status afforded certain crucial cost data and the consequent failure to
insist upon the publication of actual production costs for at least some of the high-
er-priced patented drugs. To the economist, this is a most unfortunate omission,
for such data are otherwise entirely unobtainable. On the basis of the costs sub-
mitted to the Senate Subcommittee (but not published in the record), the Report
comments upon the absence of economies of scale and the resulting "extremely
small size of plant required for economical production." ° It is certainly true that
very small firms can purchase the active ingredient for a given drug in bulk (the
so-called "bulk powder") from the large firms and tablet and sell it at prices
greatly below those which the large firms see fit to charge; there is also con-
siderable evidence that very small firms can actually manufacture their own fine
chemicals as bulk powder and sell them to small firms in bulk at prices equal to
or lower than those charged by the major firms.'° Similarly, on the basis of anti-
biotics production costs revealed to the Senate Subcommittee, but not published
in the record, Senator Kefauver said at one point in response to Eli Lilly & Com-
pany's request for confidentiality as to its production costs, "I will tell you
frankly you all have about the same cost figures. . . . For a difference of one or
two cents . . . I do not see any sense in all the secrecy. . . . Your costs, Bristol's
costs, Upjohn's costs, are all within a very, very narrow range." `~ This tends to
support the contention that there are no important economies of scale in pro-
duction, inasmuch as the level of output of Bristol's product was almost twice as
great as Lilly's. Granted that the ethical drug firms can purchase bulk chemicals
from heavy chemicals firms which do themselves produce under conditions of
economies of scale, within ethical drugs proper, there seems to be no evidence
of anything but roughly constant returns to scale in the production process. The
most important economies of large size seem to lie in the area of large-scale sell-
ing and advertising, to be discussed below. In the absence of important economies
Cortisone Quest: The Right Process Bug. Chem. Week, Aug. 25, 1952, as quoted in
Hearings on Administered Prices. pt. 14, at 8291.
6 Gaden. Fermentation. Chem. Engr.. April 1956, p. 159.
FTC. Economic Report on Antibiotics Manufacture 118 (1958).
8 Production in 1958 of the most important synthetic corticosteroid hormones (the sec-
ond largest category of ethical drugs by sales volume) was, according to the U.S. Tariff
Commission. as follows: cortisone, 2.15 tons; hydrocortisone, 3.21 tons; prednisone, .63
tons; prednisolone, 1.4 tons; methyiprednisolone, .5 tons. Hearings on Administered Prices,
pt. 14. at 8285-86.
Subcomm. Report 4.
10 According to Dr. Phillip Berke, president, Formet Laboratories, a small producer of
drug bulk powder, with estimated sales of about $500,000 for all products, could in 1959
sell bulk prednisone at prices competitive with those charged by Merck and Pfizer, firms
with over $200,000,000 In sales of all products. Hearings on Administered Prices, pt. 14, at
8056.
l11~7, pt. 24, at 14125-26.
PAGENO="0403"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1953
of scale in production, however, it might be conceivable that the ethical drugs
industry could approximate to the condition of pure competition.
Patent protection is available in the drug field for process improvements and
for the development of new products. Although a product patent cannot be ob-
tained without an accompanying process patent, this is no obstacle to product
patents. "Patent medicines" are by no means peculiar to the proprietary drugs
field. Unrestricted patent privileges on drug products are, however, virtually
peculiar to the United States. Most countries do not award product patents for
drugs. Only 28 of the 77 countries for which the Senate Subcommittee could ob-
tain patent law data granted pharmaceutical product patents. Twenty-five of
these countries have provisions for compulsory licensing. Only Panama, Belgium,
and the United States are without any such limitations.12 In the United States,
the definition of a patentable drug has been broadened in recent years. Prior to
1940, naturally occurring substances could not be patented; product patents on
cortisone and hydrocortisone, for example, were impossible to obtain. In 1946 a
patent was obtained on the antibiotic streptomycin, on the grounds that, although
it was a product of nature, it was transitory in nature, had never been isolated,
and its therapeutic use was unknown.'3 In 1955 a patent was issued for tetra-
cycline, even though this compound was jointly produced in the yields from the
process for making chlortetracycline, a previously patented drug, and even
though tetracycline had previously been marketed by another firm (which did not
receive a patent) ~ Patents, therefore, are becoming increasingly easy to obtain
in the ethical drugs industry. Trademarks, in the drug industry as elsewhere, may
be employed to extend a privileged market position for a larger period of time
than the term of the relevant patent. A good example is the case of the barbitu-
rate phenobarbital, which the patent holder sold as "Luminal", at $9.80 per ounce.
After the patents expired in the 1930's, competitors produced phenobarbital for
sale at $.85 to $90 per ounce, but for years "Luminal" continued to sell at $9.80.'~
The Food and Drug Administration has numerous responsibilities with regard
to ethical drugs, chief among which are the control of the composition and quality
of ethical drugs, the investigation of the safety of new drugs, the certification of
antibiotics and insulin, and the prevention of the illegal distribution of prescrip-
tion drugs. The laws are very complex. For the purposes of this paper, it may be
sufficient to note that the purely economic impact of such regulation is, first, to
supplement the quality control of individual firms through the authority of the
Food and Drug Administration to inspect samples of ethical drugs (and its more
limited authority to inspect production facilities) and, second, to prevent the
marketing of dangerously lethal new drugs.1° Violative samples of drugs may be
simply confiscated and destroyed, or legal proceedings may be instituted, and
civil and criminal penalties may be imposed by the court, including fines large
enough to put a small firm out of business. The certification of new drugs, if done
with proper regard for the standards necessary to insure drug safety, can be a
time-consuming process. Economically, the consequences of regulation are there-
fore to provide an additional guarantee to the consumer that all drugs on the
market are of acceptable quality and to delay the introduction of new drugs in
the interests of safety.
II. DRUG INDUSTRY RESPONSE TO THE SPECIFIC ECONOMIC FRAMEWORK
The ethical drug industry's response to the given economic framework may be
interpreted as a successful strategy to exploit the monopolistic potentialities in-
herent in trademarks, the patent privilege, and the great inelasticity of market,
in such a way as to overcome the potential for competition implicit in the absence
of economies of scale and, to a lesser extent, in the presence of governmental in-
spection to insure the safety of all drugs placed on the market.
"The vast majority of European drug discoveries have been made In countries where no
patent protection was available. Germany, Switzerland, Sweden, Austria, the Netherlands,
and Italy are without product patents. France had none until 1959. Great Britain allows
product patents on new drugs, but not on combinations of known drugs, and all such
patents are subject to compulsory licensing at reasonable royalties. Subcomm. Report
105-6.
"Id.,at 141.
14 Id. at 146.
`~ Hearings on Administered Prices pt. 18, at 10433.
`~ Testimony of Food and Drug Administration Commissioner G. P. Larrick, id., pt. 22,
at 12110-36.
PAGENO="0404"
1954 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The most prominent monopoly element in the market, the patent privilege, may,
as is well known, be employed in many ways to limit competition beyond what is
explicit in the specific grant of an individual product or process patent. In the
chemicals industries the necessary expedients are of easier access than is gener-
ally the case. Original patent grants may be extended indefinitely by a judiciously
timed succession of improvement patents.17 Improvement patents in chemicals are
particularly readily devised, especially where there are process patents on a con-
tinuous-flow production method. On the other hand, it is relatively easier to "pat-
ent around" existing patents by engineering small chemical changes in product
composition. Whenever the product or process developed as a result of patenting-
around efforts is more efficient than that developed by the original patent holder,
a situation results w-hich requires cross-licensing.18 The negotiating of cross-
licensing agreements has as an attendant hazard the introduction of the possi-
bility of such meetings of the minds as may be incident to the making of agree-
ments involving the mutual compromise of patent monopoly positions. Patents
in the drug industry may bring about few absolute monopolies (few patented
drugs are uniquely efficacious, even for a single disorder) but may provide the oc-
casion, and the economic basis, for the development of a greater sense of corn-
munity of interest in matters of price and production. So much is speculation.
However, the terms of the actual patent licensing and cross-licensing agreements
obtained by the Subcommittee, together with the evidence of price uniformity
(not only between different brands of the same drug but also among different
drugs in the same field of therapy), lend some support to such speculation.19
It is of course claimed by the drug makers that patents promote competition.
Several relatively large drug firms have asserted that patents are a small coin-
pany's best friend.20 By this it is meant that the company in question gained
monopoly power in some market and was enabled to out-grow its small, competi-
tive beginnings.21 This is consistent with the theory implicit in much current anti-
monopoly legislation: that the remedy for the existence of monopoly power on the
part of the large firms in a given market is to confer a similar degree of mo-
nopoly power on the small firms.
Really small firms, on the other hand, have testified that a patent is merely
an invitation to costly litigation and the only advantage a really small firm can
take of a useful patent is to sell it to a larger rival, usually for a lump sum
paid-up royalty.22 Even if a small firm were willing to face the hazards of litiga-
tion, it would lack the funds necessary to introduce even a hypothetical new
product of superlative efficacy because of the needed outlay on sales promotion.23
17 Insulin is an excellent case in point. See subcomm. Report 141.
18 This is often the case where the production of a final product requires as inputs a
number of intermediate products, some of which may be capable of chemical synthesis as
well as of organic cultivation or fermentation. An important example is the case of the
broad-spectrum antibiotic, tetracycline, independently discovered by several drug firms at
about the same time. Bristol Laboratories used a method of direct fermentation to produce
tetracycline. Lederle had patented an earlier antibiotic, chlortetracycline ("Aureomycin")
anti had never licensed any competitor. Lederle found that by dechlorinating chlortetra-
cychine. the resulting compound, tetracycline, had unusual antibiotic properties. The
chemical dechlorination process was more efficient than direct organic fermentation, but
Bristol apparently never applied to Lederle for supplies of chlortetrncycline, presumably in
the interest of secrecy. or in anticipation of refusal. Pfizer had done tetracycline research,
and had filed a patent application, but found that Lederle had also done so. Pfizer con-
vinced Lederle that Pfizer's claims were superior; Lederle, however, had patents on chlor-
tetracycline, which Pfizer needed as an input for tetracycline. A mutual agreement was
made, insuring both Pfizer and Lederle of a market for tetracycline, regardless of tile
award of the patent. Hearings on Administered Prices, pt. 24, at 13848. 13697-701.
10 The currently pending Federal Trade Commission complaint, FTC v. Cyanamid, con-
cerns such a point in connection with the tetracycline patent negotiations mentioned just
above. The FTC charged the respondents with a conspiracy in connection with the award-
ing of the patent to Pfizer, and with a conspiracy in restraint of competition, operating
through the subsequent patent licensing arrangements. Subcomm. Report 145-47.
20 Testimony of Dr. P. I. Bowman, president of Bristol Laboratories. Bristol failed to
get tile tetracycline patent, but the mere award of a license (in 1956) seems to have been
lucrative. Bristol's profits after taxes as a per cent of net worth averaged 2.9 per cent for
the four years prior to licensing, but 10.3 per cent for the four years following. Hearings
on Administered Prices, pt. 24, at 13835, 13858.
21 Testimony of Henry Hoyt, president of Carter Products. In the five years prior to the
marketing of its patented tranquilizer, meprobamate ("Miltown"), Carter had earned 16.8
per cent after taxes on net worth. In the next five years (1955-1959) the rate of profit
after taxes averaged 41.0 per cent. Id., pt. 16, at 9113, 9170.
22 Testinlony of Seymour Blackman, executive secretary, Premo Pharmaceutical Prod-
ucts. Id., pt. 14, at 8254.
93y of Myron Pantzer, vice-president of Panray Corporation, Id., pt. 16, at
PAGENO="0405"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1955
And the most certain evidence of the lack of that sort of competition which is
beneficial to the consumer is the rigidity of the prices of patented antibiotics,
compared with the steady price decline displayed by the few unpatented anti-
biotics, such as the earlier penicillins.24 The extent to which patents are used to
limit the number of producers of a given drug can briefly be illustrated by coni-
paring the number of producers for 42 drugs, in the categories of hormones, dia-
betic drugs, tranquilizers, sulfa drums, and antibiotics. Of these 42 drugs, which
account for over half of all ethical drug sales four were unpatented, while 38
were produced under patent or patent-application licensing agreements. Of the
unpatented drugs, two were made by three firms, and two by seven firms. Of the
patented drugs. 24 were made by a single producer, 8 were made by two firms,
4 by three firms, and 2 by five firms.25 Twenty-four of 42 drugs or 57.1 per cent,
were therefore produced by a single patent holder. The concentration of sales is
not as great as the concentration of production, but 16 of the 24 patent holders
who license no others to produce, also license no others to sell, and therefore
market their products as patent monopolists. One firm which licenses another
to produce for it licenses no other sellers, and thus constitutes a seventeenth
patent monopolist. Of the remaining eight products for which there is a pro-
duction monopoly, two producers have licensed only one other seller, three pro-
ducers have licensed two other sellers, and the two other three producers have
licensed six, seven, and eight other sellers.26 Even in an industry without econo
mies of scale, patents can act as a barrier to entry and enable monopoly positions
to be attained.
An inelastic industry demand curve does not in itself insure monopoly power
on the part of the seller, if there is workable competition among sellers. In the
absence of economies of scale, small firms could compete with large firms on
a price basis. To minimize the effect of suCh competition, the ethical drug in-
dustry has taken measures to influence the normal means of dissemination of
market information so as to prevent the physician from becoming effectively
aware that there are lower -priced sellers in the market, to prevent the -recogni-
tion of lower priced alternatives to higher priced drugs wherever the aware-
ness may exist that there are lower priced drugs, and to convince the physician
that all lower priced drugs are of dangerously poor quality. These efforts to
maximize the degree of imperfection of market information have been extremely
successful, in large measure because of the nature of the market in which drugs
are bought and sold. It is essential at this point to examine that market in
some detail.
The direct buyers in this market consisted, in 1959, of some 2,800 drug whole-
salers, 53,000 retail pharmacies with 110,000 pharmacists, 9,000 public and pri-
vate general hospitals, and 4,000 veterinary hospitals.~ The individual patient is
the final consumer, but the physician acts as his purchasing agent when writing
the prescription. The physician may write the prescription in terms of the generic
name of the compound, or by use of a particular firm's brand name. The retail
pharmacist may fill generically written~ prescriptions by using any firm's brand
of the drug, but any prescription written in terms of brand names must be filled
with the brand called for, unless the pharmacist obtains specific permission from
the physician to make a "substitution" of another brand. Unless the pharmacist
obtains permission in each instance, he is subject to punitive action by the state
pharmacy boards in 44 states under the "substitution" doctrine.28 It must be
admitted, however, that the courts are not always willing to concur in this doc-
trine. In Michigan State Board of Pharmacy v. B. L. Casden the state court
dismissed the contention that any meaningful "substitution" had taken place
St For the ten-year period 1951-1960 the bulk price of penicillin dropped from $2.50 to
$0.21 for ten million units: the bulk price of streptomycin dropped from $3.24 to $0.36 for
ten grams; the prices of the patented antibiotics chlortetracycline ("Auteomycin"), oxy-
tetracycline ("Terramycin"). chioramphenicol ("Chloromycetin") and tetracycline (intro-
cluced in 1955 and sold as "Polycycline." "Achromycin," "Steclin," "Tetracyn," and "Pa n-
mycin") remained constant at $5.10 per hott1e of sixteen bbiliirram capsules, until Au-
gust 1960 (one month before the scheduled hearings on antibiotics) when they were re-
duceti by 15 per cent. Id. pt. 24, at 13664. It should be observed that all of these anti-~
biotics are produced by the same general process of fermentation, with the exception of
chioramphenicol. which is produced by a cheaper synthetic process.
Subcomm. Report 67.
Id. at 67-68.
Testimony of Dr. Austin Smith, president of the Pharmaceutical Manufacturers Asso-
ciation. Hearinge on Administered Prices, pt. 19, at 10728.
26 Letter from H. S. McNeil, president of McNeil Laboratories, id., pt. 21, at 11713.
PAGENO="0406"
1956 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
when the identical chemical substance was sold under two different brand
names.~°
Most hospitals operate under the formulary system, which authorizes the hos-
pital pharmacist to fill prescriptions with any generic equivalent of the pre-
scribed drug, regardless of the brand name written by the physician. Such for-
mularies operate under agreements involving the written consent of all resident
physicians. In practice it also applies to visiting private physicians treating hos-
pital patients. The drug trade associations have attempted to challenge this
practice, but without marked success.3° The lack of success here stands in
marked contrast to the effectiveness of the campaign of the National Phar-
maceutical Council in influencing state boards of pharmacy to adopt "substitu-
tion" laws. "Substitution" used to refer to the dispensing of a different chemical
substance than the prescription required. The National Pharmaceutical Council
undertook to influence state boards of pharmacy to adopt antisubstitution laws
which defined substitution in terms of brand names. In 1953 only four states
had any sort of antisubstitution law-s, but by early 1959 the efforts of the
National Pharmaceutical Council had increased the number to 44,31
Hospitals and dispensaries in the Armed Forces operate under the Military
Medical Supply Agency, which purchases on the basis of competitive bids. The
bids received often fall well below the wholesale price of the drug in some lines.
The Veterans' Administration customarily enters into negotiations with indi-
vidual suppliers, although it has also solicited competitive bids, particularly
after the Military Medical Supply Agency had demonstrated how this could
lower prices.
The market consists, in addition, of some 170,000 practicing physicians with
the M.D. degree (allopathic physicians), 12,000 osteopathic physicians, and 16,000
veterinarians, all of whom act as purchasing agents for the drug consumer.~2
A vast selling effort is directed at these purchasing agents, primarily at the allo-
pathic physicians, to create brand preferences which will result in the writing
of brand-name prescriptions. The types of selling effort include (1) advertise-
ments in over 300 medical journals, (2) direct mail advertisements, (3) the
employment of chemists and pharmacists as highly trained itinerant salesmen,
and (4) the sponsoring of medical conventions and meetings, in connection with
the exhibition of wares. For the 22 largest firms in the ethical drugs industry.
selling expenses were in 1958 the second largest component of costs, averaging
24.8 per cent of sales, as compared with a cost of goods sold of 32.1 per cent of
sales. Average selling expenses were therefore 77.3 per cent of the cost of goods
sold. Sixteen companies had a ratio of more than 75 per cent, nine had a ratio of
more than 100 per cent, and four had a ratio of more than 150 percent. Total sell-
ing outlays amounted to about ~3,200 per physician per year.33
The result of this volume of selling effort on the structure of the market has
been to eliminate price competition except in the hospital and government bid-
ding markets, and to substitute product differentiation and brand preference.34
Any physician who wishes to put himself to the trouble of looking up the generic
name of a compound advertised by brand name can write any prescription ge-
nerically. The retail pharmacist may then fill the prescription with any available
brand. For patented drugs there is little likelihood that there will be any great
variation in price, although there are a few such instances.3~ For unpatented
drugs, however, the unadvertised, generically named products of small firms may
sell for 10 per cent or less of the price of the large firms.
~ Text of Wayne County, Michigan circuit court opinion no. 301799, id., pt. 21. at
ii 761-62.
~° Memorandum iS of the National Pharmaceutical Council, 1957, id., pt. 21, at 11835-36.
~1 Subcomm. Report 236.
~ Testimony of Dr. Austin Smith, Hearings on Administered Prices, pt. 19. at 1072S.
~ Data submitted by each firm to the Subcommittee. Subcomm. Report 31.
~ The extent to which advertising. salesmanship. and the bestowal of free samples is
regarded by drug makers as capable of creating lasting brand preference is indicated in an
article by the vicee-president of Abbott Laboratories, where it is stated : "Practically tIle
only ethical drur products whose sales will not be benefited greatly by sampling [the dis-
tribution of free samplesi are the rare medicameats distinguished by advantages so great
that in simple justice to the patient their actual use is called for." Downs, Advertising, in
Drug Research and Development 4S5 (Smith & Herrick ed. 1948).
Reserpine, a tranquilizer patented by CIBA. is sold under its brand name, "Fierpasil.'
at a wholesale price of 84.50 per hundred .25 milligram tablets. CIBA has licensed several
other firms, some of whom make sales of the bulk powder to small companies who in turn
can undersell the large firms' advertised brands. The 1959 Drug Topics' Red Book listed 41
firms selling this product. Two firms were selling at over $4.00 per bottle (CIBA's $4.50
was the highest price), two were selling from $3.00 to $4.00 per bottle, two were selling
from $2.00 to $3.00 per bottle, eleven were selling from $1.00 to $2.00 per bottle. and 24
were selling at less than $1.00 per bottle. The lowest price, Winsale's $0.45. was exactly
10 per cent of CIBA's price. Hearings on Administered Prices, pt. 18, at 10595.
PAGENO="0407"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1957
It is conceivable that some degree of price competition might gain a foothold
in the prescription drug market. This is nOt likely to occur, however, even where
generic name products are sold in competition with brand name products, for
several reasons. First, the physician is probably only infrequently very price-
conscious.3° It has even been argued that the physician who prescribes the more
costly drugs will be accorded the greater prestige. Second, drug advertisements
virtually never mention prices, and the various physicians' reference manuals
never give competitive prices.37 It is, in fact, apparently illegal for a generic
name product price catalog to compare its prices with those of identical prod-
ucts sold under brand names, or even to mention any of the brand name equiva-
lents of a generic name product.38 Third, the pharmacist has no economic interest
in selling the cheapest brand of drug, since the absolute magnitude of his cus-
tomary markup of 66% per cent over invoice cost will thereby be minimized.
Indeed, it may even happen that the pharmacist will buy generic name drugs
and sell them at brand name prices, if the product itself is not clearly identified
as to manufacturer, and if the policing policies in use among druggists are not
adequate. Fourth, and perhaps most impoi-tant, the large firms who do the adver-
tising, support the trade associations, and hire the detailmen, make every effort
to disparage the products of their smaller competitors who sell at lower prices.39
There is probably no other industry in existence where the disparagement of
the quality of lower priced products can so completely substitute for actual price
competition.4°
Physicians are the chief targets for such "educational" disparagement efforts,
but the pharmacists and the general publiè are also exposed to it whenever oppor-
tunity allows. Several representatives from the large drug firms, and from the
Pharmaceutical Manufacturers Association, showed no reluctance to use the
Hearings as a forum for the disparagement of drugs selling at low prices.
The intent behind these various efforts is clearly to deprive the physician of
access to all the information necessary tO allow him to function as a competent
purchasing agent for his patient, and to influence favorably his attitude toward
brand names. To the extent that the physician, who does not pay, can be con-
sidered to have a demand curve for the drugs he prescribes, the efforts of the
drug makers can be interpreted as attempts either to give the physician's demand
curve an upward slope within the relevant region of drug prices, or to reduce
his demand to zero for prices lower than those charged by the major firms. The
demand curve of the patient is perhaps nearly vertical up to prohibitively high
prices if he trusts the judgment of his physician; at any rate, it will be very
inelastic. The patient has no alternative but to purchase the medication at the
specified price, or go without, and thus forego the benefit of the medical advice
he is paying for.
No one wishes to be exposed to the hazards of inferior drugs, regardless of
their price. Is there reason to suspect that lower priced drugs are of poorer
quality than higher priced drugs? Actually, there is very little to be gained by
cutting corners in production and quality control, and much to be lost if such a
practice is detected by Food and Drug Administration inspectors. The reasons
why no quality differences should be expected to exist between the drugs of large
and small sellers may be briefly listed. (1) The same standards of identity.
potency, and purity are enforced with regard to all makers for drugs listed in
the United States Pharmacopoeia and the National Formviary. The Food and
Drug Administration is empowered to inspect any drug produced by aimy firm
at any time, and to impose criminal penalties upon violators. (2) The Food and
Drug Administration has concentrated its drug regulation efforts on the smaller
firms. From January, 1950, until June, 1960, 8,376 samples of drugs produced by
the 22 major firms were examined, and 8,621 samples from some 1.200 smaller
firms were examined. The larger firms produced 87 per cent of all ethical drugs,
~ Mr. Blackman of Premo testified that it once took him two and one-half hours of argu-
ment to convince one physician of his acquaintance that the physician was, in fact, acting
as a purchasing agent for his patient. Id., pt. 14, at 8223.
~ Detailmen are also instructed to ignore prices. One doctor persuaded several of his
colleagues to interrogate detailmen on drug~prices and costs. If pressed, they would gen-
erally quote prices, but no one ever mentioned a cost figure. Id., pt. 18, at 10455,.
3815., pt. 14, at 8212-13.
~ An almost prototypical utterance of this genre is the characterization of the lower
priced drug seller as providing "the unfair competition of the unscrupulous one who
manufactures under unsanitary conditions, cheapens his product with low quality in-
gredients, and falsely labels his product as being better than it is Statement of
Dr. Austin Smith, president of the Pharmaceutical Manufacturers Association, iS., pt. 19,
at 10738. ,
4° See the testimony of Dr. Solomon Garb, 15., pt. 18, at 10476-77.
PAGENO="0408"
1958 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the smaller firms only 13 per cent.41 The likelihood that a specific drug prepara-
tion made by any small firm will be inspected is therefore about 6.9 times as great
as in the case of any large ñrm. It must be conceded, however, that during this
period the average large firm had 381 samples inspected; the average small firm,
7.2. Each large firm is virtually certain of one or more samples per year; a small
firm may hope to go uninspected for a year or more. (3) Most small firms buy
the drugs in bulk form from large firms, and merely tablet and bottle the finished
dosage forms. Pejorative comments by large firms, therefore, often imply at least
some criticism of their ow-n bulk drugs. (4) The United States Pharmacopoeia
sets tolerance intervals above and below 100 per cent of labelled strength for the
active ingredient in many drugs. If the lower limit is, for example. 05 per cent, it
has been claimed that the producer might aim of 95 per cent strength and thereby
achieve economies. For a small firm, however, the cost of goods sold may average
about tw-o-thirds of sales ~ and the cost of raw materials may constitute about
half of the total cost of goods sold ~ so that raw materials costs may be in the
neighborhood of one-third of total costs for a small firm. Hence a reduction of
5 per cent in raw material input w-ould cut total cost by only about 1% per cent.
or by much less than it would require to begin to justify a procedure which would
result in turning out a substantial number of tablets below the 95 per cent low-er
limit of tolerance.44 (5) It is claimed that low priced sellers economize on quality
control. Again, there seems to be little room for cost savings, for the cost of
quality control is probably no more than 21/2 to 3 per cent of the total of all
costs 45-surely not enough to justify a procedure which would make the violation
of the drug laws a certainty.
Perhaps the best argument w-hy quality differences might be expected to exist
between higher and lower priced drugs is the contention that the Food and Drug
Administration is understaffed and cannot make enough inspections. It is true
that insufficient inspections are made with regard to both large and small firms.
From January, 1950. to June, 1960, 7,699 samples of drugs produced by large
firms ($10 million or more in annual sales) and 9,298 samples from small firms
were inspected. During the same period, 84 incidents of irregularities developed
in connection with drugs made by large firms, 79 of which were handled by the
"drug recall" procedure and five of which led to legal actions. For small firms,
690 such irregularities developed, 206 of which involved drug recalls, and 484
of which led to legal actions. The ratio of legal actions is almost 100 to one,
while the ratio of drug recalls is only five to two. Clearly, those irregularities
involving large firms are much more frequently negotiated than those involving
small firms.44 For the large firms, there was a ratio of 1.1 percent of irregulari-
ties to total samples inspected. For the small firms, the ratio was 7.4 percent.
The difference is a matter of degree; both figures are too high. The remedy is
not to prohibit or to discourage by propagandistic activities the sales of drug at
41 Data presented by Food and Drug Administration Commissioner G. P. Larrick, id,
pt. 22. at 12147.
~ The only data available are 68 per cent for Panray Corporation. Id., pt. 16, at 9375.
4~ The only data furnished by any one company for any product show a 52 per cent ratio
for Carter's meprobamate. Id., pt. 16, at 9157, 9161.
~ The products of large firms commonly sell at prices as much as several hundred per
cent above those charged by small firms, as has been indicated in the case of reserpine.
Small firms can undercut large ones by economies in selling costs without any risk of
violating the food and drug laws. Large firms spend about 25 per cent of sales revenue in
sales promotion, according to data submitted to the Senate Subcommittee by the major
firms during the hearings. It is clearly preferable to save 20 to 25 per cent of total cost
by foregoing sales promotion in favor of price competition, than to save 12/s per cent in
raw materials cost and risk a violation that might result in a fine large enough to put a
very small firm out of business.
~ Panray Corporation's quality control costs were 3 per cent of sales revenue; its profits
before taxes were about 10 per cent of sales. Hence quality control for this small firm
accounted for about 2.7 per cent of total cost. That this is adequate is evident from the fact
that no samples of Panray products have been judged violative by Food and Drug Ad-
ministration inspectors. Hearings on Administered Prices, pt. 16, at 9375, 9378.
~ It must not be assumed that irregularities involving legal action were more serious,
on the average, than those settled by drug recall. If anything, the reverse is true. The nature
of 46 violations taken to court, for 32 large and small firms appearing in the record of the
hearings, had to do almost entirely with deficiencies or excesses of the active ingredient;
only three samples in 46 contained adventitious deleterious substances. Of 285 drug recalls,
most were for lack or loss of potency. but several very serious irregularities appeared. For
small firms, there were four instances of contaminated drugs, and one Instance where
the drug was suspected of causing agranulocytosis. For large firms, there were six instances
of contaminated drugs. and one instance of suspected agranulocytosis, three instances where
the drugs had caused liver damage or jaundice, one instance of induced acidosis, and one
instance (again a recall, not a legal action) where polio vaccine had caused widespread
paralytic poliomyelitis. Id., pt. 22. at 12137-61.
PAGENO="0409"
COMPETITIVE PROBLEMS I~ THE DRUG INDUSTRY 1959
lower prices. This would merely increase the monopoly power of the large firms
in those few markets where price competition exists, and would therefore in-
crease that part of the profits of the large firms which constitute moflol)O1Y re-
turns. The obvious remedy is to provide adequate inspection. Food and Drug
Administration Commissioner Larrick testified at the hearings that adequate
ethical drugs inspection could be obtained at a cost of an additional $3.418,000.~°
For the 22 largest firms in the industry in 1958, net profits before taxes
amounted to $562 million.48 A simple computation may throw some light on the
nature of the alternative. A reduction of net profits before taxes of $7121 million
would cut federal income tax receipts by 52 percent of this amount, or by $3703
million. The net gain would therefore be the $7.121 million saved by drug cus-
tomers, less the $3.703 million lost in tax~receipts, or the needed $3418 million.
If adequate drug inspection could establish confidence in lower priced drugs to
the extent that the resulting competition would lower major thug firm prices
and profits by as little as 1.27 percent before taxes, the savings realised would
pay for the expanded enforcement program. To this, the substantial benefits ob-
tained by the elimination of inferior drugs must be added.
Nowhere is the policy of fostering imperfection of market information so much
in evidence as in the controversy surrounding the issue of prescription by brand
name instead of by generic name. Drug advertisements must, according to law,
mention the appropriate generic name, but the generic name is universally given
in much less prominent type face,4° and is sometimes concealed by its appearance
in an unlikely place in the advertising copy, or else is suppressed in favor of the
full and formidable chemical nomenclature. These generic names do' not lend
themselves to practical use. Drug makers are free to designate the generic name
for any new compound they market, and generally do so in such a way as to
insure minimum use of the generic name in favor of the brand name.5° To take
a single example from among thousands, the compound with the brand name
"Darvon" is burdened with the generic name dextropropophyphefle hydrochlo-
ride.5' Dr. Frederick H. Meyers of the University of California epitomized the
predicament of the medical profession by commenting that `. . . unless these
generic names are disciplined, the United States Pharmacopoeia will eventually
become the dictionary of a nonsense language." 52
Such generic names are at the very least inappropriate, since the generic
name need serve no other purpose than simple identification of the compound.
The definitive and adequately descriptive name is the chemical name, which is
a verbal summary of the molecular structure, and which may considerably ex-
ceed in length even the most prolix examples of generic names. The generic
name is usually merely an essentially unenlightening encumbrance, useless to
chemist, pharmacist, and physician alike. An interesting contrast is provided
by the procedure for designating new names for insecticides. Rules exist which
require that names be short, distinctive, easily spelled, and in conformity w-ith
accepted scientific usage.53 No such rules exist in the case of pharmaceutical
preparations for human use.
The efforts made to suppress the use of generic names have been very effective.
Medical schools teach doctors to prescribe by generic name.54 in practice, hiysi-
clans have proved vulnerable to the usual advertising appeals, ~nd about 88 per-
cent of all prescriptions are written in terms of brand names.'~ By this means,
88 percent of the prescription market is~removed from the sphere of price com-
petition.
~° Id., pt. 22, at 12132.
4° Data submitted by firms to the Subcommittee. Profit estimate obtained by multiplying
total sales by the 25.8 per cent weighted average sum of profits after taxes and taxes.
Subcornm. Report 31.
5° Dr. Austin Smith of the Pharmaceutical Manufacturers Association was at one point
during the hearing unable to locate the generic name on an advertisement for a drug caJed
"Dimetane" until given a magnifying glass by the Subcommittee counsel. Hearings on
Administered Prices, pt. 19 at 10931.
5° Drug-naming procedures are described in id, pt. 21, at 11499, 11675, 11868.
51 Id., at pt. 21, at 11775-76.
52 Id., pt. 18, at 10401.
~ Testimony of Dr. C. 0. Wilson of Oregon State College, id., pt. 21, at 11505.
~° A survey sent to the 82 medical schools in the United States in 1960 brought 77 replies.
Sixty-four schools teach only generic terminology ; three teach generic and brand names
together; and ten used brand names only under certain circumstances, such as the monopo-
lization of a patented generic compound by a single firm. Subcomm. Report 226.
5° 1958 National Prescription Survey, Pharmaceutical Extension Service, Rutgers Uni-
versity, as reported in Hearings on Administered Prices, pt. 15, at 8776.
PAGENO="0410"
1960 CO1\'IPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The industry's reaction to the presence of Food and Drug Administration
regulation has been apparent primarily in the area of the certification of new
drug applications. In order to secure permission to market a new drug, the
applicant must submit a sufficient amount of clinical and experimental data
to establish that there is no significant danger connected with the proper use
of the drug, i.e., that its acute toxicity is sufficiently low to let it be considered
a safe drug. It takes time to conduct sufficient experiments and carry out
enough clinical studies to determine probable toxicity in general use. It takes
still more time to have the application studied and processed. In drug market-
ing, many firms are often working on the same product at the same time, and
e~c'h desires to cut the period between discovery and marketing to an abso-
lute minimum, for the order of priority in market appearance usually deter-
mines the relative sales ranking for different brands of a given drug. Conse-
quently, the motivation is to limit experimental and clinical work to the mini-
mum acceptable level,57 to skip stages in product development, such as the
pilot-plant stage. and to influence the staff of the Food and Drug Administra-
tion in such a way as to facilitate rapid approval.57
The response of the ethical drug industry to its specific economic frame-
work may be summarized as follows: the employment of the patent privilege
to erect barriers to entry into the production of specific patentable drugs, the
substitution of product differentiation for price competition, and the use of the
accompanying techniques of sales promotion to minimize the impact of the
price competition that might be offered by smaller firms. This is accomplished
in three ways: the vast bulk of advertising done by the major firms tends, by
its mere magnitude, to obscure the very existence of small, non-advertising
generic sellers; the employment of opaque brand names for advertised drugs
makes it formidably difficult for buyers to detect the existence of lower-
priced, generic equivalents; and the campaign of disparagement renders sus-
liect the quotation of a low price. No price competition need ever develop for
patented drugs; for non-patented drugs, product differentiation, the adoption
of deliberately confusing nomenclature, and the waging of a never-ending
campaign of disparagement against low-priced drugs can effectively substitute
for price competition, and prevent small, low-priced sellers from taking over
any appreciable amount of the prescription drugs market, even though the
absence of economies of scale Or other barriers to entry will permit small sellers
to undersell large by remarkable margins. This has been accomplished in the
face of the presence of governmental inspection to insure the quality of all drugs
nerely by the extension of the policy of disparagement to include the adequacy
of the Food and Drug Administration's facilities for making inspections.n
III. EVALTJATION OF MARKET PERFORMANCE IN THE ETHICAL DRUGS INDUSTRY
A. Varieties of competition among drugs
There seem to be three main dimensions of competition: price competition,
product competition, and product differentiation. While price competition is not
~ This Is naturally deplored by physicians and medical educators. Dr. C. D. Leake of
Ohio State University noted pertinently: "There is no shortcut from chemical laboratory
to clinic, except one that passes too close to the morgue." Id., pt. 18, at 10418.
Dr. Barbara Moulton. formerly with the Food and Drug Administration, testified that
excessive deference was at times shown to impatient applicants. She testified that if the
medical officer in charge of the evaluation of a new drug application Is not satisfied as to
the evidence of its safety, the applicant will frequently make an appointment with the
medical director. She continued: "I have known such conferences to be followed by an
order to the medical officer to make the new drug application effective, with the statement
that the company In question has been evaluating new drugs much longer than the medi-
cal officer, and should, therefore, be in a much better position to judge their safety." Id.,
addiction danger warning on the label of a tranquilizer was refused by her superior, whom
pt. 22, at 12025. She related an experience of her own in which her request to place an
addiction danger warning on the label of a tranquilizer was refused by her su'~erior. whom
she quotes as saying. "I will not have my policy of friendliness with Industry interfered
with." I'd., pt. 22 at 12032.
5S The large firms, acting through trade associations, have been Instrumental in con-
tributing to the inadequacy of governmental inspection. Prior to 1953, the Food and Drug
Administration had broad powers to inspect plants as well as products. Tbe Factory In-
spection Amendment of 1953, supported by the major firms, made it possible for drug
makers to refuse to allow inspection of significant phases of drug operations. Large as well
as small firms have not hesitated to avail themselves of this privilege. Testimony of
G. P. Larrick, Id., pt. 22, at 12113.
PAGENO="0411"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 1961
apparent, closer study reveals that even in the case of sales to retail druggists,
there is some price competition via quantity discounts and the shipment of "free
goods" in excess of the invoice quantities ordered. In the hospital purchases
markets, bids are frequently requested, and in some product lines, sharp price
competition may develop.
Product competition proper is more easily achieved in the chemical industries
than in some other fields. Research effort, whether domestic or foreign, public
or private, may culminate in the disc'ovei'y of a new compound with important
therapeutic applications. If one firm is awarded the patent, other firms will at-
tempt to modify the molecular structure of the compound in order to discover
a "different" and patentable therapeutic agent which will in all probability be
of use in the treatment of the same classes of disorders as the original drug,
but which will hopefully be more potent, or less toxic, or will at least yield a dif-
ferent variety and incidence of concomitant side effects.
Product differentiation is seen in its purest form when several firms obtain
licenses from the patent holder and then proceed to market the identical com-
hound under different names, each of which must then be intensively advertised.
The choice of a name for such a drug constitutes an exercise in product differ-
entiation at a higher level than is usually encountered in the commodity market,
since the name typically abandons or repudiates description of the good and
refers instead to nothing outside itself; stressing only its abstract (and pre-
sumnably unique) identity.59 Product differentiation efforts are of course also en-
listed in behalf of products which are~ physically different from each other.
Since "ethical" drugs are ethical in the sense that they cannot be bought over
the counter or advertised to the public at large, the market consists of the
private physicians and hospital pharmacists who can order them. This is a
relatively small and well-defined market which can be intensely saturated by
advertisements and the employment of itinerant salesmen. There is, in fact,
some evidence that the intensity itself, at least in some quarter.6°
The role of research as the driving force behind both the search for new prod-
ucts and the devising of minor molecular modifications on old ones should be
considered separately at some length. Lack of space, however, allows only a
brief summary. Spokesmen for the major drug firms invariably defend the
height of their prices, or the gap between computed costs and market prices,
by reference to the vast sums spent on research in the interests of advancing
the cause of health and medical science. It is also asserted that in the absence
of the patent incentive, research would disappear. Both statements may be
questioned. It may be doubted that they can be reconciled. It is probably difficult
to convince an economist that the primary purpose of drug research is not to
increase profits. This is economically desirable only if no cheaper way can be
found to insure an adequate supply of new and improved drugs. That research
is the monopoly of the large and profitable firms, or that the most profitable
firms do most of the research, may be questioned. That research outlays are a
major or even a very significant factor in the total cost picture for most major
drug firms may further be questioned. Finally, that the patent privilege is a
necessary incentive to elicit truly productive drug research may be disputed.
As a brief summary of the evidence in regard to these issues, the following facts
must suffice:
1. The size of the research budget as a per cent of the sales dollar varies
inversely with the size of the firm. For a group of 22 large firms and two
representative smaller firms, the most profitable firm (Carter, with a profit
before taxes of 43.8 per cent of sales) had the smallest research budget (2.7
percent of sales), while the least~ profitable firm (Panray, 10 per cent of
sales) is estimated to have had the highest relative research budget (15
per cent of sales), during the year 1959.
~ An example may make the contrast with the usual type of product differentiation by
brand name more apparent. Dr. Solomon Garb of the Albany Medical College of Union
University submitted that if the drug makers took over the manufacture of canned beans,
then rather than selling "Pfizer's Beans" or "Parke, Davis and Co. Beans" they would
prefer to coin novel and unique anagrams such as "Sneabs" or "Nabes," or adopt unde-
scriptive slogans like "Lo Cals" or "Hi Pro's" and abandon entirely the generic noun. Id.,
pt. 18, at 10481.
~° See particularly the testimony of Drs. Bowes, Console, Seidell, Weinstein, Bean, Garb
Leake, and Meyers, id., pt. 18.
PAGENO="0412"
1962 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
2. For the 22 largest drug firms in 1959, research outlays comprise the
smallest part of the budget. Of the sales dollar, 32.1 cents represented cost
of goods sold, 25.8 cents was net profits before taxes, 24.8 cents went for sell-
ing costs, 11.0 cents for general and administrative expenses, and only 6.3
cents for research.6' Selling costs average about 394 per cent of research
costs. If the cost of drugs is excessive, it is not because of the magnitude of
the research budget.
3. The growth of a large American drug firm has typically depended but
little upon the quality of its own research. The largest firm Pfizer did not
begin to grow rapidly until 1941, when it w-as 92 years old, and then its
growth was largely in the nature of a w-indfall gain. Pfizer was one of only
a few firms with experience in fermentation methods when w-artime demand
arose for fermentation-produced penicillin (discovered in England). The
same is true for Bristol, which owes its present position largely to the pur-
chase in 1943 of a small firm with some experience in fermentation tech-
niques.62 Smith, Kline, and French ow-es its growth not to its own research,
but to its ability to take advantage of promising Europeon discoveries. To
a lesser extent, this is true of Carter. Additional examples could readily
be given.
4. More basically new drugs have been found in European countries with-
out drug product patents than in the United States with its unrestricted
drug patent privileges. On the basis of an inclusive list (including molecular
modifications as well as basic new- discoveries) drawn up by the Senate Sub-
committee staff and revised by the Pharmaceutical Manufacturers' Associ-
ation, since 1886, 82 drug discoveries have been made in countries without
product patents, compared with 79 in the United States, only (30 of which
~sere found in the laboratories of drug firms. Fifteen drugs were found in
foreign countries w-ith drug patents.63 Hence, 75 drugs were found by Ameri-
can commercial firms and in foreign countries with product patents, w-hile
101 drugs have been found in countries without product patents, and by
noncommercial American investigators.
5. Most American drug firm research is not basic research and is not very
fruitfu1.~ With regard to four drug categories, American industry is respon-
sible for the basic discovery in only one area, cortisone in the corticosteroid
field. In oral antidiabetic drugs. the basic drug was found in Germany; in
tranquilizers in France, Switzerland, and England; and in antibiotics, in
England. Minor modifications have been made by American firms in all four
areas. for it is in this sort of research (or more properly, product develop-
ment) that the domestic industry excels: the further exploitation of an
already accomplished basic discovery. It is in this sort of product develop-
ment-research that the risks are least.' Such activities in this country have
been fruitful, as in the field of broad-spectrum antibiotics, where important
new drugs were found, but no basic research was involved. The basic dis-
coveries bad been made in England in 1929. the problems of large-scale pro-
duction had been solved by government scientists during the second world
u-ar, and the observation w-as current that many molds occurred in nature.
some of which might be capable of yielding new antibiotics. All that re-
mained to be done w-as to collect and screen soil samples. a tedious, trial-
and-error procedure of a routine nature, w-hich the technicians involved have
61 Subcomm. Report 31. -
62 Hearings on Administered Prices, pt. 24, at 1382o.
~ Subcomm. Report 116-18.
64 Dr. H. J. Weinstein, former medical director of Pfizer, testified that much drug firm
research "has little relationship to what most people engaged in academic and research
activities would consider to be scientific research." He gave as examples the practices of
molecular manipulation, the aimless proliferation of combinations of existing drugs which
lack flexibility and compound the problems of dosage and toxicity, and the "battle of the
additives" where much energy is devoted to showing that the addition of certain substances
to an existing drug will intensify its activity. Hearings on Administered Prices, pt. 18,
10243-44. Dr. Console testified: "I think the majority of it [research] is in that cate-
gory . . . with many of these products. it is clear while they are on the drawing board
that they promise no utility. They promise sales." Id., pt. 18, at 10380.
65 Dr. A. D. Console, former medical director of Squibb, testified that "they [the drug
firms I stress that there are many failures for each successful drug. This is truo since it is
the very essence of research. The problem arises out of the fact that they market so many
of their failures." He continued, "I doubt that there are many other industries in w-hicli
research is so free of risks . . . with a little luck, proper timing, and a good promotion
program, a bag of asafetida with a unique chemical side chain can be made to look like a
wonder drug." Id., pt. 18, at 10372-37~3.
PAGENO="0413"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1963
sometimes referred to as "making glorified mud-pies." ~ Furthermore, since
no basic research was involved, the returns to more and more intensive
exploitation of the same method have approached zero. No important anti-
biotic has been found in the United States since 1955, and the same is true
in other drug fields.
B. Degree of product substitutability among, drugs
With this background, which cannot pretend to do justice to the many com-
plexities of the research issue, it is possible to turn to the evaluation of market
performance in the four drug product categories investigated during the Senate
Subcommittee Hearings, corticosteroid hormones, tranquilizers, oral antidiabetic
drugs, and antibiotics. These four categories accounted for about 42 per cent of
all ethical drug sales in 1959. The corticosteroids market will first be described,
and will then be used as the point of departure for parallel comparisons with
other drug product categories. By virtue of the emphasis which it places upon
the continuous obsolescence of existing corticosteroids through the development
of new synthetic steroid hormones eligible for patent protection, this market may
be taken as typical of ethical drugs as a whole.
The important corticosteroids in 1959 were seven in number: cortisone, hydro-
cortisone, prednisone, prednisolene, methyiprednisolone, triamicinolone, and dexa~
methasone. All of the corticosteroids are to some extent substitutes for one
another in the treatment of many disorders. Striking evidence of this is avail-
able on the basis of a 1959 study by the Pharmaceutical Manufacturers Associa-
tion, showing that all seven of the major cor-ticosteroids were prescribed in every
one of the seven disease groups requiring such therapy, with but two exceptions.67
No one of these drugs i-s apparently characterized by advantages so great that its
actual use is always exclusively in order in a certain disease group. Obviously, for
individual diseases or for individual classes of patients this may not be the case,
for the disease groups are not very narrow-ly defined, and even within the limits
of a single disease, patients of different ages or with different degrees of prone-
ness to complications might require different specific drugs. Nevertheless, it is
striking that in no instance is as much as -two-fifths of any disease group treated
by a single drug. In such a situation it may well be the case that no drug has any
commanding advantage over all the others for any major disease group; hence
the patterns of demand may be greatly influenced by sales efforts, and the prod-
ucts may be differentiated in terms of their claimed freedom from side effects,
their relative potency, their prol)able impact on certain types of patients, and in
fact upon anything but their relative prices.
Similar evidence is available for the tranquilizers market from the same
source. This time, each of the seven major tranquilizers is prescribed in every
one of the five disease groups requiring such therapy.68 The maximum usage in
any disease group is 41.4 per cent. If all drugs were prescribed at random for all
disease groups, each would be used in 14.3 per cent of -all treatments. In only four
instances is the frequency of use more than twice what one would expect on the
basis of random choice. Substitution among drugs is apparently as much -the rule
in tranquilizers as in corticosteroids.
The oral antidiabetics drug market is more readily described. There are three
such drugs, tolbutamide, chioropropamide' and phenformin. All seem to be fairly
close substitutes, and tolbutamide (Upjohn's "Orinase") and chiorpropamide
(Pfizer's "Diabinese") have been subjected to strenuous competitive product
differentiation programs. The market shares follow the order of priority of mar-
ket appearance, the earliest to appear having the largest share. These products
niay not be perfect substitutes, however, for it has been claimed that chiorpro-
pamide is effective in some cases where tolbutamide is not,69 and there is evidence
to show that side effects are more frequent and severe with chlorpropamide.7°
It is impossible to summarize the extent of possible substitution among anti-
biotics in brief compass. The main areas in which substitution is possible are
among the five "broad spectrum" antibiotics, all of which are considered inter-
changeable in many uses, and all of which are highly advertised; and between
~ Subcomm. Report 124.
67 Hearings on Administered Prices, pt. 19, at 10826. (One drug was prescribed In only
five groups.):
68 Id.. pt. 19, at 10827.
69 Testimony of Dr. C. J. Donovan, formerly of Upjohn', id, pt. 20, at 14070-72.
70'Testimony of Dr. Henry Dolger, Mount -Sinai Hospital of New York, id, pt. 20, at
11146.
PAGENO="0414"
11964 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
penicillin, which is unpatented and not highly advertised, and the broad spec-
trum antibiotics. Within its range of activity, penicillin is usually more potent
than the broad spectrum drugs.1 But micro-organisms tend to become resistant to
certain antibiotics; many of the newer antibiotics are said to be effective against
strains resistant to the older ones. Among the broad spectrum drugs, chlor-
amphenical may currently be regarded as somewhat distinct from the others, in
that fewer strains have become resistant to it, and for two reasons. First, the
use of chlora.mphenicOl was discouraged to some extent in the early 1950s when
publicity was given to the incidence of its lethal side-effects; hence its less fre-
quent use during that period provided less occasion for the development of
resistant strains. Second, this drug resembles penicillin more than do the other
broad-spectrum antibiotics in that microorganisms do not easily develop re-
sistance to it ;72 chloramphenicOl is, in fact, the most lethal of the broad spectrum
antibiotics to micro-organism and patient alike.78 The incidence of side effects is
often more important than the spectrum of primary efficacy of the antibiotic as
far as drug demand is concerned. Advertising may influence prescription demand
by stressing the alleged absence of side effects as well as therapeutic efficacy.
The absence of any advertising may conversely prevent the wide-scale use of
extremely efficacious antibiotics. Bacitracin is a case in point. The drug is so
fiee of side effects that some of its preparations may be bought over the counter.
The patent is ow-ned by the federal government, and royalty-free licenses may be
obtained. The only ethical drug firm currently producing this drug, Pfizer, also
sells its own patented antibiotics, tetracycline and oxytetracycline, and does no
sales promotion for bacitracin, save of a negative sort. Originally the Food and
i)rug Administration restricted its use to hospitals, but later removed the re-
striction, but as of September, 1960, Pfizer (despite protests from the discoverer
of the drug) continued to print the former restriction on bacitracin labels and
brochures.74
C. Pi-odiict competition in individilal drug markets
The earliest of the corticosteroids, cortisone, was first synthesized in 1944 by
Merck, and commercial production began in 1948.~ The drug first proved its
effectiveness in relieving the symptoms of rheumatoid arthritis and related clini-
(al syndromes, although concomitant side effects presented complications. Corti-
sone is a naturally occurring substance in the human body (adrenal cortex) and
no product patent could be obtained. Process patents presented no real barrier to
entry, and Merck soon had competition from a number of large and small firms.76
Squibb discovered fludrocortisofle in 1953, by modifying the hydrocortisone
molecule. Squibb obtained a patent, sells the drug under the name "Florinef"
and has apparently licensed no other sellers. More potent than its parent, its
side effects are also more severe.77
Prednisone was the next corticosteroid to be found, along with the related
compound prednisolone. These discoveries are credited to the Syntex Corporation
of Mexico in 1955. By nature still further modifications of the cortisone molecule
they are generally felt to constitute some actual improvement over the parent
substance. They are used to treat the same disorders, and have about five times
the antirheumatic activity of the parent steroids.78 No patent has yet been
awarded for either drug, and several large firms and many small firms sell both
drugs.
Some medical authorities hold that no essential advances have been made in
the corticosteroids field since prednisorie and prenisolone were introduced. Up-
john found methyiprednisOlofle in 1957 as a further modification of the pred-
71 L. S. Goodman & E. Giliman, The Pharmacological Basis of Therapeutics 1322--23
(1056).
72 Testimony of Dr. Maxwell Finland of Harvard University, Hearings on Administered
Prices, pt. 24, 13947-49.
~ In a nation-wide survey of all severe antibiotic reactions from late 1953 to early 1957,
25 deaths from severe skin reactions and blood duscrasias were associated with the use -
of chiorampheflicol. and only two deaths from the same causes were associated w-ith tile
use of the tetracyclines. L. Meyler. Side Effects of Drugs 129-30 (1960).
~ Testimony of Dr. F. L. Meleney of the University of Miami, Hearings on Adminis-
tered Prices. pt. 24. at 14195.
~ Testimony of Dr. H. C. Kendall of Merck. id., pt. 14, at 8018-20.
~ Testimony of J. T. Connor, president of Merck. id., pt. 14, at 8025-26.
Modern Drug Ejlcyclopedia and Therapeutic Index 471 (B. P. Jordan ed. 1958).
78 Meyler, op. cit. supra note 73, at 148.
PAGENO="0415"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1965
nisolone molecule, and claimed that it had i!ewer side effects than other cortico-
steroids, while having the same general field of use. Upjohn obtained the patent
and then licensed Schering to produce the drug.
Lederle brought out triamcinolone in 1958, another molecular modification of
cortisone, with a potency about equal to that of methylprednisolone, with the
same general therapeutic applications as the other steroids, and with the same
claim of fewer side effects. Lederle obtained the patent, and licensed only Squibb.
The latest modification of the cortisone molecule to be marketed is dexa-
inethasone, brought out by Merck and Schering in 1958. This drug is said to be
the most potent of the corticosteriods, being about seven times as potent as
prednisone,7° and is also claimed to have fewer side effects. No patent has yet
been awarded. Merck, Schering, and CIBA were producing this drug in 1959.
A Survey of worldwide experience with these newer drugs prompted the opin-
ion that their side effects are".. . certainly no less than those of prednisone and
prednisolone. .. . With the newer drugs, digestive accidents, fractures, adrenocor-
tical inhibition, and indfections are still the main dangers. In contrast to the
manufacturers' statements, risks are in no way eliminated by the new deriva-
tives." 88 Since it takes years to determine the true incidence of side effects for a
given drug, medical opinion is as yet inconclusive, but the most frequently ex-
pressed view is that all the corticosteroids will prove to have the same side effects
in the long run.8' Thus product improvement has here taken the form of product
competition by means of trial-and-error molecular manipulation, and product
differentiation by means of massive selling efforts, rather than by revolutIonary
advances which would make all previous corticosteroids obsolete.
In other drug markets, the development of new products also relies heavily
upon the devising of molecular variations on former products. In oral anti-dia-
betic drugs, tolbutamide was found in Germany and licensed by. Upjohn for sale
in the United States as "Orinase." Upjohn carried out extensive clinical testing,
and the drug was released for general prescription use in the United States in
June, 1957. Other firms conducted research in order to produce a patentable sub-
stitute compound. Lilly and Pfizer came up, with chiorpropamide, a slight molec-
ular modification of tolbutamide, and Pfizer began to market this product as
"Diabinese" in November, 1958. Lilly withdrew its patent application in favor of
Pfizer; at present no patent has yet been awarded. United States Vitamin and
Pharmaceutical Corporation later discovered phenformin, a molecular modifica-
tion of certain compounds subjected to study by many scientists prior to the dis-
covery of insulin in 1920, and marketed it as "DBI." 82 Tolbutamide appears to be
unusually free from side effects. Chlorpropamide, on the contrary, is not; nor is
phenformin.83 Some physicians are of the opinion that advertising, rather than
therapeutic superiority, is responsible for the sales of the later oral antidiabetic
drugs.
In the tranquilizers market, the basic research on the more potent tranquilizers
was done largely in France, with the development of certain phenothiazine com-
pounds with sedative effects. Smith, Kline and French was licensed to sell chior-
promazine ("Thorazine") and prochlorperazine ("Compazine") in the United
States. "Thorazine" was released for sale in 1954 as an anti-emetic; the tran-
quilizer properties were advertised only later when it was found that a market
existed in the mental institution area. "Compazine" was introduced in 1956 as a
tranquilizer, and is a phenothiazine compound similar to chlorprornazine. The
market for tranquilizers grew, and so did the number of such drugs. Promazine
was marketed as "Sparine" by Wyeth in 1957; again the drug is a molecular
modification. Amusingly, Wyeth began to advertise promethazine hydrochloride
("Phenergan"), an antihistamine originating in France, as a tranquilizer.84
78 Report from Pharmaceutical Manufacturers Association, Hearings on Administered
Prices, pt. 19, at 10845.
80 Meyler, op. cit. supra note 73, at 148.
81 Testimony of Dr. Russell Cecil, consulting medical director of the Arthritis and Rheu-
matism Foundation, Hearings on Administered Prices, pt. 14, at 7985.
82 Subcomm. Report 123.
83 The use of chiorpropamide is associated with liver damage and other mischief. Dr.
Henry Dolger, in testifying at the Hearings, concluded: "I deplore the premature prosecu-
tion of unnecessary and inadequately investigated modification of sound drugs and their
indiscriminate and unsubstantiated promotional campaigns." Hearings on Administered
Prices, pt. 20. at 11149.
841n 1952, 20 per cent drowsiness was noted as a side effect. In 1959, it was stated in an
advertisement that "Phenergan relieves apprehension, relaxes the patient, and produces
light sleep." Id., pt. 16, at 9272.
PAGENO="0416"
1966 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
Other molecular modifications followed: perphenazine ("Trilafon"), trifluopera-
zine ("Stelazine"), and trimepazine ("Temaril") ~ In the mild tranquilizers
field, a Dr. Berger had discovered mephenesin, a muscular relaxant, in England,
and came to the United States to patent a molecular modification, meprobamate,
as a tranquilizer, which was marketed by Carter ("Miltown") and Wyeth
("Equanil") in 1955. Both products were highly advertised, a condition of
Wyeth's license from Carter being that Wyeth would spend at least 20 percent
of total meprobamate sales revenue in promoting its sale.8° Smith, Kline and
French found itself with the potent tranquilizers and no mild tranquilizer; it
then began to advertise "Compazine" as a mild tranquilizer. "Compazine" is
actually much more potent than "Thorazine." 87 Medical opinion inclines to the
view that none of the later modifications of the original phenothiazines has been
a marked improvement. Dr. Lehmann testified: "There hasn't been a very much
better one than the very first ones that came out, in the six or seven years of
frantic research since then." 83 No diminution in the incidence of side effects
could be demonstrated.88
The situation in the antibiotics market is largely the same, but of greater
complexity. The discovery of penicillin in England in 1929, and the further
discovery of streptomycin at Rutgers University in 1943, demonstrated the pos-
sibility that many naturally occurring molds might have antibiotic properties.
Many drug makers set themselves the task of the trial and error screening
of thousands of soil samples. The first success was encountered at Yale Uni-
versity in 1947 under a Parke, Davis grant, the antibiotic chioramphenicOl
("ChloromyCetin") being isolated. Lederle next discovered chlortetracycline in
1948. and named it "Aureomycin." Pfizer then found oxytetracycline in 1949,
and designated it "TerramyCin." Pfizer managed to elucidate the chemical struc-
ture of oxytetracycline, and with the molecular structure known, molecular
manipulation became possible in antibiotics, although to a more limited extent
than is possible with drugs which can be made synthetically. In 1955, Pfizer
developed tetracycline as a molecular modification of chlortetracycline.°0 The
screening of soil samples continued. Lilly in 1952 discovered erythromycin. Pfizer
in 1953 brought out a closely related analog, carbomycin, which affected the same
bacteria as erythromyCin, but which proved less effective in human disease than
in the test tube. and which was finally withdrawn from the market in 1990.81
In 1956. Pfizer introduced another closely related analog of erythromycin, oleando-
mycin: and in 1957, a modification of oleandomycin, triacetyloleandomycin, which
was advertised as a great advance, since the same oral dose as in oleandomycin
llrocluced somewhat higher concentrations of the drug in the blood. To offset this,
Lilly in 1958 modified its original erythromycin to market it in the form of its
propronyl salt, which is said to produce a higher concentration of this drug in
the blood than triacetyloleandomycin.°2 Hence, five drugs were produced to serve
the purpose of one (since slightly higher doses of erythromycin could serve the
same purpose as the newer derivatives). Another broad spectrum antibiotic was
brought out by Lederle in 1959, demethylchlortetracyCline, or "Declomycin," a
sPght modification ef Lederle's chlortetracycline. Fewer side effects than other
broad spectrum antibiotics were claimed. Not all physicians agree.83 The later
Tenlclilins are patentable variations of the earlier original penicillin w-hich the
British discoverer did not bother to patent. In 1952, four American firms made a
substitution of benazthine for procaine in the penicillin compound; the parties
~° Subcomm. Report 121.
86 Testimony of Irene Till, Hearings on Administered Prices. pt. 16, at 9208.
~ Testimony of Dr. Heinz Lehmann of Verdun Protestant Hospital, Montreal, id., pt.
16. at 9068-73.
83 Id., pt. 16. at 9029.
89 Physicians employed by drug firms who testified sometimes found it easier to ignore
side effects than one suspects their drug users could. Dr. Gibson of Merck went so far as
to introduce his aesthetic preference for facial contours as the determinative criterion for
evaluating the side effect of "moon face" encountered in steroid therapy. In commenting
noon this side effect (caused by redundant retention of cellular fluids, a circumstance con-
clucive to additional strain on the cardiovascular system), Dr. Gibson testified: "When
those people {malnourished patients] got a moonface I think it is a desirable effect." Id.,
pt. 14. at 8188.
O~) Subcomm. Report 116-17, 124.
81 Testimony of Dr. H. F. Dowling of the University of Illinois, Hearings on Adminis~
tered Prices. pt. 24, at 14167.
92 Id., pt. 24. at 14167-68.
93 Several of the physicians who testified during the Hearings professed to see little dif-
ference among the broad spectrum of antibiotics. See id., pt. 18.
PAGENO="0417"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1967
agreed to concede priority to Wyeth in return for suitable cross-licensing assur-
ances. In 1951, Lilly secured a patent which embraced phenoxymethyl penicillin
among other compounds. Lilly apparently did not recognize its value; not until
an Austrian firm applied for an American patent on it did Lilly learn of its
potential. Lilly then entered into cross-licensing agreements with that firm, and
did not market its product until 1955, four years after discovery. The next
penicillin, phenoxyethyl penicillin, is a slight modification of phenoxymethyl
penicillin, was found in England, and is marketed under four different brand
names in the United States.°4
D. Price competition in individua' drug mairkets
In general, the presence or absence of price competition in drugs depends
upon the presence or absence of patent protection. Even where some price com-
petition exists, however, it is rendered all but ineffective in the prescription
market because of the inability of the lower price of the smaller producer to
overcome the disadvantages of the obscurity inherent in small size and the dis-
paragement associated with low price. In such cases, effective price competition
is limited to the hospital formulary and government bid markets. Unpatented
drugs compete with patented drugs in two out of the four markets under consid-
eration, there being no important unpatented tranquilizers or oral antidiabetic
drugs.
In the corticosteroid hormones market, cortisone and hydrocortisone were
ruled ineligible for patent protection, and no patents have yet been issued for
prednisone, prednisolone, and dexamethasone; all other corticosteroids have been
patented. For cortisone, the price was $20 per gram in October, 1951, but intense
price competition forced the price down to $5.48 per gram in 1954, at which level
it has since remained constant.95 But the case of prednisone is particularly instruc-
tive and merits review at some length. Five major drug firms filed patent appli-
cations for prednisone, and an interference proceeding was declared by the
Patent Office. The parties involved undertook to make interim arrangements
among themselves, involving the payment of "interim royalties" to one of the
firms, and as a result all five firms began selling prednisone at the identical price
of $17.90 per bottle of one hundred 5 milligram tablets. The Syntex Company of
Mexico later became a party to the interference proceedings but was not allowed
to participate in the interim arrangements because of its known reputation for
price competition.96 Syntex then retaliated by selling the drug in bulk form to
small drug firms in the United States, and two of the major firms met its lower
price for bulk sales. Syntex then cut prices still lower, from $10.01 per gram in
1957 to $2.36 per gram in late 1959.°~ The major firms' prices at wholesale for the
finished drug remained constant at $17.90 per bottle, but small firms were selling
at prices as low as $1.75 per bottle by 1959.98 The lower-price brands, however,
were unable to obtain more than eight or nine per cent of the prescription market.
In the absence of extremely imperfect market information, such great differences
in price would be impossible. The larger firms apparently found it much more con-
genial to disparage the products of the lower priced sellers than to meet their
competition.
Had they cut prices, however, there would have been a great deal of room
for price competition. The actual full cost of production for a bottle of one
thousand 5 milligram tablets of prednisone is only $8.99.~~ This cost is but 5.07
per cent of the equivalen wholesale price of $179.00.
No price competition whatsoever developed among the patented corticosteroid
hormones. Drugs with increasing potencies per gram appeared from year to year,
but the price per tablet of each new drug was set so as to achieve a price per
equivalent therapeutic dosage unit identical with those of earlier drugs with
different dosage units per ta'blct. In 1958, however, two firms applied for patents
upon dexamethasone and marketed it (while patent interference proceedings
were still pending) at a price ten per cent below that being charged for its pat-
"Subcomm. Report 124-25.
~ Hearings on Administered Prices, pt. 14, at 7884.
9815., pt. 14, at 7920.
~ IS., pt. 14, at 8045.
98 Subcomm. Report 17.
9° Hearings on the Drug Industry Antitrust Act before the Antitrust Subcommittee of
the House Comm. on the JudIciary, 87th Cong., 2d Sess. 15 (1962). (Hereinafter cited as
"Hearings on Drug Industry Antitrust Act.")
81-280 O-68-pt. 5-27
PAGENO="0418"
1968 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
ented subStitutes.10° No small firms entered production, however, because of the
higher costs involved than in the case of prednisone. Here, lower prices could
be quoted by major firms in spite of appreciably higher costs because of the very
large profit margins made on substitute patented drugs.
In the oral antidiabetic drugs market, no comparisons between the prices
of patented and unpatented drugs can be made in view of the ainence of the
latter. Costs may, however, be compared with prices in the case of "Orinase"
(Upjohn's name for tolbutamide). The Senate Subcommittee staff estimated
the cost of producing one thousand 500 milligram tablets to be $6.86, which is
only 8.0 per cent of the wholesale price of $83.40. Upjohn, however pays a royalty
of 7.5 per cent of the sales price to the German patent holder, which increases
the cost to 15.5 per cent of the wholesale price.10'
None of the tranquilizers is unpatented, but price competition has been allowed
to develop in the case of a single drug, reserpine. CIBA obtained the patent, but
elected to take the unique course of `licensing the patent widely, and also allow-
ing the licensees to sell the drug in bulk powder form.'°' The licensees sold the
bulk powder to small firms which fabricated their own capsules and proceeded
to cut prices by as much as 90 per cent below CIBA's price. CIBA charged a
wholesale price of $39.90 per bottle of one thousand 25 milligram tablets; other
firms quoted prices as low a Panray's $2.65.'°~ The cost of production of such a
bottle is a matter of record: sixty-three cents.'°4 This is only 1.6 per cent of the
wholesale price.
Price competition in the antibiotics market is brisk for unpatented drugs, such
as the earlier penicillins, and for streptomycin, the patent for which was licensed
by Merck to several other firms. Price competition is, however, extremely slug-
gish for the later patented antibiotics.
Penicillin declined in price by about 85 per cent between 1945 and 1948, and
by almost 90 per cent between 1949 and 1952.'°~ Between 1952 and 1960 the price
fell by another 80 per cent, although there was a marked increase from 1955
to 1957. Increasingly efficient production methods, coupled with the persisting
competitive pressures resulting from free entry, made possible such a decrease
in the price of this extremely important drug. Streptomycin prices declined by
95 per cent, 60 per cent, and 88 per cent for the same intervals, or by about the
same amount as penicillin, and for the same reasons.
Broad spectrum prices show very different patterns. Lederle set an initial price
of $15.00 for sixteen 250-milligram capsules of chiortetracycline. This may have
been above the profit maximizing price; within two months it was cut to $10.00,
the same price at which Parke, Davis entered the market in March. 1949. On
February 1, 1950, both sellers cut prices by 20 per cent, to $8.00. In April, 1950,
Pfizer entered the market with oxytetracycline, selling at $8.40, 5 per cent above
its rivals. In May, 1950, these rivals cut their prices by 25 per cent, to $6.00, but
Pfizer did not meet the cut until November. On September 27, 1951, Pfizer cut
its price by 15 percent to $5.10, and by November 1 its rivals had met the price
cut. Tetracycline and demethylchlortetracycline came later, and sold also for
$5.10. No further price cuts were made for almost nine years, when Pfizer in
August, 1960, again acted as the price leader by initiating a 15 percent "trade
discount" adjustment which eventually became a 15 per cent price cut to $4.34~106
(This price cut came one month before the Senate Hearings on antibiotics were
scheduled.)'°' From 1948 to 1949, broad spectrum prices fell by one-third; from
1949 to 1951, by 49 per cent, and from 1951 to August, 1960, by zero per cent. It
100 Hearings on Administered Prices, pt. 14, at 8020.
101 Subcomm. Report 20.
102 Reserpine is the name given to the active substance isolated from the root of Rau-
wolfia Sepentina, a species of tropical plant the medicinal properties of which had been
known in India for centuries. CIBA was therefore perhaps uncertain as to the validity of
its patent.
103 Hearings on Administered Prices, pt. 16, at 9369.
10~ Hearings on Drug Industry Antitrust Act 16. (Cost data from McKesson and Robbins.)
105 FTC, Economic Report on Antibiotics Manufacture 164, 179-80 (1958).
100 Hearings on Administered Prices, pt. 24, at 13664.
`~ Mr. Duncan of Lederle may have failed to sense the humor of the situation at one
point during the Hearings. He testified that Pfizer had cut its prices on a Saturday in
order to "steal a march" on the industry. The following colloquy ensued:
"Senator KEFAUVER: You mean after ten years of operation, they would suddenly steal
a march on you `1"
"Mr. DUNCAN: Yes, Senator. These things happen very rapidly. If you can get any kind
of an advantage on your competitors, you try to do so." (Hearings on Administered Prices,
pt. 24, at 13728.)
PAGENO="0419"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1969
can be maintained that since 1951 the costs of producing broad-spectrum anti-
biotics have fallen by about as much as the cost of producing penicillin, since
the production methods employed are broadly the same.108 A decline of 90 per cent
in penicillin prices since November, 1951 is to be contrasted with a 15 per cent
decline for broad spectrum antibiotics. Even if broad spectrum costs fell only
half as much as penicillin costs, the burden on the drug consumer due to the
absence of pure competition in broad spectrum antibiotics would be an extra 30
percentage points in price.
The Senate Subcommittee staff made production cost estimates for chioram-
phenicol, and made public Bristol's tetracycline production costs. It cost Bristol
$1.07 to produce a bottle of one hundred 250 milligram tablets which sold at
wholesale for $30.00. Production costs were therefore 5.5 percent of the wholesale
price.100 Bristol paid a royalty of $1.03 to Pfizer, increasing Bristol's costs to 8.9
percent of the wholesale price."0 The estimated cost of producing a bottle of one
hundred 250 milligram capsules of chloramphenicol (the production and sale of
which is monopolized by Parke, Davis) is $1.52, or only 5.0 percent of the whole-
sale price of $30.00.'1'
Limitations on space prevent adequate treatment of antibiotics price competi-
tion in the government bid market. For unpatented drugs, competitive bidding has
reduced prices steadily. For the broad spectrum antibiotics, only sporadic com-
petition was manifest until foreign bids were accepted for tetracycline at less
than half domestic bid levels. Price competition was absent until foreign firms,
not encumbered by restrictive patent agreements, were allowed to bid.'12
iv. CONCLUSIONS
It seems evident, particularly on the basis of experience in the corticosteroids,
tranquilizers, and antibiotics markets, that genuine price competition among
ethical drugs is effectively prevented, for the most part, by the existence of
product patent privileges. Patent or patent application holders may exercise
restrictions on output, and the resulting high level of prices used to finance sell-
ing campaigns which contribute to the otherwise serious imperfections of market
information and make it impossible for small sellers of generic name products
to obtain any significant share of the retail prescription market, even for those
products not (or not yet) protected by patents.
It is possible that the ethical drug industry might be highly competitive, per-
haps almost purely competitive, in the absence of patent protection and the brand
name monopoly or oligopoly it makes possible. The only important economy of
scale in most drug markets seems to be that of large scale advertising and distri-
bution. The extent to which the costs of financing a sizable selling effort can in-
crease the price of a product is well illustrated by the experience of the Panray
Corporation in attempting to market its own brand of the tranquilizer reserpine,
which is highly advertised by CIBA under the brand name "Serpasil" and is sold
by CIBA at a wholesale price of $39.50 per bottle of one thousand 25-milligram
tables. Panray found it impossible to sell to druggists for less than $20-$21 under
its own brand name, "Serpanray." After it abandoned its selling efforts, it was
able to make a profit by selling reserpine under its generic name at $2.~5 per
1,000 tablets."1
Simple yet radical remedies are required for the institution of competition in
the industry. The first step is the abolition of the patent privilege as it applies to
drug products. Most of the countries in the world with patent laws grant no
patents on drug products as a matter of public policy. Provisions should be made
for the compulsory licensing, at reasonable royalty rates, of drug process patents.
It is often claimed that such measures would destroy the incentive necessary for
108 Subcomm. Report 82. Chloramphenlcol costs have probably declined by still more
since Parke, Davis discovered a way of producing this drug synthetically.
~ on Drug Industry Antitrust Act 17.
110 HearIngs on Administered Prices, pt. 24, at 13876.
luId. pt.24 at13971-78.
112 The hypothesis of the absence of real price competition Is afforded some support by
the testimony of Admiral Kplckerbocker. The decision to seek foreign bids was made
following a meeting of Military Medical Supply Agency officials with Pfizer representatives,
during which meeting, according to Admiral Knickerbocker's testimony, "Mr. Cooney
[Pfizer's sales representative] In rather an unguarded moment, made the statement that
the price of tetracyline would stay where It was until Pfizer did something about It."
Id., pt. 24, at 13819. Had all firms been yoked by patent restrlclons, his posilon would
have been stronger.
218 TestImony of Myron Pantzer of Panray, Id., pt. 16, at 9366-68.
PAGENO="0420"
1970 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
pharmaceutical research. A major shift of pharmaceutical research endeavor
from private firms to public, university, and foundation channels could scarcely
avoid resulting in gains. Private firms carry on relatively little basic research,
and most of their so-called research budgets would probably better be described
as applied research, product development, and product promotion. A large pro-
portion of some very scarce human resources is wasted in the competitive duplica-
tion of research programs carried on concurrently by virtually all the major firms
as is attested to by the virtually simultaneous discovery of many of the cortico-
steroids and antibiotics. In the absence of economies of scale, the removal of
patent protection would induce new entry of many firms producing under con-
stant cost conditions. The best remedy for a vertical or nearly vertical industry
demand curve is a horizontal industry supply curve.
The second step is to make Food and Drug Administration inspection and
regulation more adequate to meet the needs of the drug buyer and prevent his
exploitation. The Food and Drug Administration should be given sufficient addi-
tional funds to allow it to inspect adequately the operations and products of all
drug makers, large and small. It should further be given authority and addi-
tional funds requisite for the control of all drug advertising, and the labelling
laws with respect to drugs should be amended to require that all ethical drugs
be identified solely by the generic name, in conjunction with the name of the
seller employed as an adjectival modifier. All advertising as well as labelling
should be undertaken with respect to the generic name, coupled with the name
of the seller, rather than concentrating upon the promotion of quasi-chemical
syllable assortments. All advertisements should prominently feature suggested
retail prices.
Both measures would tend to promote the same ends. The abolition of product
patents would lead to the elimination of the monopoly profits which are neces-
sary to finance intensive selling efforts and wasteful product-modification ori-
ented research, and the departure of such profit possibilities would lessen the
need for such advertising and the motivation for such research. The reform of
drug nomenclature and advertising would enable physicians more easily to be-
come aware of the nature of the drugs they prescribe, of the existence of lower-
priced drugs, and of the precise nature of the alternative available in the drug
market. The insuring of the adequacy of drug products inspection would render
specious all activities devoted to the disparagement of lower-priced drugs.
Uncompromising measures, rigorously enforced, are indicated in order to halt,
and in time to remedy, the existing misallocation of resources in excessive selling
efforts, duplicative research and product development programs, and exception-
ally high profit levels-all of which are characteristics of the industry and much
of which is paid for by individuals who can ill afford to do so.
PAGENO="0421"
Rep rinted from
The Journal of
Industrial
Economics
Volume XII JULY 1964 No. 3
HENRY STEELE Patent Restrictions and Price Competi-
tion in the Ethical Drugs Industry.
BASIL BLACKWELL
BROAD STREET OXFORD ENGLAND
Price: Thirteen shillings and sixpence net or $1.90. Annual subscription
Thirty shillings or $4.20
1971
PAGENO="0422"
PAGENO="0423"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1973
policy recommendations volunteered by witnesses varied all the way
from a request that drug patent protection be extended from the
present 17 years to a period of 25 or 30 years,4 to a suggestion that
some personal violence against certain drug makers and their
lobbyists might be in order.5 In part, however, the inconclusiveness
of some of the economic aspects of the record is due to the `trade
secret' status afforded certain crucial cost data, and the consequent
failure to insist upon the publication of actual production costs for
some of the higher-priced patented drugs. To the economist, this is a
most unfortunate omission, for such data are otherwise entirely
unobtainable.
This paper is directed to the assessment of price competition among
ethical drugs, and the extent to which such competition is prevented
by the patent privilege. Before undertaking a detailed investigation
of price competition in the varioñs drug markets, however, it is
desirable to describe the factors influencing supply and demand for
ethical drugs.
II. ETHICAL DRUGS: THE MARKET CONTEXT
The most striking features of the ethical drugs market are the
extreme inelasticity of demand and the monopolistic restriction of
supply, with the element of rivalry (rather than price competition)
taking the form of product differentiation through research and
development.
As regards the nature of demand, the prescription drugs market is
unique. Only a physician can order such drugs, but the patient must
make the payment. Chemically identical drugs may be offered for
sale under different names at widely varying prices, but the physician
has no direct motivation to prescribe the lowest-cost brand, or even
to become aware of prices at all. Furthermore, the demand for a
drug which is efficacious in a given disease does not simply consist
of the total demands of all individuals suffering from that disease
at a given time. Instead, the relevant market for such a drug is
composed of the total effective demands of all individuals who can be
persuaded to consult physicians, and who suffer (or think they suffer)
from whatever disorders physicians may be inclined or induced to
prescribe the given drug in connection with. `Ethical drugs' by
~ Testimony of Dr. Chauncey D. Leake of Ohio State University, ibid., Part x8, p.
10433.
~ Testimony of Mr. Mike Gorman, executive director of the National Committee
Against Mental Illness. In commenting upon a speech by F. C. Brown, president of
Schering, and then president of the American Pharmaceutical Manufacturers Association,
he concluded: `In fact, Mr. Brown went much further than Marie Antoinette, and he still
has his head and his profits, too', ibid., Part i6, p. 8995.
PAGENO="0424"
PAGENO="0425"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1975
range of products than is allowed in the original grant. A series of
improvement patents may indefinitely prolong the life of an original
patent, as has occurred in the case of insulin,9 and these are usually
easy to obtain. But it is also easy to `patent around' existing chemical
patents, and if such efforts result in devising improvements on
products made or processes employed by original patent holders, the
logical solution is cross-licensing. Cross-licensing negotiations, in-
volving as they do the mutual compromise of patent monopoly
positions, supply the motivation for a greater sense of community
of interest in price and production policies, and serve to further limit
competition.'° Drug patents may bring about few absolute monopo-
lies, since few patented drugs are without effective substitutes, but
the licensing and cross-licensing agreements prevalent among patent
holders facilitate a high degree of market control from the supply
side. Lack of space prevents effective documentation of patent abuses,
but it may be mentioned that of the forty-two most important
patented drugs, twenty-four are produced by only a single supplier."
Not all ethical drugs are protected by patents, and since there
seem to be no important economies of scale in production for such
drugs, it would appear possible for small firms to compete. Hence the
larger firms have taken measures (i) to confuse the normal flow of
market information in order to prevent physicians from knowing
of lower-priced sellers in the market; (2) to prevent the identification
of lower-priced equivalents of higher-priced drugs, and (3) to
persuade the physician that all lower-priced drugs are of hazard-
ously low quality. These efforts have been quite successful. Small
firms with little or no advertising budgets cannot make their presence
known in the deluge of major firm drug propaganda.'2 Marketing
tactics produce economies of scale in advertising drugs, where none
exist in producing them. The second objective is accomplished by
making the physician brand name conscious, and by devising and
advertising generic names of drugs in such a way as to minimize
the use of cheaper equivalents. Generic names are designed to be
lengthy and complex; brand names are brief and euphonious.
Generic names, by law, must be included in all advertising, but are
usually printed in minute type face and located in anomalous places.
Brand names are given great prominence and are advertised inten-
9lbid.,p. `4'.
10 The currently pending Federal Trade Commission complaint, FTC v. American
Cyanamid, et. al., charges the respondents with such a conspiracy in connection with the
tetracycline patent negotiations, ibid., pp. 545-7. See also the discussion of the prednisone
cross-licensing agreements given below.
11 Ibid., p. 67.
12 Selling expenses amounted to $3200 per physician in 5959, and constituted 24.8 per
cent of sales revenue and was 77.3 per cent as large as the cost of goods sold, ibid., p. 3!.
PAGENO="0426"
1976 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
sively. If a physician prescribes by generic name, the pharmacist
may fill the prescription with any brand of that drug; if he pres-
cribes by brand name, the pharmacist must dispense only the brand
called for. Unless he obtains specific permission from the physician,
prescribing another brand makes him liable to prosecution by state
pharmacy boards under the `substitution' law.'3
The third objective is accomplished by a continuing program of
disparagement of lower-priced drugs on the part of the major drug
firms through their trade associations and salesmen, especially
detailmen.'4 In no other industry is it likely that disparagement of
lower-priced products has been able to supplant price competition
so completely. Food and Drug Administration regulation has been
surmounted by also disparaging the adequacy of FDA inspection.
III. PRICE COMPETITION AMONG ETHICAL DRUGS
Very little useful quantitative data exist outside the material
published in the Hearings, which concerned themselves with four
drug categories, antibiotics (about 20 per cent of total ethical drug
sales), corticosteroid hormones (i o per cent), tranquilizers (~ per
cent) and oral antidiabetic drugs (3 per cent). The corticosteroids
market is easily the most representative of the four, since the pattern
is not as clear in antibiotics nor as complete in tranquilizers or oral
antidiabetics. Consequently, chief emphasis will be given to the
corticosteroids market as a paradigm of the effect of patents and
other drug marketing institutions on price competition.
A. Corticosteroid Hormones
There are seven closely related corticosteroid hormones which are
used in the treatment of a number of related disease groups, particu-
larly those with inflammatory symptoms of a chronic nature. All
seven are to some extent substitutes for one another in the treatment
of many disorders.'5
Cortisone, the first corticosteroid hormone to be discovered, was
found in ~ and commercially produced in 1948. Its price history
13 Among pharmacists, `substitution' formerly signified the dispensing of a chemical
substance other than that required by the physician. The National Pharmaceutical
Council engaged in a crusade to induce state boards of pharmacy to draft antisubstitution
laws defining substitution in terms of brand names, and succeeded in increasing the
number of states with such laws from four to forty-four from ~ to 1959, ibid., p. 235.
14 Part i8 of the Hearings contains much testimony to this effect by physicians and
medical educators. See particularly the testimony of Dr. Solomon Garb, Hearings, op. cit.,
Part i8, pp. 10480-5.
15 A study made in 3959 showed that all seven of these drugs were prescribed in every
one of the seven disease groups requiring corticosteroids therapy, Hearings, op. cit., Part
39, p. 10825.
PAGENO="0427"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1977
reveals a pattern characteristic of a new product in the chemical
industry, with initially almost prohibitively high prices falling rapidly
to reach a plateau. Merck and Company first developed cortisone,
and although unable to obtain a product patent, it managed to
keep a pure but transitory monopoly of production until 1952, largely
because it alone had processing facilities. In the absence of patent
protection, the cost savings from process improvements were passed
on to buyers:16
July 1949 $200 per gram Aug. 1950 $50 per gram
Dec. 1949 150 ,, ,, Oct. 1950 28 ,,
Mar.1950 135 ,, ,, Apr. 1951 24
May 1950 115 ,, ,, Oct. 1951 20 ,, ,,
July i~~o 95
But by 1953, Merck's brand, `Cortone' held only 54.8 per cent of
the market. Intense price competition continued through 1954, at
which time Merck's prices had dropped to $5.48 per gram. Merck,
Schering (`Cortogen'), and Upjohn (`Cortisone') have produced
virtually all of the United States supply of cortisone since 1954, and
each has charged $5.48 per gram at wholesale since that time.17
Hydrocortisone prices fell commensurately with cortisone prices
until 1956, when all producers quoted identical wholesale prices of
$7.99 a gram.'8
A third corticosteroid hormone, prednisone, was first discovered
by Syntex in 1955, but several other firms were working on corti-
costeroids, and discovered the compound at about the same time.
(Merck claims to have produced prednisone shortly after their
discovery of hydrocortisone, but failed to realize its potentialities.19)
Schering, Merck, Upjohn, Pfizer and CIBA all filed patent applica-
tions on prednisone, and an interference proceeding was declared
by the Patent Office. Syntex later became a party to the inter-
ference, basing its claims on a 1951 patent covering a class of products
which Syntex claimed included prednisone and prednisolone. The
interference has not yet been resolved by the Patent Office, but the
parties involved undertook to make interim arrangements among
themselves. Apparently Schering could convince Upjohn, Merck and
Pfizer that it had the best chance of obtaining the eventual patent.
In 1955, Merck, Upjohn and Pfizer separately agreed on cross-
licensing arrangements with Schering, each involving the payment
16 Testimony of J. T. Connor,~ ibid., Part i~, p. 8030.
17 Ibid., Part 14, p. 7884.
18 Ibid., Part 14, p. 7885.
19 Testimony ofJ. T. Connor, Hearings, op. cit., Part iz~, p. 8027.
PAGENO="0428"
1978 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
to Schering of unconditional interim royalties of 3 per cent of sales,
for a period of 3 years, in exchange for the right to make prednisone
and prednisolone and market it only in finished dosage form. The
royalties were `interim' in the sense that they could be collected by
Schering until an interference was declared at the Patent Office,
after which time they would simply accrue, but would not be paid
to Schering until and unless a patent was issued to Schering. Parke,
Davis signed a largely similar agreement in ~ and CIBA fol-
lowed in 1958. Such licenses were granted on the basis of patent
applications, rather than on patents held.2° The president of
Schering defended the practice on the grounds that this was the
only way to insure that all parties involved would be licensed to
continue production, no matter who was awarded the patent.2'
This is an admission that failure to pay Schering interim royalties
might jeopardize the granting of a license if Schering did obtain the
patent, and the size of the interim royalty is a measure of the degree
to which the other companies considered Schering's patent claims
to be superior to theirs. All of the cross-licensing agreements (except
with Upjohn) provided for the sale of prednisone and prednisolone
to third parties in finished dosage form only, thus preventing bulk
sales of the finished powder to competitors who might tablet the
powder and sell it, either generically or under their own brand
names, at their own prices. Upjohn, although not bound by such a
restriction, has in fact made no sales of bulk powder except to
Schering.22 Schering's president saw nothing unusual in the interim
royalty feature. Merck's president disagreed in principle, but con-
ceded that there was probably nothing illegal about ~t.23
From the very beginning, the wholesale prices of both prednisone
and prednisolone charged by Schering, Merck, Upjohn, Pfizer,
CIBA and Parke, Davis have been identical, at $17.90 per bottle of
one hundred s-milligram tablets.24 The sixth party to the inter-
ference, Syntex, requested a license under Schering's process in 1955.
A Schering spokesman is said to have refused because Syntex had a
reputation `for knocking the pants off prices'.25 Syntex then began
to make sales of prednisone bulk powder (i.e. the finished product
in bulk powder form) in the United States in late 1956. Schering
20 Data taken from cross-licensing agreements submitted to the Suncommittee by
Schering, ibid., Part 24, pp. 7918-20.
21 Testimony ofF. C. Brown, ibid., Part i~, p. 7928.
22Report, op. cit., p. 151.
23 Hearings, op. cit., Part 24, p. 8096.
~ Data from American Druggist Blue Book, as reported in the Hearings, ibid., Part i~,
p. 7884.
25 Ibid., Part 24, p. 7920. Reference is to a remark allegedly made by Irving Jurow
of Schering (who disavows the phrase) and reported in Fortune, August 1958.
PAGENO="0429"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1979
brought suit for infringement under the process patent, and Syntex
retaliated by suing both Schering and Merck for infringement of
its 1951 product patent. Schering thought enough of the validity of
the Syntex patent to agree in 1958 to pay Syntex a lump sum paid-
up royalty in return for a release from all claims, present or future,
under the 1951 patent.26 Merck made a similar agreement with
Schering at about the same time.27 However, the bulk powder price
structure for prednisone and prednisolone had already begun to
deteriorate by the time such agreements were accomplished. When
Syntex began to make its `uncontrolled' bulk shipments into the
United States, Pfizer early in 1957 violated the terms of its cross-
licensing agreement with Schering and made bulk sales of its own
powder to third parties, meeting Syntex's low price. Merck soon
followed suit, but Upjohn continued to sell in bulk only to Schering.28
This competition in the bulk market steadily forced bulk prices
downward, from levels which were initially quite low in relation to
the wholesale price of the tableted and bottled powder. The bulk
prices charged by Syntex declined from $io.oi per gram in the first
quarter of 1957 to $2.36 per gram in the third quarter of 1959.
Merck and Pfizer are said to have met Syntex's price; after they
entered the market at lower prices they made sales at Syntex's
expense, and the records show that sales by Syntex declined, despite
the great reduction in its price.
In March 1959 Schering and Syntex made an agreement, under
the terms of which Schering agreed, in the event that it was awarded
the prednisone patent, to license Syntex (i) to make prednisone and
sell it in bulk form to pharmaceutical manufacturers for use as a
chemical intermediate in the manufacture of products other than
prednisolone, or of those so closely related to prednisone or predniso-
lone as to constitute their equivalents, and to pay a 6 per cent royalty
on such sales; (2) to make and sell prednisone in finished form under
its own label at a 6 per cent royalty rate; and (~) to sell prednisone
in bulk to licensees of Schering, who in turn would pay royalties
directly to Schering. If Syntex were to be awarded the patent,
Schering would receive a license to make and sell prednisone at a
royalty rate of 3 per cent.29
This agreement contrasts with those made by Schering with
Merck, Pfizer, Upjohn, CIBA and Parke, Davis. The latter agree-
ments provided for immediate áross-licensing, the payment of a
26 Testimony of F. C. Brown, ibid., Part 54, PP. 7921-2.
27 Text of Merck-Schering agreement, ibid., Part i~, pp. 8364-g.
28 Ibid., Part 14, p. 8095.
29 Text of agreement submitted by Schering to Subcommittee, ibid., Part 15, pp.
8878-9.
PAGENO="0430"
1980 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
3 per cent interim royalty (5 per cent for Parke, Davis), such interim
royalties to be in effect for 3 years (the earlier agreements with
Merck, Pfizer and Upjohn stipulated that in the event of an inter-
ference proceeding,3° interim royalty payments would be suspended
and allowed to accrue until a patent was actually issued; the CIBA
and Parke, Davis agreements called for interim royalties for 3 years
after the date of the agreement itself); and, finally, all of the agree-
ments (except with Upjohn) restricted sales to finished dosage forms
only.3'
The purpose of restricting sales to finished forms only is clearly
to prevent any shipment of bulk powder being made available to
non-licensed competitors (chiefly the smaller firms who take the bulk
chemical, tablet and bottle it, and sell it under their own brand
name, or more frequently, under the generic name) who might
make use of such an `uncontrolled' supply to introduce price
competition into the market. It is by no means unusual in the drug
trade to find licensing agreements predicated upon the basis of patent
applications rather than issued patents, but it is somewhat unusual
to have interim royalties paid on the basis of applications alone.
Counsel for the companies pointed out that there was nothing illegal
about interim royalties, since one firm might very well pay another a
royalty in order to use a secret process for which no patent was
sought. In such circumstances, the consideration for the royalty
would be the disclosure of an otherwise unknown process. Schering's
agreements with Pfizer, Upjohn, Merck and perhaps some of the
others, could not be so felicitously construed, since all firms had been
doing intensive research in the field and were roughly on the same
footing.
It was generally agreed, however, by Schering and its licensees,
that it would be impossible to enforce the restrictions on the form in
which the good is sold, until a valid patent had been issued.32
Nevertheless, a pattern of action which can be described as volun-
tary compliance was observed until Syntex began making its ship-
ments of bulk prednisone, at which time Pfizer and Merck violated
the letter of their agreements with Schering by following suit. The
initial effort at voluntary compliance may conceivably have been a
gesture of goodwill toward Schering, or it may have been a collec-
tively acceptable way of dealing with the problems posed by highly
profitable prices (which could bear a great deal of shading) coupled
30 The prednisone and prednisolone interferences were declared in December i 956 and
December 0957 respectively, ibid., Part i~, p. 8093.
31 Licensing agreements submitted to the Subcommittee, ibid., Part `5, pp. 8360-4.
32 of F. C. Brown, ibid., Part i~, pp. 7926-8; testimony of J. T. Connor,
Part i~, p. 8094.
PAGENO="0431"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1981
with the probable excess capacity on a product which as yet had no
patent protection.33 When Pfizer and Merck violated their agree-
ments by making bulk sales, Schering took no action, and its presi-
dent indicated in the hearings that he realized that this part of the
agreement was not enforceable.34
Price competition ensued. Small companies could obtain the bulk
powder from Syntex, Merck, Pfizer or from a small domestic
producer, Formet, and package their own finished dosage forms with
no restrictions until such date as a patent might be issued. If a patent
were to be issued to any firm other than Syntex, these smaller pro-
ducers could expect to be eliminated from the market immediately.35
Until such time, however, small firms were free to cut prices as far
below those of the major firms as they could. The lowest wholesale
price reported in the hearings ($1.75 per hundred 5-milligram
tablets) was less than one-tenth of the $17.90 charged by Merck,
Schering and Upjohn.
Given perfect market information, differences of 90 (or of iooo)
per cent in the prices of identical products would be impossible.
The larger firms found it much more congenial to disparage the
products of lower-priced sellers than to quote lower prices to meet
their competition. Nor was any price reduction necessary in the
retail druggist prescription market, where well over 90 per cent of the
sales were made by the three large firms. It may be appropriate to
inquir~, in the perhaps extreme case of prednisone, what substance
there may be to charges that lower-priced producers sell inferior
prednisone.
Prednisone could be purchased under the generic name in late
1959 from a number of smaller firms, for prices ranging from $1.75
to $12.00 per hundred ~-mil1igram tablets at wholesale. It could also
be purchased from Schering as `Meticorten', from Upjohn as `Delta-
sone', from Merck as `Deltra', or from Parke, Davis as `Pa~acort'
~ Excess capacity is very likely to have existed. Dr. Philip Berke of Formet Laboratories
testified that an investment of only four or five million dollars would easily be sufficient
to supply the entire world demand for prednisone and prednisolone, ibid., Part 14,
p. 8056.
~ Ibid., Part `4, p. 7928.
~` It has not been the practice of the major drug firms to license smaller companies.
Blackman of Premo, a small firm selling prednisone and other drugs at low prices under
generic names, stated that he could produce the newer, and as yet unpatented, cortico-
steroid dexamethasone in the interval before a patent was issued, but that after its
issuance his investment in production facilities would be worthless. He declared: `With
as much assurance as any human being can muster, I feel we would not get a license',
ibid., Part i 4, p. 823!. Dr. Berke of Formét Laboratories, which is apparently the only
small firm in the United States making prednisone and prednisolone in bulk form, was of
the same opinion, and indicated that he would apply for a dexamethasone license `tongue
in cheek'. He further indicated that it was cheaper to produce prednisone and prednisolone
than dexamethasone; hence he could afford to produce the former while patent pro-
ceedings were still pending, but not the latter, ibid., Part 14, pp. 8057-8.
PAGENO="0432"
1982 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
at $17.90. Claims that brand name differences mean quality
differences must be dismissed. Prednisone is listed in the United
States Pharmacopoeia, which specifies standards of purity, potency and
identity which must be met by all makers, and there is no therapeutic
gain in producing drugs to a higher purity exceeding them.36
They are enforced by Food and Drug Administration inspection.
No producer of prednisone, large or small, had ever been accused of
irregularities in connection with prednisone products as of the date
of the hearings. For prednisone, all sellers must stay within the
potency tolerance, between 90 and 110 per cent of the level claimed
on the label, whether their products are branded or are sold generic-
ally. Since the cost of the active ingredient is typically about one-
third of total factory cost, there can be no appreciable savings in
aiming at the lower limit. Furthermore, the firm that sells the pro-
duct may not manufacture the bulk powder, or even do the tableting.
Parke, Davis, in 1958, made none of its own bulk powder, while the
smaller companies bought their bulk powder largely from Pfizer and
Merck (who are reported to have met Syntex's lower prices) as well
as from Syntex and Formet. Further evidence of the reliability of at
least some of the lower-priced sellers is given by the fact that both
Premo and Chase were allowed to make sales to the Military
Medical Supply Agency, which makes thorough inspections of the
facilities of all low bidders on government drug purchase contracts.37
No brand name preference is given for purchases by government
agencies or by hospital formularies. Here, small firms may compete
with large firms in terms of price, and large firms, especially on
government contract bids, may have to cut their prices drastically,
and in such a way that something may be learned of their costs and
pricing policies. Sales were negotiated by the Veterans' Administra-
tion with Merck and Schering in February 1958 at a price of $136
per bottle of one thousand s-milligram tablets of prednisone. In
March 1958 competitive bidding was instituted; Merck cut its bid
to $95, Schering dropped to $68, but a small firm, Panray, obtained
the contract with a low bid of $38.50, or 28.3 per cent of the price
negotiated the month before. Competitive bids submitted to the
Military Medical Supply Agency at about the same time showed
Merck bidding $75, Schering $79.74, and another small firm, Chase,
obtaining the contract with a low bid of $41.50. In April 1958 Merck
and Schering entered their previous bids, and Premo was the low
bidder at $38.40. In December of the same year, Schering cut its
36 Testimony of Dr. Walter Modell of Cornell University, ibid., Part 21, p. i i6io.
`~ Testimony of Captain H. R. Fahlbusch of the Military Medical Supply Agency,
ibid., Part 21, p. 15547.
PAGENO="0433"
COMPETITIVE PROBLEMS IN, THE DRUG INDUSTRY 1983
price to $23.62 and obtained the contract, underbidding smaller
firms. At this time, Schering was charging $170 per thousand at
wholesale to druggists, implying a discount of 86.7 per cent to the
government.38 In February 1959 Premo was again the low bidder,
with a price of $20.98. At that time, Premo was selling at $26 per
thousand to the druggist, with the comparative discount to the
government at 19.3 per cent.39
It is interesting that quoted wholesale prices and bids tend to vary
inversely with the size of firm. The three large firms charged identical
prices of $17.90 per hundred s-milligram tablets. Prices of the
smaller companies named varied from Lannett's $12.00 to Premo's
$2.35. A Washington chain drugstore dealer reported to the Sub-
committee that he had been able to purchase the same goods at
$1.75 from an unidentified `first line' company, and the 1959 Drug
Topics Red Book listed twenty firms selling at prices below $3.oo.~°
Such facts lead one to the question: what is the relationship between
size of firm and cost of production? The presidents of Merck,
Schering and Upjohn testified that the sales they made to the
government were either not profitable, or did not yield a sufficient
profit to cover the cost of an adequate research program. The
smaller firms testified that they definitely made a profit at their
prices. Syntex, however, spent about half again as high a percentage
of sales revenue upon research as did the larger firms, and Formet
spent relatively nearly twice as much.4' A comparison of selling
costs is more instructive. Premo (the only small firm for which sales
costs are available) devotes 2.0 per cent of its sales dollar to selling
activities while the weighted average for the three large firms was
24.3 per cent.42 There is some reason to suspect that Premo might
spend more of its revenues on selling efforts than do other small firms.
Premo once attempted to increase its sales force, hire detailmen and
compete with the large firms, using their own methods. In 1955 the
company's sales outlays increased to almost 7 per cent of sales, but
their campaign failed because of `the tremendous increase in the
advertising dollars spent by our large competitors, to the extent that
our efforts appeared, in the market place, as a mere spark in a vast
conflagration'.43 Advertising outlays were cut, but may yet remain
above the levels for small firms which never attempted large-scale
selling campaigns.
38 Data submitted by the Military Medical Supply Agency to the Subcommittee,
ibid., Part 28, pp. 11551-3.
~ Data from Premo 8959 Hospital Price Catalog, as reported in the Hearings, ibid.,
Part i~, p. 8709.
40 Report, op. cit., p. 87. 41 Hearings, op. cit., Part i~, pp. 8064, 8301.
42 Report, op. cit., p. 3!. ~ Hearings, op. cit. Part 84, p. 8215. -
81-280 0-68-pt. 5-28
PAGENO="0434"
1984 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Production costs for prednisone are only a small fraction of the
wholesale price. Rather than make public the actual production
costs of individual firms, the Senate Subcommittee staff estimated the
cost of a bottle of one thousand 5-milligram tablets of prednisone at
$13.61 .44 This cost estimate was attacked by industry witnesses as
being too low, as not including selling and research costs, etc., etc.,
despite its being made unmistakably clear in the record that the cost
estimate referred only to production costs. Actually, this cost estimate
is quite generous. It is based on a bulk powder cost obtained by
using the prices charged by bulk producers for sales of bulk powder
to other firms, and hence includes not only their actual direct and
allocated costs, but also some margin for profit. To determine the cost
of tableting and bottling, firms which perform such functions on a
contract basis were contacted, and the median cost quotation (not
the lowest) was employed. The actual production cost of McKesson
and Robbins is only $8.99 for such a bottle.45 This cost is only 5.07
per cent of the equivalent wholesale price of $179 per thousand,
and only 3.17 per cent of the retail price of $ 283.33.46
Price competition departs and then reappears on the scene with
regard to the three newer corticosteroids. When Upjohn introduced
methylprednisolone in 1957 (`Medrol'), it first had the market to
itself~, and only later licensed Schering, which had no chance to
market its brand until 1959. Being 25 per cent more potent than
prednisolone (a standard dosage form is the 4-milligram tablet
as compared with the s-milligram prednisolone tablet), its price per
milligram was increased by exactly 25 per cent, so that the price per
tablet is the same, and the introduction of the more potent product
did not give rise to any price competition.47 As a new drug, however,
~ Report, op. Cit., p. 34.
~ Testimony of Dr. Herman Nolen, president of McKesson and Robbins, in Hearings
Before the Antitrust Subcommittee of the Committee on the Judiciary, House of Representatives,
Eighty-seventh Congress, Second Session, Washington, Government Printing Office, I 962,
p. `5.
46 While the present paper is not concerned directly with the retailer's markup of
66~ per cent, it certainly helps to contribute to the magnitude of the final price to the
consumer. It is well known that retail druggists, like retailers everywhere, dislike price
competition in practice, and favor high markups-the role of the National Association of
Retail Druggists as one of the prime movers behind the `Fair Trade' laws is notorious-
and the complaints that high markups have failed to produce satisfactory profits for
retail druggists is simply a consequence of the freedom of entry in the field and the over-
capacity, low turnover rates and higher average unit costs which results from new entry
in response to the high unit margins. It is pertinent to call attention to this point in view
of the possibility that pressure may be exerted by organized druggists on the manu-
facturers to discourage any tendencies toward price cutting, or at least that the drug
manufacturers may be aware of such distributor sentiment when contemplating possible
price cuts. It must be admitted, however, that no shred of evidence on this point was
offered (or sought) during the whole of the hearings.
`~7 Data from American Druggist Blue Book, 1959-1960, as reported in the Hearings,
op. cit., Part 14, p. 8324.
PAGENO="0435"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1985
methyiprednisolone began to take over a notable share of the
market, and by 1958 `Medrol' had i8.i per cent of the new pres-
cription market, while the leader in 1956, `Meticorten', had fallen
from 53.9 per cent to 30.7 per cent.48
In 1957 Lederle introduced triamcinolone, a corticosteroid
identical in potency to methylprednisolone, and licensed it to Squibb.
Both firms set prices at a level identical with that of prednisone,
prednisolone and methylprednisolone. Lederle's brand, `Aristocort',
had 23.8 per cent of the new prescription market in 1958, while
Squibb's brand, `Kenacort', had 8.9 per cent.49 Eighty-five per cent
of the bulk powder, however, was made by Squibb, and only i~ per
cent was made by Lederle. As a consequence of Lederle's prior and
intensively advertised appearance in the market, it would appear that
physicians' brand preferences crystallized in favor of Lederle's
product so predominantly that Squibb had no outlet for much of its
production except to sell it in bulk to Lederle.
In 1958 Merck and Schering introduced dexamethasone, a corti-
costeroid with a potency about six and two-thirds as great as
prednisone, the standard tablet containing .75 milligrams instead of
~ milligrams. For the first time since the prices of cortisone and
hydrocortisone reached a plateau, price competition was introduced
into the market when Merck priced its `Decadron' at retail at 27
cents per tablet, 10 per cent below the prices of the other four
competitive corticosteroids. Schering later marketed its brand,
`Deronil', at the same price as Merck. Merck, intending to regain its
earlier predominant place in the market, gave its dealers additional
quantity discounts amounting to as much as 10 per cent.50 A patent
interference was declared between Merck and Schering. Merck
subsequently licensed CIBA and Organon of Holland. CIBA then
began to market dexamethasone as `Gammacorten' in 1958. While
dexamethasone has yet to be patented, no small firms seem to be
producing this compound, because of the relatively high costs in-
volved. In fact, it may be the case that the profit margins to the
producers are considerably lower than in the case of the other
corticosteroids. Sales of bulk powder by Merck to CIBA were made
in 1958 at $6~ per gram. If the bottling and tableting charges used
previously are added to the bulkpowder cost, a cost of $72.69 per
gram is obtained, which amounts to 33.8 per cent of the wholesale
price of $214 per gram.5' (The comparable figure for prednisone is
8.7 per cent.) This is only slightly more than the average ratio of
48 Data from the market research staff of Merck and Company, ibid., Part 84, p. 8028.
~ Ibid., Part i~, p. 8028.
50 Testimony ofJ. T. Connor, ibid., Part 84, p. 803!.
5' Ibid., Part 84, p. 8324.
D
PAGENO="0436"
1986 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
32.1 per cent for cost of goods sold to total sales for the twenty-two
largest firms in the industry, as cited above. Dexamethasone there-
fore combines a lower price with much higher costs-a phenomenon
impossible in pure competition.
It is inappropriate to attempt to trace in corresponding detail the
course of price competition, product by product, and in the light of
existing patent licensing agreements, in the other drug group mar-
kets. They all reveal the same general features, hence only a sketch
will be given of prices, the relation of prices to computed costs and
competitive bids, and of patents, in the other three markets.
B. Oral Antidiabetic Drugs
The oral antidiabetics market may readily be summarized in this
regard. Each of the three products is either patented or produced
under patent application arrangements. Upjohn set the price of
`Orinase' in such a way as to be almost exactly equal to the price of
insulin on a per dose basis, at 13.9 cents per 500-milligram tablet,
comparable to 14.0 cents for 10 units of insulin.52 This price was not
based on cost, for the cost of insulin is no doubt considerably greater
than the cost of `Orinase', or tolbutamide, since animal pancreas
is undoubtedly more expensive to obtain and process than the in-
dustrial chemicals required to produce tolbutarnide.53 Yet the greater
convenience of administration of an oral drug over an injectable
drug would probably support a higher price than the insulin equiva-
lent. The price was simply set equal to insulin's price, apparently in
deference to the industry's tradition of pricing new products at (or
very near) the price of existing drugs used to treat the same dis-
order. (The price set is actually below the total cost of insulin
therapy, including syringe and needle.) The price of insulin itself
had not been changed since 1947. Had the practice of medicine not
changed, we might still be paying second-century A.D. prices for the
remedies of Galen.
Pfizer priced its drug, chlorpropamide (`Diabinese') at i~ cents
per 250-milligram tablet, but since Pfizer's drug is somewhat more
potent than tolbutamide, the daily cost may run somewhat less.
The third drug, phenformin, is made by one of the smaller firms,
U.S. Vitamin. Phenformin is still more potent than chlorpropamide, a
~o-mi1ligram tablet costing i 2 cents. The range of possible daily
cost (minimum to maximum recommended dosage) is from 14 to ~6
cents for tolbutamide, from 7 to 30 cents for chiorpropamide, and
from 12 to ~ cents for phenformin.54 The drug buyer seems to have
52 Testimony of Dr. Upjohn, ibid., Part 20, pp. 11037-9.
~ Report, op. cit., p. g~. ~ Hearings, op. cit., Part 20, p. 11120.
PAGENO="0437"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1987
real price alternatives, for once, but because of the priority in
appearance of the higher priced drugs, and their intensive advertising,
sales seem to be directly related to the height of daily average cost.
The Subcommittee staff computed the estimated production cost
for one thousand 500-milligram tolbutamide tablets (twenty bottles
of fifty each) on the basis of bulk powder transactions plus contract
tableting and bottling charges, as in the case of prednisone. The
production cost estimate is $6.86; the royalty (71 per cent of the
selling price) is $6.24. The wholesale price of twenty bottles is
$83.40 and the retail price is $139.00 Upjohn's production cost is
therefore 8 per cent of the wholesale price; the production cost plus
the royalty brings Upjohn's total estimated cost of production and
royalties to 15.5 per cent of the wholesale price and to io.6 per cent
of the retail price.55 Production cost estimates for the other two drugs
are not available. The only bid on a Military Medical Supply
Agency contract that is available for any such drug is Upjohn's
successful bid of $4.00 for 18,432 bottles of fifty 500-milligram tablets.
Since the wholesale cost for one bottle is $4.17, it can be seen that in
the absence of lower-priced competition from smaller firms, a large
order from the government, with its attendant cost savings, could be
obtained at only a 4 per cent quantity discount.56
C. Tranquilizers
Price competition among tranquilizers varies with the type of
tranquilizer. There are two types, the `potent' type used chiefly
in mental institutions, and the `mild' type used for non-hospitalized
patients troubled with anxiety.57 In the potent group, the first
drugs to be introduced were chiorpromazine and reserpine in 1954.
Both drugs were priced at wholesale at $3.03 per bottle of fifty
tablets. Later during the year CIBA introduced its brand of reserpine
at a price for fifty tablets 25 per cent below its existing competitors
for that drug (Squibb and Riker), $2.25. When Smith, Kline and
French introduced prochiorperazine in 1956, they priced it at $3.03
per fifty tablets, the same price they charged for their chlorpromazine.
When Wyeth introduced promazine hydrochloride in 1957, it did
so at a price of $3.00 for a bottleof fifty tablets.58 In an interesting
~ Report, op. cit., p. 20.
56 Hearings, op. cit., Part 21, p. 11551.
~ As is true in most drug categories, much substitution among drugs is possible in
most disease groups treated by tranquilizers. A 1959 study by the Pharmaceutical Manu-
facturers Association shows that all seven of the major tranquilizers (both mild and
potent) are prescribed in all five major disease groups requiring such therapy, ibid.,
Part i~, p. 10827.
58 Ibid., Part i6, p. 8887.
PAGENO="0438"
1988 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
move in 1958, Wyeth changed the wholesale price of its drug pro-
methazine hydrochloride, which was originally an antihistamine with
drowsiness-inducing properties, from $4.45 per hundred 25-milli-
gram tablets to $5.94, thus making it competitive with its recent
tranciuilizer promazine hydrochloride.59 An antihistamine was thus
made a tranquilizer in response to changing market demand by
changing the advertising to stress what were previously considered
side effects as the desirable end result, and by increasing the price in
recognition of the increased marginal revenue possible in the more
lucrative market for tranquilizers. Later potent tranquilizers were
priced at about $3.00 for the equivalent dosage. Apparently reser-
pine was considered to be a sufficiently distinct product that CIBA's
25 per cent lower price did not call for price reductions for other
potent tranquilizers. CIBA, perhaps uncertain as to the validity of
its patent (reserpine is the name given to the active substance
isolated from the root of Rauwolfia Serpentina, a species of tropical
plant, the medicinal properties of which had been known in India
for centuries), issued licenses to nine other firms in the United
States, eight of which were allowed to sell in bulk powder form.
These firms made sales to smaller firms, which in turn proceeded to
cut prices down to as much as 90 per cent below CIBA's levels.
CIBA continued to advertise its brand of reserpine and its price
held up; other major firms dropped their advertising of this pro-
duct, by and large, and their prices were cut to about $1.91 per fifty
tablets at wholesale.6° No other potent tranquilizer has ever experi-
enced a cut in price.
Carter was first in the mild tranquilizer market with mepro-
bamate, which it has sold since 1955 at $2.60 per bottle of fifty
400-milligram capsules at wholesale. Carter licensed Wyeth to sell
meprobamate under Wyeth's brand name, and Wyeth charged the
same $2.60, although its royalty payment to Carter would be 13
cents, and its raw material~cost (purchased from Carter) would be
25 cents greater.6' Carter licensed its very profitable product only
because it did not have the selling facilities which larger drug firms
such as Wyeth had. Little competition to meprobamate has devel-
oped. Phenaglycodol is sold by Lilly at the same price to the drug
buyer as meprobamate.
The computed cost for reserpine production is $2.48 per bottle of
one thousand 25-milligram tablets.62 CIBA's wholesale price for this
quantity is $39.50; other firms quote prices as low as Panray's
$2.65. For CIBA, estimated production cost is 6.3 per cent of its
5~ Hearings, op. cit., Part i6, p. 9274. 60 Ibid., Part i6, pp. 9387-90.
61 Report, op. cit., p. i8. 62 Hearings, op. cit., Part i6, p. q436.
PAGENO="0439"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1989
price; for Panray, 93.6 per cent. It must be the case, however, that
the computed cost figure is much too high, for in competitive bids
to the Military Medical Supply Agency Panray has won bids at as
low as $0.51 per thousand tablets on a 4080-bottle order, and CIBA
has failed to get contracts on bids as low as $0.52-I .3 per cent of
the wholesale price. When asked how CIBA could afford to bid
successfully at $o.6o in February 1959 Mr. T. F. D. Haines, presi-
dent of American CIBA, testified: `. . . we didn't anything like re-
cover our out-of-pocket costs . . . it was perhaps a mistake that we
did that'.63 This makes it all the harder to explain CIBA's later and
lower unsuccessful bids. (Raw materials costs for one thousand 25-
milligram tablets should be about 30 cents.) The president of Mc-
Kesson and Robbins later volunteered his firm's production costs for
reserpine: $0.63 per thousand tablets, or 1.6 per cent of the whole-
sale price.64
The computed production cost for meprobamate to Carter is
$7.32 for twenty bottles of fifty 400-milligram tablets, or~ the basis
of a 20,000-bottle order. For Wyeth, the cost would be $12.40 plus
$2.60 royalty. (Carter agreed that the Subcommittee staff's cost
estimate was correct for its own production.) Both Carter and Wyeth
obtain $52.00 at wholesale for twenty bottles; a retail drug buyer
would pay $108.40 for this amount. Carter's production cost is 14.1
per cent of the wholesale price, Wyeth's is 24.0 per cent, and with the
royalty to Carter, 28.0 per cent. (Carter's production cost is only
6.8 per cent of the retail price.)65 Bids to the Military Medical
Supply Agency from Carter and Wyeth ranged from $22.50
for five hundred 400-milligram tablets in February 1958 to $1 9.845
in February i 960, or from 90 to 79.4 per cent of the wholesale price
for this amount. In the absence of any domestic competition, the
Military Medical Supply Agency purchased abroad in June 1960,
obtaining a low bid of $3.95 from a Danish firm, a bid only 29 cents
above Carter's production costs, and about 8o per cent below the
bids of Wyeth and Carter.66 Danish costs may be below American
costs, but Carter could at least have met this bid had it seen fit
to do so.
D. Antibiotics
Price competition in the antibiotics market is intense in unpaten-
ted drugs, such as the earlier penicillins, and streptomycin, the patent
63 Ibid., Part i6, p. 9430.
64 House of Representatives Hearings, op. cit., p. i6.
~ Report, op. cit., p. i8.
66 Testimony of Rear Admiral W. L. Knickerbocker of the Military Medical Supply
Agency, Hearings, op. cit., Part 24, pp. 13784-5, 53794.
PAGENO="0440"
1990 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
for which was licensed by Merck to several other firms. Price com-
petition is, however, considerably less in evidence for the later
patented antibiotics.
Table I shows how progressively more efficient production
methods and continual competitive pressures consequent upon free
TABLE I
PENICILLIN, STREPTOMYCIN AND BROAD SPECTRUM
ANTIBIOTICS
PRICES,
1945-60
Date (year; month
for price changes)
Penicillin
(io mil-
lion units,
Strepto-
n~ycin
(io
grams,
Broad
spectrum
antibiotics
capsules)
(i6 250-milligram
Chlor-
Chlor-
Oxy-
Tetra-
Demet/~yl-
chlor-
bulk)
bulk)
tetra-
amphe-
tetra-
cycline
tetra-
cycline
nicol
cycline
(~ brands)
cycline
(iso mg.)
S
S
S
$
S
S
$
1945 May
6o.oo
1946 May
36.00
,6o.oo
1947 January
21.00
45.00
1948 July
13.00
10.24
1948 December
9.50
8.19
15.00
1949 February
9.50
8.19
10.00
1949 March
7.50
5.00
10.00
10.00
1950 February
4.75
3.15
8.oo
8.oo
1950 April
1950 May
1950 November
4.75
4.75
4.75
3.15
3.15
3.15
8.oo
6.oo
6.oo
8.oo
6.oo
6.oo
8.40
8.40
6.oo
1951 September
1951 November
3.75
2.50
3.15
3.15
6.oo
5.10
6.oo
5.10
5.10
5.10
1952
1.15
3.24
5.10
5.10
5.10
1953
.95
1.70
5.10
5.10
5.10
5.10
1954
.75
1.70
5.10
5.10
5.10
5.10
1955
1956
.44
.52
.90
.75
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
1957
1958
1959
1960
.79
.70
`28
.21
.88
.88
.38
~36
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
5.10
Source: Federal Trade Commission, Economic Report on Antibiotics Manufacture, Washing-
ton, 1958, Section II, pp. 164, 179, i8o, 192; Senate Hearings on Administered Prices in the
Drug Industry, Part 24, p. 13664.
entry made possible the precipitous decreases in the prices of the
extremely important drugs penicillin and streptomycin. `Broad
spectrum' antibiotics prices have acted very differently. Lederle
quoted an initial price of $15.00 for sixteen 250-milligram capsules
of chlortetracycline, possibly overestimating the inelasticity of
demand. Within 2 months, Lederle reduced its price to $io.oo, at
which price Parke, Davis entered the market in March i~. ~n
February 1st, 1950, both sellers cut prices to $8.oo. In April 1950
a new rival, Pfizer, entered the market with oxytetracycline, selling
PAGENO="0441"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1991
at $8.40, 5 per cent above its competitors. In May 1950 Lederle and
Parke, Davis reduced prices to $6.Oo, but Pfizer did not meet the
cut until November. On September 27th, 1951, Pfizer lowered
prices to $~. 10, and by November ist Lederle and Parke, Davis had
met the price cut. Tetracycline and demethylchlortetracycline came
later, and were also priced at $~.io. Prices were then stable and
rigid for almost 9 years, when Pfizer in August 1960 again assumed
the role of price leader by announcing a i~ per cent `trade discount'
adjustment which eventually became a i~ per cent price cut, to
$4.34.67 (This price cut was initiated only i month before the
Senate Hearings on antibiotics were scheduled to take place.) From
1948 to 1949, broad spectrum prices dropped by 33 per cent; from
1949 to 1951, by 49 per cent, and from ~ to August 1960 not at
all. It is reasonable to suppose that since 1951 the costs of producing
broad spectrum antibiotics have deClined by approximately as much
as the cost of producing penicillin, the production methods employed
being largely identical.68 A comparison of the decline of 90 per cent
since November 1951 in penicillin prices, with the 15 per cent
decline for broad spectrum antibiotics measures the effect of res-
triction of entry in broad spectrum antibiotics.
The Senate Subcommittee staff made production cost estimates
for tetracycline and chloramphenicol, and in 1961, during the House
of Representatives hearings, Senator Kefauver made public Bristol's
actual production costs. Bristol incurred production costs of $1.67
per bottle of one hundred 250-milligram capsules (in comparison
with $2.88 as estimated by the Senate Subcommittee staff in the
Senate hearings). The price of such a bottle to the druggist is $30.60;
to the consumer, $~i.oo. Bristol's'production cost is ~ per cent of
the price to the druggist; the addition of royalties to Lederle and
Pfizer of $2.15 paid per bottle brings the cost up to 12.5 per cent.
(Bristol actually received an average of $25.27 per bottle for such
sales in 1958; hence its production costs and royalties totalled ~i
per cent of its average price received.) Production costs are about
28.7 per cent of the wholesale price for Upjohn, which purchases
bulk tetracycline from Bristol; its total production costs and royalties
are $9.30, or 30.4 per cent of the price to the druggist. For Pfizer,
~ Mr. Duncan, the manager of Lederle, construed Pfizer's action as vigorous and
aggressive price competition. He testified that Pfizer had cut its prices on a Saturday
in order to `steal a march on the industry'. This explanation apparently surprised the
Subcommittee chairman:
Senator Kefauver: `You mean after ten years they would suddenly steal a march on
you?'
Mr. Duncan: `Yes, senator. These things happen very rapidly. If you can get any kind
of advantage on your competitors, you try to do so', ibid., Part 24, p. 13728.
68 Report, op. cit., p. 82. Chloramphenicol costs have probably declined by still more
since Parke, Davis discovered a way of making this drug synthetically.
PAGENO="0442"
1992 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
which holds the patent and pays no product royalties, and which
reportedly has a lower cost production process, the total costs of
production might perhaps be as low as $1.50. If Pfizer pays the
same royalty as Bristol for the use of Lederle's process, its production
costs plus royalties would be about $2.75. For Upjohn, Bristol and
Pfizer, the ratio of production cost plus royalties to the price to
the druggist might be about 30 per cent, i~ per cent and 9 per cent.
The gap between estimated production cost and price for chlor-
amphenicol is predictably handsome. One bulk powder transaction
did take place, between Parke, Davis and an Italian firm, Farmitalia,
at $30.00 per kilogram. Using this price for the raw material, and
Upjohn's actual costs for capsuling, finishing, and packaging, a cost
of $1.52 for a bottle of one hundred 250-milligram capsules is
obtained. This is 5.0 per cent of the price to the druggist of $30.60,
and 3.0 per cent of the retail price to the consumer.69
The bids received by the Military Supply Agency reflected the
status of market competition. For penicillin, competitive bidding was
the rule, the price for 250,000-unit potassium penicillin G tablets
falling from $1.29 per hundred in April 1957 to $o.6~ in June 1960.
As in the case of prednisone, small firms selling penicillin regularly
charge much lower prices than large ones. For this particular pro-
duct, the smallest firm (sales of less than $ ioo,ooo) had the second
lowest price ($3.30); the second smallest firm (sales of less than
$i ,ooo,ooo) had the lowest price; and three of the largest firms
(sales of $i~o to $250 million) had the highest price ($12.00).7°
For broad spectrum antibiotics, the picture is entirely different, as
can be seen from Table II.
For this entire period the price to the druggist on all these drugs
was $30.60, and the price to the consumer, $~i.oo. Lederle and
Parke, Davis began the period of rigid prices to the trade by quoting
a ~o per cent discount to the Armed Services Medical Procurement
Agency (the predecessor to the Military Medical Supply Agency);
Pfizer allowed about a 51 per cent discount off the druggist's price.
Apparently the three firms recognized the high degree of sub-
stitutability among their drugs, for prices fell for each seller with
69 Harry Loynd, president of Parke, Davis, objected at length that the Farmitalia
price was much lower than American production costs. In order properly to appraise the
reliability of Loynd's assertions, it is advisable to read the whole of his quite remarkable
testimony; nevertheless, a consideration of the import agreement is instructive. Farmitalia
agreed to supply Parke, Davis with up to 30,000 kilograms ofchloramphenicol during i 960
at ~30 per kilogram. ThroughJuly iith, however, only 6ooo kilograms had been imported.
If this price were such a bargain, why did Parke, Davis not take greater advantage of it?
The suspicion is that Parke, Davis might be able to make it for less in Detroit. However,
even if the price were twice as high, it would only increase the cost per i oo-capsule
bottle to $2.31, or 7.5 per cent of the sales price, Hearings, op. cit., Part 24, pp. 1397 i-8.
70 Report, op. cit., p. 84.
PAGENO="0443"
COMPETITIVE PROBLEMS IN~ THE DRUG INDUSTRY 1993
regard to its previous price until April 1956 when Parke, Davis
repeated its previous bid of $12.50. At that time Lederle's price was
$ii.oo on the first sale of tetracycline hydrochloride, pricing it at the
same level as its chiortetracycline. Lederle was low bidder, and was
much surprised to find that Pfizer and Squibb had both bid $19.58,
and Bristol, $ I8.97.~' For the next 2 years Lederle always bid
BROAD SPECTRUM
AGENCY AND TO
*
ANT
THE
TABLE II
IBIOTIC PRICES TO THE ARMED SERVICES MEDICAL PROCUREMENT
MILITARY MEDICAL SUPPLY AGENCY, NOVEMBER 1951-JUNE 1960
(ioo capsules, ~o milligrams)
Date
Chlortetra- Chioram-
~ycline phenicol
Oxytetra-
cycline
Tetracycline-
hydrochloride
(ASMPA)
1951 November
1952 March
1953 June
5954 January
1954 May
1955 March
1956 April
1956 October
15.30 15.30
15.00
12.50
~
12.00
11.00 52.50
15.00
12.84
11.47
10.97
11.00
(MMSA)
1957 February
1958 February
1958 March
5958 April
1958 June
1958 November
5959 June
1959 August
1959 December
1960 May
1960 June
11.00
12.50
~
11.25
11.00 11.25
11.25
11.25
~
10.97
10.75
10.11
17.24
19.19
17.24
17.15
14.36
17.15*
8.15
6,6
5.62
* Low bid-not accepted.
Source: Federal Trade Commission, Economic Report on Antibiotics Manufacture, Washing-
ton, 1958, p. 194 (i~~i -~6 data); Senate Hearings on Administered Prices in the Drug Industry,
Part 24, pp. 13779-82; 13791-2 (1957-60 data).
$19.58. On the second bid for tetracycline hydrochloride, Pfizer
met Lederle's previous bid of $11.00, but Lederle and Squibb were
both at $19.58. Pfizer, like Lederle, then learned that it was not
necessary to make `ridiculously low' bids. Its prices on the next two
bids went up to $17.24 and $19.19. In April 1958 Parke, Davis cut
its chioramphenicol price by 10 per cent, to $i 1.25, after a con-
ference with the Agency purchasing officer on the subject of its
71 Duncan of Lederle conceded that he had made a mistake in simply pricing tetra-
cycline hydrochloride at the same level as chiortetracycline. He characterized the pressures
of competitive bidding in terms of carelessness: `In other words, one gets a little sloppy in
bidding for this kind of business. You sometimes simply bid a ridiculously low figure',
ibid., Part 24, p. 13690.
PAGENO="0444"
1994 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
rigid price.72 In June 1958 the tetracycline hydrochloride price
dropped back to $17.24, where it stayed for three consecutive
awards. In November Squibb cut its bid to $17.15, to undercut
the others, all of which bid $17.24. InJune 1959 Lederle bid $17.24,
Bristol and Squibb bid $17.15, but Pfizer was low bidder at $14.36.
Pfizer's bid is surprisingly low,73 and it encouraged Military Medical
Supply Agency officials to believe that prices had at last fallen.
Such expectations were disappointed, for in August ~ three
bids were received at $17.15. Negotiations failed to shade the price,
and the contract was cancelled because of the Agency's judgment
that the price was unreasonably high. Foreign bids were received,
and an Italian firm was awarded three successive contracts at
$8.15, $6.i6 and $5.62 in 1959 and 1960. Rather than lower its
price to compete with foreign bidders, Pfizer first protested the
Italian procurements (the Comptroller General rejected Pfizer's
contentions as being without merit) and then attempted unsuccess-
fully to buy out the Italian firm.74 Had Pfizer chosen to engage in
price competition, it could in all probability have undercut the
Italian firm on all three bids.
The case of tetracycline hydrochloride prices is rather puzzling in
that the product is a close substitute for three other drugs sold at
lower prices to the government. Perhaps the industry wished to
receive higher prices for the new drug, and a certain amount of trial
and error bidding was necessary to facilitate the learning process.
It is interesting to note that some sort of division of the total tetra-
cyclines market in total dollar sales was accomplished-even at
varying prices-during the period October 1956 through October
1959, or from the first award until the date of the first foreign pro-
curement. During this period Pfizer (the patent holder) received
46.6 per cent of all contracts in dollar terms; Lederle received i 7.8
per cent; Bristol, 17.6 per cent; Squibb, 17.5 per cent; and Upjohn,
only 0.5 per cent. This may indicate nothing more than the varying
72 Ibid., Part 24, p. 13782.
~` Senator Kefauver expressed surprise that Lederle would continue to bid $17.24
after the last sale had been awarded at S 57.55. Mr. Duncan replied that he was against
quoting lower bids because'.. . we had been doing that, and all that happened was that
the price just went lower time after time until it got down to an uneconomic level, and we
decided that we would simply establish that Federal Government price, and stick to it,
which is what we did', ibid., Part 24, p. 13694. It maybe doubted that `uneconomic levels'
implied losses. On the basis of the cost data obtained by the Subcommittee, and assuming
that Pfizer paid the same royalty on Lederle's process that Bristol did, that Lederle's and
Pfizer's production costs were somewhat lower than Bristol's, and that Squibb's costs
were the same as Upjohn's, Lederle's production costs and royalties might be $3.60;
Pfizer's, S4.oo; Bristol's, S5.o3, Squibb's and Upjohn's, S9.3o. This might aid in ex-
plaining the lower initial bids of Lederle and Pfizer.
74 Testimony of Commander Arnold Wales, U.S. Navy Supply Corps, ibid., Part 24,
p. 13815.
PAGENO="0445"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1995
incidence of excess capacity or excess inventories, but does not
reflect either relative production or relative sales. In 1958, Bristol
produced 36 per cent, Lederle 33, and Pfizer 31. Sales are not
available, but Bristol sold about ode-third of its output to Upjohn
and another third to Squibb,75 so that relative sales might be:
Lederle, 33 per cent; Pfizer, 3'; Squibb, i ~; Bristol, I 2~ and
Upjohn, i i, unless Pfizer purchased significant quantities from
Lederle, as it is entitled to do by agreement. It seems likely that,
although prices fluctuated, real price competition never entered this
market until foreign firms, not encumbered by restrictive patent
agreements, were permitted to bid.76
IV. CONCLUSIONS AND POLICY RECOMMENDATIONS
Evidence has been presented to, show that effective price com-
petition among ethical drugs is seriously limited by the patent
privilege. Holders of patents (or sometimes merely of patent applica-
tions) may legally exercise restrictions on output and maintain
prices at levels that are extremely high relative to production costs.
The resulting gross profit margins are employed in large measure
to finance enormous advertising and sales promotion campaigns
which contribute materially to the already grave imperfections of
market information. By this means, small sellers of generic name
drugs are deprived of the physician's attention, and cannot obtain
any significant share of the prescription market, even though they
may be selling at prices which are a small fraction of their larger
rivals'. This applies not only to products protected by patents or
patent applications, but to virtually all advertised drugs. Monopoly
profits from the sale of patented drugs thus finance advertising
campaigns which extend monopoly power into other drug markets
which may not be protected by patents.
Drug firm monopoly and oligopoly could perhaps be rather readily
supplanted by workable competition if two simple but radical
reforms were effected in the institutional structure of the drug
market: the abolition of the patent privilege as it applies to drug pro-
ducts, and the expansion of the powers of the Food and Drug
~5 Data supplied by Bristol, ibid., Part 24, p. 13907.
76 Admiral Knickerbocker offered some interesting testimony which may tend to
illuminate the chiaroscuro pattern of competition in the bid market from a somewhat
different perspective. The decision of the Military Medical Supply Agency to accept
foreign bids was not reached until after a meeting with Pfizer representatives, during
which, Admiral Knickerbocker testified, `.. . Mr. Cooney [Pfizer's sales representative],
in rather an unguarded moment, made the statement that the price of tetracycline would
stay where it was until Pfizer did something about it', ibid., Part 24, p. 13819. Had the
network of patent control been absolutely worldwide, this alleged `unguarded moment'
might not have cost Pfizer any sales.
PAGENO="0446"
1996 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Administration to insure fully adequate drug plant inspection
and control of drug advertising.
It is something of an anomaly that the United States patent laws
permit drug product patents; about two-thirds of the countries in
the world prohibit them on grounds of public policy, and all but
three of those which do permit such patents have provisions for
compulsory licensing. Drug product patents should be abolished;
drug process patents should be subject to compulsory licensing at
reasonable royalties. This may or may not reduce the volume of
research done by private drug firms. In turn, a possible diversion of
drug research effort from private to public channels may or may not
improve the efficiency of resource allocation in this sector. In many
foreign countries, drug research has been immensely productive in
the absence of drug patents. In the United States, patent protection
has allowed prices to be maintained at very high levels and has
detrimentally affected the nature of research. With prices enor-
mously above costs (defined as raw material cost plus production cost
plus a competitive rate of profit on the investment in productive
facilities), the pressure to obtain more volume is inevitably diverted
into lavish selling expenses, and into the `research' of molecular
manipulation.77 Basic research by drug firms may be of question-
able productivity, but the high salaries paid absorb all too large a
fraction of those very scarce human resources qualified to engage
in basic biochemical and pharmacological research. It is common
knowledge in the industry that each major firm's research programs
duplicate those of their rivals; witness the near-simultaneous dis-
covery of many antibiotics and corticosteroid hormones by two or
more firms. The total productivity of drug research efforts would
arguably be increased by a partial shift from private to public
and university channels.
The Food and Drug Administration should be given sufficient
additional funds to allow adequate inspection of all drug producers,
large and small. It should also be given the authority to regulate
all drug advertising and labelling, with a view toward eliminating
brand names in favor of identification by generic name plus the
name of the seller.
Such reforms, properly implemented, might for the first time
77 Abundant evidence was presented during the Senate Hearings to show that much
drug product `research' is misnamed, and could more appropriately be referred to as
product development and product promotion. Dr. A. D. Console, former medical
director of Squibb, testified to this effect, adding: `I think the majority of it [drug firm
research] is in that category. . . with many of these products, it is clear while they are
on the drawing board that they promise no utility. They promise sales', ibid., Part i8,
p. 10380. See also the testimony of Dr. H. J. Weinstein, former medical director of Pfizer,
ibid., Part i8, pp. 10243-4
PAGENO="0447"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1997
make the ethical drug market workably competitive. The elimination
of patents would (in the absence of economies of scale in production)
lead to the entry of new firms producing under conditions of approxi-
mately constant costs. The inelasticity of demand would thus be
offset by highly elastic supply. The existing economies of scale in
selling would, in time, be eliminated by advertising reforms and
the disappearance of those monopoly profits which have motivated
and financed extravagant selling efforts. The advertising reforms
would also remedy the imperfection of market information, allowing
physicians to become aware of the precise nature of the price and
product alternatives, and disparagement efforts would wane as the
funds necessary to finance them dwindled due to increasing com-
petitive pressure on prices, and as the manifest adequacy of FDA
inspection rendered the substance of such contentions obviously
specious. S
RICE UNIVERSITY S
HOUSTON, TEXAS
PAGENO="0448"
1998 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
THE FORTUNES OF ECONOMIC REFORM LEGISLATION: THE CASE OF THE DRUG
AMENDMENTS ACT OF 1962
(By Henry Steele 1)
I. INTRODUCTION
On April 12, 1961, the late Senator Estes Kefauver introduced a bill (S. 1552,
87th Congress, first session) to amend the antitrust and food and drug laws with
regard to the prescription drugs industry. This bill, the outcome of extensive
hearings held by the Subcommittee on Antitrust and Monopoly of the Senate
Judiciary Committee in 1959-1962 regarding administered prices in the drug
industry, embodied provisions to prevent monopolistic abuses in drug marketing.
The Report issued by the Subcommittee established the existence of considerable
monopoly and oligopoly power on the part of major drug firms, and indicated
the extent to which such market power was obtained by virtue of patent privi-
leges, excessive advertising, and other marketing practices which tended to
augment the imperfection of market information in the interests of the larger
firms.2
IT. ANALYSIS OF DRUG MARKETING PRACTICES
An economist would briefly summarize the drug marketing situation as fol-
lows: demand for many prescription drugs is almost perfectly inelastic. Supply
can be restricted by drug firms which secure product patents and refuse to li-
cense them. The resulting enormous profit margins on such products (factory
costs may be less than 10 percent of wholesale price 3) stimulates entry, which
take the form of imitative research activity on the part of other drug firms,
aimed at devising a substitute for the patented drug which they can patent and
advertise as superior to the original. Rivalry is thus diverted from price com-
petition to product differentiation, as rival patented drugs attempt to take over
the prescription market for a given group of disorders by intensive advertise-
ment to doctors (the only agents qualified to prescribe) through direct mail,
itinerant salesman (detailmen), and the sponsoring of medical conventions.
High unit profit margins can finance large (and often imitative, duplicative,
and wasteful) research programs, overpowering advertising campaigns, and
still yield very high profits on investment. Since costs are low, and economies
of large scale production apparently unimportant, small firms might be able to
compete with large firms on a cost basis, but they suffer from two disabilities.
Being unable to finance sales campaigns, their products do not come to the at-
tention of the prescribing physician. Furthermore, the detailmen employed by
the major firms make a practice of disparaging low-priced drugs, inducing the
physician to equate low price with low quality. At best, the physician may not
be very price-conscious, since he does not pay for the drugs he prescribes.
The situation is aggravated by weaknesses in patent laws and practices, and
in the regulations governing drug advertising. Patent protection is absolute as
regards a given compound, but weak as regards slight variations from that
compound. The game of "molecular manipulation" is a popular one with drug
firm research personnel: the goal is to devise a derivate compound which has
the same therapeutic effect as some primary patented compound, but which
by virtue of its marginally different chemical structure can be patented as a
different drug and advertised as a "new and superior" healing agent.4 Patents
can also be obtained for naturally occurring substances if never previously
isolated, for processes occurring in nature, such as fermentation, and even for
combinations of existing patented drugs. But since most major drug firms have
research programs which substantially duplicate each other, near-simultaneous
discovery of the same drug is a not infrequent occurrence. When several patent
applications covering the same compound are received by the patent office, an
interference is declared, and the parties attempt to assist in resolving the
question of priority to the satisfaction of the patent office. There is, however,
1 The author is assistant professor of economics at Rice University.
2 Study of Administered Prices in Tue Drug Industry, Report of the Subcommittee on
Antitrust and Monopoly of the Senate Judiciary Committee, Pursuant to Senate Resolution
52, Eighty-Seventh Congress, First Session, Washington, D.C., Government Printing Office,
1961.
3lbid.,pp.15-16.
4Ibid., p. 16.
PAGENO="0449"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1999
no barrier to private settlement of interferences by agreements among the
parties themselves, involving concession of the patent to one firm in return for
granting licensing rights and perhaps other privileges to the other firms. The
patent office rather welcomes such private settlements in view of the pressure
of work on its staff, but the antitrust division has been critical of the nature of
many private interference settlements.5
Advertising abuses relate chiefly to drug nomenclature and to the quantity
and quality of sales efforts. Four hundred ~ or more new drugs are marketed each
year in the United States. Each new drug should be given a generic name to
identify it; individual firms marketing the drug may confer upon it their own
brand names. If a physician prescribes a drug by generic name, the pharmacist
may fill the prescription with any brand of the drug; if the prescription is by
brand name, only that firm's brand may be dispensed. Evidence indicates that
generic names are designed to minimize use (excessive length, complexity, and
clumsiness) while brand names are brief and memorable. Regulations require
that generic names be displayed in all advertisements, by they are often
obscured by the use of microscopic type face, by being concealed in obscure
places in the advertisements, and are sometime simply omitted.6 Some drugs
have no generic name; others have more than one. This deliverately cultivated
confusion is evidently intended to suppress use of generic names in favor of
brand names.7 By this means, trademarks are made to supplement patents as
monopolistic devices. The same patented drug may be sold by ten different
licensees under ten different highly advertised brand names, and each licensee
through sales efforts may succeed in differentiating in the mind of the physician
this physically homogeneous drug. In those rare cases where there is no patent
protection, generically-named products sold by small firms may be priced at ten
per cent or less of the price of the majOr drug firms, but price competition is
rendered impossible unless prescriptions are written generically. However, the
detailmen have been very successful in their disparagement campaigns; surveys
show that almost 80 per cent of drug prescriptions are written by use of brand
names.8 The quantity of drug advertisements is overwhelming in itself, includ-
ing not only propaganda but also free drug samples, trinkets (toy urinals,
chinese dolls, head cushions, etc.) and gifts such as golf balls engraved with
the recipient's name. But the quality of advertising claims is the greatest
obstacle to the enlightened practice of~ medicine. Dozens of examples of mis-
representation were unearthed at the Senate hearings; one example must suf-
fice: a drug firm mailed to physicians advertising copy showing X-ray photos
clearly designed to imply a dramatic recovery in a patient's condition before
and after the use of the advertised drug. Upon inquiry to the firm's medical
director, it developed that the two X-rays were of entirely different persons with
qualitatively different disease conditions, and that neither had ever used the
drug being advertised.9
Ill. INDICATED REFORMS
To an economist, the abuses in drugs seem amenable to relatively simple re-
forms. Drug product patents should be abolished; drug process patents should
be subject to compulsory licensing at reasonable royalties. Generic names should
be simplified, the use of brand names should be eliminated, and firms should be
required to advertise and sell their products under labels giving the generic
name of the drug, followed by the name of the firm, e.g., chloramphenicol-Parke,
Davis instead of "Chloromycetin." Food and Drug Administration drug plant
inspection authority and funds should be increased in order to guarantee the
safety of all drugs on the market, rendering specious all disparagement cam-
paigns. Control of the quality of drug advertising should be made truly effective.
If rigorously enforced, these reforms should suffice. Absence of patent protec-
tion and mandatory use of generic names would allow price competition between
large and small firms; excessive profits would disappear, and with them would
disappear the ability to carry out wasteful and duplicative "research" programs,
5lbid., p.48.
6 study of Administered Prices, op. cit., p. 234.
Firms may, however, on occasion find generic names of use as when the necessity
arises for sending out circulars to warm physicians about side effects. Hearings on S. 1552,
Part 5, p. 2938.
8lbid., Part 18, pp. 10481-10482.
Ibid., Part 7, pp. 3301-3310.
81-280 0-68-pt. 5-29
PAGENO="0450"
2000 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
and to advertise so intensively as to prevent the products of smaller firms from
coming to the attention of the physician. Much of drug research would thereby
be shifted from private to public and university channels, with a probable in-
crease in the efficiency of such efforts, since the patent-induced incentives to
engage in duplicative research, molecular manipulation, and promote unpromis-
ing drugs in the interests of their sales potential rather than their efficacy,
would be removed.
IV. S. 1552: FROM BILL TO ACT
A. OriginaZ provisions
Senator Kefauver, however, did not contemplate so sweeping a reform of
existing practices, perhaps in view of the political difficulties. His original bill
did, nevertheless, incorporate a number of valuable reforms, particularly of
patents and drug nomenclature. The salient provisions of the bill may be listed
as follows:
(1) Drug patents shall take effect as of the date of the new drug applica-
tion (or in the case of drugs for which no new drug application is necessary,
the date for filing the patent application) and shall grant unconditional
monopoly power for a period of only three years, after which such patents
are subject to compulsory licensings at a royalty not exceeding 8 per cent of
the sales price.
(2) No patents are to be issued for compounds merely embodying molecu-
lar manipulations of existing drugs, or combinations of existing drugs, unless
the Secretary of the Department of Health, Education and Welfare deter-
mines that such a drug has a significantly greater therapeutic effect than
the unmodified drug, or of the combined drugs when taken separately.
(3) The Sherman Act should prohibit private drug patent interference
settlements or any other private arrangements whereby parties concede
patent priority claims in return for royalty-splitting agreements, differen-
tially favorable royalty rates for participants relative to third parties, or
limitation of licensing to parties to the agreement.
(4) The Food and Drug Administration should be given authority to
establish generic names of drugs, with a view toward simplification and
increased use of such names.
(5) The Food and Drug Administration shall pass upon the efficacy as
well as safety of drugs.
(6) The Secretary of the Department of Health, Education and Welfare
should publish and distribute to physicians and hospitals copies of drug
firm package inserts which describe the action of drugs and give data on
dosage, contraindications, and side effects. (At present, such inserts go by
only to pharmacists, not physicians.)
(7) All drug firms must be licensed by the Secretary of the Department of
Health, Education and Welfare, and must submit to inspections of plant and
equipment. (Plant inspection regulations are also to be made more adequate.'
(8) All drug advertising shall include (a) the generic name in type face
as large as the brand name, (b) a statement of the conditions for which
the drug is an effective treatment, and (c) a statement of all side effects and
other warnings.
The reforms contemplated are rather modest. The provision for making patents
effective as of the date of filing a new drug application is intended to eliminate
any incentive for delay on the part of the applicant in patent proceedings, par-
ticularly in negotiating interferences in the interest of prolonging effective patent
protection, not only for the successful applicant, but also for the entire group
of interim licensees. The three-year term of absolute patent monopoly is intended
to allow the successful firm to recoup its research costs during an initial monop-
oly period of high prices and no rivals; the compulsory licensing provision dur-
ing the next 14 years is tempered by allowing the patent holder to charge an S
per cent royalty, a very high royalty by drug industry standards, the usual rates
being between two and six per cent. The prohibition of drug combination or
molecular manipulation patents in the absence of evidence of superior efficacy
is a laudable (although administratively difficult) attempt to halt wasteful du-
plicative research efforts.
The proposed Sherman Act amendment is rather strict in that it entirely pro-
hibits private drug patent interference settlements, by making it illegal to with-
draw any pending drug patent application. This is desirable in that private settle-
PAGENO="0451"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2001
ments in effect transfer the determination~ of priority of invention from the Pat-
ent Office to the rival parties themselves. Such private agreements often contain
stipulations which restrain trade. One party is awarded the patent, and as a
condition of surrendering their claims, the other parties are guaranteed licenses,
and may even begin to pay royalties before the patent is issued. The agreements
usually contain restrictive provisions under which the parties agree not to sell
to outsiders, not to sell the drug in bulk powder form (which might get into the
hands of non-licensed dealers who could tablet and bottle it and sell it at low
prices), and to accept other limitations on their marketing practices. In addition,
all of the parties to such agreements will almost invariably sell at identical
prices.10
Most of `the other proposed reforms would alleviate abuses in drug marketing.
Although the use of brand names is not outlawed, reforms in generic names
would be a step in the right direction. The requirement that the FDA pass on
the efficacy `as well as on the safety of drugs is a very reasonable one, but one
which was bitterly fought by some drug firms during the original hearings on
administered prices. In the absence of such a requirement, drugs which are not
clearly harmful, and for which various physicians have written "testimon-
ials" supporting a new drug application, may be allowed on the market, at which
point the market success of the drug depends (`at least for a while) more on the
skill of the sales department than on the intrinsic therapeutic merits of the
compouii~L Several physicians testified at the hearings that not a few drugs cur-
rently on the market are absolutely useless.11 The sixth provision is aimed at one
of the many paradoxes in `drug marketing. Drug makers are required by FDA
regulations to print up a leaflet describing the uses, dosages, and side effects
of each prescription drug, *but the leaflet need only be included in the drug
package, which goes to the pharmacist, who, having no use for it, throws it away.
The physician, to whom such information is of the most vital importance, has
no ready access to this information, but must routinely rely on advertisements
and detailmen. Distribution of all such material to physicians is obviously an
imperative necessity. Finally, the provisions for licensing all drug makers and
for requiring more adequate inspection of drug plants is intended to insure the
quality of all drugs, generic and brand name, and thus hopefully to reduce the
effectiveness of generic drug disparagement an'd increase the physician's willing-
ness to prescribe by generic name.
B. Senate hearings on 5. 1552
The hearings on S. 1552 were held between July 19(11 and February 19(12. The
battle lines were rather tightly drawn up on most issues.12 In the medical pro-
j~ession, the American Medical Association (AMA) betrayed dedicated opposition
to every section of the bill upon which they took an official position; on the other
hand, of the eleven medical educators who appeared either individually, or as
representatives of such groups as the American Public Health Association, ten
favored the proposed reforms and an eleventh expressed opposition to the patent
provisions, an area perhaps not strictly within his professional competence.
Other physicians appeared in their capacities as administrators `of hospitals or
health insurance plans, and indicated,' general approval. Ten patent attorneys
and other patent spokesmen testified; nine of them, including representatives of
the American Bar Association, The National Association of Manufacturers, the
American Patent Law Association, etc., found little or no virtue in any `part of
the `bill. The tenth, an attorney from a small town in a mid-western state, and
representing only himself, regisitered substantial agreement with the aims of the
bill. The Pharmaceutical Manufacturers Association opposed most of the bill's
provisions, but showed limited agreement in some areas; other drug industry
witnesses were not nearly so favorable.
Testimony was given by two economics professors. One appeared at the request
of the Subcommittee as an expert on the economics of patents, and gave testi-
mony to the effect that the patent provisions of S. 1552 would probably not cause
10 Report on 5. 1552 op. cit., p. 45.
11 See, for example, .tTeariflg8, op. cit., Part 1, p. 285.
12 Battle lines were also drawn up within the Subcommittee itself, the majority (often
represented by Kefauver alone) In favor of reform, and the minority (represented by Dirksen
and Hruska) apparently concerned with obstruction of the proceedings and harrassment
of witnesses favorable to the bill. Hruska in particular was adept at the latter, not
scrupling to Insinuate Communist leanings in those with whom he disagreed. See Hearings,
op. cit., Part 3, pp. 1410-1411.
PAGENO="0452"
2002 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the ruin of the drug industry if adopted; indeed, he saw- many faults in the
present patent system, and suspected it w-ould be possible to increase the effi-
ciency of drug research by transferring some of it from private to public
auspices. The other was retained by counsel for the Pharmaceutical l\ianufac-
turers Association, and he could find no evidence of any need for Kefauver's
bill.
Testimony was also obtained from a number of labor and consumer groups.
Representatives from the AFL-CIO, the UAW, and the Internationa.1 Union of
Electrical, Radio and Machine Workers supported the bill, as did spokesmen for
Consumers Union, the National Consumers League, and the Cooperative League
of the United States. Two representatives of retail Pharmacy appeared. One
spoke for the American Pharmaceutical Association, a group of pharmacists. He
articulated a viewpoint consonant w-ith that of the drug makers, and expressed
great concern that the government not take the unwise step of distributing drug
information to physicians. Another pharmacist, a former teacher and state phar-
macy and drug law enforcement officer, found considerable merit in many of the
reforms. The last parties to testify were advertising agency representatives who
predictably saw no apparent need for any advertising reforms.
It is likely that the most influential testimony w-as that given by the American
Medical Association, the various patent law- groups, and the Pharmaceutical
Manufacturers Association. It is instructive briefly to review the character and
merits of the testimony given by these groups.
1. Testimony of the American Medical Association
The AMA made no recommendations in regard to the antitrust law amend-
ments, but took the view that none of the other proposed reforms were defen-
sible, including the requirement that the FDA pass on the efficacy of drugs.
Their spokesman, Dr. Hugh Hussey, recognized the need for certain improve-
ments in regard to drug nomenclature and physician information. The AMA
preferred, however, to carry out the reforms itself, in cooperation with the drug
industry, but with no participation by any public bodies.n One may be for-
given for entertaining the view- that the AMA position is simply a defense of
the revenues it obtains from drug firms for advertisements in its journals. The
roots of AMA opposition are more ramified and complex, but it is easy to trace
the influence of advertising revenues, as seen against the background of prior
AMA actions.
First, the AMA reform program was adopted only five weeks prior to the sched-
uled appearance of its representatives before the Subcommittee, and the imple-
mentation of the program was to be gradual, extending over two or more years.
Five years previously, however, a similar reform program in response to similar
legislative demands, was outlined by the AMA, proposing cooperation between
itself and the drug industry to control misrepresentation in advertising, but it
entirely failed of implementation.~ Second, the AMA has become increasingly
dependent upon drug advertising for its own financing. In 1949, medical journal
advertising revenues comprised about 31 per cent of all AMA revenues: in 1955,
about 44 per cent; and in 1900, a little over 50 per cent. Total advertising rev-
enues, however, been augmented by royalties received from the leasing of the
rights to use mailing lists of physicians.15 The sums received have increased
-from small amounts to about 5.6 per cent of total revenues by 1960. Hence in
1960 about 56 per cent of AMA revenues came from drug firm advertising efforts.
Third, the AMA in recent years has become increasingly permissive in its
attitudes toward advertising standards, and it is likely that the increase in its
advertising revenues is in good part attributable to this. The period of increasing
leniency coincided with that of increasing advertising revenues. The measures
by which this more lenient policy was evolved, or from which it can be inferred,
may best be described and interpreted in chronological order.
13Ib Part 1, PP. 47-49.
14 Ibid., Part 1, p. 341. Testimony of Dr. Allan M. Butler, Professor Emeritus at Harvard
University.
15 Obviously a royalty on direct mail advertisements ($2 per thousand mailings) creates -
a direct financial interest in maximizing the volume of such traffic, and is hence undesirable
in even greater degree than policies to increase advertising in the AMA journals; There is
evidence that advertising standards in AMA journals are still higher than those in most
(but not all) other medical journals, but there is no AMA control over the quality of
direct mailings. The royalty income feature of AMA finances tends to justify in part the
accusation of James Carey, president of the International Union of Electrical, Radio, and
Machine Workers, that "The AMA, in our opinion, is just a business." (Ibid. Part 5,
p. 2731.) It is ironic that, while most students of the AMA compare it to a trade union, a
leading trade union spokesman sees It as a business.
PAGENO="0453"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2003
In 1950, a rule (adopted in 1905) that only the inventor of a new drug could
use his brand name of the drug in advertising in the AMA journals, was dropped.
In 1952 the publication of a handbook, Useful Drugs (issued periodically since
1917) was discontinued. (The 1961 Ai\'IA proposals contemplated the publication
of a similar book which, however, would include all drugs and would thus
lack the discrimination between useful and less than useful drugs characteristic
of the earlier publication.) In 1955, major advertising policy changes were made
in response to a survey by Ben Gaffin and Associates addressed to the problem of
increasing AMA advertising revenues. Before 1955, the AMA Council on Drugs,
an expert technical body, had effective control over advertising. Only Council-
approved drugs could be advertised. Approval required the obtaining of a "Seal
of Acceptance" which was granted only after the drug firm had submitted to
the Council all the data it requested, including both favorable and unfavorable
reports. (By way of contrast, at that time there was no requirement that all
test data, favorable and unfavorable, be sent to the FDA with new drug appli-
cations.) The Council on occasion inspected drug factories. It exercised some con-
trol over generic names, since only generic names approved by the Council could
be used in AMA journal advertising. The Council examined and passed on
all advertising.
In 1952, Ben Gaffin and Associates began a survey to determine how AMA
journals could get more advertising. It noted that AMA advertising had in-
creased only 3 per cent since 1948, while other medical journals had increases
of 40 per cent. The survey indicated that while physicians and small drug firms
thought well of AMA advertising control policies, large firms were critical, par-
ticularly of the "Seal of Acceptance" program. The Gaffin study recommended
that the Seal of Acceptance program be dropped, and the advertising controls
be liberalized.16
In 1955, several major changes were made in advertising policy. Advertising
control was taken out of the hands of the Council on Drugs. The Seal program
(in effect in that form since 1929) was abolished. The Council lost its influence
over generic names by this means, and also its power to elicit unfavorable as
well as favorable evidence on drugs. The Council lost all control over advertis-
ing.17 While rules requiring generic name advertising were being abolished, edi-
torial changes were made which drew increased attention to brand names.18 It is
instructive to note that the 1961 reform proposals of the AMA did not include
reinstatement of the control over drug advertising on the part of the Council
on Drugs. Indeed, the preparing its presentation for the hearing, it did not even
seek the advice of the Council, its own expert advisory body on the subject matter
to which the hearings were addressed. This fact was testified to with dismay
by more than one member of the Council.19 Had the Council been influential
in drafting the AMA's testimony, it is hardly conceivable that the AMA would
have gone on record as opposing the requirement that a drug need be efficacious
in order to merit a new drug permit.
During 1955 a microbiological laboratory, established in 1949 to establish
purity standards for antibiotics and other drugs, was abandoned. And in 1959 a
chemical laboratory, founded in 1906 to formulate and develope standards for
drugs, was abandoned because of overwork and lack of finances.20
16 Ibid., Part 1, PP. 102-103; Part 2, pp. 490ff. Part II of the survey includes the
following: ". . . while the possibility of increasing advertising revenue by several million
dollars per year is a good motive for putting into effect the information gained from these
two studies, there is an even more important reason for so doing. . . . the AMA has an
opportunity to assume leadership in improving some $130 million worth of medical adver-
tising per year." It thus becomes apparent that it was the duty of the A?~tA to enrich itself.
17 Mr. Stetler of the AMA denied that these changes had occurred in response to Gaffin's
survey. Gaffin himself was convinced otherwise. In another drug survey, he links his
recommendations and the AMA's actions: "The survey of pharmaceutical advertisers
played a part in bringing about a number of policy changes . . . [including] the eventual
dropping of the 58-year old council seal of acceptance program." Ibid., Part 1, p. 125. If
the AMA disagreed with Gaffin's diagnosis, they took no disciplinary action. Gaffin continued
to conduct their surveys throughout the 1950's.
18 Dr. Hussey volunteered that "Trade names, formerly listed at the end of monographs,
also have been placed immediately after the nonproprietary titles of monographs and on the
front page of the Journal to more readily catch the eye of Journal readers." Ibid., Part 1,
p. 106.
19 Ibid., Part 1, pp. 216, 375.
20 The Council of Drugs also suffered from insufficient finances, its appropriations in-
creasing from $135,000 in 1950 to $156,000 in 1954 (the last year of active Council adver-
tising control) and declining to $75,000 in 1960. Curiously, the public relations department
budget increased prodigiously, from $102,000 in 1950 to $494,000 in 1958. Ibid., Part 1,
pp. 126-127.
PAGENO="0454"
2004 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
It seems safe to conclude that on the basis of the recent record, the role of
the AMA in contributing to high standards of drug advertising has steadily
diminished. The adoption of the joint AMA-drug industry program outlined by
Dr. Hussey would provide at best a questionable safeguard for the drug
consumer.
2. Witnesses Testifytng on Patent Provisions
The great majority of the testimony on the proposed patent reforms revealed
the presence of a tropismic conservative reaction against any modification of
the patent system, and the absence of any great evidence that the witnesses
had studied the concrete operation of the patent system in the framework of
the drug industry.boa The presentation given by Joseph Jackson, the chairman
of the patent, trademark, and copyright law section of the American Bar As-
sociation, may serve as an illustration. Speaking for the American Bar Associa-
tion (ABA), Jackson reported that the ABA Board of Governors had adopted
five resolutions in regard to S. 1552: (1) disapproval of the antitrust pro-
visions limiting private patent interference settlements; (2) disapproval of
the requirement of "non-obviousness" regarding the patentability of a drug
product; (3) disapproval of the requirement of proof of significantly greater
therapeutic effect in order to qualify molecular modifications for drug patents;
* (4) disapproval of the distinctive treatment of drug patents with regard to
the date of effective patent protection; (5) disapproval of compulsory licensing
for drug patents.21
When Mr. Jackson was examined on his testimony, it developed that these
resolutions, purporting to speak for the ABA, a `body with a membership of some
102,000, had been drafted in Saint LouIs, about ten days before the Hearings, by
a group of 150 to 200 patent lawyers, with orders to "act with great aecelera~Uon"
because of `the "emergency." ~ These resolutions were then submitted to the
Board of Governors of the ABA (no member of which is a patent lawyer) and
were PrOmulgated by them more or less over the head of the House of Delegates,
the representative deliberative body of the ABA. (This is an extraordinary pro-
cedure which is nevertheless permittof `by the constitution of the ABA.) Further-
more, members of the ABA were apparently not given notice of this action by
their Board of Governors, as evidenced by the surprise of several members of the
Subcommittee staff belonging to ABA. It further developed that the ABA group
required less than `an hour and a half to reach conclusions on matters which the
Subcommittee and its staff had been studying for over two years. Jackson ad-
mitted that there had been no discussion of the economics of the `drug industry
and its relation to patents, no study of costs, no systematic consideration of prof-
its, no attention to concentration or to the interdependence of major firms, no
concern with entry conditions, no inquiry into price policies, and not even any
consideration of the nature of incentives and the quality of research effort in
drugs. Jackson explained that none of these matters had been considered, since
"If we had discussed all these subjects you are presenting, we would still be in
Saint Louis. We would not be here with ally resolutions at all." ~ But that is
precisely the point. It may be inferred that the only `aim of the ABA action was
to go on `record as condemning S. 1552 in time for the Hearings, regardless of
the factual merits of the bill.
The patent provisions stand or fall depending upon their application to the
specific cireumstances of the drug industry; the entire patent system is not at
stake. But patent attorneys in testifying almost universally took the position
that any amendment to the patent laws in respect of a particular industry would
necessarily imperil the patent laws with regard to all other industries. Jackson
no doubt faithfully reported this attitude on the part of the ABA group when,
in response to an Observation by the Subcommittee counsel that `S. 1552 was
limited in its application to the drug industry alone, "We were afraid we would
be faced with a special antitrust law and a special patent law for butter and
eggs, and another one far milk and beer, and so on in different areas."24 One gets
tsa There Is no evidence of any articulate "grass roots" support for the opponents of the
patent provisions of S. 1552. Symptomatic of the testimony in behalf of the patent status
quo is a communication from an individual styled "Clair V. Johnson, Newfane, Vermont,
Patent Lawyer," who roundly condemns the bill. It appears that Mr. Johnson Is also a
director of U.S. Vitamin and Pharmaceutical Corporation. (Part 3, p. 1601).
21Ibid., Part 3, p. 1473.
~ Ibid., Part 3, p. 1~72.
~ Ibid., Part 3, p. 1481.
24Ibid., Part 3, p. 1480.
PAGENO="0455"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2005
the impression from the main drift of tile patent testimony that this rather naive
contention is not advanced merely as a straw man, but that this consideration
alone was enough to close the minds of patent attorneys regarding S. 1552. A
further observation by Jackson leads one to wonder why his group should have
debated even as long as an hour and a half on their resolutions: "There was
only one man of the whole group who had anything good to say about the
law . . . he said it might not be politically expedient to show the full extent of
our disapproval."
Had tile ABA group referred to the Subcommittee's Report on the drug in-
dustry, they would have discovered that it is not at all unusual for nations to
make exceptions of their drug industries in regard to patent privileges. Of
some 77 nations with patent laws, 49 absolutely prohibit drug product patents
on grounds of public policy, and 25 others have provisions for compulsory li-
censing. Only Panama, Belgium, and the United States allow unrestricted drug
product patent privileges.~° Indeed, to judge by a comparison of drug discoveries
by drug firms in countries with and without product patents, it is `by no means
clear that the patent incentive is necessary to elicit productive drug research.
The patent privilege restrictions embodied in S. 1552, although more liberal, are
closely related to those of Germany, long one of the world leaders in drug re-
search. German patent law denies drug product patents, but drug processes may
be patented, and such patents cover the products made by those processes. If,
however, alternative processes are devised to produce the same drug, the drug
in question then fails to retain its effective protection. Germany permits un-
restricted drug process patents for a period of three years. After that, compul-
sory licensing is required, with royalties of between 5 and 10 percent, as
determined by the decision of a special tribunal.27 Professor Machlup of Princeton
University testified that not only has the existence of such a patent law in
Germany failed to halt productive drug research, but that its expediency and
equity is not even questioned.3°
One ominous tendency which came to light at the patent hearings concerns
certain evidence that pressure groups are attempting to weaken the protection
which drug buyers currently enjoy under the patent laws of many nations of the
world. Professor Machiup testified that in recent years several countries, in con-
sequence of pressure by industrial groups, have made their patent laws more
favorable to such industries.29 It is ironic that at the time when efforts are being
made to `bring United States drug patent policy into line with the more enlight-
ened practices of other industrial countries, some of these very countries are
experiencing a retrograde tendency. In 1949, England amended its patent law to
allow drug products to be patented, but required compulsory licensing. France
adopted a largely similar law in 1960.~°
Drug and other chemical interests appear to have been active in connection
with a certain diplomatic conference held in LiSbon in 1958 for the purpose of
revising the International Convention for the Protection of Industrial Property.
Prominent among the United States delegation were Roland Libonati, a Repre-
sentative from New Jersey (where many drug firms have plants) and P. J.
Federico, examiner in chief of the United States Patent Office, who testified
at the hearings and was at pains `to take issue with the Subcommittee's conten-
tions on the relative strength of patent protection in countries with and without
product patents. The Lisbon convention agenda contained an item proposing
to require all countries adhering to the, Oonvention to grant patents for chemical
products, including pharmaceuticals. The United States delegation sponsored
this resolution, and it fell to Mr. Federico's lot to expedite the proceedings. Tile
resolution did not pass-12 nations voted against it-'but observers were some-
what surprised to find that those voting in favor included the delegations from
eight countries which proiibited drug product patents. After the resolution
failed, Germany introduced a resolution to recommend that member countries
study the question with a view toward revising their patent laws. This resolu-
tion passed. Such efforts a't "study" may be bearing fruit: France, one of the
countries opposing the original resolution in 1958, adopted its own drug product
patent law in 1960, and the Scandinavian countries and Finland are trying to
work out a common patent law which will extend to drug products.
Ibid., Part 3, p. 1480.
2~ Study of Administered Price8, op. cit., p. 106.
~ Hearings on S. 1552, Op. cit., p. 1385.
28 Ibid., p. 1385.
29 Ibid., Part 3, p. 1371.
3°Ibid., PartS, pp. 1201, 1216.
PAGENO="0456"
2006 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
3. Testimony in Behalf of the Pharmacentical Manufactnrers Association
The influence of the drug makers lobby in behalf of aborting reform efforts
was not limited to their presentation of testimony at the hearings, although
a survey of their manifold efforts a influencing-and even drafting-the legisla-
tion on to control their own activities is beyond the scope of this paper.zoa Suffice
it to say that the legislators must have found the PMA testimony the most con-
vincing case with which they were presented, since the PMA recommendations
agreed to a great extent with the Act as it was finally passed. Mr. E. N. Beesley,
chairman of the PMA, presented his views as follows: (1) agreement with the
the provision that FDA should pass on the efficacy of drugs, but judged on the
basis of "substantial" and not "preponderant" evidence; (2) agreement that dis-
tribution to physicians of full information on drugs is important, but that the
provisions of S. 1552 in this respect are unnecessary, in view of the AMA's pro-
motion of a cooperative program with PMA which will handle these details;
(3) qualified agreement that some more systematic approach to the designation
of generic names is desirable, but a preference for the announced AMA program
of voluntary cooperation with PMA on generic names, with the concession that if
such voluntary means do not achieve results, the FDA may designate a generic
name; (4) disagreement with the compulsory licensing of drug manufacturers,
but a readiness to compromise on simple registration, together with an expanded
program of factory inspection by the FDA; (5) complete rejection of all phases
of the patent program: condemnation of compulsory patent licensing, disap-
proval of the restrictions of the patenting on molecular modifications and drug
combinations, and disagreement w-ith the contemplated restrictions on private
patent interference settlements, but with the concession that (as suggested by
Patent Commissioner Ladd and Assistant Attorney General Loevinger) firms
be required to file private patent interference settlements and other relevant data
at the Patent Office for public inspection.3'
The PMA testimony revealed a supporting willingness to compromise in regard
to some parts of the bill, but the economic impact of these concessions would be
but slight. It is also to be noted that while much is made of the prospects for co-
operative programs with the AMA, nothing is included which would bring ad-
vertising claims under any direct control. The prospects for constructive reform
in these respects are made doubtful not only by the past record of the AMA but
also by certain recent activities of organized drug makers in relation to attempts
by a group of physicians, acting independently, to increase the standard of drug
advertising.~
The PMA showed a clear intent to resist economic reforms which would reduce
drug profits. Mr. Beesley testified directly to this effect: `Deletion of patent and
trademark protection . . . would increasingly dilute the intensive competition
30a Kefauver, angered at the opposition of the subcommittee minority, and apparently sus-
pecting it of connivance with the drug lobby, took the politically risky step of confiding in
a journalist, Richard Harris, who was then doing a study of the drug industry. Kefauver
made a detailed first hand report of the day by day political maneuvering. The resulting
chronicle, summarized in the New Yorker, March 28, 1964, pp. 46 if., makes fascinating
reading. It appears that PMA's initial reaction to Kefauver's bill was divided. One faction
reportedly preferred to buy outright the necessary votes to defeat the bill; another thought
it wiser to influence legislation in the making, conceding some points in order to moderate
reformist pressures and prevent the eventual passage of truly substantial reforms. It
appears that, under White House auspices, representatives of PMA were asked personally
to aid in drawing up an alternative drug bill to replace that of Kefauver. At the subse-
quently notorious "secret meeting" at which the substitute bill was drawn up, the only
persons present were PMA lobbyists, legislative draftsmen from the Department of Health,
Education, and Welfare, and counsel fbr the Subcommittee minority. Not a single senator
was present. Nor had Kefauver, or any other member of the subcommittee majority been
informed of the meeting. Fortunately, this attempt to railroad through an impotent bill
was nullified when Kefauver succeeded in drawing public attention to it and the Adminis-
tration saw fit to repudiate the substitute bill.
~` Ibid., Part 4, nn. 1997-2007.
~` The AMA seal of acceptance had been valued by some of the more responsible drug
firms, large and small. With the demise of this program, some firms then supported a group
of physicians who undertook to establish drug advertising standards and to review adver-
tising copy, and who later became known as the Physicians Council. This group awarded
approved advertisers with an insignia which could be displayed in advertisements. As
the program became more widely accepted, the industry began to view it with suspicion,
and in February 1959 the medical section of the American Drug Manufacturers Association
(now part of PMA) adopted a resolution calling upon its members to refuse cooperation
with the Physicians Council. The resolution claims that such outside review ould
disrupt the normal, prompt, and effective processes of ethical drug promotion," a state-
ment with which the Physicians Council would probably not find fault. The financial sup-
port of the Council disappeared, and they abandoned their insignia program. Testimony
of Dr. Julius Richmond of New York State University, Thid., pp. 361-368.
PAGENO="0457"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2007
for superiority in discovery and manufacturing; it would leave price competition
as the only remaining form of competition, and this would be a tragic result in
an area so vital to the public health." ~ This is precisely the view of the subcom-
mittee majority, although expressed from the point of view of an antipodal set
of values. The Subcommittee was at great pains to prove that the "intensive com-
petition for superiority" (i.e., in profits) stimulated by patent availability is
wasteful, duplicative, and supports high prices, but that the replacement of prod-
uct differentiation by price competition would be tragic only to excessive profit
levels, and would improve resource allocation in the drug market and eliminate
the inequity of burdening the sick with high prices.
The drug industry tried to rebut these presumptions of noncompetitiveness by
presenting a great deal of economic evidence regarding (1) price index data pur-
porting to show a decrease in drug prices from 1949 to 1960, and (2) firm con-
centration and turnover in firm rank order of sales volume in various drug sub-
markets. Professor Markham of Princeton, retained by counsel for the PMA,
presented a price index of prescription drugs, showing a decline of 7.6 percent
for the period 1949-1960, or about one-half of one percent per year, compounded
annually. This evidence, if valid, would argue for a remarkable degree of price
stability in the industry, and might sustain the Subcommittee's loose usage of the
loose term, "administered prices", one agreed-upon characteristic of which is
that they do not typically increase during infiations to the extent that one would
expect on the basis of short-run profit-maximizing theory. But what is the cause
of this slight decline? Dr. Blair of the subcommittee staff pointed out that pat-
ented drugs almost never change in price, while non-patented drugs typically
experience a substantial price decline over time, and suggested that the chain
index might reflect the presence of a number of patented drugs with fixed prices,
and a number of unpatented drugs which declined in price. He asked further
information be submitted with regard to the relative price changes of patented
and unpatented drugs separately, and also for other information, including a
tabulation of the drugs and prices employed in the index. Apparently the com-
putations on patented versus unpatented drug price trends were never submitted,
although the other material requested was supplied. Anyone who takes the trouble
to look through the basic price data used in computing the index will find that
of the 329 drugs included, no fewer than 193, or 59 percent of the total, had
experienced no price change during the, period since their introduction in the
index (not all drugs in the index were sold during 1949; many appear oily in
subsequent years); 95 drugs (28 percent of the total) had experienced price in-
creases (i.e., price in 1960 higher than price in initial year of appearance), and
only 41 drugs (13 percent of the total) experienced price decreases.34 This tends to
confirm the hypothesis of price stability. It is of further interest to note that
the BLS index of the prices of drug and pharmaceutical materials declined by
32.7 percent during this time, while its index for the cost of retail prescriptions
increased by 27.8 percent.33 It is perhaps evidence of lack of competition that
prices of drugs dropped only 7.6 percent while prices of inputs dropped 32.7
percent; it is also curious that prescription costs increased 27.8 percent during
this period if prescription drug costs actually fell by 7.6 percent, in view of the
facts that (1) most prescription drugs are priced by druggists by adding a fixed
markup (66% percent) over wholesale price, and that (2) if anything, the labor
and other compounding costs of prescriptions should have declined since 1949
as a result of the marked decrease in the percentage of prescriptions which re-
quird compounding by the pharmacist, rather than the transfer of capsules to
different bottles.
The concentration and turnover ratio data presented by Markham do indicate
very considerable turnover among drug firms in relative sales rank. The con-
clusions reached as regards concentration as a whole must be interpreted in the
light not only of the usual imperfections of concentration ratio data (especially
the implicit assumption of independence of action among firms) but also the fact
that the data refer to sales of the finished product rather than to production of
the bulk powder and there are great differences between concentration of these
two in drugs. In 1958, for example, of the 51 most important prescription drugs,
27 were made by only one firm, 18 by two or three firms, and six by four to
~ Ibid., pt. 4, p. 2003.
~4Ibid. pt. 4, pp. 2576-d through 2576-k.
~ Ibid., pt. 4, pp. 2469-2470.
PAGENO="0458"
2008 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
seven firms; with regard to sales, only 17 were sold by only one firm, 12 by two
or three firms, and 22 by four to thirteen firms.~ Production of patented drugs is
usually completely monopolized by the patent holder (unless there has been an
interference settlement, with mutual licensing among contestants), but the sale of
such drugs may not be monopolized. The patent holder thay not have a sufficiently
developed marketing network, enough detailmen, etc., to fully exploit the produc-
tion monopoly, and may hence license other firms with wider marketing facilities
to sell the finished product, such licensees tableting and bottling the bulk powder
which the patent holder sells to them. For example, Carter licenses meprobamate
("Miltown") to Wyeth in order to increase its total sales.~ While it is illegal
for patent holders to dictate the pricing policies of licensees, price competition
between such parties is almost unknown. But it is quite consistent with continual
pure monopoly control of drug production that several of the licensees of the
patent holder may alter their relative positions in the market from year to year.
The turnover data presented by Markham do not indicate the extent to which this
may be the case. The presumption made at the hearings that most of the turnover
is induced by firms developing new products remains only a presumption; Mr.
Mannis of Arthur D. Little, a participant in Markham's study, wasn't quite sure.~
Markham placed great emphasis on the amount of turnover, but to assess the de-
gree of workability of competition evidenced by such turnover, it is necessary to
determine whether it was brought about by price competition, by genuine product
competition, or by mere product differentiation with its accompanying wastes in
the drug industry. The evidence (admittedly uneven in quality) of some ten
thousand pages of Senate hearings indicates that a very small weight be placed
on the first factor, a quite moderate weight on the second factor (less important
since 1955), and the preponderance of weight on the last. But Markham seems to
believe that turnover in itself is a good thing, and didnot attempt to analyze its
causes.n
According to Markham, the purpose of his study is "a thorough and objective
appraisal of all economic aspects" of the drug industry.40 But the study was spon-
sored by PMA, and PMA had its schedule. Markham continued, "We have begun
by organizing our research schedule so as to center attention on the specific
issues raised by the Subcommittee's report no. 448, entitled `Administered
Prices-Drugs,' published on June 27." (Markham's study began in July.) At the
session of hearings when PMA presented its case in December, 1961, the study
was apparently not yet complete. One of the issues slighted was price competi-
tion, although this issue had certainly been raised in the Report. During the
examination on his testimony, Markham conceded that he had not examined the
question as to whether or not any drug firm had experienced a change in relative
sales rank because of price competition,'~ and Mannis also confessed to ignoring
the role of price competition in effecting changes in concentration.'2 Markham
agreed that price competition is of paramount interest to the consumer, but con-
cluded his contribution to the Hearings with this statement: "I have riot made
any careful study of the workability of competition in the ethical drugs Industry.
I was examining primarily these partIcular issues that seemed to be important."3
While a careful study of the workability of competition would admittedly take
longer than the five months Markham had been able to devote to it prior to his
~ Study of administered prices, op. cit., p. 67.
~ Ibid., pp. 17-18.
~ Mr. Mannis testified, "Although I don't know, I would guess that back In 1951 or
perhaps in the 1953 era, the leading product was quite different from the product that
is leading the field now." Hearings, op. cit., pt. 4, p. 2102.
: ~
`1lbid., Part 4, p. 2096. Other attempts to support the existence of price competition were
also less than convincing. Professor B. V. Rostow of Yale University, testifying in behalf
of PMA, was asked to give instances where patented drug prices had weakened. His "best
example," upon closer inspection, pertained to an instance where prices had temporarily
declined, not because of competition among different patent-licensed manufacturers, but
as a result of manufacturers disciplining their own wholesalers (who had sold at low
prices to hospital formularies and in other competitive bid markets) by undercutting
their prices, then cutting off their supplies, and later restoring the original prices. Ibid.,
Part 4, pp. 2081-2082.
42Mr. Mannis attributed such changes as had occurred to "competition between and
among products based on their merit. Ibid., Part 4. p. 2102. This would be more acceptable
If genuine product competition-even at Identical prices-were prevalent In drugs; but
when product differentiation becomes the road to sales, "merit" becomes what is measured
by a DODularity contest.
~3fbi~., pt. 4, p. 2111.
PAGENO="0459"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2009
appearance, it is quite another matter to seem to suggest that the workability of
competition is not an important issue. On the contrary, the chief purpose of S. 1552
was to make the drug industry more workably competitive; it was not designed,
for example, to stimulate or retard turnover in sales rank as an end in itself, or
otherwise. But the point need not be labored.
*B. Senate Judiciary Committe Action
S. 1552 was reported out of Senator Eastland's Judiciary Committee as a much
more innocuous document than when it was submitted.4" All the patent pro-
visions had been rejected. The other contemplated reforms were considerably
attenuated. The request for public control over generic names on the part of the
Secretary of the Department of Health, Education, and Welfare survived in
modified form: under certain circumstances, the Secretary is to hold hearings
and establish generic names for certain drugs. The request that the FDA pass
on the efficacy of drugs became a provision that drug efficacy claims be sup-
ported by substantial evidence, but the wording adopted limited its jurisdiction
only to initial claims made subequent to approval. The licensing requirement
became mere registration, but the FDA was granted greater power to inspect
plants. The requirement that advertisements contain generic names and state-
ments on efficacy and side effects was entirely rejected.
Kefauver and other members of the Subcommittee majority noted that the
amended bill would perhaps improve the quality of drugs, but would have little
if any effect on prices, and indicated their intention to propose several amend-
ments to the bill in order to salvage some of its economic impact. First, they
proposed that all drug patent license agreements be filed with the Commissioner
of Patents. It will be recalled that 1~i~IA had expressed willingness to make pri-
vate patent interference settlements a part of the public record; the amendment
would not require public filing, but would extend to all patent licensing agree-
ments. Second, it was proposed that the compulsory patent licensing feature of
the original bill be retained, but that it be invoked only where wholesale drug
prices exceeded 500 per cent of factory costs. Third, it was pointed out that the
wording of the current bill would not restrict drug claims to demonstrated
efficacy after the approval of a new drug application, and requested the necessary
modification in wording. Fourth, a compromise on the contents of drug adver-
tisement was put forth: generic names need appear in type face only half as
large as brand names, and summaries, rather than complete statements, of efficacy
and side effects should be included."
C. Congressional action
On August 3, 1962, the late President Kennedy communicated to Senator East-
land a list of amendents for the Committee to consider, relating chiefly to drug
safety. The Committee met a few days later, and in addition to adopting some of
the amendments suggested by the President, accepted Kefauver's amendments re-
garding advertising content and efficacy claims, but rejected his patent amend-
ments. The bill was then transmitted to the House of Representatives, where
further amendments were proposed. A conference report of October 3, 1962 shows
that the conference between the Senate and House managers of the bill resulted
in the substantial adoption of the Senate version, with only minor changes, except
that the requirement that the Secretary of HEW publish and distribute to
physicians all the detailed information required by law to be included in the drug
package was dropped.'~ No reason is given for omitting this important pnovision.
In the absence of information regarding motive or influence, the deprivation of
easy access by the physician to the best information on drugs can only be inter-
preted as an act of disinterest misanthropy.
v. CONCLUSIONS AND POLICY RECOMMENDATIONS
The amended bill was passed by both houses and became law on October 10,
1962, as the "Drug Amendments of 1962," more frequently referred to as the
Kefauver-Harris Act. It will serve to improve the quality of drugs, but its Impact
on competition will be slight. Requiring evidence of efficacy will, if properly en-
Ma Rt. on S. 1552, op. cit., pp. 1-8.
"IbId., pp. 36-48.
~ Drug Amendments of 1962: Conference Rept. No. 2526, House of RepresentatIves, 87th
Cong., second sess., Washington, D.C., Government Printing Office, Oct. ~, 1962 pp. 10-18.
PAGENO="0460"
2010 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
forced, keep off the market a certain number of useless but harmless drugs which
presently encumber it. To the extent that drug firms are tempted to patent and
push such inefficacious drugs in order to take advantage of the susceptibility of
physicians to massive advertising, refusal to certify them will prevent further
market confusion and reduce the amount of research devoted solely to pro
vithng vehicles for sales promotion.40 But most research is probably not aimed
at pure sales promotion of useless compounds; far more resources are wasted in
duplicating existing drugs via molecular manipulation, and the failure of this
part of S. 1552 to be enacted keeps incentives for such "research" as great as ever.
Competition between generic and brand name products usually exists only
where the drug is not patented, i.e., in only a small part of the prescription
drugs market. The ability of generic drug firms to compete with brand name
firms is slightly increased by the provisions for registration and Inspection
of plants. This should give physicians more confidence in the quality of generic
name products and increase their willingness to prescribe by generic name,
but in view of the demonstrated ability of drug firm detailment to disparage
generic producers, it is doubtful if the fact of more adequate FDA inspection,
in itself, will eliminate the cultivated distrust of generic drugs on the part
of physicians. In the past, detailmen have also disparaged the scope and the
adequacy of FDA inspection, and ii rc~mains to be seen if they will have similar
success even under the new program. The only sure way of eliminating disparage-
ment is to eliminate the army of over 15,000 detailmen; eliminating monopoly
profits will largely bring this about. The provisions giving the Secretary of
HEW authority to establish generic names supplies the basis for sweeping
reforms resulting in shorter and simpler names physicians can remember, spell,
and hence can prescribe. But much depends upon the aggressiveness with which
the Secretary carries out this mandate. At present, few generic names do not
need~ reforming, but a universal housecleaning in nomenclature is probably
not to be expected. Yet it would be most disappointing if the Secretary were only
to exercise his power in those relatively few cases where no name exists, or
more than one.
The remaining reforms of advertising also proceed in the right direction, but
not far enough. The requirements for summaries of side effects and efficacy
in all advertisements will make for a more enlightened practice of medicine,
but will not have direct effects on drug economics. The requirement that generic
names be included in all copy will help, but the more such names are simplified,
the more it will help.
To an economist concerned with questions of public policy, the congressional
performance in regard to S. 1552 is discouraging, but not entirely so.. The Act
effected significant medical, if not economic, reforms. It now seems appropriate
that renewed efforts be applied on the part of the public to induce congress to
pass a more adequate act abolishing drug product patents and requiring generic
name labeling. But the obstacles should not be underestimated. The influence
of opposing groups is considerable, and the performance of several of these at
the Hearings left much to be desired from the standpoint of concern for the
public interest in health. Congress was apparently more sensitive to the need
of protecting the consumer's physical health than his economic well-being; in
fact the prime reason that any action was taken on the bill may have been its
coincidence in time with the Thalidomide tragedy in late 1961, with the great
publicity given the embryo-deforming effects of a new sedative drug. But it
would seem preferable for the public to conduct a continuing campaign for
more effective drug industry legislation, rather than to have such reform
legislation come about piecemeal as a by-product of a series of tragic and avoid-
able blunders on the part of pharmaceuticals makers.
~ Hearings on administered prices, op. cit., pt. 18, p. k0379 contains interesting testimony
by Dr. A. D. Console, former medical director of Squibb, who in reply to Senator Kefauver's
question as to how much drug firm research produces nothing worthwhile and is not in-
tended to, answered: "I think the majority of it is in that category . . . with many of
these products, it is clear while they are on the drawing board that they promise no
utifity; they promise sales."
PAGENO="0461"
Income Opportunities arTd Physician Location
Trends in the United States
BY
HENRY B. STEELE and GASTON V. RIMLINGER
Reprinted from THE WESTERN ECONOMIC JOURNAL
Vol. III, No. 2, Spring, 1965
2011
PAGENO="0462"
2012 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
INCOME OPPORTUNITIES AND PHYSICIAN LOCATION TRENDS
IN THE UNITED STATES*
HENRY B. STEELE and GASTON V. RIMLINGER
Rice University
INTRODUCTION
In a highly industrialized society professional manpower is one of the most
important resources, which implies that trends in its geographic location have
important economic and social implications. The spatial distribution of physicians
in relation to population has been highly uneven in the United States for at
least several decades.' Since this situation must be considered less than ideal from
both the economic and the social points of view, the trends in the location of
physicians should be of particular interest. The present article seeks to answer
three basic questions. First, is there any tendency toward change in the distribu-
tion of physicians, and if so, is the long-run tendency toward a more even or a
more uneven distribution pattern? Second, what is the role of physician location
trends in the distributive pattern? Third, to what extent are the observed changes
in physician location accounted for by such variables as population and per capita
income, which may be taken to reflect income opportunities for physicians? It
should be borne in mind that a more even distribution of physicians in relation
to population can come about either because more physicians move to areas with
a physician shortage or because more people move to areas with a relative physician
surplus.
The time period covered by the analysis is from 1950 to 1959. This is a rather
short period, but its choice is dictated by the fact that adequate and comparable
data on physician location are available only for these two dates. The data are
from the health manpower surveys of the U.S. Public Health Service.2 The
location trends over this period will be related to the following variables: (1) the
degree of urbanization, (2) the regional shift in population, (3) the regional per
capita disposable income at the beginning of the period, and (4) the increase in
regional per capita income during the period. The choice of these variables is again
limited partly by the availability of data and partly by the nature of the analysis,
which is primarily statistical. A priori, the most important omission among vari-
ables would seem to be the income of physicians. Relevant ~comparative regional
~ We are indebted to Mrs. Gloria Shatto for assistance in statistical computation. Respon-
sibility for the article is solely that of the authors.
An analysis of this distribution can be found in 0. V. Rimlinger and H. B. Steele, "An
Economic Interpretation of the Spatial Distribution of Physicians in the U.S.," Southern Economic
Journal, July 1963, PP. 1-12.
2The 1950 data are from M. Y. Pennell and M. E. Altenderfer, Health Manpower Source Book,
section 4, U.S. Public Health Service (Washington, D.C.: U.S. Government Printing Office, 1954);
and the 1959 data are from W. H. Steward and M. Y. Pennell, Health Manpower Source Book,
section 10, U.S. Public Health Service (Washington, D.C.: U.S. Government Printing Office, 1960).
PAGENO="0463"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2013
data on physician income are available only for the year 1949.~ An attempt will
be made to investigate the significance of the physician income level at the
beginning of our period in a separate comparison of standard metropolitan areas.
TRENDS IN PHYsIcIAN~ DismiBul-loN AND LOCATION
A brief introductory comment on the meaning of the physician distribution
may be in order. Concern with physician distribution is based on the problem of
access to medical services. The number of physicians in relation to population in
a given area may be used as an index of the availability of physician services, but it
should be noted that this is only an approximation. Availability depends on the
relationship of the supply and demand for services. The demand is determined
not only by the size of the population, but also by such interrelated factors as the
level and distribution of income, the age structure of the population, the possession
of health insurance, the level of educatiOn, and the degree of urbanization. The
supply of physician services depends similarly not only on the number of physicians
but on the amount of time they devote to patients and their effectiveness in treat-
ing them. An area with a high concentration of retired physicians or with many
physicians engaged in research or teaching naturally has a smaller supply of
services for patients than a similar area where all physicians are engaged in active
private practice. Unfortunately, we lack the data necessary to take these factors
into account. The present analysis is based on the total number of physicians in
an area rather than on the amount of service they provide. The areas used are for
the most part groups of similar counties within a state; in one case standard metro-
politan areas will be used. In each state counties are combined into the following
groups adopted by the health manpower surveys: greater metropolitan, lesser
metropolitan, adjacent, isolated semi-rural, and isolated rural.4 Although some
information regarding income differentials is lost by aggregating the 3069 counties
of the United States into a little over 200 groups, the group is considered a prefer-
able area concept as far as the market for medical services is concerned.
We turn now to the question of whether there was a change in the distribution
of physicians between 1950 and 1959 and the direction of any such change. No
change in distribution would mean that every area experiences the same relative
increase or decrease in the physician-population ratio. In other words, if the
national physician-population ratio increases by 5 per cent, an increase of 5 per
cent of this ratio in every area of the country would leave the over-all distribution
unchanged. Given the constant internal migration of the population and the
entry and exit of physicians into and from professional life, it would be a very
unusual coincidence if there were no change in distribution over a nine-year
period. Hence, the question of whether there was a change in distribution is by
itself not very meaningful. It is the extent and the direction of change that are
`These data are used below in the analysis of 19 standard metropolitan areas. For the source
of the physician income data, see W. Weinfeld, "Income of Physicians, 1929-1949," Survey of
Current Business, July 1951, pp. 9-26.
4For a precise definition of each of these, see Penriell and Altenderfer, op. cit., section 4,
pp. 18-20.
PAGENO="0464"
2014 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
important. Physicians are most highly concentra~ed in high-income and in urban
areas. After experimenting with several approaches, it developed that the most
informative way to put the question of change in distribution is to inquire whether
the degree of concentration of physicians in these areas has increased or decreased.
The answer depends on how concentration is measured. It is interesting, and some-
what unexpected, that relative concentration in higher income areas decreased
between 1950 and 1959, while at the same time the physician-population ratios in
more highly urbanized areas increased. To clarify the meaning of these findings
a more detailed examination is necessary.
To measure the change in relative concentration with respect to regional
income, the writers constructed two Lorenz curves relating physicians to popula-.
tion, one for 1950 and one for 1959, as shown in Chart I. On the horizontal axis
CHART I
LORENZ CURVES FOR DISTRIBUTION OF PHYSICIANS BY RANK ORDER
OF COUNTY GROUP PER CAPITA DISPOSABLE INCOME'
0
0
C
C,
C,
Per cent of total population
Total physicians, 1950 and 1959; physicians in active private practice, 1959.
PAGENO="0465"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2015
the population of the United States is arrayed in rank order of county group per
capita disposable income,5 with incomes rising from left to right. The vertical
axis represents the total number of physicians in the country. It will be noted that
the Loren.z curve for all physicians in 1959 lies closer to the diagonal than the
corresponding curve for 1950, indicating a change toward a more even distribution
of physicians with regard to regional income. It is interesting to note also that
when only physicians in private practice are considered, the distribution in 1959
is more even than when all physicians are taken into account. Unfortunately, we
do not have the necessary data to analyze the change in the distribution of private
physicians between 1950 and 1959.
TABLE 1
CHANGE IN PHYSICIAN CONCENTRATION IN RELATION TO
PER CAPITA REGIONAL INCOME, 1950-1959
Population
ranked by
income
Per cent of
physicians
in 1950
Per cent of
physicians
in 1959
Per cent
change
1950-1959
Lowest 10%
4.0
4.8
+ 19.8
10-20
5.4
6.1
+ 13.5
20-30
6.7
8.4
+ 24.5
30-40
8.0
8.1
+ 0.9
40-50
8.9
9.6
+ 7.8
50-60
12.0
10.4
- 13.3
60-70
12.6
11.8
- 6.1
70-80
11.6
12.2
+ 5.3
80-90
13.2
13.9
+ 5.1
90-100
17.5
14.9
- 15.1
A clearer picture of the change in concentration can be obtained by examining
Table 1, which shows the gains and losses of the regionally stratified income
groups. On the whole the change was clearly in favor of lower income areas.
The top 10 per cent of the income-ranked population suffered the greatest relative
loss of physicians. The population with the greatest gain was in the third decile
from the bottom of the income range. Of course, physicians still remain con-
centrated in high-income areas. In 1950 the bottom 25 per cent of the population
had 13.0 per cent of the physicians and the top 25 per cent had 36.5 per cent of
the physicians. In 1959 this had improved to 14.4 per cent for the bottom quartile
`on population and per capita disposable income are from Sales Management Annual
Survey of Buying Power, May 1951, and May 1960. The income for the county group is a weighted
average. The Sales Management staff terms its income measure "effective-buying income," but
observes the same definition the U.S. Department of Commerce employs in defining disposable
income,
81-280 0-68--pt. 5-30
PAGENO="0466"
2016 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
and 35.5 per cent for the top quartile. The top 10 per cent of the populatipn still
had three times as many physicians in 1959 as the bottom 10 per cent, but in
1950 the top 10 per cent of the population had over four times as many physicians
as the lowest 10 per cent. This improvement in the physician-population ratios
of the lower income relative to the higher income areas should not be interpreted
to mean that physicians relocated from high to low income regions. A more
likely explanation is the tendency of the population to move to higher income
areas, and the changes in regional income levels.
A valid argument can be made that for a general evaluation of the distribution,
the concept of relative distribution developed above is more meaningful than the
measurement of distribution based on degree of urbanization. Relative distribu-
tion takes into account the proportion of the population that gains or loses by a
change in distribution. On the other hand an analysis in terms of the degree of
urbanization shows what happens to specific social segments of the population
and gives us another way of looking at the problem of distribution. It also has the
advantage of isolating one of the important causal factors in the location of
physicians and the relocation of the population as a whole. The results of this
approach for the period 1950-1959 are summarized in Table 2.
An important fact which can be noted in Table 2 is that in none of the areas
did the number of physicians increase as much as population or decrease as little
as population. This means that the over-all decrease in the country's physician-
population ratio between 1950 and 1959 left all aggregated county groups with
fewer doctors in relation to population. However, not all areas shared equally in
this relative loss of physicians. There was a redistribution in favor of the more
highly urbanized areas and against the rural areas. In isolated rural areas popula-
tion decreased slightly (.7 per cent), but the number of physicians decreased about
six times as fast (4.4 per cent). This is the meaning of the 5.9 coefficient of elas-
ticity of physician mobility.6 This coefficient merely describes a numerical rela-
tionship and is not intended to indicate a causal significance of the same magni-
tude. Isolated semi-rural areas fared even worse than isolated rural areas, even
though they had a slight increase (1 per cent) in the absolute number of physicians.
Their problem is that population increased 11 times faster than the number of
physicians. The other three county groups had somewhat comparable changes in
physician-population ratios. Most favored were the greater metropolitan areas,
where the relative increase in the number of physicians was equal to 70 per cent
of the increase in population, as indicated by the elasticity coefficient of .7. In the
lesser metropolitan areas, the relative increase in the number of physicians was
second largest in the country, but the relative increase in population was the
largest. As a result the lesser metropolitan areas had a rate of physician increase
that was about half as fast as the rate of population increase. In the adjacent
areas, where the rate of physician increase was much smaller, the rate of popula-
tion increase was relatively still smaller. The result was a rate of physician increase
equal to 60 per cent of the population increase.
6 Computed before rounding the percentage changes.
PAGENO="0467"
TABLE 2
CHANGE IN PHYSICIAN DISTRIBUTION IN RELATION TO DEGREE
OF URBANIZATION, 1950-1959
Per cent
change in
County group physicians
(1)
Per cent
change in
population
(2)
Average population
elasticity of
physician mobility
(1) ± (2) i
(3)
Per capita
ncome in 1950
(4)
Per capita
income in 1959
(5)
Per cent
increase
in income
(6)
Average per
capita income
elasticity of
physician mobility
(1) -~ (6)
(7)
Isolated
rural -44 - 7 59 $ 849 $1,468 72 9 06
Isolated
semi-rural 1.0 113 .1 1,046 1,506 44.0 .02 LTJ
Adjacent 9.8 16.4 .6 1,287 1,975 53.5 .18
Lesser
metropolitan ... 15.6 32.1 .5 1,439 1,888 31.2 .50
Greater
metropolitan ... 18.7 27.9 .7 1,623 2,185 34.6 .54
I.
PAGENO="0468"
2018 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Table 2 shows also some relationships between changes in physician-population
ratios and regional per capita incomes. Column (7) shows the average per capita
disposable income elasticity of physician mobility and reveals a somewhat sur-
prising negative coefficient for isolated rural areas. A 1 per cent increase in per
capita disposable income in an isolated rural region would be associated with a
decrease of .06 per cent in the number of physicians. With increasing degrees of
urbanization, the income elasticity increases: for isolated semi-rural areas, an
increase in per capita disposable income of 2 per cent will increase the number
of physicians by .02 per cent; the ratio increases to .18 per cent, .50 per cent, and
.54 per cent for adjacent, lesser metropolitan, and greater metropolitan county
groups.
There is a clear tendency toward higher income areas having smaller declines
in the number of physicians in relation to population. This would seem to con-
tradict the earlier finding about a change in relative distribution in favor of lower
income areas. However, there is no necessary inconsistency, for two quite different
ways of measuring changes in distribution are involved. Nevertheless, the apparent
inconsistency does highlight the difficulty of describing changes in distribution in
an unequivocal manner. Some questions may be raised also with respect to the
relevant time period for which income is measured. This is especially the case
when there is a change in the income rank over time. For instance, in 1950 lesser
metropolitan areas had the second highest average income of the five county
groups, but in 1959 they were in third place, behind adjacent areas. If the causal
effect of income operates with a lag, the initial income level is more likely to be
relevant. If it operates through anticipation, the final income level might be more
appropriate, or even the rate of increase in income over time. Multiple regression
and multiple correlation analysis is necessary to determine the respective signifi-
cance of these different factors. This is the task of the following section.
ANALYSIS OF PHYSICIAN LOCATION TRENDS
The preceding discussion has shown that changes in physician location over
time are an important aspect of the regional redistribution of physicians relative
to population. A complete analysis of changes in distribution would have to
account for the movements of the population as well as the movements of phy-
sicians. In this paper, however, we are concerned only with the movement of
physicians. Population shifts will be taken as an independent variable that has a
causal effect on the location of physicians. One would expect that an increase
in population in a given area, especially if it is through migration, makes it easier
for physicians to open new practices and induces them to locate there. Of course,
once a doctor has established his practice in a given area and built up his clientele,
he is not likely to move. The highly personal nature of the doctor-patient rela-
tionship is undoubtedly an important barrier to mobility. On the other hand,
new doctors are constantly entering the profession while others are retiring and
leaving vacancies. It is mainly through these entries and exits that physicians are
relocated, which makes it important to add a substantial time dimension to the
PAGENO="0469"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2019
analysis of their mobility. It is the mobility of the group rather than that of the
individual that counts.
In addition to increases in population, Table 2 suggests that the regional level
of income may act as a strong inducement to new physicians. The simple prospect
of a wealthy clientele might have much drawing power. People in the higher
income brackets might be in the habit of paying their medical bills more promptly
than persons with low incomes. But if the drawing power is based primarily on
the physician's expectation to be able to earn a high income by charging higher
fees, it will be greatly dampened if the high-income area already has a very large
number of doctors in relation to population. On the other hand, a high concen-
tration of doctors (provided that it does not affect income too adversely) might~be
attractive because it offers more opportunity, for leisure. It is reasonable to assume
that where there are many doctors it is easier to find vacation or weekend replace-
ments than where there is a shortage.
The results of an attempt to evaluate the relationship of changes in the
regional number of physicians to changes in a selected set of variables are pre-
sented in Tables 3 and 4, which show the coefficients of regression, and of correla-
TABLE 3
REGRESSION COEFFICIENTS PERTAINING TO PERCENTAGE CHANGES IN THE
NUMBER OF ALL PHYSICIANS IN Six SECTORS, 1950-1959 a
County
group
Regression
constant
(a)
Coefficient of
per cent change
in population
(b')
Coefficient of
per capita p
income in 1950
(c)
Coefficient of
er cent change
in income
(d)
Isolated rural
-40.43
.64
.02
.03
Isolated semi-rural
- 29.34
- .22
.40
.05
Adjacent
48.26
1.42
- .50
.14
*
Lesser metropolitan
56.72
1.00
- .08
.29
Greater metropolitan
48.86
.92
- .01
.61
.
`~
~
Coefficient of
physician
income in 1949
*
Nineteen standard
metropolitan areas
50.33
.96
.01
.00"
a The value of the coefficients is derived from the linear regression equation:
`b~P di~I
- = a + + c It(o) + ~, where N represents the number of physicians; I, the
Nt(o) Pt(O) it(o)
regional per capita income; P, population; and t(o), the initial year of the period.
b Less than .005.
PAGENO="0470"
TABLE 4
income ~
change ~
(R,.4) ~
.82
.06"
CORRELATION COEFFICIENTS PERTAINING TO PERCENTAGE CHANGES
IN THE NUMBER OF ALL PHYSICIANS IN SIX SECTORS, 1950-1959
County
groups
TOTAL CORRELATION
PARTIAL CORRELATION
SIMPLE CORRELATION
All
variables
(R1.,,,)
Population
and income
level
(R1.,,)
Population
and income
change
(R,.,4)
income level
and
income change
(R1.,4)
Population
(R,,,)
Income
level
(R1.,)
Isolated rural
.33
.16
.20
.14"
.14
.06"
Isolated semi-rural
.63
.61
.59
.28
- .57
.26
Adjacent
.87
.82
.82
*
.08"
.82
- .01"
Lesser metropolitan
.93
.90
.92
.88
.87
-32
Greater metropolitan
.67
.48
.63
.64
.44
- .37
.53
Population
and physician
income
Physician income
and
income change
Physician
income
Nineteen standard
metropolitan areas
.95
.95
.95
.17
.95
.17
"Not significant at the .05 level.
PAGENO="0471"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2021
tion of these variables. The tables relate the average rate of change in the number
of physicians in each of the county groups to (1) the average rate of population
change, (2) the average level of absolute income in 1950, and (3) the rate of
change of the average per capita income during 1950-1959. These national
averages are simple arithmetic averages of state data, which in turn are weighted
averages of county data within each county group. In the case of the 19 standard
metropolitan areas, the averages are simple arithmetic averages, but the level of
income of the population in 1959 is replaced by the average level of physician
income in 1949, which adds a fourth variable to the analysis. The fifth variable
is the degree of urbanization, which is treated as a qualitative variable. It could
be included in the regression and correlation equations by means of dummy vari-
ables, but we chose the alternative of computing separate regressions and correla-
tions for each county group and for 19 standard metropolitan areas. This has the
effect in each case of keeping the degree of urbanization constant and provides a
basis for comparison between urban and rural areas.
One of the obvious facts indicated by Table 3 is that the impact of the popu-
lation and income variables depends greatly on the degree of urbanization. It is
well known that the rural environment :tends to discourage and the urban
environment to encourage the location of physicians. There are many factors,
especially differences in cultural and professional advantages, which account for
this differential in drawing power. But here we are concerned only with variables
that affect the income opportunities of physicians. To what extent do they have
a differential impact in the different environments? We might like to know
whether improved income opportunities would work in attracting more physicians
to rural areas and roughly how much of ati improvement would be required. Of
course, our data can only give crude indications of what the answer to these
questions inight be.
Let us compare now the effects of our variables in the different environments.
Column (b) of Table 3 shows the percentage increase in the number of physicians
associated with a 1 per cent increase in population in the different county groups.
On the whole the effect of population change is quite strong, but it varies greatry
between areas. It is greatest in the adjacent areas, where a 1 per cent increase in
population is associated with a 1.42 per cent increase in the number of physicians.
Somewhat puzzling is the case of semi-rural areas, where a 1 per cent increase in
population is associated with a .22 per cent decrease in the number of physicians.
In lesser metropolitan areas, population and physicians tend to increase at the
same rate, while in more urbanized areas the rate of physician increase is somewhat
smaller than the rate of population increase, and in isolated rural areas it is much
smaller. The basic distinction is clearly between rural and nonrural areas.
Assuming the observed association to be of a causal nature, we could say that
`A few counties changed from one group to another over the nine-year period in question,
usually moving from lesser to greater urbanization categories. In these cases the county was
included in the category which it occupied in 1950, on the assumption that its condition in the
initial year had a greater impact than its classification at the end of the period. The total effect
of such changes, however, is negligible since only 34 out of 3069 counties were involved.
PAGENO="0472"
2022 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
in rural areas population increases tend to lower the physician-population ratio
substantially, whereas in metropolitan areas these ratios would be little affected
and in adjacent areas they would increase.
The extent to which the average income level in 1950 is associated with the
increase in the number of physicians is shown in column (c) of Table 3. The
variations between areas are again very great and this time rather unsystematic.
In isolated rural areas, a $100 higher per capita income in 1950 is associated with a
2 per cent increase in the number of physicians over the 1950-1959 period. On
the other hand, in isolated semi-rural areas, a $100 higher income is associated
with a 40 per cent increase in physicians, and in adjacent areas with a 50 per cent
decrease. These erratic results raise questions about the significance of the
observed association, which will be explored below. With respect to the 19
standard metropolitan areas, the observed relationship indicates a 1 per cent
increase in the number of physicians over 1950.4959 for a $100 higher level of
physician income in 1949.
Column (d) of Table 3 describes the associations between increases in phy-
sicians and increases in regional per capita income over 1950-1959. In the 19
standard metropolitan areas a systematic relationship is for practical purposes
nonexistent. But it is interesting that in the other areas the association becomes
constantly closer with higher degrees of urbanization. In isolated rural areas a
1 per cent increase in income is associated with only a .03 per cent increase in
the number of physicians, whereas in greater metropolitan areas it is associated
with a .61 per cent increase in physicians.
We have now measured the relationships between changes in selected vari-
ables and corresponding changes in the number of physicians. But it must be
remembered that all these changes are averages and that their significance depends
on the scatter of the original values about the regression line. An indication of
this significance is presented in Table 4, which contains the correlation coeffi-
cients for the population and income variables. On the whole the correlations
are highest in the 19 standard metropolitan areas and the lowest in the isolated
rural areas. As a matter of fact, the variables considered in this study, other than
the changes in population, are of little use in explaining changes in the number of
physicians in isolated rural areas because the simple correlation coefficients do not
appear to be statistically significant. In isolated semi-rural and in greater metro-.
politan areas the situation is slightly better. The index of total correlation R1.234
for the semi-rural areas is .63, which makes the index of total determination
R21234 = .40. In other words, 40 per cent of the observed change in the number
of physicians in the semi-rural areas over the nine years is "explained" by popula-
tion change, the initial income level, and the change in income. In greater metro-
politan areas 45 per cent of the change is explained by these variables. However,
in adjacent, lesser metropolitan, and the 19 standard metropolitan areas, the
corresponding percentages are 66, 86, and 90, respectively.
Among the variables examined, population change is the one most closely
correlated with change in the number of physicians. Total correlation is generally
PAGENO="0473"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2023
not greatly affected by dropping either the initial level of income or the change
in income as independent variables. In the case of the adjacent, the lesser metro-
politan, and the 19 standard metropolitan areas, total correlation is not greatly
affected even if both of these variables are dropped. Only in lesser metropolitan
areas is there a high simple correlation between change in income and change in
physicians. In these same areas the combination of income level and income
change also has the highest correlation with changes in the number of physicians.
The level of physician income at the beginning of `the period has only a very low
correlation with the subsequent increase in physicians. Finally, it should be noted
that, on the whole, the change in physicians is more closely correlated with
changes in per capita income than with the initial level of per capita income.
This is especially the case in the lesser and the greater metropolitan areas.
SUMMARY AND CONCLUSIONS
It has been shown that the two most important factors affecting the change
in location of physicians over time are the regional degree of urbanization and the
increase in population. The extent to which an area is urban or rural, or adjacent
to an urban area, influences its attractiveness to physicians aside from what
is happening to its population or to its level of income. An increase in population
has an effective drawing power in urban but not in rural areas. In the isolated
rural areas as a whole both population and physicians decreased, but there was
no meaningful correlation between these two changes. Rather surprisingly, the
level of per capita income at the beginning of `the period turned out to have little
effect on physician location. Nor were physicians induced to locate on the basis
of higher average physician incomes. Increase in per capita income seems to have
been a, factor in attracting more physicians but only in lesser and greater metro-
politan areas. The fact that the isolated rural areas had the highest rate of increase
in per capita income did not prevent a decrease in the number of physicians.
Even when initial per capita income and increase in income are combined, their
effect is strong only in lesser and greater metropolitan areas. This indicates that
on the whole the income of `the population is a less significant factor than we had
originally suspected.
However, in evaluating the meaning of these findings it is important to bear
in mind that they refer to all physicians, rather than only to those engaged in
private practice. The Lorenz curves in Chart I show that there is a substantial
difference between the distribution of all physicians and that of private physicians.
It is possible therefore that if a comparison over time were made of physicians in
private practice, the income of the population would be a more important factor.
At least this was suggested by our analysis of the distribution of private physicians
in 1959.8 It is quite clear that a more disaggregated analysis is necessary to obtain
more precise and more reliable results. For instance, it might be worth while to
break down the increase in population into increase through migration and natural
increase. It might be profitable also to analyze changes in the number of physicians
See Rimlinger and Steele, op. cit.
PAGENO="0474"
2024 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
by examining both entries and exits in a given area. The age structure of the
physicians would be important in this respect, since it affects both their income
and their exit from the profession. Additional insight would be gained by studying
separately the location trends of specialists and general practitioners. Table 5
shows that the degree of urbanization is a much stronger factor in the location of
specialists than in the location of general practitioners.
TABLE 5
PRIVATE NON-HOSPITAL PHYSICIANS PER 100,000 POPULATION
IN THE UNITED STATES, 1959
County group
General
practitioners
Full-time
specialists
Isolated rural
Isolated semi-rural
Adjacent
Lesser metropolitan
Greater metropolitan
42.8
46.5
46.4
42.5
51.6
2.6
22.9
20.0
55.4
64.4
Sou1~cE: Health Manpower Source Book, section 10, p. 12.
A final comment may be in order with regard to the general problem of
physician distribution. It was shown that the relative inequality of distribution
based on regional per capita income has decreased while the urban-rural inequal-
ity has increased. This is consistent with the findings of the physician location
trends. Since the pull of population increase is stronger than the attraction of per
capita income, a redistribution in favor of lower income areas is possible. The
change in distribution, however, seems to be dominated by the movement of
population rather than the movement of physicians. Population tends to move
to higher income areas and physicians follow but at a slower rate. This has the
long-run tendency of reducing the concentration of physicians in higher income
areas, while at the same time it is not necessarily inconsistent with a continuation
of great urban-rural inequalities. But it must be remembered that an increasingly
smaller percentage of the total population will be in areas with low physician-
population ratios. Complete elimination of higher physician-population ratios in
higher income areas is not likely to occur, nor is it necessarily desirable from an
economic point of view. How fast the equalization of distribution will proceed
depends in part on how the population redistributes itself and how fast physicians
follow the movement of the population. Obviously, if population increases in
higher income areas did not attract more physicians, the relative concentration of
physicians in these areas would be more quickly eliminated.
PAGENO="0475"
Reprinted from SOUTHERN ECONOMIC JOURNAL
Vol. XXX, No. 1, July, 1983
Printed in U.S.A.
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
I. INTRODUCTION
The number of physicians in relation to
population varies widely between localities
and regions in the United States. This pat-
tern of distribution is related to differences
in regional per capita income levels and
raises questions not only from a public
health but also from an economic point of
view. To what extent and in what manner
do economic factors influence the distribu-
tion of physicians? What kind of economic
behavior of physicians is suggested by their
spatial distribution? Do the economic forces
at work within the given institutional ar-
rangement tend toward a distribution that
is consistent with efficient use of health
manpower resources?
The purpose of this article is to seek an
answer to these questions.' It begins with a
description of the empirical relationship
between the ratios of physicians to popula-
tion and regional per capita incomes. Several
hypotheses are then developed to explore
the economic variables that may account
for the observed relationship. Finally, the
* We are indebted to Professor Edgar 0. Ed-
wards and Dr. George J. Benston for their helpful
comments on an earlier draft and to Mr. Vincent
Tarascio for his assistance in compiling statistical
information. Responsibility for the article is
solely our own.
1 Similar questions may be raised with regard
to distribution between specialties. None of these
questions has been adequately explored by econ-
omists. See, however, Rashi Fein, "Studies on
Physician Supply and Distribution," American
Journal of Public Health, May 1954, pp. 615-624;
and M. Friedman and S. Kuznets, Income from
Independent Profe8sional Practice (New York:
National Bureau of Economic Research, 1954).
2025
Rice University
hypotheses are subjected to empirical tests,
insofar as is possible with the data available
on physician incomes, physician visits, and
medical expenditures. Although relevant
data in this area are limited, the article
does bring together in a concise form a large
amount of statistical information from both
public and private sources.
II. THE DISTRIBUTION OF PHYSICIANS
Any analysis of the spatial distribution of
physicians poses the problem of defining
the geographic area served by physicians.
Following the practice of the Health Man-
power Survey,' we shall assume that the
"county group" is the relevant economic
area. A county group combines counties
with similar population concentrations
within a state; these counties are often but
not always geographically contiguous.' The
2 The Health Manpower Surveys used in this
study are in M. Y. Pennell and M. E. Altenderfer,
Health Manpower Source Book, Section 4, U. S.
Public Health Service (Washington, D. C.: U. S.
Government Printing Office, 1954); M. Y. Pennell
and M. E. Altenderfer, Health Manpower Source
Book, Section 5, U. S. Public Health Service
(Washington, D. C.: U. S. Government Printing
Office, 1954); and in W. H. Stewart and M. Y.
Pennell, Health Manpower Source Book, Section
10, U. S. Public Health Service (Washington,
D. C.: U. S. Government Printing Office, 1960).
`Five groups are used and they are designated
as follows: (1) greater metropolitan, where the
populations exceed one million; (2) lesser metro-
politan, where populations range from 50,000 to
one million; (3) adjacent counties, those which
are contiguous to metropolitan areas; (4) isolated
semirural areas, for counties with at least one city
of 2,500 people; and (5) isolated rural areas, for
the remaining rural counties. Residents of ad-
jacent areas are no doubt served on occasion by
AN ECONOMIC INTERPRETATION OF THE SPATIAL
DISTRIBUTION OF PHYSICIANS IN THE U. S.~
GASTON V. RIMLINGER
AND
HENRY B. STEELE
PAGENO="0476"
2026 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
600 IOPO 1500 2000 2500 3000
CHART I PER CAPITA EFFECTIVE BUYING INCOME 1959
data presented in Chart I thus are based on
the number of physicians, the size of the
population, and the per capita income of the
population in each county group for the
entire country in 1959.~ Each observation is
for a county group, of which there are one
to five in each state.
The distribution is stated in terms of
privately employed physicians, both general
practitioners and specialists, per one hundred
thousand persons in a county group. This
measure excludes all federal physicians, all
retired or inactive physicians, and all full-
time hospital, teaching, and research physi-
cians. In an analysis of the role of economic
variables in the spatial distribution of physi-
cians these exclusions seem to be clearly
desirable. The income measure to which
the physician-population ratio is related in
Chart I is the county group per capita net
effective buying income.6
metropolitan physicians, which tends to under-
state the effective physician-population ratio for
adjacent areas and to overstate it for metropolitan
areas. Isolated areas are more likely to be self-
sufficient.
`The data for Chart I are from Health Man-
power Source Book, Section 10.
`Net effective buying income is identical with
"disposable personal income" as computed by
the Department of Commerce. See Sales Manage-
ment, May 1961, pp. 54-56. The physician-popula-
tion ratios and the per capita incomes are weighted
averages of all counties of a given county group.
The most significant aspect of the distri-
bution presented in Chart I is the tendency
of the physician-population ratio to rise
with increases in regional per capita income.6
For instance, at the $1,000 per capita income
level the average ratio is about 45 physicians
per 100,000 population, but at the $2,000
level it is around 90 physicians per 100,000
population. This doubling in the number of
physicians in relation to population is not
due solely to changes in income. Low income
county groups are almost exclusively rural
while high income county groups are metro-
politan areas. Income and environment
change together, which makes it impossible
to separate their impacts. There are valid
reasons for believing that both higher in-
comes and urban environment attract more
doctors, but we are concerned primarily
with the question of how larger incomes may
account for higher physician-population
ratios.
III. THEORETICAL INTERPRETATION
This section develops hypotheses to ex-
plain the rise of the physician-population
ratio with increases in per capita income.
The relationship is mildly curvilinear as in-
dicated by the equation p = 21.24(2.058)' for the
regression line fitted by least squares.
160
~15C
0
a
~I2C
0
09C
0
6c
Sc
PAGENO="0477"
COMPETITIVE PROBLEMS
A. Linear Model
Let us begin with the simple case in which
it is assumed that all physicians are inter-
ested in maximizing money income. We may
further assume either that all physicians
are of equal ability, or that patients have
no way of determining physician ability; in
either event, there are no quasi-rents at-
tributable to ability or reputation. It is also
assumed that all physicians are in the habit
of charging what the traffic will bear, and
that this amount is a fixed per cent of the
patient's income.7 Finally, it is assumed that
there is no price competition among physi-
cians and that the number of patient visits
per physician is a function of the physician-
population ratio alone, which rules out any
relationship between per capita income and
the number of patient visits per time period.
On the basis of this set of highly simplified
assumptions we shall set up a theoretical
model using the following symbols:
a = proportion of income spent on the
average visit, which is constant
b = number of visits per person per unit
of time, which is also constant
Y = national per capita income
Yr = regional per capita income
N = total number of physicians
P = population
R = P/N, national physician-population
ratio
Rr = regional physician-population ratio
y = income of physicians, which in
equilibrium is the same for the
entire country
The equilibrium income of physicians is
equal to the aggregate expenditure on physi-
cian services divided by the total number of
physicians.
- (abY)P abY
NR
In any given region we have
The percentage will vary among different
types of visits, but it is not necessary to introduce
this complication.
IN THE DRUG INDUSTRY 2027
(2)
abYr
y=~
and
(3) Rr=~Yr
y
By; thus determining the relationship be-
tween Rr and Yr we have defined the spatial
distribution of physicians in relation to
population and regional per capita income.
This distribution pattern is illustrated
in Figure 1 for various total numbers of
physicians in the country. The number of
physicians on the graph increases from N0
to~ N3 while the corresponding physician
income decreases from ys to y~. The physi-
cian-population ratio in any area is a linear
function of regional per capita income and is
represented by a straight line starting from
the origin. The larger the total ninnber of
physicians (N), other things being equal, the
smaller the average physician income (y)
and the greater the slope (ably) of the line
representing the distribution of physicians.
Several characteristics of this distribution
may be noted. (1) If physicians and patients
behave as the model assumes they do, the
physician distribution will be a direct conse-
quence of the regional distribution of income.
(2) The absolute difference in the number
Fin. 1.-Linear Model of Physician Distribu-
tion. Regional Physician-Population Ratio (Rr)
varies directly with Regional Per Capita Income
(Y~). The total number of Physicians (N) deter-
mines the level of physician income (y).
PAGENO="0478"
2028 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
of physicians in relation to population be-
tween any two areas depends on the differ-
ence in income, the proportion of income, a,
spent on each visit, the number of visits per
person per unit of time, b, and the total
number of physicians, N. However, (3) the
relative distribution of physicians, as shown
by a Lorenz curve, is independent on the
magnitude of either a, b, or N.
The effect on distribution of an increase
in N can easily be demonstrated graphically.
In Figure 1 an increase in the number of
physicians from N1 to N3 raises the physi-
cian-population ratio in area Yr, from R1
to R3 and in area Yr, from R2 to R4. The
absolute increase in the ratio is greater in
area Y~, than in area Y,! , since R4 - R2 >
R3 - R1, but the relative increase is the
same in both areas, since R2/R1 - R4/R3.
The percentage increase in the physician-
population ratio in any area will be the same
as the percentage increase in the total num-
ber of physicians.8
The distribution of physicians in this
situation could therefore not be made more
equal by simply increasing the number of
physicians; it would in fact become less
equal in an absolute sense. Nor would a
willingness on the part of physicians to
reduce their fees in relation to income result
in a more equal relative distribution, unless
fees were reduced to zero. A change in fees
or in the number of visits per person would
8 This can be demonstrated mathematically by
showing that the elasticity of the regional physi-
cian-population ratio with respect to changes in
the number of physicians equals one:
dR? /dN
ERr,N1/ ~
(abY)P
By substituting -~- for y, we get
have the same effect on physician distribu-
tion as a change in the number of physicians,
since they all affect the slope of the distribu-
tion line.
B. Nonlinear Models
It is not consistent with known facts to
assume that the number of visits per person
remains constant as per capita income in-
creases. As a rule higher incomes are corre-
lated with larger numbers of visits, although
the absolute increase is not likely to be very
large. We can easily take this relationship
into account by expressing the number of
visits as a function of income. Let
b'Y = number of visits per person per
unit of time in the country as a
whole, and
= / regional number of visits per
person per unit of time.
The regional physician income then becomes
ab'Y~2
(4)
and the regional physician-population ratio
ab'Y2
(5) = -f-
The relationship between regional income
and the physician-population ratio is now
nonlinear, and the distribution lines are
curves convex to the origin.
Other things remaining equal, the intro-
duction of a positive relationship between
visits and income makes the physician
distribution more unequal, in the sense of
pushing the Lorenz curve to the right. How-
ever, as before, an increase in the number of
physicians would still leave the relative
distribution unaffected and increase the
physician-population ratio everywhere by
the same percentage.9
As in the previous case, the elasticity of the
physician-population ratio with respect to changes
in number of physicians equals one. From equation
(5) we get
Y~ N
}7~ p
y
dR~ Y~
and
We can now show that:
dR~/dN Y~N YrN 1 -1
R~ / ~ - YP* R~ YP YTN/YP -
PAGENO="0479"
COMPETITIVE PROBLEMS IN
There may also be some question as to
whether the fee remains a constant propor-
tion of income as income changes. An in-
creasing proportion would have the same
effect as that described above for an increas-
ing number of visits. A decreasing propor-
tion, which is more likely, would tend to
bend the curves in the opposite direction.
We can also get a nonlinear relationship
between physician-population ratios and
income, and hence a distribution that tends
to be more unequal than that of our first
model, if we abandon the assumption that
physicians try to maximize only income.
For the sake of simplicity, let us return to
our initial assumptions of constant fees in
relation to income and a fixed number of
visits per person regardless of income. In
equilibrium, all doctors have identical in-
comes, but those in low income areas work
very long hours, as indicated by the low
physician-population ratio, while those in
high income areas have considerable leisure.
Since there is freedom of movement between
areas, it seems likely that some doctors will
be tempted to leave the low per capita in-
come areas and move to higher per capita
income areas, where the same physician
incomes are being earned with less work.
This would result in a spatial redistribution
of physicians and physician incomes. At
the new equilibrium, each doctor would
maximize the utility derived from both
leisure and income. Some doctors may
actually move from higher to lower income
areas as the influx into high income areas
lowers income there and increases leisure.
But if doctors in general are willing to work
Hence
d& 1~?~
dN
harder only if they get a higher income,'°
there will be at the new equilibrium fewer
doctors, earning higher incomes, in low per
capita income areas and more doctors, earn-
ing lower incomes, in high per capita income
areas. Somewhere in the intermediate per
capita income range the physician-popula-
tion ratio and the physician income would
remain unchanged. The extent of the re-
distribution depends on the physicians'
income-leisure preference pattern.
Figure 2 illustrates the impact of income-
leisure substitutions for three different num-
bers of physicians in the country and initial
levels of physician income. The convex
curves represent the new equilibrium physi-
cian-population ratios in relation to per
capita income of the population. In this case,
the high income areas have increased the
ratios and the low income areas have reduced
them. How a change in the number of physi-
cians, in fees, or in the number of visits
would affect the distribution of physicians
under these circumstances cannot be deter-
mined a priori without specifying the physi-
cians' income-leisure preference pattern.
10 Technically, their marginal rate of substitu-
tion of income for leisure is less than zero and
leisure is not an inferior good.
THE DRUG INDUSTRY 2029
Fie. 2.-Linear and Nonlinear Models of Phy-
sician Distribution contrasted. Regional Physi-
cian-Population Ratio (R?) increases non-linearly
with increasing Regional Per Capita Income (Y?),
for various assumed total numbers of Physicians
(N).
Thus
dR? /dN dR~ N
ER?.N=~~~/ ~;~i~i;
~2 1
Y2P Y?2/Y2P - 1
PAGENO="0480"
2030 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
C. Role of Mobility
The analysis thus far indicates that, as
long as the fee rises with income and the use
of physician services per person is fixed,
physicians in low income areas will tend to
have either less leisure or less income than
physicians in high income areas, but not
both. This follows in part from the assump-
tion of mobility and equal ability. If these
assumptions are relaxed the result no longer
necessarily holds. Absence of mobility from
low to high income areas may leave doctors
in these areas with both less income and less
leisure than in high income areas. On the
other hand, a preference by doctors for high
income areas has the same effect as a substi-
tution of leisure for income; it tends to in-
crease leisure and reduce income in high
income areas and cause the opposite in low
income areas.
The relaxation of the assumption of equal
ability may also affect the results obtained.
Specialists, who repzesent a group with
higher than average qualifications, are likely
to draw higher fees than general practi-
tioners. Since they choose to locate mainly
in high income metropolitan areas, they may
raise both the leisure and the average physi-
cian income in these areas.
D. Summary and Conclusion
The main determinants of physician
distribution, given the regional distribution
of income, were found to be (1) the relation
of fees to patient income, (2) the relation
of demand for physician services to patient
income, and (3) the behavior of physicians
with respect to price competition, income
maximization, desire for leisure, and geo-
graphic mobility. The degree of inequality
of physician distribution tends to be greater
the greater the increase in fees and in de-
mand for services for a given increase in
interregional income and the greater the
inclination of physicians to substitute leisure
for income.
An interesting insight provided by the
theoretical analysis is that so long as ex-
penditures on physician services rise with
increases in income, and physicians do not
engage in price competition but try to
maximize income, changes in the total
number of physicians, or in the general level
of fees, or in the level of demand for services,
do not affect the relative distribution of
physicians. However, increases in the num-
ber of physicians, in level of fees, and in
level of demand would increase the absolute
physician-population ratio differentials be-
tween high and low income areas and in that
sense would make the distribution more
unequal.
IV. EMPIRICAL ANALYSIS OF DISTRI-
BUTION DETERMINANTS
We shall now examine to what extent
available evidence on physician behavior,
fees, and demand for services supports the
preceding theoreti~cal explanation of physi-
cian distribution.
A. Income Maximization, Leisure, and Mo-
bility
In investigating the behavior pattern of
physicians we shall make the assumption,
which will be substantiated later, that
medical expenditure per person (fees times
visits) rises with income. This will tend to
give us an uneven distribution of physicians
in favor of high income are~ss, unless physi-
cians ignoi~e income. There is no evidence
of such behavior. Nor are there data avail-
able that would allow a positive test of the
income maximization hypothesis. The only
feasible approach is to test the behavior
pattern for its consistency with income
maximization and to find out in this manner
whether substitution of leisure for income
and lack of mobility should be considered
significant factors in the observed distribu-
tion of physicians. Lack of mobility is in-
tended to cover all reasons for not moving
other than income and leisure.
If leisure is valued by physicians, and if
they are fairly mobile, we should find higher
physician incomes where leisure is low and
PAGENO="0481"
COMPETITIVE PROBLEMS
lower incomes where leisure is high. This
pattern seems to emerge in a test of thirty-
two major cities." The test, however, is not
based on a direct measure of leisure but
assumes that the higher the physician-
population ratio the greater the amount of
leisure available to physicians. For the
cities tested, high physician incomes are
correlated with low physician-population
ratios (low leisure) and lower physician
incomes with higher ratios (more leisure).
If, on the other hand, leisure does not
increase with higher physician-population
ratios, low incomes are not compensated by
more leisure. This would most likely indicate
a lack of concern with leisure and would
make lack of mobility a significant factor
in the distributive pattern. In this instance
lack of mobility makes the distribution more
uneven than it would have been under in-
come maximization. If physicians tried to
maximize income, there would be a redistri-
bution from the low income (high ratio)
areas in favor of the areas with higher in-
comes and fewer doctors.
An investigation of data on physician
visits in eight major cities supports the
hypothesis that lack of mobility rather than
desire for more leisure is the major factor in
the distribution of physicians among cities.1'
11 This test uses 1949 income data for physicians
and 1950 data on physician distribution. The
regression on physician income (y) of the physi-
cian-population ratio (x) can be expressed by the
equation: y = 22.106 - .556 x, where y is expressed
in terms of thousands of dollars per year and x is
the number of physicians per hundred thousand
population. The coefficient of correlation is .808,
and is significant at the .05 level. Income data are
from W. Weinfeld, "Income of Physicians, 1929-
1949," Survey of Current Business, July 1951, pp.
9-26. Physician data are from Health Manpower
Source Book, Section 4.
12 The test is concerned only with physician
visits outside of hospitals, for a period extending
from July 1957 to June 1959. If we omit Boston
(which is very unusual in that its ratio of physi-
cians not in private practice to total physicians is
by far the highest of the eight cities), the relation-
ship between per capita visits (x) and the
physician-population ratio (y) may be expressed
by the regression equation: y = -4.57 + 28.83x.
The coefficient of correlation is .963. Data on
visits are from U. S. Public Health Service, Health
IN THE DRUG INDUSTRY 2031
Physicians in those cities having the higher
physician-population ratios do not appear
to have more leisure than physicians in
other cities. Hence, in the case of the 32
cities considered above, it was probably not
a desire for greater leisure but some obstacle
to mobility that kept physicians from mov-
ing to towns where there were fewer doctors.
A careful examination of the available
data leaves a clear impression that desire
for leisure is not a strong motivating force
in the physicians' choice of location. Their
low valuation of leisure is reflected in the
fact that they tend not to reduce the length
of their workweek as their income increases.
The percentage of doctors working 49 hours
or more per week is about the same at all
income levels and ranges between 70 and
85." It might be, of course, that physicians
at higher income levels take longer and more
frequent vacations.
When we compare physician incomes
among communities of greatly differing
sizes it becomes extremely difficult to deduce
a rational behavior pattern. We find the
lowest physician incomes in the smallest
and largest size communities, those with
under 2,500 and over 1,000,000 population.
These are the areas where physician-popula-
tion ratios tend to be the lowest and the
highest respectively. We must invoke special
kinds of immobility to account for the fact
that physicians in very small communities
tend to have less income and probably also
less leisure than their colleagues elsewhere.
Perhaps those in the profession who are
least qualified tend to locate there. The high
concentration of physicians in very large
cities, despite lower incomes, may be partly
explainable by cultural and professional
Statistics, Series C, No.6, p. 25. The high apparent
correlation may be partly spurious in view of the
presence of the population factor in the denomi-
nators of both the dependent and the independent
variables. However, the statistical issue raised by
this kind of ratio correlation remains unsettled at
present. See J. R. Meyer and E. Kub, "Correlation
and Regression Estimates when the Data are
Ratios," Econometrica, October 1955, pp. 400-416.
13 Health Manpower Source Book, Section 5, pp.
79-81.
81-280 0-68--pt. 5-31
PAGENO="0482"
2032 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
advantages and by opportunities to practice
certain specialties not available elsewhere,
which constitute another form of immobility.
In our estimation, improved mobility,
in the direction indicated by income maximi-
zation, would result in a more even distribu-
tion of physicians among large cities, but
might very well result in a more uneven
distribution between large cities and smaller
communities. Whether either redistribution
would be desirable in terms of improving
the relative availability of medical services
depends, of course, on the structure of the
demand for services.
B. Demand for Phyrfcian Services
The demand for physician services is
relevant both as a factor affecting the dis-
tribution of physicians and as a measure of
the utilization of physician services. Under
perfectly competitive conditions, assuming
that doctors are of equal ability and wish
to maximize income, there would be regional
differences in physician-population ratios so
long as there are regional differences in
demand for services. Under these conditions,
however, differences in ratios would corre-
spond exactly to differencesin use of services.
In equilibrium physicians everywhere would
have the same income and provide the same
amount of services. No region would suffer
from a relative shortage of physician serv-
ices, if use of services could be taken as an
index of need. On the other hand, if there
are substantial regional differences in use of
services per physician, without correspond-
ing differences in physician income, we may
have an indication of maldistribution of
physicians.
We shall examine first the factors deter-
mining the use of physician services and
then the regional use of services in relation
to regional physician-population ratios.
The major determinants of the use of
physician services are income, age, sex, race,
education, place of residence, and possession
of health insurance. We need not bother with
the use of services as related to sex in a
regional analysis. All the other factors tend
to be correlated with income. The effect of
a person's family income on the number of
physician visits per year is an increase of 24
per cent from families with annual incomes
under $2,000 to families with incomes of
$7,000 and over.'4 In both the "under
$2,000" and the "$2,000-$3,999" income
groups, the number of visits per person per
year is the same, namely 4.6; in the "$4,000-
$6,999" income group it is 5.1, and in the
"$7,000 and over" group, 5.7. These are
averages for entire income groups; the
differences are greatest if only children and
young persons in different income groups
are compared, and smallest in comparisons
of only middle-aged persons.
The tendency toward larger number of
visits as income rises was detected also in a
comparison of cities with different levels of
per capita income. The relationship between
annual per capita physician visits and per
capita effective buying income was studied
for eight standard metropolitan areas. The
coefficient of correlation wa~ found to be
.619 and the coefficient of determination
.383. This correlation is significant at the .1
level.
The reasons why persons with high in-
comes visit physicians more frequently than
those with low incomes are no doubt more
complex than merely greater ability to pay.
The visit differentials should be related to
the economic value of good health, percep-
tion of illness, and easy access to physicians.
These factors are in turn correlated with
economic opportunity, education, and place
of residence. In 1958-59, whites visited their
physicians 35 per cent more often than non-
whites. For families whose head had a college
education, the number of visits per person
per year was 56 per cent higher than for
families whose head had under five years of
education and 42 per cent higher than for
families whose head had five to eight years
of education.1' For the country as a whole
`~ Health Statistics, Series B, No. 19, p. 20.
"Ibid., p.13.
PAGENO="0483"
COMPETITIVE PROBLEMS
the number of visits per capita in urban
areas (towns with 2,500 inhabitants or more)
was about 39 per cent higher than in rural
farm areas, and about eight per cent higher
than in remaining rural areas.
Another factor correlated with income is
the possession of health insurance,'6 which
tends to increase the demand for medical
services. People who possess health insurance
tend to have a higher number of physician
visits, higher hospital admission rates and
higher surgical rates than people without
insurance at comparable income levels.'7
This is true not only for services that are
covered by insurance but also for those that
are not covered. People seem to be more
inclined to see a physician if at least some
of the potential expense is prepaid.
It has been established that both the
physician-population ratio and the number
of visits per person increase with income.
The question now is whether the demand
for services and the relative supply of phy-
sicians increase in the same proportion,
which would mean an uneven distribution
of physicians but an even distribution of
work loads. The available data indicate
that regional differences in use of physician
services are of a lower order of magnitude
than differences in physician-population
ratios. We computed the number of non-
hospital visits per private non-hospital phy-
sician by multiplying the regional average
number of visits per person by total regional
population and dividing by the regional
number of physicians. This measure can be
taken as a reliable index of work load, for
the physicians concerned, unless there are
16 Possession of any kind of health insurance is
about three times more frequent in families with:
incomes of $7,000 and over than in families with
incomes of under $2,000. See U. S. Department of
Health, Education, and Welfare, Medical Care
Financing and Utilization, Health Economics
Series No. 1 (Washington, D. C.: U. S. Govern-
ment Printing Office, 1962), Table 87, p. 99.
17 0. W. Anderson and J. J. Feldman, Family
Medical Cost and Voluntary Health Insurance:
A Nationwide Survey (New York: McGraw-Hill,
1960), pp. 73, 183, 194.
IN THE DRUG INDUSTRY 2033
important regional differences in the amount
of: inhospital visits per physician. The re-
sults, presented in Table I, show a large
work load differential between the region
with the highest and the lowest number of
visits. Physicians in the West South Central
area (Arkansas, Louisiana, Oklahoma,
Texas) have 74 per cent more visits of all
kinds outside of hospitals than those in New
England and 37 per cent more than those
in the Middle Atlantic region. Although
New England may be a special case, there
is a general tendency for the number of
visits to increase with decreases in the physi-
cian-population ratio. Apparently, doctors
in areas with low physician-population ratios
either work longer hours or work more
intensively. Contrary to what seemed to be
indicated in the earlier comparison of metro-
politan centers, the census area comparison
indicates a tendency toward a possible
redundancy of physicians in high ratio areas
and toward a shortage in low ratio areas.
However, census divisions are much less
homogeneous than metropolitan areas, and
TABLE I
REGIONAL PsIYSIcIAN WoRE LOADS
IN THE U. 5., 1959
Region
Non-
hospital
physicl'an
In~ne
capita
Physi-
ci~s0~r
popula-
Average
income
~
eraiphy-
1949
New England
Middle Atlantic
Pacific
East North Cen-
tral
West North Cen-
tral
Mountain
EastSouthCentral
South Atlantic
West South Cen-
tral
3,817
4,847
5,284
5,694
5,986
6,042
6,286
6,499
6,630
2,396
2,543
2,502
2,337
1,978
1,994
1,424
1,804
1,764
113
120
112
83
79
84
65
74
75
$ 9,442
9,574
12,782
12,158
11,961
11,214
11,325
11,131
11,794
Sources: Statistical Abstract of the U.S., 1960 Table 6; 1961,
Table 419. Health Manpower Source Booh, Section 10, Table 2;
Health Statistics, Series C, No. 6, Table ii; Survey of Current
Business, July 1951, Table 12, p.19.
PAGENO="0484"
2034 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
10
TABLE II
PITY5ICIAN HOSPITAL WORK LOADS: 1959
Area
.
~
population
Hospital
discharges
ph~ian
Patient
hospital
Phjsic~n
Northeast
West
North Central
South
149.0
128.2
103.4
91.4
71.3
94.7
113.6
127.5
728
729
966
930
Sources: Health Manpower Source Book, Sec. 10, pp. 5-6;
Health Statieticz, Series B-No. 32, pp. 18-19.
there is at least the possibility that in higher
income areas physicians spent more time
taking care of patients in hospitals than in
low income areas, which would tend to offset
the disparities noted above.
But the opposite is more likely to be the
case; that is, in low income regions doctors
have probably higher hospital work loads
than in high income regions. It was possible
to compute the relationships between the
number of physicians and the number of
hospital discharges and the number of pa-
tient hospital days for the four major census
divisions of the country. The physicians
used for this purpose include all active
nonfederal physicians in both private prac-
tice and in the hospital service. The results
of the computations are presented in Table
II, which shows the tendency of higher
hospital work loads in areas with low physi-
cian-population ratios. The regional uneven-
ness in work loads observed above thus
tends to be accentuated when hospital work
is taken into account.
We may also conclude that while differ-
ences in physical demand for physician
services may account in part for the observed
regional differences in physician-population
ratios, they are not a sufficient explanation
of existing disparities in the supply of physi-
cians.
C. Fees for Physician Services
It was shown theoretically that high
physician-population ratios in high income
areas may be due in part to the fact that
fees are based on per capita income rather
than on the supply and demand for physician
services. It is not necessary, of course, that
there be a complete absence of price competi-
tion, as was assumed in the theoretical
model. A modest amount of price competi-
tion may well exist, which would keep fees
from rising as much with increases in regional
income as they would otherwise.
Some indication that fees are lower in low
income areas is implicit in the data presented
above in Table I, unless the relationships
between regional physician incomes changed
considerably between 1949 and 1959. Table
I implies that the average income per visit
in New England and the Pacific was between
35 and 40 per cent higher than in the south-
ern areas. The East North Central area
income per visit was about 20 per cent
higher than in the South.
There are no relevant data available on
physician fees,18 but we can remedy this
deficiency in part by using another indirect
method. The available data on spending and
on number of visits make it possible to
compute an index of fee differentials between
income groups and regions. For the United
States as a whole we have for a twelve-
month period in 1957-58 an estimated $25
spent on physician services per capita by
families earning under $2,000 annually;19
this income group is reported to have had
4.6 visits per capita during a twelve-month
period in 1957-59, which indicates an ex-
penditure of 85.43 per visit. By the same
process we obtain an expenditure of $6.93
per visit for the 87,500 and over income
18 The Bureau of Labor Statistics publishes
data on certain types of fees paid by comparable
wage and salary earner families in twenty major
cities. These data, unfortunately, are not suitable
for our purposes, since we are interested in com-
parisons between income groups.
19 One third of $75 spent on all health services.
See 0. W. Anderson, P. Collette, and J. J. Feld-
man, Family Expenditure Patterns for Personal
Health Service, 1953 and 1958, Health Information
Foundation Research Series No. 14 (New York:
Health Information Foundation, 1960), Table 4a,
p.S.
PAGENO="0485"
class. The highest income class paid on the
average 28 per cent more per visit than the
lowest. This is not a measure of price dis-
crimination in the ordinary sense, since it
does not refer to a single service but to a
composite bundle of services bought by
different groups. There may be differences
in kind as well as quality between bundles.
There may be more homogeneity i,n a
comparison of expenditures of people living
in similar environments. A calculation for
farm families only gave the following ex-
penditure differentials: under $2,000 family
income, $4.01 per physician visit; $2,000-
3,999 income, $4.35 per visit; $4,000-6,999
income, $4.74 per visit; $7,000 and over
income, $5.71 per visit.'5 In this case, the
highest income group paid 42 per cent more
per visit than the lowest. It is noteworthy
that the heaviest discrimination occurs be-
tween the highest and the next highest in-
come group. A computation by region shows
farm families in the South paying $4.13 per
visit against $4.82 in the North Central area.
In spite of the necessarily approximate
nature of our statistical results, all indica-
tions are that prevailing pricing practices
contribute to the observed differences in
physician-population ratios between areas
with different income levels and between
urban and rural areas.
v. CONCLUSIONS
It has been shown that high income areas
have substantially larger numbers of pbysi-
cians in relation to population than low
income areas. If we assume that in a wealthy
country the entire population should have
reasonable access to physician services, the
implications of this i~ncome-related differ-
ential deserve careful consideration.
Although there are other factors making
high income areas attractive to physicians,
20 Expenditure data from J. L. Pennock, "Farm
Medical Care Expenditures," Public Health Re-
ports, April 1958, p. 290; visits are adjusted for
rural level.
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2035
there are definite economic incentives work-
ing in the same direction. Persons with
higher incomes tend to visit physicians
more often and have more expensive visits.
The larger number of visits per capita in
higher income areas tends to justify a certain
unevenness in the distribution of physicians
in relation to population, but high income
areas tend to have more physicians than
would be called for by this difference in use
of services. The ability to charge higher fees
to more well-to-do clients makes possible a
high concentration of physicians in high per
capita income areas without commensurate
loss in physician income. In this sense, the
widely defended practice of charging accord-
ing to ability to pay may have undesirable
consequences in terms of the spatial avail-
ability of physician services.2' The avail-
ability of services should be considered not
only in terms of access to a physician but
also in terms of the amount of the physi-
clan's time available to the patient on each
visit. It was shown that where the physician-
population ratio is low, physicians tend to
have very large numbers of visits which are
almost necessarily brief.
The economic implications of the observed
wide regional differentials in physician work
loads depend on how health is valued. If we
value health from a strictly economic point
of view, it is in all probability more advisable
to spend a lot of physician time on high
income executives, who presumably have a
high productivity, rather than on low income
unskilled workers. In this case, the observed
spatial distribution of physicians may well
be optimal, or perhaps there should still be
more physicians in high income areas and
fewer in low income areas. This approach,
however, becomes very awkward when we
extend it to the children of various income
groups. To the extent that the children have
comparable productive potentials, economic
21 Within a given area, of course, this practice
may make medical care available to some low
income groups who would otherwise be deprived
of it.
PAGENO="0486"
2036 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
logic suggests that they should have com-
parable access to medical care.
If we value health from a social point of
view and put the same value on the health
of all individuals, the economic problem
becomes essentially that of adjusting the
available supply of physicians as efficiently
as possible to the existing need. In the
absence of any data on actual need, let us
use expressed need, as represented by the
number of visits. On these assumptions it
can be argued that the existing pattern of
physician distribution involves an ineffi-
cient allocation of medical manpower re-
sources.
To the extent that wide disparities in
regional work loads involve substantial
differences in the amount of time a physician
can devote to each patient, there is a tend-
ency toward inefficiency. The assumed social
values make the social utility of a compara-
ble visit the same for all individuals. From
the medical point of view, it is probably
reasonable to assume that the effectiveness
of a visit does not increase proportionally
with the length of the visit beyond a fairly
short minimum period. Physician resources
made available for long visits are therefore
not employed most efficiently. This does not
even take into account the fact that where
long visits are possible, the visits themselves
may be less urgent. Further, where physi-
cians are overloaded, the pressure of time
to accommodate many patients may impair
a doctor's ability to function with optimal
efficiency. Finally, there may be economic
inefficiency involved with regard to patients,
who may have to waste a great deal of time
in the waiting rooms of overloaded doctors.
An efficient spatial allocation of physician
manpower would therefore be one which
tends to equalize the average allowable
time per visit between all areas.
A few brief observations are in order also
with regard to lack of mobility, which is
clearly an important factor in the existing
distribution pattern. It is rather doubtful
that the existing distribution would be more
equal if physicians were more mobile, in
the sense that they would be more inclined
to maximize either income or some combina-
tion of income and leisure. Many of the
poorer areas of the country would undoubt-
edly be even worse off if all physicians made
their location decisions with strict economic
rationality.
Economic incentives and the lack of mo-
bility also have implications for the improve-
ment cf the geographic distribution of physi-
cians. It is unlikely, at least in the short
run, that the distribution can be made
significantly more even for the country as a
whole by simply increasing the manpower
output of our medical schools. It is very
doubtful whether an increase, perhaps even
a substantial increase, in the supply of physi-
cians would automatically solve the distribu-
tion problem, as some medical authorities
seem to think.22 Such a measure might result
in an increase in the number of physicians
in areas where there is a shortage, but this
would tend to be accompanied by an even
greater increase in areas where there is
already a relative excess supply of doctors.
This does not promise to be an economical
approach to the problem of distribution.
On the other hand, there is good reason to
believe that the unevenness in the distribu-
tion of physicians will diminish automati-
cally, in the long run, along with the tend-
ency toward regional equalization of per
capita income.
22 See for instance the statement by Dr. P. R.
Hawley in Committee on Medical Care Teaching,
Readings in Medical Care (Chapel Hill, N. C.:
University ~f North Carolina Press, 1958), p. 27.
PAGENO="0487"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2037
Senator NELSON. I would like to submit for the record at this point
a letter from E. B. Anderson, Hoffman-La Roche Co.
(The letter referred to follows:)
HOFFMANN-LA ROCHE INC.,
Nutley, N.J., January 2, 1968.
Hon. GAYLORD NELSON,
Chairman, Senate Monopoly Subcommittee,
U.S. Senate, Washington, D.C.
DEAR SENATOR NELSON: I want to thank you for the opportunity which you
have given us to meet with your staff concerning recent statements which have
been issued on the prices paid by certain government authorities for two Roche
products. It was most helpful to review together the data which had been
accumulated.
Hoffmann-La Roche Inc. shares your belief in the desirability of uniform
pricing policies under which government agencies purchasing equal quantities
of the same drug product would receive the same price. For some time Roche
has had in effect a pricing policy designed to meet this objective, under which
any Federal, State or local government which purchases from Roche is charged
the same price for the same quantity of the same product.
On December 19, 1967, you issued a public statement and a price table which
referred to the prices of several drug products, two of which are sold by
Hoffmann-La Roche Inc., chiordiazepoxide hydrochloride and sulfisoxazole. Since
several figures contained in the statement and in the survey data released at
the same time are incorrect, I am writing, to you to provide the correct informa-
tion, as I am sure you want the record to be as accurate as possible. We also
believe that some explanation of other items in the statement and the price
table would permit a more meaningful comparison of prices for these two
products.
In addition to the statement in the hearing, there was an additional state-
ment by your office released on Decrembér 24, 1967, relating to the prices paid
by various government purchasers for LIBRITJM capsules (chlordiazepoxide
hydrochloride). This later statement, as reported in the New York Times of
December 25, 1967, and several other papers, also contains some incorrect figures
and conclusions which we are sure you would like to see corrected.
In several instances the statements and the table have confused prices for
25 my. capsules of chlordiazepoxide hydrochloride with prices for 10 mg. capsules
of the same drug. As a result, the prices for different dosages have been com-
pared, which is obviously misleading.
A. Chiordiazepoccide hydrochloride
We have included hereafter the items in the statement and the price table of
December 19, and the statement of December 24, relating to chlordiazepoxide
hydrochloride which are not in accord with the facts, or which we believe leave
an erroneous impression which should be corrected.
1. In regard to the City of Des Moines, the table states that the city pays
the highest price of $45.00. While it does not state that Roche sold the product
to Des Moines at this price, some persons may have so inferred. Actually, there
are no direct sales of chiordiazepoxide hydrochloride to the City of Des Moines
by Hoffmann-La Roche Inc. Since Roche has had no direct dealings with Des-
Moines for this product, the city's purchase of this product would have been
from a wholesaler or retailer. As in the case of Des Moines, the cities of Cleve-
land and Paterson do not buy directly from Roche. Thus, the prices paid by
those cities for Librium are presumably paid to wholesale or retail distributors.
Roche has made some direct sales to county hospitals where these cities are
located, but the prices in the table and statement are not the prices which we
have charged for these sales. Broadlawns Polk County Hospital (Des Moines)
and Highland View Hospital in CuyahOga County (Cleveland) have purchased
Librium 25 mg. capsules in bottles of 500 for $89.60, which is our price to direct
government purchasers which do not buy the 5,000-capsule bottle or in quanti-
ties large enough .to support a contract. Preakness Hospital in Passaic County
(Paterson New Jersey) has purchased Libriurn 25 mg. capsules in bottles of
50 capsules at a price of $4.40 per bottle which is our price for that size bottle.
The District of Columbia, we understand, purchases its requirements of Librium
from a Veterans Administration depot.
PAGENO="0488"
2038 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
I am sure you understand that the manufacturers of many different types of
products cannot sell directly to all end-users, and, as a result, the prices paid
by end-users will vary. In addition to direct sales to large-volume users, it is
essential that Roche sell through drug wholesalers and retailers in order to serve
adequately all of the small-volume purchasers. We, of course, have no control
over the prices charged by other persons who distribute our products after we
sell to them.
2. The information you received from the City of Philadelphia stated that
Philadelphia paid $18.50 for 500 25 mg. chlordiazepoxide hydrochloride capsules.
Our contract with Philadelphia signed on June 12, 1967, in fact specifies a price
of $28.50 per 500 25 mg. capsules for a quantity in excess of 100,000 capsules. The
price quoted in the statement, namely $18.50, refers to bottles of 500 10 mg.
capsules for a quantity in excess of 250,000 capsules. In addition, it was stated
that New York City paid $28.50 per 50025 mg. capsules of chlordiazepoxide hydro-
chloride. That was our previous price to New York City. The current contract,
which was effective October 1, 1967, calls for $25.50 per 500 25 mg. capsules, based
on a quantity in excess of 500,000 capsules.
3. The statement noted that the Defense Supply Agency paid $13.20 per 500
25 mg. capsules of chlordia.zepoxide hydrochloride. Actually, the price of $13.20
per 500 capsules is for 10 fig, capsules under a contract for over 20 million
capsules. Our contract with the Defense Supply Agency for capsules of the 25 mg.
strength calls for a price of $24.00 per 500 capsules in a total purchase of over
3,000,000 capsules.
4. Your information that Philadelphia got a better price than the Veterans
Administration for 25 mg. capsules of chlordiazepoxide hydrochloride is incor-
rect. As noted above, Philadelphia paid Roche $28.50 per 500 25 rng. capsules,
based on the purchase of over 100,000 capsules. The Veterans Administration
paid $21.75 per 500 25 mg. capsules based on a contract purchase of over 5,000,000
capsules. As can be seen, the Veterans Administration pays a lower unit price
for capsules of the equal strength because it buys in greater volume.
5. The article also states that "Los Angeles paid a premium price even
though it bought three times the quantity Philadelphia bought." This statement
is incorrect. As the following table shows, for 10 mg. capsules Los Angeles paid
Roche a lower price based on a quantity purchase of over 1,000,000 capsules than
Philadelphia paid based on a quantity of over 250,000; and for 25 mg. capsules
Philadelphia paid Roche $28.50 based on a quantity of 100,000 and Los Angeles
paid $25.50 based on a quantity of 500,000:
Philadelphia
10 milligrams $1850
25 milligrams 2850
Los Angeles
10 milligrams 16.50
25 milligrams 25. 50
6. The prices liSted in the article for the cities not covered above are also con-
sistent with Roche's policy of equal prices for government buyers which purchase
equal quantities of the same drug from Roche. The cities of Wichita, Kansas;
Newark, New Jersey; and Grand Rapids, Michigan, all have paid $39.60 per
500 capsules of the 25 mg. strength of LIBRI1JM. This is the Roche price for a
bottle of 500 25 mg. capsules, to direct government purchasers who do not buy
large sizes, such as the bottle of 5,000 capsules, or annual quantities large enough
to justify a contract purchase.
B. $ulfjsoa,azole
The statement and price table of December 19 also contain information on
sulfisoxazole which we believe may lead to some erroneous inferences and there-
fore further explanation.
1. In regard to the City of Des Moines, the table states that the city paid $25.30
per bottle of 1,000 0.5 grain sulfisoxazole tablets. Roche does not sell this product
directly to Des Moines. Purchases of Roche's sulfisoxazole by the City of Des
Moines would have been from a wholesaler or retailer, or Des Moines may have
purchased sulfisoxazole from another company.
2. The table quotes a price of $6.90 for a bottle of 1,000 0.5 gram sulfisoxazole
tablets to the City of Philadelphia. The current Roche contract for sulfisoxazole
with Philadelphia, dated~ June 12, 1967, calls for a price of $9.00 per bottle of
1,000 tablets based on a quantity of 1,000,000 tablets.
PAGENO="0489"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2039
3. The prices cited in the table of $8.35 and $8.50 per 1,000 0.5 gram sulfisoxazole
tablets paid to Roche by the Defense Supply Agency and the Veterans Adminis-
tration, respectively, are below that paid *by Philadelphia because they are
supplied under contracts for very much larger quantities.
As you can see in the above items, much of the information upon which you
relied regarding the prices for Roche products is inaccurate. We believe that,
left uncorrected, these inaccuracies reflect adversely on the reputation and poli-
cies of Roche, and upon your Subcommittee. Accordingly, we respectfully request
that the records of your Subcommittee be corrected to reflect the facts regarding
the Roche prices and policies as stated herein. We appreciate the cooperation
which both you and your staff have given us in reviewing this data.
Very truly yours,
E. B. ANDERSON,
Vice President.
Senator NELSON. The committee will adjourn until next Thursday,
January 25, in room 457.
(Whereupon, at 1 :15 p.m., the hearing in the above-entitled matter
was adjourned, to reconvene on Thursday, January 25, 1968, in room
457.)
PAGENO="0490"
PAGENO="0491"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
THURSDAY, JANUAB,Y 25, 1968
U.S. SENAIT,
MONOPOLY SUBCOMMIITEE OF THE
SELECT COMMrITEE ON SMALL BUSINESS,
Wathi'ngton, D.C.
The subcommittee met, pursuant to recess, at 10:10 a.m., in room 457,
Old Senate Office Building, Senator Gaylord P. Nelson (chairman of
the subcommittee) presiding.
Present: Senator Nelson.
Also present: Benjamin Gordon, staff economist; James H. Gross-
man, minority counsel; Susan H. Hewman, research assistant; and
William B. Cherkasky, legislative director, staff of Senator Nelson.
Senator NELSON. I now open: the hearings of the Monopoly
Subcommittee.
Our witness today is Dr. William Comanor.
Dr. Comanor, we appreciate very much your taking the time to
come here today to present your testimony to the committee. I have
read through your testimony and I think it represents a very thought-
ful, valuable contribution to the record of the hearings that we are
conducting here on the pricing structure in the drug industry.
I notice that you have a biographical sketch. That will be printed in
full in the record preceding your testimony. If you would like to just
state extemporaneously your background and your present position,
then you may proceed with your statement.
I trust that you have no objection to interruptions for questions?
STATEMENT OP WILLIAM S. COMANOR, PH. D., PROFESSOR, DEPART-
MENT OF ECONOMICS, HARVARD UNIVERSITY, CAMBRIDGE,
MASS.
Dr. COMANOR. Not at all
(The biographical sketch follows:)
(JIIERIODIUM VITAE
Name: William S. Comanor.
Address:
Residence-10 Fernald Drive, Cambridge, Massachusetts, 02138.
Office-1746 Cambridge Street, Cambridge, Massachusetts, 02138.
Date of birth: May 11, 1937.
Birthplace: Philadelphia, Pennsylvania.
Marital status: Married, no children.
Nationality: United States.
2041
PAGENO="0492"
2042 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Education:
1951-1955 Central High School, Philadelphia, Pennsylvania.
1955-1957 Williams College, Williamstown, Massachusetts.
1957-1959 Haverford College, Haverford, Pa., A.B. in Economics.
1959-1963 Harvard University, Cambridge, Mass., Ph. D. in Economics.
1963-1964 London School of Economics, London, England.
Thesis title: "The Economics of Research and Development in the Pharmaceuti-
cal Industry."
Fields of interest:
Industrial Organization and Public Policy.
Economics of Science and Research.
Statistics and Econometrics.
Economic Theory.
Honors and Fellowships:
Sophomore Honors, 1957.
High Honors in Economics, 1959.
Phi Beta Kappa, 1959.
Harvard University Fellowship, 1959.
National Science Foundation Fellowship, 1963.
Sloan Foundation Research Grant, 1966.
Professional experience:
Teaching Fellow, Harvard University, 1961-1963.
Resident Tutor, Quincy House, Harvard University, 1961-1963, 1964-1965.
Economist, Caterpillar Tractor Co., Peoria, Ill., Summer 1960.
Economist, United Research, Inc., Cambridge, Mass., Summer 1961.
Research Assistant, Professor Richard E. Oaves, Harvard University, Sum-
mer 1963.
Instructor in Economics, Harvard University, 1964-1965.
Special Economic Assistant to the Assistant Attorney General, Antitrust
Division, U.S. Department of Justice, Washington, D.C., 1965-1966.
Assistant Professor of Economics, Harvard University, 1966.
Consultant, Booz, Allen, and Hamilton, Inc., 1966-1967.
Consultant, Antitrust Division, U.S. Department of Justice, 1966.
Publications:
"Research and Competitive Product Differentiation in the Pharmaceutical
Industry in the United States," Econornica, November 1964.
"Review and Technical Change in the Pharmaceutical Industry," Review of
Eccrnomics and Statistics, May 1965.
"The Drug Industry and 1~Iedical Research: The Economics of the Kefauver
Committee Investigation," Journal of Bnsiness, January 1966.
"Competition and the Performance of the Midwestern Coal Industry," Jour-
nal of Indnstrial Economics, July 1966.
"Vertical Mergers, Market Power, and the Antitrust Laws," American Eco-
nomic Review, May 1967.
"Advertising Market Structure and Performance," Review of Economics
and ~tatisti~s, November 1967 (with Thomas A. Wilson).
"Market Structure, Product Differentiation, and Industrial Research,"
Quarterly Journal of Economics, November 1967.
Senator NELSON. Go ahead, Doctor.
Dr. COMANOR. My name is William S. Comanor. I am presently as-
* sistant professor for economics at Harvard University. I received my
A.B. from Haverford College in 1959 and my Ph. D. from Harvard in
1964. During the academic year 1965-66, I served with the Department
of Justice as Special Economic Assistant to the Assistant Attorney
General in charge of the Antitrust Division. I currently serve as a
consultant to the Antitrust Division of the Department of Justice.
My field of special interest within economics is industrial organiza-
tion, problems of competition and monopoly, and government policy
toward business. For sometime now, I have been especially concerned
with the economics of the pharmaceutical industry. My Ph. D. thesis
was entitled "The Economics of Research and Development in the
Pharmaceutical Industry." From my research, I have published three
PAGENO="0493"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2043
articles dealing with the economics of the drug industry. The first is
entitled "Research and Competitive Product Differentiation in the
Pharrnaceut1cal Industry in the United States."1 It appeared in the
1964 issue of Economica. The second is entitled "Research and Tech-
nical Change in the Pharmaceutical Industry," 2 and appeared in the
May 1965, issue of the Review of Economics and Statistics. The third
is entitled "The Drug Industry and Medical Research: The Economics
of the Kefauver Committee Investigation,"3 and appeared in the Jan-
uary 1966, issue of the Journal of Business, published by the Univer-
sity of Chicago.
I have copies of all of these articles. You may wish to include t.hem
in the record.
Senator NELSON. The articles you have mentioned will be printed in
the record at the conclusion of your testimony, along with the supple-
mental study you have furnished the committee, entitled "Advertising
Market Structure and Performance," ~ which I understand is to be
published.
Dr. COMANOR. That is right. I will mention that.
Senator NELSON. That study will also be printed at the conclusion
of your testimony.
Dr. COMANOR. I would like to report this morning on some findings
from a study which Prof. Thomas A. Wilson, of the University of
Toronto, and I have been carrying out on the relationship between
advertising and profit rates in the American economy, and then I
want to comment on some of the implications of the study for competi-
tion and public policy toward the pharmaceutical industry.
The statistical results are contained in a paper entitled "Advertising
Market Structure and Performance," which will be published in the
November, 1967, issue of the Review of Economics and Statistics. I
am afraid that although economists advocate efficiency for others,
they are frequently as lax in their own operations as anyone else, and
the November, 1967, issue of this review is not scheduled for publication
until February and March of t.he current year.
THE STATISTICAL FINDINGS
The core of this paper is a statistical analysis of the relationship
between advertising and profit rates across a wide range of consumer
goods industries. Our analysis was limited to this sector of the economy
for it is only here that advertising seems to have a major competitive
impact.
Even among consumer goods industries, advertising seems far more
important in some industries than others. In most industries, adver-
tising expenditures account for a relatively low proportion of total
revenues while in a handful of others, they reach very high propor-
tions. Thus, in the 41 consumer goods industries examined in our study,
25 had advertising-sales ratios below 3 percent, eight had ratios be-
tween 3 percent and 6 percent, and only six industries had ratios which
exceeded 6 percent.
1 See article, p. 2069, infra.
2 See article, p. 2078, infra.
~ See article, p. 2086, infra.
~ See article, p. 2092, Infra.
PAGENO="0494"
2044 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In the latter group, perfumes had an advertising-sales ratio of 15
percent, cereals and drugs were approximately 10 percent each, soap
9 percent, beer 7 percent, and soft drinks slightly more than 6 percent.
At lower levels, cigarettes and wines had advertising-sales ratios
of about 5 percent.
These findings suggest that the economic effects of advertising
expenditures are not likely to be felt generally throughout the economy
but rather are limited to a small subset of industries.
Senator NELSON. Doctor, I notice that you comment on this later
on in your testimony, but the figure that you use for expenditures for
advertising of drugs is 10 percent of sales.
What you are saying is that the total expenditure for advertising
for drugs equals 10 percent of the total sales of drugs, is that correct?
Dr. COMANOR. That is right.
Senator NELSON. And in this figure, you are using both prescription
and nonprescription; that is, proprietary drugs.
Dr. COMANOR. That is right.
Senator NELSON. What is the total amount spent on prescription
and proprietary drugs?
Dr. COMANOR. I have not looked at the most recent figures, it is
something in the order of $2.5 billion, $3 billion, but this could be
checked.
Senator NELSON. This figure confuses me. My memory is that we
have had testimony that somewhere around $4 to $4.5 billion is spent
on prescription drugs alone in this country.
Dr. COMANOR. It has been some time since I looked at the specific
figure that you are asking. I may be completely wrong. I would like to
check the numbers before I answer t.he question.
Senator NELSON. Do you have those figures in your study?
Dr. COMANOR. The figures which you ask are n'ot contained in the
printed paper which I am presenting to you, but I have them in
Cambridge. I can certainly send them to you.
Senator NELsoN. I would appreciate your getting these figures
for the committee, because it seems to me quite a critical statistic. If
I recall the testimony correctly, $4.5 billion is now spent on prescrip-
tion drugs; that is one figure. I would like to know what the figure
is on proprietary drugs. Then the next aspect of the question is, it
seems to me, the critical aspect of advertising in the drug industry;
that is, advertising for prescription drugs, which is directed only to
the medical profession. I would like to know what percentage the
advertising cost as a percentage of sales for prescription drugs. Again
the figure that we have here is that $800 million is spent on advertising
of prescription drugs, not counting detail men; $8(X) million as a
percent of $4.5 billion is a much higher advertising figure than the
figure you use of 10 percent.
I may be wrong in my recollection on the statistics.
Dr. COMANOR. The 10-percent figure applies both to proprietary as
well as prescription drugs. As I note later in my statement, it applies
to the average values for the years 1q54 through 1957. So we are
dealing with a period over 10 years ago.
PAGENO="0495"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2045
Senator NELSON. Your statistics apply to what period?
Dr. cOMANOR. They apply to average values for the years 1954
through 1957. I state later in my testimony precisely why this period
was chosen. They are not the most recent figures that one could find
on this particular percentage in the pharmaceutical industry.
Senator NELSON. The reason it seems to me that it would be valu-
able to have the statistics separated out with a separate set of statistics
for proprietary and prescription drugs is that they present two very
different problems in terms of the difficulty in entering the market-
place that you talk about. It seems to me that it is one kind of problem
to get into the marketplace and compete in the sale of a prescription
drug when all of the advertising of that drug is aimed solely at the
medical profession and quite another problem to get into the market-
place with a proprietary drug where all the advertising, or almost all
of it, is aimed at the general public.
Dr. COMANOR. I agree with you completely. I think these two sepa-
rate industries create very different problems for public policy. This
figure was used because the data was obtained from the Statistical
Division of the Internal Revenue Service and the IRS publishes data
for the two industries combined. The only tax return data that are
available to the general public for this particular industry combine
both proprietary and prescription drugs. So that although for public
policy purposes, I agree with you completely, for the purposes of our
study, we were forced to use the combined industry.
Senator NELSON. That is because the profit figures-
Dr. COMANOR. The profit figures also apply to the combined indus-
try. This is the way the IRS publishes tax return data.
Senator NELSON. But the statistics are available, are they not, for
the total amount of proprietary drugs sold and the total amount of
prescription drugs sold?
Dr. COMANOR. That is right, but we wanted to undertake a study
dealing with consumer goods industries generally throughout the
economy.
Senator NELSON. I realize your study was not directed at precisely
all of the points that are being raised here. But for purposes of the
record, if the statistics are easily available, I would appreciate it if you
would submit the most current statistics that you can get on total sales
of prescription drugs and what percent of sales is represented by
advertising for prescription drugs and the same for proprietary
drugs. I think it will probably show that the percentage of advertis-
ing to sales of prescription drugs is much higher than this average of
10 percent that you have given us for all drugs.
Dr. COMANOR. I will try to get the statistics for you.
Senator NELSON. -If it is easily available to you, if you would submit
that for the record in the next week or 10 days so we could print it.at
this stage of the testimony, I would like that.
Dr. COMANOR. Fine.
PAGENO="0496"
2046 COMPETITIVE PROBLEMS LN THE DRUG INDUSTRY
(The material referred to, subsequently received, follows:)
HARVARD UNIvimSITY,
DEPARTMENT OF ECONOMICS,
Cambridge, Mass., February 8, 1968.
Senator GAYLORD NELSON, -
Chairman, Monopoly Subcommittee,
Select Committee on Small Business,
U.S. Senate, Washington, D.C.
DEAR SENATOR NELSON: During my testimony before the Monopoly Subcom-
mittee on January 25, 1968, you requested that I submit some additional infor-
mation to be included in the record.
In the1963 Census of Manufacturers, Pharmaceutical Preparations (Standard
Industrial Classification 2834) were divided into two classes: ethical and pro-
prietary. Domestic shipments of ethical preparations were $2,054,897,000, while
domestic shipments of proprietary preparations were $836,166,000. I am afraid,
however, that total advertising expenditures in these two industries are not
available from public sources. For the same year, Internal Revenue Service
statistics report that total business receipts in the drug industry (IRS Minor
Industry 2830) were $4,505,870,000 while total advertising expenditures for the
same firms were $405,007,000, or 9.0% of sales. For the following year, 1964, the
IRS listed a smaller number of firms in this industry, and business receipts for
these firms were $4,175,331,000, while total advertising outlays amounted to
$461,648,000, or 11.1% of sales.
The drug industry, as defined in IRS statistics, include sales of both ethical
and proprietary products. Furthermore, the IRS statistics include advertising
outlays and revenues for entire firms even though some firms may have con-
siderable operations which are not related to the productions and distribution
of drugs. Census statistics, on the other hand, are based on data gathered at the
establishment level, and therefore include only business operations which deal
specifically with a particular industry.
As I noted in my testimony, the Statistics Division of the Internal Revenue
Service accepts the definition of advertising which is made by individual firms in
their tax returns. Different firms, moreover, appear to list different categories of
expense under different headings, so that gifts for doctors and medical students
are included in advertising expense by some firms while others list these outlays
under promotional expense. Furthermore, I understand that contrary to my
statement on January 25, the cost of samples are sometimes included in adver-
tising expense. Thus, it appears that reported outlays on advertising understate
the true volume of advertising and promotional effort in the drug industry even
apart from the salaries paid to detail men.
Sincerely,
WILLIAM OOMANOR.
Dr. COMANOR. In most industries, even in the consumer goods sec-
tor, advertising probably plays a relatively minor role, although it
appears to play a major role in a few others.
One further feature of these statistics is that the subset of indus-
tries with high advertising-sale~s ratios is comprised entirely by those
which produce consumer nondurables rather than consumer durables.
Industries which produce electric appliances or radio and television
sets spend relatively little on advertising as a proportion of total sales.
And yet, it is precisely these industries which are generally considered
to produce highly complex products about which consumers might
need considerable information.
In the statistical analysis, we examine the joint effect of advertising,
together with a number of other market structure variables, on profit
rates. The additional variables are the degree of market concentration,
the rate of growth of demand, and estimates of two entry restricting
factors: the extent to which production economies of scale exist in the
industry, and the total amount of capital required for entry. These
additional variables were introduced into the analysis although our
PAGENO="0497"
Profit rate Advertising sales
ratio
1. Soft drink 10.0 6.2
2. Malt liquors 7.2 6. 8
3. Wines 7. 3 5. 2
4. Distilled liquors 5. 0 2. 1
5. Meat 4.6 .6
6. Dairy 7.9 2.2
7. Canning 6.4 2.9
8. Grain mill products 7. 0 1.9
9. Cereals 14.8 10.3
10. Bakery products 9.3 2.9
11.Sugar 5.8 .2
12. Confectionery 10. 6 3. 5
13. Cigars 5.3 2.6
14. Cigarettes 11. 5 4. 8
15. Knit goods 3. 8 1. 3
16. Carpets 4. ~ 2. 0
17. Hats 1.6 2.2
18. Men's clothing 5. 9 1. 2
19. Women's clothing 6. 1 1. 8
20. Millinery -1. 3 . 8
21. Furs 5.7 1.0
22. Furniture 9.7 1.5
23. Screens and venetian blinds 9. 3 1. 6
24. Periodicals 11.7 .2
25. Books 10. 1 2.4
26. Drugs 14. 0 9. 9
27. Soaps 11.7 9.2
28. Paints 9.9 1.5
29. Perfumes 13. 5 15. 3
30. Tires and tubes 10. 2 1. 4
31. Footwear 7.6 1.5
32. Handtools 11. 4 4. 2
33. Household and service machinery (not electrical) 7. 3 1. 9
34. Electrical appliances 10. 3 3. 5
35. Radio, TV, and phonograph 8.8 2.2
36: Motorcycles and bicyles 5. 2 1. 1
37. Motor vehicles 15.5 .6
38. Instruments 12.0 2.0
39. Clocks and watches 1. 9 5. 6
40. Jewelry (precious metal) 5. 3 3. 2
41. Costume jewelry 1. 4 4. 0
Dr. COMANOR. A significant correlation exists between these two
variables. While high profit rates may be and frequently are asso-
ciated with relatively low advertising-sales ratios, such as instruments
which include cameras, industries with high advertising-sales ratios
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2047
primary attention was directed at the impact of advertising outlays
because they are generally considered to influence profit rates and we
wished to examine the net effect of advertising on profits after the
influence of these other factors had been accounted for.
The statistical analysis is founded largely on tax return data which
are compiled by the statistical division of the Internal Revenue Serv-
ice. It is based on industries as defined by the IRS and refers to aver-
age values for the years 1954 through 1957. Averages were taken so
that the results would not be influenced by the particular business
conditions which happened to exist in a given year, and the years
chosen refer to a complete business cycle.
The statistics on profit rates and advertising-sales ratios which are
used in the empirical analysis are presented in table A-i attached to
my statement.
(The statistics referred to follow:)
TABLE A-i-AVERAGE PROFIT RATES AND ADVERTISING-SALES RATIOS IN 41 CONSUMER GOODS INDUSTRIES,
1954-57
[In percent]
81-280 O-68-'---pt. 5-32
PAGENO="0498"
2048 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
almost without exception have relatively high profit rates. And this
refers specifically to the six industries that have advertising-sales
ratios which exceed 6 percent.
Thus, the average profit rate for those six industries was 11.9 percent
which is 65 percent greater than the average return for the 35 remain-
ing industries of 7.~ percent.
Senator NELSON. When you say "average profit rates," are you
talking about after-tax profits?
Dr. COMANOR. I am talking about after-tax profits.
While motor vehicles has the highest profit rate of any industry in
the sample, and at the same time has a relatively low advertising-sales
ratio, the next three most profitable industries-cereals, drugs, and
perfumes-are all those with very high levels of advertising
expenditures.
Senator NELSON. Does the automobile industry fit into the picture
you describe later of an industry that has some very difficult entrance
barriers-that is, the vast investment it takes to get in?
Is advertising a factor there?
Dr. COMANOR. I think most economists would agree that the auto-
mobile industry has very high entry barriers, but that these are not
accounted for by advertising expenditures as much as by the important
scale economies which exist in the industry, as well as by the very high
absolute capital requirements. In addition, the current franchise sys-
tem does tend to restrict entry. While important entry barriers exist, I
think most economists would not stress the effect of advertising in this
particular industry, although firms in the industry spend very large
absolute sums on advertising, since after all, it is one of our largest
industries. But these large absolute sums account for a relatively
small percentage of sales.
Senator NELSON. Would you agree that there is a distinction be-
tween the automobile and the prescription drug industry on at least
a couple of grounds relative to the advertising question, the automo-
bile industry having a relatively low advertising ratio to sales, with
the drug industry having a high advertising ratio to sales? Would you
agree that one of the reasons for the distinction there is that there is
a multiplicity of drugs and the person who makes the determination
about what drugs will be used and what brand will be used is the doc-
tor, whereas in the automobile industry, there are only four major
companies in this country and the consumer, by looking at the product,
makes the decision about what he will buy?
Dr. COMANOR. While this factor is certainly important, I would em-
phasize a somewhat different factor, which is that in the motor vehicle
industry, factors such as high absolute capital requirements inhibit
entry so that the auto firms do not feel compelled to allocate further
sums to restrict the entry of new firms.
In addition, while advertising promotes product differentiation in
the automobile industry, there are various other factors besides ad-
vertising which are at least as important, if not more important, in
creating product differentiation, such as the annual model changes.
In the pharmaceutical industry, where products which embody the
same chemical compound but which are produced by different firms
are now similar to one another, advertismg plays a much more im-
PAGENO="0499"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2049
portant role. It is in this context that it is interesting to note that
high advertising outlays generally are concentrated in the nondurables
field, where product differences are probably much less. Advertising
is very high, for example, in the detergent industry, although deter-
gents are probably less different from one another than automobiles.
In addition, we know that in the proprietary drug field, there are
very high advertising outlays for products such as aspirin, although
aspirin is aspirin.
Therefore, what we find is that advertising expenditures tend to be
highest where real product differences are not pronounced. And ad-
vertising tends to be relatively low where real product differences in
fact exist.
One might conclude that if there are real product differences, ad-
vertising does not play the role that jt does play when these differences
are absent. It is in this way that I think we can interpret the statistics
which indicate that advertising-sales ratios are higher in the con-
sumer nondurable sector than in the consumer-durable sector.
Senator NELSoN. I realize there are a lot of factors involved in this,
but there are several differences. There is one fundamental difference
between the automobile and the drug industry; that is, that the con-
sumer in the one case, automobiles, is qualified to make a judgment
of his own to decide whether he likes the style of the car, what horse-
power he wants, how many miles per gallon it will travel, what size
automobile he wants, and then based upon his experience in buying
cars, he is qualified to make his own judgment. However, he has no
qualifications for making any judgment whatsoever about a prescrip-
tion drug.
So they are fundamentally different problems, are they not?
Dr. COMANOR. I agree, they are quite fundamentally different prob-
lems. At the same time, we can and do find high advertising sales ratios
in industries where, in fact, the consumer does make the purchasing
decision-such as cereals and perfumes and soaps. So I would not
stress this particular factor as the primary reason why advertising
outlays are high in this industry, although I think that the fact that
the doctor is the purchasing agent for the final consumer plays a very
crucial role on the type of advertising which is undertaken and the
effectiveness of these advertising outlays.
Senator NELSON. I was not suggesting that this factor that I men-
tioned was the most important, or one of the most important, but that
it is a factor which differentiates the two products. Please proceed.
Dr. COMANOR. These preliminary observations were corroborated
by our statistical analysis where the influence of the additional mar-
ket structure variables was also considered. We found that advertising
has a statistically significant impact on industry profit rates and that
this effect is stronger than that of any of the other variables examined.
Furthermore, the magnitude of the effect is surprisingly high. In-
dustries with high advertising outlays earned, on average, profit rates
which exceeded those in other industries, after correcting for the
other variables, by nearly 4 percentage points. This difference repre-
sents, moreover, a 50-percent increase in profit rates. Since profits rep-
resent the difference between prices and costs, and since advertising
outlays is a cost to the firm making the outlay, these findings suggest
PAGENO="0500"
2050 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
that high advertising expenditures have a double effect on price levels.
Not only do they represent higher costs, but also they are associated
with higher price-cost margins.
ADVERTISING AS A BARRIER TO ENTRY
The empirical evidence described above indicates that industries
which spend substantial proportions of their revenues on advertising
tend at the same time to earn high rates of return, and I want now to
examine some economic factors which might explain this result.
In a free market economy, high profit rates, which are not accounted
for by differences in rates of growth of demand or by the degree of risk,
will not persist for a prolonged period of time in the absence of factors
which restrict the entry of new firms. If high rates of return in an in-
dustry, in firms will enter the market, and this will generally have the
effect of driving prices down to more competitive levels. High prices
and profits suggest, thereby that specific factors are present which re-
strict the entry of new firms.
Senator NELSON. The economists who have testified in behalf of the
Pharmaceutical Manufacturers Association have argued that the high
profits of the drug industry were the result of high risks.
Do you have any observations to make about that?
Dr. COMANOR. Yes; I do. I have read parts of the testimony pre-
sented by these economists.
It seems to me that to consider this, one would have to ask what we
mean by risk and in how we can mensure it. Risk involves some notion
of unexpected changes in the value of the firm or unexpected changes
in profitability. But this cannot be measured simply by the variability
in profit rates across firms in an industry. After all, there are generally
differences among profit rates. They may measure accustomed and ex-
pected and traditional differences among firms. They may measure
systematic as opposed to unexpected differences. Different firms in an
industry may have different profit rates for a number of reasons: be-
cause of economies of scale, because of other differences in efficiency.
because they operate and sell in different segments of the market, or
because of differences in consumer acceptance or product differentia-
tion.
It seems therefore that the variability of profit rates across firms in
an industry cannot be used to measure risk because these changes may
be accustomed and traditional rather than unexpected and the essence
of risk is the inability to predict what will happen in the future.
In addition, risk, however measured, is likely to be very different for
firms of different sizes, and the simple variability across firms does not
account for these differences.
This is another reason why simple variability across firms is not an
adequate measure of the degree of risk in an industry.
It seems to me that a much better measure might be the intertem-
poral difference and the variability of profit rates, across time, espe-
cially after correcting for size of firm. I note from testimony given
recently that when risk is measured in this manner, the degree of risk
for the drag industry tends to be very low.
PAGENO="0501"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2051
These statistics were presented to Dr. Mueller's testimony. There-
fore, I think it is difficult to argue, and I certainly would not, that the
drug industry is a high risk.
Senator NELSON. Did you read Dr. Mueller's testimony?
Dr. Coi~rANoR. Yes, I did.
Senator NEr4soN. Do you agree with him on the point he makes on
this question?
Dr. COi~rANOR. Yes, I do.
In the consumer goods industries where advertising is relatively
important, we frequently observe unbranded products which sell at.
prices substantially below those of highly advertised products even
though there may be little real difference between them. Because prod-
ucts which are little advertised must sell at far lower prices than those
of their established rivals, the latter can raise prices above costs and
earn high profits witihout fear of enticing the entry of new firms and
of the resulting effect on price levels.
High advertising outlays create effective entry barriers through a
number of routes. In the first place, high current levels of advertising
expenditures create additional costs for new entrants which will gen-
erally exceed those for established firms. Because of buyer inertia and
brand loyalty, more advertising messages per prospective customer
must be supplied to induce brand switching as compared with repeat
buying. And therefore, new entrants must be prepared to supply more
advertising messages per prospective customer than do their estab-
lisheci rivals.
A further disadvantage faced by new entrants is that they must
spend nearly as much in total advertising and other forms of promotion
as existing firms if their products are to compete with established and
well-known products. Since new entrants are generally small, this total
expense must be spread over fewer units of output, and therefore the
cost per unit may be quite high. Larger established firms, on the other
hand, have the advantage of being able. to spread this cost further and
often spend considerably less per unit sold. Thus, where advertising is
important, new entrants are frequently found to bear higher unit ad-
vertising costs as compared with their larger rivals.5 And this factor
again enables established firms tO set relatively high prices without
fear of the effective competition of new entrants.
Furthermore, there is evidence that the cost per advertising message
declines as the number of messages purchased increases.6 New entrants,
thus, face higher unit costs which contribute further to their relative
disadvantage.
Finally, the need to spend considerable funds on advertising will
raise the amount of capital required for entry into the market. Since
new firms frequently find it difficult indeed to raise large amounts of
In the automobile industry, for example, the two smaller firms during the 1950's were
found to spend more than twice as much on advertising per car sold as did either Ford or
General Motors; Between 1954 and 1957 Studebaker and American Motors spent annually
on nntional advertising approximately $64.04 and $57.89 per automobile sold while General
Motors spent $26.56 per unit and Ford spent $27.22 per unit. Chrysler assumed an inter-
mediate position, spending $47.76 per unit. Leonard W. Weiss, "Economics and American
Industry," p. 342.
o The extent of discounts given to large advertisers is documented in Federal Trade Corn-
mossion v. The Procter 4 Gamble Co., brief for the Federal Trade Commission in the Su-
preme Court of the United States, December 1966, pp. 12-13.
PAGENO="0502"
2052 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
capital, high advertising expenditures by existing firms, and thereby
high advertising requirements for new entrants, will further restrict
the rate of entry.
For all of these reasons, heavy advertising expenditures serve to
create substantial entry barriers. They act as an important restriction
on competition and permit established firms to charge high prices and
earn high profit rates without fear of the competitive consequences.
And the statistical findings presented earlier merely describe the extent
of the effect of advertising on competition.
THE PROBLEM OF CASUALITY
Throughout the analysis, we have assumed that the direction of
casuality runs from advertising to high profit rates, and it is necessary
to consider whether the reverse could be the case. This is especially so
since a plausible argument could be made that, since advertising re-
flects the discretionary behavior of firms, high profits could lead to
high advertising. In other words, high profits could be "spent" on
advertising.
There are a number of factors, however, which suggest that the
appropriate direction of casuality is in fact from advertising to profits.
The first is that if profits were "spent" on advertising, which is after
all an expense to the firm, higher measured profit rates would be as-
sociated with lower advertising expenditures, and lower measured
profit rates, with higher outlays. But we have observed a positive
rather than a negative relationship between these two variables.
A second and even more important reason stems from the fact that
the amount of advertising expenditures depends on many factors
besides the whims of individual managers.
Product and market characteristics make advertising a more profit-
able activity in some industries than in others, and there are few
reasons for believing that managers in some industries are better
equipped to take advantage of their opportunities than those in
others.
In fact, if we assume that managers make their decisions on adver-
tising burgets so as to maximize profits, then the differences between
industries which we observe reflect not the discretionary behavior of
individuals but rather the particular product and market characteris-
tics of member firms in the industry.
The figures presented above, which deal with the average values
across a 4-year period, may be interpreted in fact as describing the op-
timal levels of advertising expenditures in the particular industry.
And there is no reason to believe that firms with higher profit rates
will have higher optimum advertising expenditures than firms with
lower profit rates but similar market conditions.
This latter consideration is an important one for interpreting the
statistical results presented earlier. Since the analysis focuses on inter-
industry differences in advertising outlays, and since these differences
are more likely to reflect the product and market characteristics in the
industry rather than the peculiarity of individual managers, these
results have few behavioral connotations. They do not imply that an
industry can earn higher profit rates simply by spending more on
PAGENO="0503"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2053
advertising. They do not suggest that firms which currently spend
little on advertising should `double their advertising budgets if their
objective is to o'btain higher returns. But rather, they describe the
fact that firms and industries with higher optimum advertising ex-
penditures will and do earn higher rates of return than firms in less
advantageous market positions.
THE POSITION OF THE DRUG INDUSTRY
Although this study refers to the entire consumer goods sector of
the economy, we can note that the drug industry is one of those in the
sample with both high profit rates and high advertising-sales ratios.
During the period between 1954 and 1957, this industry stood in third
place out of 41 in both respects. As defined in this analysis, however,
the drug industry refers to both prescription and proprietary products.
In the prescription drug industry, advertising accounts for only a
portion of the firm's total selling and promotional budget. Detail men
and other forms of direct selling also are important, and therefore
figures on advertising alone understate the total effort in persuasion
which is carried on by the drug companies.
Senator NELSON. In your figures on advertising, you did not include
any of the costs of the detail men?
Dr. COMANOR. That is correct.
Senator NELSON. What did you include?
Dr. COMANOR. We included expenditures on advertising as reported
in tax returns.
Senator NELSON. You reported what the industry claimed as adver-
tising and what the IRS accepted as advertising?
Dr. COMANOR. That is correct.
Senator NELSON. Do you know what that included?
Dr. COMANOR. That includes expenditures on printed advertise-
ments, direct mail advertisements,, and also advertisements in medical
journals. It does not include salaries for detail men, for samples, or
things of this sort.
Senator NELSON. It did not include samples?
Dr. COMANOR. It did not include samples.
Senator NELSON. Even though samples had some advertising ma-
terial with it?
Dr. COMANOR. I think that is correct.
Senator NELSON. Did you include other expenditures such as ma-
terial furnished to medical students, stethoscopes, that sort of thing?
Dr. COMANOR. I am not absolutely certain, but I think not. I would
have to check with the IRS before I could answer these questions
completely.
Senator NELSON. I do not know what these expenditures amount to,
but we have had testimony that medical students get a considerable
amount of material, some advertising and promotional, some of them
stethoseopes and so forth.
Do the firms deduct this as a business expense?
Dr. COMANOR. This is surely a business expense. I am not certain
whether it is included in these advertising figures. My wife is a medical
PAGENO="0504"
2054 COMPETITIVE PROBLEMS IN THE DRUG ThTDUSTRY
student, and I am forced to admit that she has received a stethoscope
and a black bag.
Something thnt you might be interested to know is that one of the
drug companies provides every second-year medical student with a
black bag which is plastic. T11e same company in their fourth year pro-
vides the medical student with a black bag of real leather. She is now
going around with her plastic one.
Senator NELSON. They have somewhat less confidence in the pros-
pects for the second year student than the senior student?
Dr. COMANOR. That must be it.
Senator NELSON. Would it be easy for you to check and see whether
these items include this advertising expense in the tax column?
Dr. OOMANOR. One could check with the IRS on this question.
Senator NELSON. If you could, along with the other material you
are going to submit, it would be useful to put. it in the record.
Dr. COMANOR. I will do this.7
Advertising generally is particularly effective in situations where
consumer uncertainty or ignorance about the relative nierits of com-
peting products is relatively high. It is not so much that physicians
or other consumers consciously accept the blandishments of an adver-
tising message as much as they consider tl1e purchase or prescribing of
an advertised product as the "safe" course of action.
When considerable uncertainty exists, the choice of a highly adver-
tised product often represents minimizing the risk to the consumer that
the product will do the job for which it is being purchased. And what
becomes important is not the content of the advertising message but
simply the fact that the product is well advertised. Thus, products are
frequently described as "advertised in Life." 8
The effect of advertising on entry barriers in the pharmaceutical
industry is strengthened by an important interaction which exists
between research and selling expenditures. The relatively high level of
research in this industry has led to a rapid rate of new product intro-
duction throughout the postwar period. During the decade ending in
1960, over 3800 new products or dosage forms were introduced into
U.S. pharmaceutical markets. Of this total, slightly more than 11 per-
cent were new single chemical entities.
Given the rapid pace of new product introduction, it is not surpris-
ing that doctors have been open to the persuasion and influence of the
drug companies. At a 1959 medical conference on the evaluation of new
drugs, it was reported that "physicians are frightened, confused and
puzzled by advertising material which pushes as many as a. thousand
new drugs or combinations of drugs every year * * * several prac-
titioners at the session said they felt a. void of information about t.he
proper use of new drugs." ~ The simple pace of new product introduc-
tion has been a major factor which has increased the effectiveness of
heavy advertising and promotional expenditures.
Furthermore, most new products have received patent protection.
Thu~s, in 1961, of the 656 single chemical entities which were used for
See letter, p. 2046, supra.
~ view is similar to that taken in Donald F. Turner, "Advertising and Competition,"
speech delivered before the Briefing Conference on Federal Controls of Advertising and
Promotion, June 2, 1966, pp. 3-4.
° Herman Somers and Anne Somers, "Doctors, Patients, and Health Insurance," p. 100.
PAGENO="0505"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2055
therapeutic purposes, 377 were sold only by a single firm, which
amounted to about 57 percent of the tOtaL 10 The impact of the patent
system, however, has not been to create tight monopoly positions, since
patented products are often highly substitutable and compete with one
another. But ratl1er, it has been to foreclose, to a great extent, rivalry
between identical chemical entities or standardized commodities about
which price competition might develop. It has strengthened and en-
couraged a form of chemical product: differentiation. In the majority
of cases, therefore, advertising has been able to exploit and emphasize
the chemical differences which do in fact exist among products.
When doctors are forced to choose between different chemical com-
pounds which purportedly do the same thing, a large measure of un-
certainty is certain to exist. Even though a doctor might believe that
different compounds have similar therapeutic effects, he is never quite
sure, and the prudent course of action is to prescribe the drug which has
become well-known. And becoming well-known is, of course, the func-
tion among other things of the level of advertising and promotional
expenditures.
Mr. GORDON. Did you say the purpose of the advertising and pro-
motional activities is to prevent price c0mpetiti0~
Dr. COMANOR. After the experience of the drug industry with peni-
cillin, which was introduced in the early postwar period, when the price
dropped precipitiously in the course of 5 or 8 years-as I recall,
the price of penicillin for a standard: dosage dropped from $20 in 1943
to on the order of five cents by 1950-
Senator NELSON. $20 for-
Dr. COMANOR. I think it was 100,000 units.
Senator NELSON. From $20 to what?
Dr. COMANOR. The price dropped from $20 for 100,000 units in 1943
to 4.5 cents in 1950. The price of streptomycin behaved similarly.
At the prices which finally resulted, profits were clearly quite low.
Senator NELSON. What accounted for the drop?
Dr. COMANOR. Penicillin is a standardized commodity. A large num-
ber of new firms entered into the market, vigorous competition took
place, and prices were pushed down to costs.
Senator NELSON. Was a profit made on it?
You say the prices were pushed down to cost. That would mean no
profit.
Dr. COMANOR. It is unclear as to just what the profit margin was for
penicillin. It may very well have been that profits were low and even
nil, No figures are available on this question.
This experience, I think, had a :~ sharp impact on the industry. It
became clear that profits in the future would depend on introducing
specialized products, on restricting the growth of standardized com-
modities, and thereby restricting effective price competition. Price
competition seems to demand a number of firms who sell `the same
chemical compound, for it is only in this setting that the purchaser
may focus on price rather than the therapeutic effects of a product. It
became clear that some means of differentiating the product was re-
10 p~ Walker, "Price Levels and Market Power in the Ethical Drug Industry," paper
presented at the December 1967 meetings of the Econometric Society, Washington, D.C.,
table 3.
PAGENO="0506"
2056 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
quired. We note that throughout the decade of the 1950's, large num-
bers of new products were introduced. Most of these were protected by
patents. And also, advertising expenditures grew rapidly. We can un-
derstand the incentive for both research and advertising in the drug
industry in terms of the aim to create and promote product differenti-
ations which would prevent the development of standardized commodi-
ties and which would prevent the type of price experience which
occurred in penicillin and streptomycin.
Senator NELSON. Thank you.
Mr. GORDON. The product differentiation you just talked about is
the chief barrier to entry into the various therapeutic categories, is that
correct?
Dr. COMANOR. That is correct. It is the role of both advertising and
research in this industry to promote a form of product differentiation
which then serves to restrict entry and to restrict the development of
price competition.
In this manner, the patent system interacts with the high level of
advertising and promotion by insuring that the competition which ex-
ists in this industry will focus on chemically differentiated products,
and this factor increases substantially the effectiveness of the selling
efforts in this industry.
Were the patent system not to exist, the effectiveness of these efforts
is likely to be reduced. If new entrants were free to produce the same
chemical compound as established firms, the effectiveness of advertising
in constructing barriers to entry would be lessened, and we might
expect to find more stringent constraints placed upon the price be-
liavior of the leading firms in the industry."
Even in the absence of chemical differences among products, how-
ever, considerable uncertainty remains as to whether different brands
of the same compound are equally good. It is claimed that tradename
products which are sold by the larger companies are manufactured
under quality controls which produce higher quality products. While
there is no readily available evidence on quality control expenditures
for pharmaceutical firms in the United States, statistics are available
for Canada. Twenty-one large firms, many of which are U.S. subsidi-
aries, reported to the Canadian Government that quality control
expenditures accounted for only 1.2 percent of total sales or approxi-
mately 3.6 percent of the cost of goods sold.
Senator NELSON. What is the distinction? 1.2 percent-
Dr. COMANOR. Of total sales.
Senator NELSON. I understand the 1.2 percent of total sales and
approximately 3.6 percent of the cost of goods sold-what precisely
do you mean?
Dr. COMANOR. Production costs. You can translate cost of goods sold
as production costs, of the total production costs of the product, qual-
ity control accounts for only 3.6 percent. After these costs are adver-
tising, promotion, profits, to make up total revenues.
Senator NELSON. So when you are talking about cost of goods sold,
you mean the total cost of producing the product?
"Restrictive Trade Practices Commission, "Report Concerning the Manufacture, Dis-
tribution, and Sale of Drugs," Department of Justice, Ottawa, app. Q, p. 145.
PAGENO="0507"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2057
Dr. COMANOR. That is correct.
SenatorNELsoN. And then the sales price is different in that you
add your profits and your advertising and your promotion, is that
what you are saying?
Dr. COMANOR. That is what I am saying.
These statistics imply, then, that production costs are approxi-
mately one-third of the total selling price.
Senator N1~soN. Where do you ~et that figure?
Dr. COMANOR. We are dealing with the same expenditures on quality
control, and these outlays are taken as a fraction of two different
quantities. One is total production cOsts and the other is total revenues.
One percentage is a third that of the other.
Senator NELSON. You have several factors in here that I do not
understand. You are talking in one case about quality controls. Quality
control is not the same as production costs?
Dr. COMANOR. Oh, no; in fact, quality control is only 3.6 percent
of total production costs.
Senator N1~soN. And 1.2 percentof sales cost?
Dr. COMANOR. 1.2 percent of total sales revenues.
Senator NELSON. OK.
Dr. COMANOR. These statistics suggest that relatively small firms
should be able to invest in quality control facilities and as can be
seen, these expenditures are considerably less cos~ly to the smaller
firm or new entrant than are heavy advertising and promotional efforts.
Senator NELsoN. As to these statistics that you got from the U.S.
subsidiaries of manufacturers in Canada, does the Canadian Govern-
ment require a breakdown of these various costs?
Dr. ~OMANOR. I do not think this breakdown is reported every year.
A report published by the Restrictive Trade Practices Commission in
Canada included these statistics.
Senator NELSON. When was that report published?
Dr. COMANOR. I am afraid I dO not have the date here. Within the
last few years. I have a copy of it.
Senator NELSON. Is it a large report?
Dr. COMANOR. It is quite a large report. It must be 500 or 600
pages.
Senator NELSON. In this report, do they have the statistics on the
costs of production of a drug and the cost of quality control and the
sales and profits-all of these statistics?
Dr. COMANOR. Not per drug, but for the entire company.
Senator NELSON. All together?
Dr. COMANOR. Or it may not even have been exported for individual
companies but rather for the total of these 21 firms, only some of
which were U.S. subsidiaries.
Senator NELsoN. Did they not have any breakdown company by
company? They must have gotten the figures from some place.
Dr. COMANOR. They may not have published them.
Senator NELSON. In order to get the composite, you have to have
the individual figures.
Dr. COMANOR. They may not have published them. I am not certain
about this point. I could find the answer for you.
PAGENO="0508"
2058 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Senator NELSON. If you would give use the name of the report, I
will have the staff look into it.
Dr. COMANOR. Footnote No. 11 provides t.he reference to the report.
Senator NELSON. All right.
* Dr. COMANOR. While more work needs to be done in gathering and
analyzing statistical evidence on this problem, there is no evidence
which indicates that only branded drugs produced by large firms are
of a high quality.
THE CRITICAL ROLE OF INFORMATION
The analysis above suggests that advertising and other forms of
promotion. have an important effect on competition in `the pharmaceu-
tical industry. These outlays permit the leading firms in the industry
to enjoy `a considerable freedom from competitive contraints, and they
have been translated into very high rates of return.
Furthermore, the analysis suggests that advertising and other forms
of promotion have been effective to a large extent, because of the high
degree of uncertainty about the relative merits of competing products
which is present in the minds of t.he physician who serves as the pur-
chasing agent for the final consumer.
On these grounds, there is much to be said for providing the physi-
cian with an impartial source of information which will relieve him
of his current reliance on information provided by the drug companies.
Senator N~soN. I think this is a very good point. The fact is that a
good part of the thrust of the industry's position is that you can't rely
upon a drug unless it is produced by an established, well-known brand
name company and that you are taking a risk with drugs from other
companies. They end up convincing the physician of that.. He has be-
come convinced that there is not any such thing as generic equivalency
or that it is a very serious problem. Further, there is not any available
compendium of drugs that t.he doctor is prepared to trust, and that
can assure him that he can prescribe a generic drug and trust its relia-
bility. You are suggesting, I take it, that there ought to be a reliable,
independent source so that the doctor could select drugs that are as-
serted by an independent source to be equivalent so that he can pre-
scribe a drug that gives his patient a better price break.
Is that what you are saying?
Dr. COMANOR. That is correct. It seems to me that advertising is ef-
fective primarily because of the uncertainty which exists in the mind
of the purchasing agent, which in this case is the physician. And if
these expenditures were not so effective, we might expect a lower
amount to be spent. What is important is to reduce the effectiveness of
heavy advertising outlays by providing the doctor with an independent
and reliable source of information.
This is not to say that the drug companies do not provide accurate
information, although it may sometimes be less than complete, but
rather that information from an independent source would have an
element of impartiality which salesmen or printed advertisements
rarely achieve.
Senator NELSON. If it is less than complete, it is not accurate, is it?
PAGENO="0509"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2059
Dr. COMANOR. I agree that the drug companies do not always pro-
vide full information about the relative merits of competing drugs.
Senator NELSON. If it omits certain information to the doctor, then
it is not accurate, is it?
Dr. COMANOR. That is correct.
Senator NELSON. But what you are saying is that they do not neces-
sarily say anything incorrect, but if they omit something, the informa-
tion is not accurate?
Dr. COMANOR. That is right. There are certain types of information
which it is very important for the doctor to receive and which he
rarely receives under the current system. He rarely receives a statement
that two drugs are the same, because the incentive for the drug com-
pany is the opposite-to differentiate their product from those of their
rivals. One piece of information which it is important for him to re-
ceive is that two drugs have the same or similar therapeutic effects,
if that in fact `happens to be the case.
Senator NELSON. In the majority of cases, no company is going to
advertise that several other companies have a drug that is equivalent
to its own.
Dr. COMANOR. That is right.
Senator NELSON. So it has to come from somebody else.
Dr. COMANOR. There are certain types of information which the doc-
tor simply does not receive from advertisements. Therefore, he is left
in the dark.
And also, and `this seems to me to be equally important, it would be
open to the smaller firm and new entrant as well as the large company.
Where various firms produce the same chemical compound, the phy-
sician should be informed either that all products are equally good,
in terms of purity, strength, or some other characteristic, or that one
brand is to be preferred to another.
And where different compounds compete, impartial information
would lead to improved medical practice as well as reduce the signifi-
cance of advertising and promotion.
If a small firm or new entrant introduced a beneficial new drug,
it would compete on the basis of: therapeutic properties rather than
advertising claims. There is no evidence, moreover, that smaller firms
are at a disadvantage in research as compared with advertising and
promotion.12 On many grounds, then, there seems to be a considerable
need to provide the Nation's doctors with greater information about
new drugs which is independent of the messages they receive from the
drug companies.
Mr. GORDON. Dr. Comanor, in the footnote on page 16, you state
"The available evidence suggests, moreover, that small firms have
relatively more efficient research facilities than their larger rivals."
What evidence do you have to show this?
Dr. COMANOR. In one of the papers which I have published, I report
on a statistical analysis of the relationship between research input and
research output.
Senator NELSON. Your analysis was comparing the products result-
ing from the research?
12 The available evidence suggests, moreover, that small firms have relatively more efficient
research facilities than their larger rivals. William S. Comanor, "Research and Technical
Change in the Pharmaceutical Industry," Review of Economic8 and Stati8tiCs, May 1965.
PAGENO="0510"
2060 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. COMANOR. That is right. I carried out a study of a large number
of pharmaceutical firms and I measured the relationship between
research input and output. I measured the output of a research estab-
lishment by the number of new products introduced by the firm since
research in the pharmaceutical industry is directed largely toward new
products rather than new pr~esses. The number of new products
was multiplied by its sales during the first 2 years following intro-
duction. To make these computations, I obtained access to the Gosselin
data which provides sales by product for all products sold in the coun-
try. Research input was measured by the number of research personnel
employed by the individual pharmaceutical firm, and the relationship
between input and output was then estimated.
Senator NELSON. You did not make, I take it, any distinction between
the efficiency of the research of one company in developing a new drug
versus another company in producing a new drug? You did not make
a distinction between that and the research which produced a product
differentiation, did you?
Dr. COMANOR. No, I did not.
Senator NELSON. So whatever they produced, either as a product
differentiation or a new drug, is included in the statistics?
Dr. COMANOR. I used two measures of new products, but the most
important included only new single chemical entities. I was, therefore,
dealing with products which were new in the sense that they were
different chemical compounds which had previously been introduced
in the U.S. market.
While it is true, of course, that some of these embodied a greater
therapeutic advance than others, and it may be the case that some
embodied no or very little therapeutic advance, I did not feel capable
of deciding this question. I simply took all new single chemical
entities, multiplied by their sales in the first 2 years following intro-
duction, to measure the output of research.
When I carried out the statistical analysis between research input
and research output, which is reported in one of the papers that I have
published on the industry, there were a number of findings:
First, there was a fairly sustained and direct association between
input and output. In other words, research is not undertaken in this
industry simply in the hope of some payoff in the future, but rather
there is this fairly sustained and systematic relationship between the
amount of research undertaken and the number of new products intro-
duced by the individual firm.
Senator NELSON. When you say you are comparing the relatively
small firms versus their larger rivals, what standard did you use to
distinguish a relatively small firm and a large firm?
Dr. COMANOR. I measured the size of firm by the level of pharma-
ceutical sales.
Senator NELSON. Yes; what was the cutoff pomt?
Dr. COMANOR. There was no cutoff. It was undertaken not on~ a
discrete but on a continuous basis. I exammed the entire relationship
between small firms and large firms without any arbitrary cutoff point
to distinguish large from small.
Senator NELSON. So some of of them were relatively small?
Dr. COMANOR. Yes, and I had firms of all sizes in my sample.
PAGENO="0511"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2061
I found that economies of scale exist in very small firms, but at the
same time, diseconomies of scale exist in large firms and even in research
carried out in moderate-sized firms. From these statistical results there
appear to be no advantages from increasing the size of the firm's
research effort beyond 50 to 75 professionals. At the same time, we
can note that most of the leading firms in this industry have research
labs which are far larger than this size. The cutoff point, which was
not determined arbitrarily but resulted from the statistical analysis,
was that economies of scale appear to exist up to the level of 50 to 75
professional personnel, but become diseconomies after that level.
The position stated above is identical to that taken earlier by the
Assistant Attorney General in charge of the Antitrust Division, Mr.
Donald F. Turner, who in a recent speech discussed the competitive
effects of advertising expenditures generally throughout the economy.
Mr. Turner emphasized:
It is the extent of uncertainty about the relative merits of competing products
which contributes to the large effect of advertising, and this suggests that Govern-
ment policies be directed toward neutral vehicles of information which tend to
deal directly with the uncertainty.
He goes on to mention the Medical Letter, which is an important source of
"technical information about the therapeutic value of new drugs", and con-
cludes that "there is much to be said for providing Government funds to the
organization which publishes the Medical Letter so that its publication may be
supplied free to all doctors. In addition, the latter could be expanded to ensure
that doctors receive their first information about a new drug from this source
rather than from the lips of a detail man." ~
Mr. GORDON. Is it your opinion that this measure could offset the
huge expenditures by the drug industry on advertising and promotion?
Dr. COMANOR. I think that this type of policy, while it may not
offset fully the impact of advertising and promotion on product dif-
ferentiation and thereby on price competition, would have an im-
portant effect. I would expect that it would reduce significantly the
effectiveness of these outlays and think we could expect a higher degree
of price competition with this information provided to all doctors
than we find currently.
I agree fully with Mr. Turner's position that an impartial and in-
dependent source of information is, needed-that it would be an im-
portant step forward, both in improving the quality and standard
of medical practice throughout the country, and in reducing the anti-
competitive consequences of heavy advertising.
There are a number of alternative routes to achieving this objective.
One might be to provide Federal support to the nonprofit organiza-
tion which publishes the medical letter. Not only would this permit
a more comprehensive study of drugs and drug claims than is possible
on the basis of available facilities, hut also it would allow an expanded
Medical Letter to be sent free to all phyisicians throughout the country.
Mr. GROSSMAN. Dr. Comanor, may I ask a question?
Several of the economists who have appeared here have talked about
the fact that we should get rid of the brand name, trade name, and
just use the generic name.
Do you feel that way as well?
Dr. COMANOR. There are a number of routes to achieving the same
objective. One route, m some sense an easier route, would be to provide
~` Turner, p. 2054, supra.
PAGENO="0512"
2062 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
information to the doctor which would reduce the impact of trade
names and heavy advertising outlays.
Another route, which is perhaps a stronger step, would be to elinu-
nate all trade names altogether. I certainly feel that further informa-
tion is needed now regardless of whether or not trade names are
eliminated. At the same time, it seems to me that trade names in this
industry have become a very important source of limiting the entry
of new firms into the relevant therapeutic markets and restricting
the development of price competition. Therefore, although I have
not emphasized this problem in my statement, I would certainly sub-
scribe to this position.
Mr. GROSSMAN. If we did away with trade names, would it not be
likely that we would have a generic name (Schering), generic name
(Upjohn), and the company would promote their name? They would
find a way to get around it?
Dr. COMANOR. There is no question that the drug companies would
aim to achieve just the end which you suggest. One might argue that
their ability to differentiate their products on this basis would be
less than it is currently, and as a result, we might find a higher degree
of price competition. This is not an all-or-nothing situation; it is
clearly a matter of degree. While some measure of product differentia-
tion would remain, and might even be desirable at a small level, it would
not lead to the high prices which currently exist and the very high
rates of return.
Mr. GROSSMAN. Do you thii± doctors should be given some informa-
tion as to variable pricing such as the formulary?
Dr. COMANOR. I would like to see the price of the product included
in the Medical Letter or any sources of information provided to
doctors.
Mr. GROSSMAN. Once that is done, the doctor should be able to make
the choice? So long as he is given this information, he should be able
to make a free choice?
Dr. COMANOR. That is a very difficult question. Although price in.
formation should be provided to physicians, I do not think I would
go so far as to say that in every case, only the doctors should decide
among generically equivalent drugs-that the patient should not be
permitted to achieve generic equivalency if this were desired. I am
not prepared to say that there should be no option to the patient on
generic equivalency.
Senator NELSON. The patient does not have the option on generic
equivalency.
Dr. COMANOR. No.
Senator NELSON. The doctor does.
Dr. COMANOR. The question I think we are really getting at is
whether or not the pharmacist should be permitted to substitute
among generically equivalent drugs. I find that it is a very difficult
problem. I am not prepared to say definitely that providing since in-
formation to the physician is sufficient.
Mr. GROSSMAN. If we list various drugs under a generic heading,
be they trade names or generic names, and we also list the prices next
to them, will we have gone far enough?
PAGENO="0513"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2063
In other words, we will be giving the doctor quite a bit of informa-
tion, more than he has now.
Is this going far enough?
Dr. COMANOR. I think that is an important step forward. Whether
or not that would be sufficient is a question which is very difficult and
I am not prepared to say definitely that it is.
Mr. GROSSMAN. But it would be a step forward?
Dr. COMANOR. I think so.
An alternative proposal would be to create a new institute in the
National Institutes of Health which would be responsible for evaluat-
ing existing information on the therapeutic properties for new drugs,
carrying out their own studies where necessary, and also for providing
the accumulated information freely to all physicians.
In this manner also, an independent source of information about
both current and new drugs would, become available. By whatever
route this is achieved, however, it seems clear that the relatively small
level of expenditures required for these purposes would pay for itself
many times in terms of limiting the effects of heavy advertising out-
lays and in terms of lower prices for prescription drugs.
While a new source of information will not solve all of the com-
petitive problems which exist in the pharmaceutical industry, and
restrictions on competition will remain because of the current opera-
tion of the patent system; because of the practice of trade name pre-
scribing; and because of the problem of drug substitution or the lack
thereof at the level of the pharmacist, still Federal action in this
direction should make an important contribution toward improved
medical practice as well as toward lower prices for prescription drugs.
Mr. GROSSMAN. Independent economists have previously recom-
mended compulsory licensing as one means of making the industry
more competitive.
Would compulsory licensing lead to lower prices?
Dr. COMANOR. The patent system currently operates to provide a
form of chemical product differentiation in the sense that competi-
tion exists on the basis of different chemical entities. While it is true
that different products compete with one another, the patent system
increases the extent to which competition is based on different chemical
entities. And this always provides an important element of uncer-
tainty to the physician. He is never quite sure whether one product is
better than another.
For this reason, it increases the effectiveness of advertising. It per-
mits advertising to have a major impact on prescribing habits, on the
entry of new firms, and on price competition.
With compulsory licensing, we might expect to find new firms
enter the industry, competition take place on the basis of the same
chemical compound, and here, advertising would probably be less
effective than it is currently, especially if further information were
provided to physicians. And we might expect to find a higher degree
of price competition, and as a result, lower prices.
Mr. GROSSMAN. Now, would that lead to less research by the drug
companies?
Dr. COMANOR. With compulsory licensing, we would expect to find
different firms competing on the basis of the same che~nical com-
81-280-OS-Pt. 5-33
PAGENO="0514"
2064 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
pound. This would tend to reduce, as I have noted, the effectiveness of
heavy advertising. And we might expect to find more vigorous compe-
tition and lower prices.
An example of the effect of patent protection is the major difference
in price behavior between penicillin, where price competition has been
vigorous, and the broad spectrum antibiotics, which for a long time
were patent protected, when there was little price competition.
At the same time, we have to expect that with lower rewards from
research, it is likely that less research will be undertaken, and it is
important to recognize that there is in fact a tradeoff between indus-
trial research and lower prices.
However, it seems to me that we pay a high price for industry re-
search in terms of the high prices and profits which exist in this indus-
try. A major question for public policy concerns what we get for this
price. To answer this question, we have to think about the character or
payoff which society realizes from industrial research. We have to
consider the value of industrial research.
Industrial research is carried on within the context of strong product
rivalry. This rivalry influences significantly the type of work which is
in fact carried out. We can note that there is a large effort to invent
around existing patents, to find a new product which has similar
therapeutic properties to those already on the market. And this, of
course, is what we mean by molecule manipulations, which are pur-
ported to have the same therapeutic effect as existing products.
Also, if we examine the research efforts and output of the industry,
there appears to be a considerable effort to create new drug combina-
tions, combining different active ingredients with one another, and to
introduce new dosage forms.
As a result, we do not expect the major discoveries to be made within
the industry precisely because of the character of the research effort
which in fact is undertaken. Still, it seems to me that the minor
advances which result from the research laboratories within the phar-
maceutical industry, the modifications and variations which result, are
important elements in the process of developing a new technology.
The second major emphasis of industry research is the process by
which new knowledge is translated into marketable l)roducts. We know
that many problems remain even after a basic discovery has been made.
It has been estimated, in fact, that 50 percent of research and develop-
ment expenditures within the pharmaceutical industry goes to support
the work of synthesizing, purifying, modifying and preparing sub-
stances for testing. And we know that considerable testing is required
before a drug may be released to the public.
In addition, there is also frequently a considerable gap between
laboratory synthesis and large-scale manufacturing which is also car-
ried out within the research laboratories of the pharmaceutical
industry.
Although this might be considered development work, it is still quite
necessary. And we have only to note that although penicillin was dis-
covered originally in 1928, it was not used effectively until some years
later, until the Second World War, 1943.
This time lag is likely to b~ much shorter because of the current.
facilities of the drug industry.
PAGENO="0515"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2065
To evaluate industry research and development, however, we have
to place it within the context of the tOtal medical research and devel-
opment effort in the United States. In 1960, the industry provided
about 30 percent of total research and development outlays. At the
same time, it provided only 12 percent of the Ph. D.'s involved in
medical research and employed only about 5 percent of the M.D.'s in
medical research. This indicates, thus, that the character of research in
the industry is likely to be far different, at a lower level of scientific~
caliber, than the medical research which is carried on outside of the
pharmaceutical industry.
It seems to me, therefore, that the industry research effort cannot be
evaluated or judged by the same standards as nonindustry research. It
seems to me that the industry research effort should be considered as a,
complimentary rather than as a competitive or alternative type of
activity. There appear to be extensive resources which are devoted to)
fairly routine sets of activities.
When evaluated on its own terms, on a different basis than might be
appropriate for evaluating academic or Government medical research,
it seems to me that the industry effort, by and large, does a good job. It
has, in fact, been responsible for the large number of modifications
and improvements of existing drugs.: It has also been responsible for
accelerating the process of testing aild developing new drugs. And I
would argue that both of these provide important benefits to the
development of medical technology and medical practice in the
country.
At the same time, the question rern~dns as to whether the "price" that
society pays in terms of the opportunity costs for these results is too
high. If we examine the proposal of compulsory licensing after a short
period of time, such as the Kefauver suggestion of 3 years, it is clear
that this proposal would reduce the incentive to undertake research.
The industry certainly would have less incentive to undertake purely
duplicative research in order to enter a profitable therapeutic market.
Senator NELSON. If they had less incentive to do purely duplica-
tive research, that would be a net benefit to society; would it not?
Dr. COMANOR. It certainly would not be any loss to society.
Senator NELSON. And they would tend to do more pure research to
develop new products and that would be beneficial; would it not?
Dr. COMANOR. I think that is right. At the same time-
Senator NELSON. What I am really getting at is that we have had
testimony that much of the research is directed solely at product dif-
ferentiation, with the production of a product that does the same thing
another drug does, and in some ins~.ances not even as good a job as
another drug. I understand that the argument the industry will
make is that as a result of this kind of research and molecular manipu-
lation, we sometimes get an improved product. You recognize that,
too.
But if it did eliminate just purely duplicative research, the result
of which is to produce just another product doing the same thing as
~ drug already on the market, then it would seem to me that it would
be beneficial; would it not?
Dr. COMANOR. Yes; that is correct.:
PAGENO="0516"
2066 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
With compulsory licensing, it seems that the incentive to under-
take purely duplicative research would be clearly much less. At the
same time, the incentive to undertake expensive basic research projects
would also be much less. Basic research work is likely to be much
more costly than applied work, as well as more risky, and if the
returns from this work would be lessened through compulsory licens-
ing, the industry would probably carry out much less pure funda-
mental research than it does now.
Senator ~%ELSON. I do not follow the argument, because if they
produced a new product that. is useful, then every doctor who has a
use for that type of drug is going to prescribe it. They then could make
their profit. over some period of time and when they are forced to
license other firms they would receive a royalty from what is sold all
over the world.
Why is that not an incentive?
Dr. COMAXOR. That is an incentive, but the important question is
not simply whether an incentive exists but how much. And clearly
with compulsory licensing, the prospective gains would be less than
they are currently, because currently, they can charge high prices
throughout. With lower prospective gains, we might expect to find
less research undertaken of a very costly nature, and basic research is
probably the most costly type of work which is undertaken.
At the same time, it seems to me that with regard to basic research,
it is difficult to argue that private firms have a comparative advantage
relative to Government, foundations, and universities. Most basic re-
search in the medical area is performed outside of the indust.ry now,
imd even if the industry carried on no basic research, it is not clear
that this charge would involve heavy social costs. Basic research, as
many commentators have noted, depends on an atmosphere of freedom
of inquiry, which may not be best achieved within the confines of a
profitmaking organization.
It seems to me, then, that the comparative advantages of basic re-
search probably lies in universities and Government laboratories.
While the magnitude of industry basic research would decline, this is
precisely an area where the industry does not seem to have a com-
parative advantage. At the same time, there would still be an incentive
for the firm to differentiate its products from those of its rivals and
to create entry barriers. Furthermore, it would continue to earn high
returns during the first 3 years, and receive patent royalties thereafter.
The industry would be likely to emphasize the area where imme-
diate payoff is greatest. They would emphasize modifications which
represent relatively small advancements in medical practice.
They would also, it seems to me, be expected to emphasize the devel-
opment. and testing of new drugs which were originally discovered
outside of industry laboratories. Before a new drug is introduced, not
only must the basic discovery be made but also considerable scientific
and technical resources are required to test it under many different
conditions. This is precisely the work in which the industry appears
to have a comparative advantage. This is the sort of work in which
the payoff specifically from industry research is likely to be greatest,
and we might expect to see it continued even with compulsory
licensing.
PAGENO="0517"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2067
As you note, high gains would be obtained during the original
3-year period, and in addition, of course, gains from patent royalties
would remain.
There are important areas where drug companies probably do have
a comparative advantage in terms of the type or character of research,
but also it seems likely that this type of research effort would continue
even with compulsory licensing. Therefore, I would certainly sub-
scribe to the view that compulsory licensing, after a period of time
at reasonable royalty rates, is a desirable policy change and should
be enacted.
While it is true that the magnitude of research and development in
the pharmaceutical industry would probably decline, it seems likely
that projects of limited medical value and lowest industry compara-
tive advantage would be eliminated first.
Senator NELSON. You are talking about pure research as contrasted,
say, with applied research.
The industry does not do very much pure research now, does it?
Dr. COMANOR. There is, of course, considerable controversy over
precisely how much pure research drug firms currently undertake.
Senator NELSON. It is product-directed; is it not?
Dr. COMANOR. It certainly is.
Senator NELSON. Is it not true, by most definitions, that there is
not very much pure research done by industry as a whole in this
country as compared to applied research?
Dr. COMANOR. I think the correct statistic, throughout the entire
economy is something on the order of 10 percent, but I would have to
check to be completely accurate. But however much basic research is
done now, I would expect that less would be done with compulsory
licensing.
At the same time, I am not unhappy with that result, because as I
have noted, this is not an area where the industry appears to enjoy a
comparative advantage relative to nonindustry research.
Therefore, I would suspect that with compulsory licensing, the mag-
nitude of pure research would decline., but research with the highest
industry comparative advantage would continue. I think it iiiappro-
priate to argue that the more research the. better in this type of profit-
oriented context, and it may well be that the present research effort
carried out in the industry is sufficient or even exceeds that which is
needed for its major research functions and responsibilities;
Senator NELSON. What percentage of sales does the research effort
represent now in the industry?
Dr. COMANOR. Different sources provide different estimates. From
the statistics I have seen, it is on the order of 6 to 8 percent.
Senator NELSON. Of sales?
Dr. COMANOR. Of sales.
Therefore, the lower research which would result from compulsory
licensing seems to me to be a small price to pay for a more competitive
determination of prices and profits in this industry.
Mr. GORDON. Dr. Comanor, the relevant market of the drug indus-
try, as I understand it, is not the industry as a whole but each mdi-
vidual, therapeutic category. Would you say that is correct?
PAGENO="0518"
2068 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. COMANOR. When economists define the limits of a market, their
aim is to determine the area, whether in terms of geographic limits or
product class limits, in which prices are set, and this generally depends
on the area in which different products may be considered as highly
substitutable. And if we examine the question of substitutability in the
pharmaceutical industry, it is clear that a diuretic, which is used for
one type of ailment, cannot be substituted for an antibiotic, which has
a very different use. Given the very low degree of substitutability
among therapeutic classes, I think most economists would agree that
the relevant markets here, which define the limits within which prices
are determined, is the therapeutic class rather than the industry as a
whole.
Mr. GORDON. What do you have to say about the concentration of
output in each of these therapeutic categories?
Dr. COMANOR. If one examines therapeutic markets, the concentra-
tion ratios are very high. This is precisely because effective entry bar-
riers have been created. If advertising and research had not been able
to create effective product differentiation, we would expect high prices
to attract new firms into the market. That this has not occurred is due
to the factors which I have discussed.
Mr. Go~oN. You have stated that price competition is virtually
absent from this industry.
What do you have to say about the drive to reduce costs and pro-
duce drugs more efficiently?
Dr. COMANOR. Since production costs account for such a small pro-
portion of total costs in this industry, on the order of a quarter to a
third, there is clearly relatively little to be gained by the firm from
putting a considerable effort i~ producing drugs more efficiently. More-
over, lower production costs would have little competitive effect. With
little to be gained, it is not surprising to find a relatively small effort
devoted to reduce production costs. While I am afraid that no hard
statistical evidence is available on this question, it is likely that the
drive to reduce costs is much less than would be the case if price com-
petition were more effective and more vigorous.
Mr. GolmoN. So actually, we have a situation, do we not, where you
have not only product differentiation but also a highly concentrated
market? What would be the effect of both of these on the possibility
of price competition?
Dr. COMANOR. We observe that there is relatively little price com-
petition among highly differentiated products, although there does
seem to be some degree of price competition among those products
which are not effectively differentiated. For example., we still see effec-
tive price competition for penicillin and streptomycin.
But, as we can tell from the high overall profit rates, these must ac-
count for a relatively small share of the total number of products pro-
clucecl. During the period of the Kefauver committee investigations, it
was reported that-I think this was for 1957-the profit ra.te in the
pharmaceutical industry was double or more than double the average
of all manufacturing.
Mr. Goimox. Even if von did not have product differentiation, the
high concentration itself would tend to bar the possibility of competi-
tion in this industry?
PAGENO="0519"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2069
Dr. COMANOR. That is true., but high concentration and entry bar-
riers are related to one another. The primary reason for high concen-
tration in relevant therapeutic markets is precisely because of the
entry barriers created by product differentiation. So it is difficult to
distinguish the competitive effects of one from the other.
Mr. GORDON. Now, given the type of research conducted by the in-
dustry, and I want to highlight this point which you have already
brought up, would it be fair to say that the industry is heavily de-
pendent on new scientific knowledge developed outside the industry?
Dr. COMANOR. I think it. is clear that much of a research effort in
the industry makes use of new scientific advancements which come
about in nonindustry laboratories. At the same time creating new
drugs out of new scientific knowledge is precisely what might be con-
sidered the major research responsibility for the industry.
So I certainly would not be critical of this fact..
It seems to me this is precisely what society should expect and de-
mand from the industry, that it will utilize the research efforts from
outside, where the majority of medical research is in fact undertaken,
and translate important discoveries into marketable products.
So while much of the industry research effort is devoted to this type
of work, I feel that this is precisely what is desirable or useful from
the point of view of society.
At the same time, this type of work would probably be carried on
even under a regime of compulsory licensing after 3 years, and there-
fore, I think we have much to be gained from adopting this policy.
Senator NELSON. Thank you very~ much, Dr. Comanor. Your testi-
mony was a very useful and valuable contribution to the record. We
appreciate your taking the time to come here and testify today.
Dr. COMANOR. Thank you very much.
(The supplemental information submitted by Dr. Comanor
follows:)
[From "Economica," November 19641
RESEARCH AND COMPETITIVE PRODUCT DIFFERENTIATION IN THE
PHARMACEUTICAL INDUSTRY IN THE UNITED STATES
(By William S. Comanor)
During the past two decades expenditures in the United States on research
and development have expanded greatly. From outlays of the order of a half-
billion dollars at the start of the Second World War, they have grown to well over
$10 billion. The rapid growth of an "industry of discovery"1 has; given rise to a
large number of questions concerned with the rOle and function of industrial
research. Why do firms spend large amounts on research and development? What
is the relationship between research and market structure? What has been the
impact of public policy, in the form of the patent system, on these expenditures?
In dealing with questions of this sort it is necessary to stress the pattern of
relationships among firms within which research and development is undertaken.
In an activity characterized by high degrees of uncertainty, we should not expect
1 This article is based on the author's unpublished Ph.D. dissertation, The Economics
of Research and Development in the Pharmaceutical Industry, Department of Economics,
Harvard University, June 1963. I would like to acknowledge the assistance of J. W. Mark-
ham and R. B. Heflebower, who acted as supervisors of the original study, and also of
E. T. Penrose for a number of helpful suggestions in the writing of this article. The study
was undertaken with the aid of a research grant from the Science and Public Policy Pro-
gramme at Harvard, and I am indebted to the Programme for this support.
2 John T. Dunlop, "Introduction: Problems and Potentials", in John T. Dunlop, ed.,
Automation and Technological Change, p. 2.
PAGENO="0520"
2070 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
firms to invest so heavily in the absence of specific pressures. These pressures
arise most frequently from oligopolistic market structures in which the need for
product differentiation has been recognized and where rivalry has been limited to
variables other than price. This article puts forward an analytical framework
within which these questions may be considered, and also attempts to provide
some answers dealing with the experience of the pharmaceutical industry in the
Tjnited States.
THE GROWTH OF RESEARCH AND DEVELOPMENT
The spectacular growth of the pharmaceutical industry in the postwar period
has been founded upon a number of important medical advances and has been
associated with some radical changes in the structural and behavioral character-
istics of the industry.2 Prior to 1937 and the introduction of the early suipha drugs
the industry was composed largely of long established firms producing relatively
standardized commodities. Unlike as at present, a large proportion of medical
prescriptions were compounded by the pharmacist.4 Barriers to entry were low
and a high degree of competition prevailed.
With the introduction of penicillin and streptomycin during and immediately
after the Second World War, the nature of the industry changed. As neither of
the products was protected by patents,5 the rapid growth of demand resulting
from their introduction was accompanied by the entry of many new suppliers and
by the development of active price competition. The price of a standard form of
penicillin dropped from $20 for 100.000 units in 1943 to 4'h cents in 1050 0; the
price of streptomycin behaved similarly.7 Excess capacity was created, and it may
well be that prices fell to the vicinity of short-run marginal costs.
Following this experience, it was clear to the leading firms that their profits
in the future would depend on the development of more protected market posi-
tions to be achieved through some form of product differentiation. This would
provide the producer with substantial control over the prices of his products as
well as act as a significant barrier to entry into the relevant therapeutic market.
With the trend in the new technology towards the compounding of prescriptions
by the manufacturer rather than the pharmacist, this development seemed all the
more promising. Although increased selling expenditures would be necessary,
these were not likely to be sufficient to establish effective differentiation, for time
industry faced a relatively informed consuming public in the medical profession,
and it would be more difficult to differentiate between products consisting of
identical chemical substances. Only through the extensive introduction of new
products could significant differentiation be achieved. This would enable indi-
vidual firms to emphasize in their selling activities the improved quality of new
drugs. Moreover, if each of the leading firms followed this course of action, coin-
petition in the industry would largely cease to be founded on the sale of stand-
ardized commodities.8
The accelerated growth of pharmaceutical research and development expendi-
tures during the post-war years w-as due largely to the desire to promote the
In this study the pharmaceutical industry is defined to include those firms which pro-
duce ethical drugs, as opposed to proprietaries, and which distribute these products in
dosage-form. Pharmaceuticals are, thereby, marketed and sold only through the medical
profession and require, for the most part, a written medical prescription.
In 1961 prescriptions compounded by pharmacists accounted for only 3.6 per cent of
all prescriptions in the United States.
While a patent was issued to the inventor of streptomycin, Dr. Selman A. Waksman,
it was assigned to the Rutgers Research and Endowment Foundation, and was licensed
on a relatively liberal basis.
0 Federal Trade Commission, Economic Report on Antibiotics Manufacture, 1958, p. 230.
See United States Senate, Subcommittee on Antitrust and Monopoly, Report of tlze
,S'tudy on Administered Prices in time Drug Industry, 87th Congress, 1st Session, 1961,
pp. 81-8.
8 In this context it is instructive to note the statement made by an industry spokesman
in 1950. When considering the industry's future, he declared: "From a profit point of
view . . . the only realistic solution of this problem lies in the development of new and
exclusive antibiotic specialties. This . . . Is an exceedingly costly and vigorous alternative;
nevertheless, it is the avenue of approach being most extensively explored by certain
antibiotic houses today. This is the approach being followed by Pfizer". Statement by
John McKean, President of Chas. Pfizer & Co., quoted in Senate Report, op. cit., pp. 130-31
(italics added).
Between 1951 and 1960, research and development expenditures in the pharmaceutical
Industry grew from $60.4 million to $206.5 million, an increase of over 200 per cent.
(Pharmaceutical Manufacturers' Association, Yearbook, 1961-62, p. 168.) These figures
are measured in current dollars and are derived from surveys of P.M.A. member firms.
PAGENO="0521"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2071
introduction of new products and the accompanying extension of product differ-
entiation. Research laboratories may, in fact, be regarded as a cost of differentia-
tion. Facilities were designed for the discovery, development, and testing of
new drugs, and as a result, new products were introduced at a rapid pace. While
completely new chemical entities were the most desirable type of new product,
since these would provide the highest degree of differentiation, other forms were
also useful for this purpose, and in addition frequently had the advantage of
requiring less research input per product. Among the latter group were new
combinations of drugs already on the market, and new dosage-forms of previously
introduced products. During the decade ending in 1960, over 3,800 new products
and dosage-forn~s were introduced into U.S. pharmaceutical markets, or an aver-
age of more than 380 per year. Of these slightly more than 11 per cent. of the total
were new single chemical entities.1°
THE IMPACT OF COMPETITIVE PRE55IIIIE5
In examining t.he impact of competition on the growth of research, it is useful
to consider the rOles played by certain structural characteristics of the industry.
The most significant of these is the co-existence of a number of fairly large firms
with a sw-arm of smaller ones, which, however, account for a substantial propor-
tion of the total output of the industry. Relevant data are presented in Table 1.
As may be observed, the industry overall is moderately concentrated, with a
significant fringe of smaller firms.1' Some further important characteristics con-
cern the relatively low proportion of variable costs in total costs-in the larger
firms the proportion tends to be less than half `2-and a generally low price
elasticity of demand for pharmaceuticals.13 In a siaution of this sort, not only are
competitive price declines likely to be large but also firms could easily be com-
pelled to price below average total costs. As a result, there would be heavy pres-
sures on profits if rivalry were allowed to take the form of price competition.
TABLE 1.-PHARMACEUTICAL INDUSTRY: CONCENTRATI ON RAIl OS, 1958
Sales
Amount (millions)
Percent
Largest 3 companies
Largest 7 companies
Largest 15 companies
All companies
~384
800
1, 190
1,945
20
41
61
100
Sources: Pharmaceutical Manufacturers' Association, "Domestic Human Ethical Drug Business Survey," conducted
by Arthur Andersen & Co., 1959, mimeographed. Pharmaceutical Manufacturers' Association, "Yearbook, 1961-62,"
p. 170.
The data for the largest 15 firms are derived from a survey of sales in 1958 of 50 pharmaceutical firms, all members
of the Pharmaceutical Manufacturers' Association. Given the nature of the survey, it can be assumed that the 50 com-
panies included the largevt 15 in the industry. The figure for industry sales refers only to P.M.A. membership; it has been
maintained by the association that member firms account for over 95 porcent of total industry output.
The interaction of these two sets of factors has been significant in determining
industry behaviour and performance. Where smaller firms are relatively im
portant, one might expect to find a high degree of price competition. That this
result has not occurred can be attributed to the post-war strategy of research
and product differentiation pursued competitively by the leading firms in the
industry. Rivalry has been restricted largely to areas other than price. A high
degree of price stability on existing products has been maintained, and new prod-
ucts have been priced, for the most part,: to compete with older ones in the same
10 Paul tie Haen, New Product Survey, 1060.
11 The pharmaceutical industry falls easily into Kaysen and Turner's Type II oligopoly.
According to their definition, a Type I oligopoly is one in which the largest eight firms
account for at least 50 per cent of industry output while the largest twenty comprise at
least 75 per cent. A Type II oligopoly, on the other hand, is one in which the largest eight
firms account for at least 33 per cent. of industry output. See Carl Kaysen and Donald
F. Turner, Antitrust Policy, p. 27.
12 In a sample of twenty-two largest pharmaceutical firms, and with regard to 1958
drug operations only, costs of goods were only 43 per cent. of total costs. See Senate Report,
op. cit., p. 31.
`~ See The Economics of Research and Development in the Pharmaceutical Industry,
pp. 58-60. A similar conclusion was reached: in the Senate Report, op. cit., p. 3.
PAGENO="0522"
2072 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
therapeutic class,'4 At the same time, however, a vigorous form of rivalry has
developed, which has taken the form of what Dr. Penrose has termed "competitioox
in creativity".'5 A policy of competitive innovation seems to have been formulated
by the individual firms, and research laboratories have been established as a
vehicle of this policy. Profits have come to depend primarily on the firm's position
in the innovative race, and the cost of an organized and efficient research effort
may be regarded as a necessary expense of entering the race. Once some firms
had taken this position, moreover, others were forced to follow in order to
maintain their market positions relative to their rivals.
In this manner the growth of pharmaceutical research establishments has
been stimulated to a considerable extent by competitive pressures. Not only
has the latent price competition, arising from these pressures, impressed upon
the larger firms the need to achieve effective product differentiation, but also a
highly competitive form of research and development has been promoted. Competi-
tive factors have influenced the direction and scope as well as the size of research
establishments.
PRODUCT COMPETITION WITHIN THERAPEUTIC MARKETS
The expansion of pharmaceutical research facilities and the concomitant in-
crease in the introduction of new products have given rise to a rapid rate of
product obsolescence. New products have competed with older ones, and the
latter have been rapidly replaced by the former. By 19G0 only 31 per cent of
sales accrued from products introduced before 1951.16 As may be seen in Table
2, the sales of products introduced in a given year normally reached their peak
(as a proportion of total sales) in the second year following their introduction;
after that they declined rapidly in relative importance. New products were intro-
duced, reached their peak, and declined; and all within a relatively short period
of time."
TABLE 2.-SALES OF PHARMACEUTICAL PRODUCTS BY AGE OF PRODUCTS
Year of introduction
of products
1951 1952
Percent
of sales
1953 1954 1955
-
1956 1957 1958 1959 1960
1960 4.6
1959 7.1 10.5
1958 4.7 7.2 7.3
1957 10.1 16.8 16.7 15.1
1956 5.8 9.0 8.4 7.7 6.6
1955 9.0 18.7 15.9 11.4 10.2 8.2
1954 7.4 13.7 12.8 10.8 10.2 8.8 8.1
1953 5.4 16.0 14.8 12.1 9.8 7.2 5.8 4.7
1952 4.6 9.5 7.7 6.1 4.8 4.4 4.0 3.1 2.9
1951 1.1 1.7 1.6 1.1 .8 .7 .6 .6 .6 .6
Before 1951 -- - 98.9 93.7 83.5 67.9 55.6 45. 1 39. 5 36. 7 32.8 31.4
Total - 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Source: Arthur D. Little, Inc., "A Report on the Aspects of Concentration and Product Obsolescence in the Pharma-
ceutical Industry in the United States," p. 33.
The table is based on a sample of 33 firms, comprised, for the most part, of the larger firms in the industry. These
firms account for approximately 80 percent of industry output.
14 A similar view was expressed by the Senate Subcommittee, which reported: "When a
new product is put on the market, the customary procedure is to introduce it at or very
near the price charged for an existing drug used to treat the same general type of ailment".
Senate Report, op. cit., p. 98.
"E. T. Penrose. The Theory of the Growth of the Firm, p. 106. "In the United States
a kind of `competition in creativity' has become a dominant motif In the pattern of
competitive behavior in many industries, where consumers and producers alike are caught
up in an almost compulsive obsession for that which is `new', In the extreme case the inch-
vidual firm Is forced constantly to remould its products-to create the `new' and `improved'
either in performance or design. To a large extent the new products are superior in per-
formance: to a considerable extent they are merely new and can be sold only if the
consumer can be convinced that the `newest' is the `best'."
16 Arthur D. Little, Inc., A Repos-t on the Aspects of Concentration and Product 00-
solescence in the Pharmaceutical Industry in the United States, p. 33.
17 Somers and Somers report that the average life-span of a new drug is said to be be-
tween two and five years. Herman 1~I. Somers and Anne R. Somers, Doctors, Patients and
Health Insurance, p. 96.
PAGENO="0523"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2073
Because of the segmented nature of demand, it is misleading to consider the
market for pharmaceuticals as a single entity; the relevant market does not span
the entire industry but, rather, is limited to individual therapeutic classifications.18
And it is within these therapeutic classifications that the rapid rate of product
introduction has developed into vigorous product competition. New products
have replaced older ones, and in this process the ranking of the leading firms
in particular therapeutic markets has changed frequently. In only nine out of
twenty such markets did the same firm persist as the leading firm between 19.51
and 1960; and in most cases the same five firms did not retain completely the five
leading positions.1° While past position tends to be an advantage in determining
present market shares, it is not of crucial significance, and the measure of market
control derived from a single product appears to be limited to a relatively short
period of time.21
NEW PRODUCTS OR NEW PROCESSES
Research laboratories in the pharmaceutical industry have been concerned
largely with the introduction of new products rather than with the development
of new processes n for old products.22 Given the nature of competition, this is what
we should expect. Where product rivalry is high and the effective life of in-
dividual products is correspondingly short, firms are unwilling to invest large
amounts towards reducing production costs. By the time new techniques have
been developed, it is quite likely that demand for the product concerned will
have dropped to a relatively low level. Even with regard to products which have
fairly long lives, moreover, or to processes used to produce many products, it is
unlikely that a large research effort is undertaken in the direction of reducing
production costs. To the extent that rivalry is based on price, the introduction
of new processes will place' the firm at a' competitive advantage by enabling it
to price below its rivals without reducing its profit margins. When, however,
price behaviour is largely non-competitive, a reduction of costs will increase
profit margins without immediately affecting the firm's position relative to those
of its rivals. While this may enable the firm to undertake large'r selling or re-
search expenditures, or to' raise funds more advantageously in the capital mar-
ket, the competitive effect will not be immediate or direct.
The introduction of new products affects the competitive position of the firm
in a quite different manner. To the extent that rivalry takes the form of coin-
petition between products which are priced at the same or similar levels, the
number and character of new products which are introduced directly affect the
demand for the firm's output. If the firm introduces new products which do the
job "better", then its output and total profits may be higher even if the cos'ts
of the new products are greater than those of their predecessors and profit mar-
gins correspondingly reduced. The firm that falls behind in the race to intro-
18 Relatively high cross-elasticities of demand will be found when competing products
treat the same or similar ailments or have similar therapeutic effects. Between therapeutic
classifications, on the other hand, these elasticities will approach zero. While cases will
remain in which a specific product is required to treat a certain condition, the gap in the
chain of substitutes will appear at the boundaries of therapeutic classes. Although high
cross-elasticities of supply will exist across these boundaries, we still find it appropriate
for our purposes to define the relevant markets in terms of therapeutic classifications.
19 Data describing the market shares of the leading five firms within twenty them-
peutic markets for the years 1951 and L960 are presented in Arthur D. Little, Inc., op. cit.,
pp. 11-30.
21 This conclusion is corroborated by McKie who reports the following: "In all cases
it is clear that the period of `dominance' of any one product is short-four or five years
at most-and that a firm which fails to bring out improvements or new substitutes will
find its share of the market rapidly passing to others". James W. McKie, "An Economic
Analysis of the Position of American Home' Products Corporation in the Ethical Drug
Industry", in United States Senate, Administered Prices in the Drug Industry, Hearings
before the Subcommittee on Antitrust and Monopoly, 86th Congress, 2nd Session, 1960,
Part 17, p. 9957. The latter document will be cited as Administered Price Hearings.
21 `The distinction between products and processes cannot be based on. characteristics
of the good in question. Where intermediate products are concerned, a product for one'
firm may be a process for another. The distinction is actually relevant only within
the firm. A product is a good which is offered for sale in the normal course of business.
A process, on the other hand, is only infrequently offered for sale, and is used within the
firm to facilitate production of other commodities.
22 This emphasis was sufficiently great for a survey of resea.iich budgets of pharmaceutical
firms to allocate total outlays among four.' categories without providing one for new
process research. The categories were: new products, product improvements, new appli-
cations, and basic research. See Chemical and Engineering News, March 17, 1958, p. 52.
PAGENO="0524"
2074 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
duce new products may find its demand and profits lower than its rivals', even
if it should succeed in reducing the costs of producing its older products. This
will be the case especially if the price elasticities of demand for it~ products
are relatively low.
This explanation involves the question of profit maximization under con-
ditions of oligopoly. Where profit rates are minimal, the pressures to reduce
costs are associated with staying in business. Where profit rates are higher,
however, the pressures to reduce costs via the introduction of new techniques
are likely to be substantially less. But the pressures associated with the intro-
duction of new- products may be similar to those existing in the competitive model.
Here the introduction of competitive new products by a firm's rivals may result
in a large, and in some cases fatal, decline in the demand for the firm's output.
While the conventional theory of profit maxhuization requires that the "car-
rot" be as effective as the "stick" in eliciting certain kinds of behaviour, it ap-
pears likely that the prevention of declines of profit is more important than the
making of gains, that the maintenance of existing market shares through the
introduction of new products is more important than the reduction of cost,
and that a firm n-ill work harder for the former purpose than for the latter.
This means, merely, that in an uncertain world firms operate under some kind
of minimax strategy. On this basis, firms will emphasize new products rather
than new processes in their research efforts.
THE ROLE OF THE PATENT SYSTEM
United States law provides that patents on pharmaceuticals may be granted
for new products as well as for new processes. Process patents, however, are a
considerably weaker form of protection. As in most chemical industries, it is
comparatively easy to modify a process somewhat and thereby evade the patent.
Product patents, on the other band, are a relatively strong vehicle of protection
from competive suppliers. Once a product patent has been granted, a rival firm
cannot supply the identical compound without fear of legal proceedings. As a
result, nearly 80 per cent of medical patents are issued for new products rather
for new processes.23 Patent protection in this industry is, moreover, especially sig-
nificant since a large proportion of individual products are covered, A survey
of the industry estimated that over two-thirds of all prescription sales are for
patented drugs.2~
Although patent protection in the pharmaceutical industry takes the form
primarily of product patents. it does tend to be limited to specific chemical sub-
stances. In most cases patents cover only a single compound or a small number of
compounds. Especially when the advancement in knowledge is small, the scope
of the patent may be rather limited. In addition, there is the question of patent
specifications and the problem of anticipating all possible variants of the product.n
With regard to pharmaceuticals, it is frequently possi,ble to invent around existing
patents; to find a variant which has not been specified, obtain a patent for it, and
introduce it as a competing product.
Since a large proportion of pharmaceuticals have some degree of patent pro-
tection, entry into a specific therapeutic market requires, in most cases, some form
of scientific or chemical product differentiation. In a world of competing monopo-
lists, rivalry requires the ability to acquire a monoply position. The importance
of pharmaceutical patents, however, can easily be overstated. While monopoly
positions are conferred, patent protection does not normally confer the power to
monopolize any of the therapeutic markets. This is borne out by the high turn-
over among leading firms and the vigorous product competition within these
markets. The impact of the patent system has not been to create monopoly posi-
tions which remain active throughout the seventeen-year life of the patent, but
rather it has been to foreclose to a great extent rivalry between identical chemi-
in 1957 the industry's rate of return on investment after taxes encuallecl 21.4 per cent,
which placed it second on the F.T.C. list of thirty-nine industries. The comparable rate
reported for All Manufacturing was 11.0 per cent. See Federal Trade Commission, Report
on Rates of Return for Identical Companies in ~Selected Manufacturing Industries, lDJiO,
1947-1957, pp. 34-49. and Senate Report, op. cit., pp. 53-55.
24 In a sample of the twenty-two major pharmaceutical firms, for 1958 drug operations
only, selling expenditures reached nearly 25 per cent of total sales. Research expenditures,
on the other hand, equalled only slightly more than 6 per cent. Senate Report, op. cit., p. 31.
Arthur P. Little. Inc.. op. cit. pp. li-so. While these ra~os are probably over-stated,
the extent of the over-statement is not likely to be by much more than one-fifth.
PAGENO="0525"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2075
cal entities or standardized commodities about which price competition might
develop. It has strengthened and encouraged the high degree of chemical product
differentiation which is the primary form taken by technical change within the
pharmaceutical industry.
THE ACHIEVEMENT OF MARKET POWER
The growth of effective product differentiation has led to an appreciable in-
crease in industry profits and their maintenance at a relatively high level. In
fact, the pharmaceutical industry has `become one of the most profitable in the
American economy.26 The high level of market power, of which these profit rates
are indicative, has evolved as part of the' pattern of inter-relationship among
the major firms, and is founded upon the aChievement of product differentiation.
The crucial significance of product differentiation is that it provides the pri-
mary barrier to entry into the relevant therapeutic markets. Since effective entry
normally requires some form of technical advance, the cost and risk of research
coniprise an important part of this barrier. Joined with research and develop-
ruent, moreover, `are the extremely high selling expenditures undertaken by the
larger firms.27 Not only do these outlays accentuate the degree of differentiation
among older products, but they also raise considerably the costs associated with
launching a new product, and thereby provide a further barrier.
Entry barriers, created in this fashion, have resulted in fairly high levels of
concentration within therapeutic markets. In a group of twenty such markets,
the proportion of output accounted for by ~he leading five firms ranged from 56
per cent to 98 per cent.28 It is within these markets that decisions on prices are
made, and given such concentration ratios, `we should not expect individual firms
to disregard their own impact on market parameters. It is on this basis that
market power has ~ieen achieved.
In this manner the leading five firms in the industry have maintained control
of a large number of marke'ts, and have' created conditions within which the
rivalry among themselves will proceed on a non-price basis. What is equally
important, the achievement of product differentiation has succeeded in impeding
the entry into these markets of the large number of smaller firms which would be
likely to introduce some measure of price:' competition. Smaller firms have been
forced to rely largely on standardized and non-patented products which are
frequently non-competitive with the highly: differentiated products which lead in
most markets.
To consider further the role played by differentiation, it is instructive to
examine the behaviour of the industry in p'atent licensing. Although pre-emption
of the entire demand for a product seems most desirable for the firm, there are
a number of factors which have increased the scope of licensing agreements.
Not only may smaller firms wish to profit from the dis:tribution facilities of their
larger rivals,22 lint cross-licensing agreements may also be required to produce
drug combinations and to settle "interference" proceedings.2° In all of these cases,
however, there appears to be a definite reluctance on the part of the major firms
in the industry to license their smaller rivals even when licenses are granted to
23 See the table compiled by the United States Patent Office of all patents relating to
medicine issued during 1901. This table appears in United: States Senate, Drug Industry
Antitrust Act, Hearings before the Subcommittee on Antitrust and Monopoly, 87th Con-
gress, 1st Session, 1961, Part 3, p. 1261. This document will be cited as Antitrust Act
Hearings.
2~1bid., Part 5, p. 2621.
2S See the statement by George E. Frost, a member of the patent bar, in Antitrust Act
Hearings, Part 4, p. 2119. Frost maintained with regard specifically to pharmaceutical
patents that "the patent applicaat is rarely able to anticipate and include all variants
of his inventive concept in this document".
~ The classic example is the case of meprobamate. The patent here is assigned to
Carter Products, Inc., which markets the product under the name of Miltown. Within
a month after the issue of the original patent,' Carter licensed American Home Products
Corp., one of the industry's largest firms with extensive selling and distribution facilities,
to market the product under its own name, Equanil. Although sales of the latter quickly
exceeded their own, Carter benefited through extensive royalties from the expansion of
demand stimulated by the larger firm.
~° An "interference" is declared by the Patent Office when a number of patent applications
lay claim to the same invention. While the normal procedures in this case involve the
Patent Officein an administrative hearing to determine the true inventor, the private
settlement of' claims is widely used in the pharmaceutical industry. This commonly re-
sults in the withdrawal of all but one of the' original applications and the licensing of
all parties'when the patent is finally issued. See Senate Report, op. cit., pp. 152-54.
PAGENO="0526"
2076 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
other large firms. The motive is likely to have been the fear of instigating some
form of price competition.
THE DIRECTION OF RESEARCH
The pharmaceutical industry during the post-war period has undertaken a
heavy research effort which has been founded no a number of factors. The most
important of these was an industrial structure characterized by both competitive
and oligopolistic elements. The fact of vigorous price competition on the ~asis of
standardized commodities led to a search for an effective method of achieving
product differentiation. In addition, the development of a new technology, based
originally on scientific advances made outside the industry, favoured the move
towards a chemical form of differentiation. And finally, the existence of a strong
patent system increased substantially the degree of differentiation which could
be gained through the introduction of new products, and thereby strengthened
considerably the incentive for research. These factors, however, have not only
determined the extent of research in the pharmaceutical industry, but also they
have influenced the type of work which is pursued and the framework within
which it is undertaken.
Pharmaceutical research is conducted within a context of strong product
rivalry. As a result a substantial amount of work is done which is designed pri-
marily to invent around existing patents. While a number of important improve-
ments have been made in this fashion, this work has led also to the development
of a great many products which have pharmacological effects that are quite
similar to those of drugs already on the market.3' Moreover, a good deal of work
has gone into developing new combinations of existing drugs and new dosage-
forms. While research of this type is unlikely to result in major scientific ad-
vances, and frequently duplicates what has already been done, the minor advances
and improvements w-hich do arise are, nevertheless, important elements in the
process of developing a new technology, and are essential if its full benefits are
to be realized through the inevitable plethora of modifications and variations.
Product rivalry and the drive to strengthen product differentiation have led
also in a second direction, w-hich is concerned with the process by which new
scientific knowledge is translated into marketable products. Competitive pres-
sures have forced research directors in the industry to pay constant attention
to new knowledge arising from external sources and have ensured that promising
leads w-ill rapidly be acted upon. This form of product development is vitally
important because the major share of new scientific knowledge is likely to be
produced in laboratories outside the industry. Not only do the research expendi-
tures of the industry account for only about 30 per cent of total resources devoted
to medical research within the United States,'2 but also considerable scientific
activity is pursued abroad.
Even where the basic discoveries are made elsewhere, however, a good deal
of work remains to be done." The first problem concerns isolating or otherwise
obtaining compounds which may have promising properties. It has been esti-
mated. in fact. that approximately one-half of total pharmaceutical research
and development expenditures is used to support the w-ork of synthesizing,
purifying, modifying and preparing suitable substances for subsequent physio-
logical tests.'4 Moreover, quite extensive biological and clinical testing is required
3' With regard to products in the latter group, the term "molecule manipulation" is
commonly used because of the similarity in molecular structures which is frequently evi-
dent. An example of this feature of pharmaceutical research is provided by the case of
three competing antibiotics: tetracycline, aureomycin, and terramycin. Although there
may be some differences in their medical activity, these are not likely to be great, and the
products may be considered as substitutes. Their molecular structures, although highly
complex. are all rather similar. These structures are presented in Federal Trade Commission,
Economic Report on Antibiotics Manufacture, 1958, p. 249.
3' In 1960 the pharmaceutical industry allocated $215 million to research and devel-
opment out of total national expenditures of $715 million. Government provided $400
million while other private support amounted to S100 million. These figures refer to the
U.S. fiscal year 1960. Antitrust Act Hearings, Part 3, p. 1705.
3' Despite the original discovery of penicillin by Alexander Fleming in 1928, thorough
testing of the substance did not begin until 1939, and the development work, even under
the exigencies of war. was not accomplished until 1943. It was only after the Second World
War that penicillin became available on a significant scale to the civilian population. For
a survey of the early discovery and development of antibiotics, see Federal Trade Com-
mission, op. cit., pp. 34-45.
3' Statement by Austin Smith. President of the Pharmaceutical Manufacturers' Associ-
ation, in Administered Price Hearings, Part 19, p. 10725.
PAGENO="0527"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2077
before a new drug may be introduced. Research and development activities of
this sort are extremely costly, although there is a reasonable degree of assur-
anice of commercial viability. They demand extensive resources devoted to scien-
tific efforts which, however, may be fairly routine. Thus, a major share of phar-
maceutical research should be considered as complementary, rather than competi-
tive, to research activities outside the industry.
TABLE 3.-NUMBER AND PERCENT DISTRIBUTION OF SCIENTIFIC AND PROFESSIONAL MANPOWER
IN MEDICAL AND RELATED RESEARCH, 1960
A. DISTRIBUTION BY SECTOR
ENGAGED
Levelof training
Total M.D. and Ph. D.
Other
D.D.S.
Number 39,700 11,400 18,000
Percent:
10,300
Government 19. 7 17. 3 12. 6
34. 5
Industry 18.1 4.8 12.2
Universities and research institutes 62. 2 77. 9 75. 2
43.2
22. 3
Total 100.0 100.0 100.0
100.0
B. DISTRIBUTION BY LEVEL OF TRAINING
Percent
Number
Total M.D. and Ph. D.
Other
D.D.S.
Gnvernment 7, 800 100. 0 25. 3 29. 2
Industry 7,200 100.0 7.6 30.6
Universities and research institutes 24, 700 100. 0 35. 9 54. 8
45. 5
61.8
9. 3
Total 39, 700 100. 0 28. 7 45. 3
26. 0
Source: National Institutes of Health, `Manpower for Medical Research-Requirements and Resources, 1965-70,"
a report for the Committee on Appropriations, U.S. House of Representatives, February 1962, p. 24.
To consider this poin.t further, it is useful to examine the data on educational
levels of professional personnel engaged in medical research within and outside
the pharmaceutical industry. The relevant information is presented in Table 3.
As may be observed, 38 per cent of research professionals in the industry held
doctoral degrees while 55 per cent of government researchers and 91 per cent of
professionals within universities and research institutes had attained this level.
Moreover, although the pharmaceutical industry provided approximately 30 per
cent of total expenditures on medical research in 1960, it utilized only 5 per cent
of total personnel with the M.D. degree and 12 percent of those with the Ph. D.
degree. It appears, thus, that industry resources competed with non-industry
resources for scarce, highly trained personnel to a lesser extent than would be in-
dicated by the total size of the industry effort, and also that these resources
w'ere used primarily for alternative, non-competitive purposes.
The growth of competitive product differentiation has been associated w-ith sub-
stantial outlays on research and development; but equally significantly, it has
been associated with a specific emphasis in its direction. Research is a generic
term which covers a broad spectrum of activities, and thus the type or character
of the research activities undertaken as well as their volume are important fac-
tors in an analysis of the relationships among mnaricet structure, research and
technical change.
PAGENO="0528"
2078 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
[From The Review of Economics and Statistics, Vol. XLVII, No. 2, May 1965]
RESEARCH AND TECHNICAL CHANGE IN THE PHARMACEUTICAL INDUsTRY:
(William S. Comanor)
In recent years there has been a good deal of discussion concerning the relation-
ships among market structure, research and development, and the rate of tech-
nical change. Much of this discussion has focused on the question of whether
large firm size is a necessary condition before firms w-ill engage in research, and
whether research and development (R and D) is likely to grow more or less thali
in proportion to increases in firm size. A further set of questions deals with the
relationship between research and the rate of technical change experienced by
the firm. Can variation in the latter be explained largely by differences among
firms in the size and character of their research programs? Are economies of
scale in R and D likely to be present? What is the effect of firm size on the pro-
ductivity of a research establishment? This paper provides an empirical analysis,
concerned with these questions, of the experience of the United States pharma-
ceutical industry during the period between 1955 and 1960.
SOME MEASURES OF RESEARCh AND TECHNICAL CHANGE
The empirical investigation in this paper utilizes multiple regression techniques
on a cross-sectional basis at the firm level. It examines the impact on technical
change of a number of variables associated w-ith the character of a firm's research
and development effort. At this point, we will define the measures used to repre-
sent technical change and R and D, and describe the sample upon w-hich the
analysis is based.
Research and development establishments in the pharmaceutical industry 1 are
concerned to a very large extent with the introduction of new- products, and only
meagerly with the development of new processes.1 This is due to the fact that
the primary mothating force behind R and D outlays in this industry lies speci-
fically in the attempt to achieve scientific or chemical product differentation.3
ucts at reasonable levels of cost, these are properly included under new product
research, for we are concerned here not merely w-ith the discover or invention of
the product, but with the total R and D outlays, asseciated with its introduc-
tion. As a result, research output, in this paper, is measured entirely in terms of
new proJucts and we assume that the reesarch effort designed to produce new
processes is small and can be safely ignored.
A concern with research output in terms of new products requires considera-
tion of all products introduced w-hich are new to the firm regardless of whether
they have been introduced previously by competitors. This includes those prod-
ucts which are innovations or imitations and those which lie in an intermediate
position, embodying varying degrees of differentiation relative to products already
on the market. This last category, which in the pharmaceutical industry in-
cludes the great majority of new products, covers the entire spectrum of which
pure innovations and pure imitations are the extreme values. Our measures of
research output are, thus designed to represent the rate of new product techni-
cal change experienced by the firm.4
*This paper is based on the author's unpublished Ph.D. dissertation: "The Economics
of Research and Development in the Pharmaceutical Industry," (Department of Economics,
Harvard University, June 1963). I should like to express my appreciation to R. B. Hefie-
bower, J. W. Markham, and L. D. Taylor for their counsel and advice during the course
of my work on this project. I am also indebted to R. E. Caves, P. Dhrymes, W. E. Gustafson.
J. Schmookler, and T. A. Wilson for a number of helpful suggestions, and to P. dellaen
and H. A. Gosselin who generously provided the data used in this paper. The study was
supported financially by the Science and Public Policy program at Harvard. and the coin-
putations were carried out under grants of subsidized time on I.B.M. 7090 computers at
the Harvard Computation Center and at the M.I.T. Computation Center.
1 In this study, we shall distinguish ethical drugs from proprietaries, and shall consider
the pharmaceutical industry in terms of the former. Pharmaceuticals are marketed and
sold only through the medical professions, and require, for the most I)art, a written
medical prescription.
So great is this emphasis that a 1958 trade journal survey of research bucl~ets of
pharmaceutical firms divided expenditures into four categories without providing a cate-
gory for new process research. See Chemical and Engineering NCws (March 17, 1958). 52.
Although funds are expended to develop techniques for mass producing new prod-
1 See W. S. Comanor, op. cit., 65-85.
Technical change is considered to encompass Imitation as well as innovation and
thereby deals with the entire process by which new technology is placed into actual
practice.
PAGENO="0529"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2079
New pharmaceutical products may be divided into a number of classes con-
cerned with the innovative character of a new drug. In the first category are
the single chemical entities which have not been introduced previously into
United States markets. While these are largely innovative, they do include
salts or derivatives of older products. Other classes deal with new combinations of
active ingredients and new dosage forms of previously introduced products. These
groups contain imitative as well as innovative characteristics. At the same
time, many new products are purely imitative and merely duplicate products
already on the market. In this study, we shall measure technical change by
means of two variables. The first of these deals only with the introduction of new
chemical entities, while the second includes a large imitative component and
encompasses new products of all types.
New products vary not only in terms of the degree of innovative character
which may be associated with them, but also in terms of their economic impact.
To deal with this factor, we have weighed each new product on the basis of
its sales during the first two calendar years following introduction.5 We shall
designate Y1 to equal total sales, in their first two years, of all new chemical en-
tities introduced by the firm during the period between 1955 and 1960, and
Y2 to equal a comparable sum when "new products" assume the broader
definition.6
The scale of research and development has also been measured in two ways,
both dealing with the number of persons employed in research establishments
The first variable is the average number of professional B and D personnel
employed in 1955 and 1960. The second deals with the average number of total
personnel employed in R and D facilities at these two points in time. This latter
value equals the sum of professioal and supporting personnel.
The sample used in this study includes 57 pharmaceutical firms,8 and accounts
for nearly 80% of total pharmaceutical prescription and hospital sales during the
period between 1955 and 1900. Moreover, the sample is not dominated entirely
by large companies, but rather covers the entire range of the size distribution
of pharmaceutical firms.9 The distribution of the sample is J-shaped with one-
Sales data by product were obtained from a marketing research firm, R. A. Gosselin &
Co., Inc., which supplied information for the years 1955 through 1961. These figures are
derived from a representative sample of the nation's drug stores and hospitals. One short-
coming of these figures is that they do not represent total output but only that portion
which is sold to consumers via prescriptions and to hospitals, and thereby do not include
sales to government, corporations, or those made in physicians' offices. Prescription and
hospital sales, however, account for approximately 80% of total pharmaceutical sales.
See Lucy Kramer, "Drugs and Medicines" in Public Health Reports (Oct. 1938). 932.
We are forced to assume that our conclusions would not be substantially altered if the
remaining portion were included. Information on new products was obtained from Paul
deHaen, a consultant to the pharmaceutical industry. In addition to providing informa-
tion on year of introduction, these surveys noted the therapeutic classification of each
new product and whether the product at introduction was a new chemical entity, duplicate,
new compounded product, or new dosage form. For a more detailed discussion of the
value and limitations of these data, see W. S. Comanor, op. cit., 99-106.
6 The data on pharmaceutical sales, used in the creation of these variables, have not
been deflated for price changes but rather have been expressed in current dollars. This
step was taken because of the high degree of price stability among pharmaceuticals during
the period under consideration. See the price index computed by John M. Firestone in
United States House of Representatives, Drug Industry Antitrust Act, Hearings before
the Antitrust Subcommittee of the Committee on the Judiciary, 87th Congress, 2nd
Session (1962), 608.
Information on this subject for the years 1955 and 1960 has been published in Na-
tional Academy of Sciences-National Research Council, Industrial Research Laboratories
in the United $tates, tenth and eleventh editions. These surveys distinguish between pro-
fessional and supporting staff within R and D establishments. In addition, for those
companies in which the research effort was divided among a number of separate labora-
tories, the size and direction of research of each laboratory were, in most cases, listed
individually. Since in many large diversified firms, the pharmaceutical R and D labora-
tories were distinguished from other research facilities, it was possible to limit our data
for these firms to pharmaceutical research personnel. In the inevitable instances when
we came across laboratories in which work was done on non-pharmaceutical as well as
pharmaceutical projects, the facilities were included when it appeared from the sum-
maries provided that the major part of the effort was concerned with pharmaceutical
endeavors.
8 The sample was chosen on the basis of data availability but with the condition that
no substantial merger took place with other pharmaceutical firms between 1955 and
1960. Exceptions were macla for those mergers which occurred during 1960 and in which
the 1960 R and D survey reported separate and distinct research facilities. It was as-
sumed, in these cases, that the pre-merger firms behaved and acted independently through-
out the period.
Size, in this study. is measured by the mean value of annual prescription and hospital
sales betweea 1955 and 1960.
81-2S0-68-pt. 5-34
PAGENO="0530"
2080 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
third of the observations falling in the smallest size class.
This sample is essentially non-random, and strictly speaking, our empirical con-
clusions apply only to it and not to any larger population. Since, however, the
sample constitutes such a large proportion of industry output, and has a size
distribution similar to one which might be expected from the entire industry, we
may presume that conclusions based on an analysis of this group of firms prob-
ably can be applied to the entire industry.
THE MAJOR EMPIRICAL FINDING5
In the analysis which follows, a quadratic term for research and development
is introduced as well as a variable designed to represent the interaction between
size of firm and scale of research establishment.10 Some hypotheses concerned
with the question of economies of scale in R and D distinguish between the
productivities at a given level of research activity according to firm size. For
this reason, a variable was computed which equated the product of scale of
research and size of firm. Since R and D is measured on two bases, there are
also two measures of the interaction between research and firm size. We may
designate I~ as the interaction variable where R and D is the number of pro-
fessional, and 12 where the number of total personnel is used. In addition, we
include firm size as an explanatory variable and also a measure of output di-
versification.11 The latter step enables us to test the relationship between this
factor and the productivity of research. In the regression analysis, the measures
of technical change and R and D are deflated by size of flrm.'~
As may be seen in table 1, the model fits the data far better when research is
measured by professional rather than total R and D personnel. While we shall
have more to say later about the question of supporting personnel, it appears
that technical change is primarily associated w-ith the number of professional
investigators.
It may also be observed that diversification is negatively associated with our
measures of technical change. When new products are defined in terms of new
chemical entities, the coefficients are significant at the 99% level. With the broader
definition of new products, however, the coefficients remain negative although
there is some doubt as to their statistical significance.13 It thus appears that
diversification is more closely related to the introduction of new entities than to
new products in general. In addition, the negative dimension of the parameters
suggests that for given research and development, higher rates of technical
change will be achieved if attention is concentrated towards a few product areas.
To the extent that diversification of output denotes the scope of research ac-
tivities, it may be that inefficiencies result from R and D undertakings w-hich
are "spread too thin" and that in the context of pharmaceutical research, it is
better to work exhaustively with a limited number of problems.14
A further point is that Y1 appears more closely associated with research and
development than Y2. We would expect this to be the case because F1 deals only
with new chemical entities w-hose introduction is likely to have a relatively high
degree of research input. Since F2 includes new chemical entities as well as other
new products, the correlations observed with this variable may denote largely
the influence of new entities.
~° I am grateful to Lester D. Taylor for originally suggesting the use of an interaction
variable.
11 Diversification, in this analysis, deals only with the division of output among the
various pharmaceutical markets and not with the division between pharmaceutical and
non-pharmaceutical markets. On the basis of apparent medical usage, 40 therapeutic
markets are defined. Our first measure of diversification, Di, is the number of markets
serviced b-i- the firm which account for at least 2% of total pharmaceutical sales. The
second measure, D2, deals with the proportion of sales outside the firm's primary market,
and equals one minus the ratio of sales in the firm's largest market to total pharmaceutical
sales. The third measure, D3, is the composite of the two previous ones. It is defined as
the mean value of the products of Di and D2, calculated on an annual basis. These meas-
ures are taken from Michael Gort. Diversification and Integration in American Industry,
8-li and 23-24. In these regression equations, D3 Is the variable which is introduced.
12 This step is taken to increase the likelihood that the assumption of homosceclasticity
is satisfied.
ii We should note that Di and D2 were introduced into the analysis and provided quite
similar results to those obtained from D3. See W. S. Comanor, op. cit., Table XXXI, p. 140.
14 An alternative explanation of the negative sign of this coefficient concerns the possi-
bility that there may be selling efficiencies which result from concentrating sales in a small
number of therapeutic markets.
PAGENO="0531"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2081
The relationship between research effort and technical change appears to be
curvilinear and the quadratic term is always highly significant. The linear
term, however, is not significantly different from zero and has an alternating sign
when Y2 is used to measure technical change, but is significantly negative when
11 is used. The existence of a negative cofficient for research and development
effort in these cases does not result from the nature of the research process and
the experience of the pharmaceutical industry, but rather is due to some statisti-
cal problems associated with the data concerned with the introduction of new
chemical entities. These problems shall be discussed in the following section.
TABLE 1.-Regression and correlation analysis-Summary findings
A. RESEARCH AND DEVELOPMENT MEASURED BY PROFESSIONAL RESEARCH
PERSONNEL
.
Intercept
RD
RD2
Size
Ii
D
~R2
(1)
(2)
y1
~
b 0.422
(0. 136)
b 0. 873
(0. 222)
b -4.671
(1. 285)
-0. 060
(2. 098)
b 0.547
(0. 107)
b 0. 557
(0. 175)
b 0.0000344
(0. 0000083)
0. 0000289
(0. 0000136)
b -0.000000128
(0. 000000031)
c -0. ooooooioo
(0.000000051)
b 0.130
(0. 040)
c -0. 111
(0. 066)
b 0.40
C 0. 22
B. RESEARCH AND DEVELOPMENT MEASURED BY TOTAL RESEARCH PERSONNEL
Intercept RD RD~ Size 12 D R2
(3)
(4)
3(
Y2
b 0. 471
(0. 150)
b0.932
(0. 232)
c -1. 989
(0. 981)
1.381
(1. 513)
b 0. 112
(0. 031)
bO.118
(0.047)
b 0.0000300
(0. 0000102)
0.0000194
(0.0000157)
b -0.000000051
(0. 000000017)
-0.000000034
0. 000000027)
b -0. 120
(0. 044)
-0.100
(0. 068)
b 0. 28
0.18
Throughout this study, statistical significance for the regression coefficients is determined by one tailed
tests, and the significance of R2 by the F ratio test.
b Indicates statistical significance at the 99-percent level.
Indicates statistical significance at the 95- percent level.
THE ZERO VALUE PROBLEM
The data with which we are concerned cover a very wide range of variation in
the scale of research facilities, and deal with the six years between 1955 and 1960
inclusive. Where new products include all introductions of whatever type, this
period was sufficiently long to enable each firm to introduce at least one new
product. In the case of new chemical entities, however, 17 firms, all with rela-
tively small research establishments, failed to make even one introduction. As a
result, the value taken by Y1 for each of these 17 firms is zero. Although the zero
values are clustered at the lower end of the distribution of research, when
measured absolutely, they appear -throughout the distribution when R and D is
deflated by size of firm. If data for a longer time period had been available the
number of zeros would most likely have been considerably reduced.
The existence of 17 zero values among the dependent variables raises a number
of questions concerning the interpretation of the statistical results presented
above. Although the usual least-squares properties apply to our estimates, some
doubt is cast on the assumption that the disturbances are distributed normally,
on which our tests of significance are based. Even more important, however, is
the possibility of distortion resulting from the discontinuity which is introduced
by a substantial cluster of values at zero.
The zero problem occurs only when Yiis used as the dependent variable. In no
cases did Y2 take the value of zero. Because of this fact, it is useful to compare
regression coefficients between equations based on each variable. As may be noted
in table 1, the coefficients in all cases but one are not widely different and are
frequently quite similar. Only with regard to the first degree term for R and
I) do the coefficients diverge substantially.
To test for distortion in the equations concerned with Y1, two techniques are
used. The first is simply to eliminate the 17 zero values and compare the coeffi-
cients based on 40 observations with those based on 57. The second approach is
concerned with examining regression equations estimated from grouped data.
PAGENO="0532"
2082 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Firms are aggregated in groups on the basis of research input, and an analysis
is undertaken on group means.'5 In this manner, we are able to examine the rela-
tionship between research and technical change for firms even of the smallest
size category without encountering the problems created by zero values in the
dependent variable.'8
Table 2 presents the results of a regression analysis in which the same relation-
ship has been examined in four formats.1' As may be observed, the equations in
which the zero values are included and excluded present rather similar results.
With the exception of the linear term for R and D, the regression coefficients do
not take widely different values, although the standard errors of the coefficients
are larger with the exclusion of the zero values. These parameters are highly
significant. It does appear, however, that the significance and magnitude of the
first degree R and D coefficient is peculiarly dependent upon the zero values.
The grouped regressions present greater difficulties of interpretation. Although
the units were relatively homogeneous with regard to absolute values of R and
D, there may have been considerable variation within group boundaries in terms
of the explanatory variables used in the model. Grouping A, which has the effect
of consolidating only the smallest research establishments, provides coefficients
which are reasonable approximations of the ungrouped values. The estimates,
with the exception again of the linear coefficient for R and D, remain highly
significant; the intercept remains significant although not highly so. The second
grouping format influences the relationship to a larger extent than the first. The
coefficient of diversification becomes a good deal more negative and the standard
errors of the parameters are increased substantially. As a result, only two esti-
mates remain statistically significant.
With the exception of the first degree expression of research and development,
it does not appear that our estimated regression coefficients have been distorted
substantially by the inclusion of the zero values in the measure of technical
change. It should be recognized, however, that we have not presented any
theoretical explanation concerning the influence of these values.'8
TABLE 2-Regression analysis-Examination of the zero problem
Intercept
R. & D.~
R. & D.'
Size
.
Ii
D
Number
of obser-
vations
(1) Zero values included....
b 0. 422
(0. 136)
b -4* 671
(1. 285)
b 0. 547
(0. 107)
b 0. 0000344
b 0. 0000083
b -0. 000000128
(0. 000000031)
b -0. 130
(0.
57
(2) Zero values excluded_ -
(3) Grouping A
b 0. 690
(0. 227)
0. 424
(0. 219)
-13. 584
(10.284)
-3. 604
(3. 522)
b 0. 545
(0. 144)
b 0. 627
(0. 153)
b 0. 0000306
(0. 0000114)
b 0. 0000422
(0. 0000128)
b -0. 000000118
(0. 000000041)
b -0. 000000152
(0. 000000045)
040)
b -0. 163
(0. 055)
b -0. 169
(0. 066)
40
29
(4) Grouping B
0. 545
(0. 695)
-2. 249
(6. 625)
`0. 538
(0. 268)
C 0. 0000430
(0. 0000225)
-0. 000000114
(0. 000000088)
-0. 237
(0. 194)
16
R. & D. is measured by the number of professional research personnel.
Indicates statistical significance at the 99-percent level.
Indicates statistical significance at the 95-percent level.
15 Grouping was undertaken in two formats on the basis of non-deflated values of
professional research personnel. The first format was to lay out the entire distribution of
firms from sniallest to largest in terms of the grouping variable and mark off boundaries
so as to insure that at least 50 persons were included in each unit; grouping was begun
at the small valise end of the distribution. Twenty-nine groups were defined in this
fashion of which all but ten consisted of single firms. In the second format, six groups
were defined between the values of zero and 50, and intervals of 25 marked off through-
out the remainder of the distribution. This scheme provided 16 groups.
IC If we assume that the disturbances associated with individual firms have constant
variances, then groups of varying size necessarily introduce heteroscedasticity. To retain
the efficiency of our estimates, it has been necessary to weight all observations by the
square root of the number of firms in the group; this also includes the column of ones
used to determine the intercept.
17 Since a grouping procedure in itself influences the degree of variation "explained" by
a regression analysis on grouped observations, multiple correlation coefficients with group-
ing cannot be compared with those obtained from the original data. For this reasosi, the
R's associated with these equations have not been presented.
`~ While a more sophisticated technique exists for dealing with this matter, the corn-
putational problems involved were too great to warrant its use. See James Tobin, "Estima-
tion of Relationships for Limited Dependent Variables," in Econometrica (Jan., 1958),
24-30.
PAGENO="0533"
COMPETITIVE PROBLEMS IN, THE DRUG INDUSTRY 2083
THE INFLUENCE ON TECHNICAL CHANCE OF SCALE OF RESEARCH AND SIZE OF FIRM
In the model presented above, an interaction variable is introduced to test
whether firm size affects the shape as well as the position of the relationship
between research and technical change. Since it is necessary to deflate by size
of firm in order to deal with the problem of heteroscedasticity, this approach
is used to reintroduce a scale factor anti thereby examine differences in the
form of the relationship as between small firms and large. In addition, an addi-
tive expression denoting firm size is used tO consider the impact of this variable
on the equation's position.
As may be observed in table 1, the linear expression of firm size is positively
associated with research output. Since new products have been weighted by their
sales during their first two years after introduction, we should expect our
measures of technical change to be influenced by such factors as distribution
facilities, selling effort, and firm reputation. To the extent that these factors
are correlated with firm size, the size variable introduced into the equations will
represent their influence, and the coefficient will be positive. Firm size appears
to have acted to increase the gains resulting from new product introduction, and
these gains increased more than proportionately with size of firm.
Our findings imply further that the relationship between research and devel-
opment and the rate of technical change is substantially influenced by size of
firm. Not only is the relationship shifted in total by differences in this variable,
but also the functional form of the relationship is dependent on these values.
Towards the lower end of the size distribution of firms, economies of scale in
research seem to be present, while diseconomies are likely when firm size becomes
large. It appears, thus, that we cannot deal with the question of economies of
scale in R and D without considering the size of the firm within which the re-
search is undertaken.1°
These findings may be considered further by dealing with our primary rela-
tionship in a different format. If both sides of equation 1 in table 1 are multi-
plied by 2, the relationship may be expressed in the form:
Y=a+bR+cR2 (1)
where
b=-4.671-0.000000128 $2, (2)
a is a function of S and D, and c cquals 0.547.
Thus
~=b+2cR. (3)
Since b is inversely related to 8, is larger for smaller values of S. For given
values of R, declines with the square of S. These results imply that for the
pharmaceutical industry, the marginal productivity of professional research per-
sonnel is inversely related to size of firm.
Using the functional form of the relationship presented in (1), the elasticity of
Y1 with respect to R was calculated to three points in the distribution of ~
19 One additional point concerning our measures of technical change should be mentioned.
Since these variables have been deflated by size of firm, they may be taken to represent
the proportion of sales accounted for by new products. If larger firms sell relatively
more standardized or older products than smaller firms, then we should expect the deflated
ratios denoting technical change to be inversely related to firm size, and this fact would
have little to do with the productivity of research establishments. It appears, however,
that new products account for as large a proportion of total sales in large firms as in
small ones. The proportion of total sales accounted for by products introduced within
the preceding five years was computed for each year between 1955 and 1960 for each of
the 57 firms included in the sample. The simple correlation coefficient (r) between the
mean proportion for each firm and the corresponding measure of firm size was + 0.18.
The simple correlation coefficients between deflated values of l'i and F2 and firm size
were also computed. The coefficients were +0.14 and +0.08 respectively. Since it does
not appear that larger firms sell, on the average, proportionately less new products than
smaller firms, we should not expect this consideration to distort our findings.
SYR
20 The elasticity was calculated as -h..
PAGENO="0534"
2084 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Since this elasticity is defined as the percentage change in research output for a
unit percentage change in R and D, holding S constant, it will describe the find-
ings on economies of scale in research which are implicit in the equation results.
Elasticities were computed at values of S equal to 1,000, 10,000, and 50,000 (phar-
maceutical sales in thousands of dollars) ~ When S equaled 1,000, the calculated
elasticity was 1.39. At S equal to 10,000 and 50,000, the corresponding figures
were 0.61 and 0.54. While there are likely to be increasing returns to scale in R
and D at low values of 5, decreasing returns seem to be the case when S becomes
moderately large.n 23
SOME ALTERNATIVE HYPOTHESES
Up to this point we have proceeded on the assumption that the direction of
causation runs from research and development to `technical change, and have
implied that the degree of diversification affects the rate of technical change.
It is necessary, however, to consider some alternative hypotheses. It may be
that successful new product introduction influences the extent of research activi-
ties rather than or in addition to the opposite alternative. A rapid rate of tech-
nical change may stimulate the firm to expand the scale of R and D facilities;
accelerated technical change may result in higher profit rates which enable the
firm to increase its investment in research and development. Similarly, it may
he hypothesized that the rate of new product introduction is the vehicle by which
diversification proceeds. If these alternative hypotheses are correct, we should
expect that technical change at the start of the period would be better correlated
with R and D and diversification at the end, than if the lead-lag relationship
were reversed. Our original hypotheses would postulate that R and D and
diversification should lead the variables denoting tecimical change. To investi-
gates these questions, we shall consider our regression equations with alternative
lead-lag structures.
The empirical findings concerned with this matter are presented in `table 3.
The research and development variables denote the number of professional
research personnel employed in either 1955 or 1960, while the technical change
and diversification variables span the period which is designated. The size vari-
able remains the same as in the previous analyses.
The results are striking. With a lead-lag structure consistent with our original
hypotheses, all coefficients are significant at the 99% leveL With the opposite
lag structure, the R2's are not significant, the parameters of diversification are
smaller `than their standard errors and alternate in sign, and the coefficients for
size, interaction, and the quadratic expression for research are not significant.
There is, however, some positive linear association between technical change
in 1955-1957 and research in 1960 when the variable Y2 is used, but this re-
lationship may be relatively weak for it fails to appear in the regression equa-
tions deallng with the entire six-year period. We may conclude, `therefore, that
the apparent lead-lag structure is not inconsistent with our original hypotheses.
21 Corresponding values of R were determined by taking the mean value of that variable for firms in the
neighborhood of the appropriate value of S. For S equal to 1,000, firms were included between S equal to
500 and 1,500; for S equal to 10,000, the corresponding figures were 5,000 and 15,000; while for 50,000, tile firms
with Sbetween 40,000 and 60,000 were included. The three values of R w-hich were determined in this fashion
were 13.1, 59.2, and 353.3, respectively. D was computed at its mean value throughout.
22 One qualification of the above results should be mentioned. It may be that large firms undertake more
"basic research" than smaller firms. Although it is assumed that the returns from this activity are obtained
through its advancement of the rate of new product technical change, the time horizon in this case may be
substantially longer than that of "applied research and development" so that the six-year period under
examination may not be long enough to measure accurately the returns. As a result, the level of research
output for larger firms would be understated relative to smaller firms. It does appear that larger firms spend
somewhat larger proportions of their R. & D. budgets for "basic research." For 1959, National Science
Foundation data for industry group "Drugs end Medicines (SIC 283) disclose that firms with total em-
ployment over 5,000 undertook 50% of total industry funds for the performance of R. & D. while these firms
accounted for nearly 63% of industry "basic research." Tile difference between tile elasticities is, therefore,
likely to be somewhat overstated.
23 One further implication of the functional form of (1) should be mentioned. Since b is negative and c is
positive, the function, at given values of 5, originally declines, reaches a minimum, and then increases.
This minimum is reached at values of R equal to ~ In addition; since b is an inverse function of 5, the
values of R at which the minimum is reached increases with S. At values of S equal to 1,000, 10,000. and
50,000 the minimum is reached at values of R equal to 4.4, 16.0 and 298. Actual values of R will, in nearly all
cases, be greater than these values. At the extreme upper tail of tile distribution of 5, however, the marginal
productivity of research estimated l)y equation (1) in table 1 will be negative.
PAGENO="0535"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2085
TABLE 3.-Regression and correlation analysis-Various lead-lag structures
Intercept
RD
RD2
Size
Ii
D R2
1955
1955
1955
1955-1957
(1) Fi (1958-1960)_...
(2) 1"i (1955-1957)_U
bO.357
(0. 100)
0.125
(0. 076)
b-3.360
(1. 025)
1900
-1.193
(0. 751)
1955
bO.351
(0. 083)
1960,
0.0861
(0. 0519)
1955
0.0000159
(0. 0000049)
0.00000649
(0. 00000485)
b_0.0000000693
(0. 0000000212)
1960
-0.0000000172
(0. 0000000152)
1955
b--0.109 b031
(0. 032)
1958-1960
-0.0164 0.13
(0. 0205)
1955-1957
(3) F'2 (1958-1960) - - - -
b 0. 594
(0. 122)
b -3. 898
(1. 252)
1960
b 0. 465
(0. 102)
1960
b 0. 0000171
(0. 0000060)
b -0. 0000000810
(0. 0000000259)
1960
~-0. 118 b 0. 32
(0. 039)
1958-1960
(4) F'2 (1955-1957)~..
`0.311
(0. 140)
b4343
(1. 387)
0.0053
(0. 0959)
0. 00000309
(0.00000895)
0. 0000000012
(0. 0000000280)
0.0131 0.18
(0. 0378)
R and D is measured by the number of professional research personnel.
Indicates statistical significance at the 99 percent level.
Indicates statistical significance at the 95 percent level.
THE ROLE OF SUPPORTING PERSONNEL
At this point we shall examine the question of whether a large ratio of sup-
porting personnel to professional staff substantially increases the efficiency of
a research facility. The relevant variable is defined a's the mean ratio in 1955 and
1960, and the analysis is carried out by introducing it into our regression
equations.
The cofficients of the ratio of supporting to professional personnel are always
negative and in no instances are they statistically significant.24 Thus, we may
conclude that supporting personnel play only a minor role in determining the
magnitude of research output. High ratios of auxiliary staff to professionals
do not seem to increase the productivity of the professional researchers. These
results corroborate our earlier observations that the number of professionals
seems to be a better measure of research input than the total number of It and
ID personnel.
THE EFFECT OF GROWTH IN RESEARCH FACILITIES
In this section, we examine the influence on technical change of rapid expan-
sion of R and D activities. Some observers maintain that research efficiency is
likely to be impaired if the rate of expansion is large.n To consider this position,
we define two variables and introduce them separately into the regression
equations. The first variable, which is designated G~, is the ratio of total research
personnel in 1960 `to the corresponding value in 1955. The second variable G2,
is the ratio of 1960 to 1955 levels of professional research personnel.
The coefficients of G~ and G-~ are not statistically significant although they
are negative and thereby do take the hypothesized sign.2° As a result, the hy-
pothesis was restated and examined in a different format. It may be that rapid
expansion of research facilities does not affect the rate of technical change
continuously. Inefficiences may be created only when growth is pushed above
some threshold rate. To investigate this position, three dummy variables are
defined according to whether the research establishment showed no growth,
moderate growth, or rapid growth, between 1955 and 1960, and the first two
dummies are introduced into the regression equations.27
24 When the ratio of technicians to professionals is introduced into equations 1 through
4 of table 1, the estimated coefficients and standard errors are: (1) -0.181, 0.144; (2)
-0.149, 0.220; (3) -0.086, 0.139; and (4) -0.161, 0.211. (In equations 2 and 4, the
non-significant linear RD variable was omitted when these estimates were obtained.)
~ Mansfield has reported: "There are considerable costs involved in a very rapid ex-
pansion of a firm's R and D department, the importance of which was stressed in inter-
views with various executives." Edwin Mansfield, The Ewpenditures of the Firm on Research
and Development (mimeographed), 4.
26 When G~, is introduced into our original regression equations, the estimated coefficients
and standard errors are: -0.124, 0.126 in the case where technical change is measured by
Fi, and -0.168, 0.185 when F2 is used. The corresponding estimates for G2 are: -0.0229,
0.0958 and -0.061, 0.153. In these calculations, RD is measured by the number of pro-
fessional research personnel.
~ The dummy variables are defined on the following basis: If, during the period, the
laboratory remained stable or declined in size, the firm is listed in the first category; If
the facility increased in size but by less than 100%, the moderate growth grouping is
designated; while if the research establishment at least doubled in size, it is considered
to have experienced rapid growth. The three categories are defined in terms of increases
in both total and professional research personnel.
PAGENO="0536"
2086 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
An F test is used to deal with the question of whether the dummy variables
significantly affect the rate of technical change. While the dummies based on
increases in professional personnel do not lead to significant values of F even
at the 50% level, those based on the expansion of total research personnel resu~1t
in test values which are significant at the 75% level when F1 is used to measure
technical change. In addition, the estimated regression coefficients founded on
the latter set of dumniies take the expected sign and relative order of rnagnitude.IS
These findings point out that any inefficiencies resulting from rapid expansion
of research facilities are more likely to be associated with increases in the total
size of the establishment rather than particularly related to increased hiring
of professional personnel. It should be emphasized, however, that empirical
support for this hypothesis exists only if we are willing to accept statistical
findings as the 75% level of significance. As a result, this conclusion is highly
tentative.
CONCLUDING STATEMENTS
In this paper, we have investigated the relationship between research and
development and the rate of new product technical change in the pharmaceutical
industry. From the empirical evidence, there appears to be a fairly sustained
association between research input and new product output. Within this industry,
research expenditures are not undertaken merely with the hope of some distant
but unknown returns, but rather with the expectation that profitable gains will
accrue within a reasonable period of time.
Our analysis also provides some evidence that in the pharmaceutical industry,
there are substantial diseconomies of scale in R and D which are associated
with large firm size; and that these disadvantages are encountered even by mod-
erately sized firms. One implication of this finding is that an actively enforced
pro-competitive policy in this sector is not likely to dampen the rate of technical
change and may well stimulate it. While little is known about the extent `to
which this result is applicable to the economy at large, it does appear that there
are grounds for considerable doubt as to the position that large firm size is always
a necessary condition for rapid teelmical advance.
[From The Journal of Business of the University of Chicago, Vol. XXXIX, No. 1, Part I,
January 1966]
THE DRUG INDUSTRY AND MEDICAL RESEARCH
THE EcoNoMics OF THE KEFAUVER COMMITTEE INvE5TIGATI0N5*
(By William S. Comanort)
In the course of recent American politics, few confrontations have been more
lively and more prolonged than that w-hich was waged between the pharmaceuti-
cal industry and the Senate Subcommittee on Antitrust and Monopoly (The Ice-
fauver Committee) .~ In December, 1059, the committee opened hearings on the
~ The dummy variables denoting "no growth" and "moderate growth" are introduced into
equations (1) and (2) of table 1. When the dummies describe the rate of expansion in
total research personnel, the estimated coefficients and standard errors are, for equation
(1) : 0.371, 0.220 and 0.209, 0.184, and for equation (2) 0.399, 0.336 and 0.159, 0.297.
The coefficient of the "no growth" dummy in equation (1) is statistically significant at
the 95% level. Furthermore. the intercept, which encompasses the effect of the third clas-
sification, is reduced from 0.422 to 0.199 by the introduction of the dummies into equation
(1).
*This article is drawn in part from my unpublished Ph.D. dissertation, "The Economics
of Research and Development in the Pharmaceutical Industry" (Department of Economics,
Harvard University, June 1963). I am grateful to J. W. Markham and R. B. Hefiebower,
who acted as supervisors of the original study, and to Carl Kaysen for valuable counsel
in the writing of this paper. I have also benefited from comments by Oswald H. Ganley. Mary
Lee Ingbar, Lars Sandberg, and Harvey M. Sapolsky, and from the secretarial assistance
of Anne Caneles. This study was undertaken with the support of the Science and Public
Policy Program at Harvard.
tlnstructor of economics, Harvard University. Currently on leave and associated with
the Antitrust Division, U.S. Department of Justice. This article was written at Harvard
and in no way does it reflect the views of the Department of Justice.
1 The pharmaceutical industry in the Congressional proceedings, and therefore in this
paper, is limited to firms that produce ethical drugs, as opposed to proprietaries, and that
distribute these products in dosage form. Pharmaceuticals, thus, are marketed and sold
through the medical profession and require, for the most part, a written medical pre-
scription.
PAGENO="0537"
COMPETITIVE PROBLEMS IN ~ THE DRUG INDUSTRY 2087
ii~dustry, ard 11,000 pages later, testimony was still being recorded. Now that
the record has been closed and the dust from the contest `has been `allowed time
to settle, it is useful to consider some of the questions raised in `the proceedings.
This paper reviews the evidence on drug research.
I
A major point at issue between the industry and the committee concerned the
medical value of industry research. Although forced to admit that expenditures
on research were large,2 the coinniit'tee charged that little of social value came
from industry laboratories. It maintained that nearly all the important new
discoveries in recent years had been derived from research work performed out-
side the industry and that commercial laboratories are concerned primarily with
"molecule manipulations" or new drugs therapeutically quite similar to drugs
already on the market.3 The committee implied that industry research was highly
duplicate and that a large portion of this work could be eliminated without
reducing very much the flow of important new drugs. In defense of its efforts,
the industry argued that from industry laboratories come most of the new
drugs that are extensively utilized in current medical practice.4
These opposing views on the nature and significance of the pharmaceutical
research effort `collided in what may be termed the "battle of the lists." This
contest was based on competing lists of new drugs prepared by the committee
and the industry. One of the major points of `difference between the two is that the
committee tended to concen'trate on drugs that embodied what it considered to
be major therapeutic advancements at the time of introduction. As we would
expect. the conunittee and i'ts lists emphasized the role played by foreign and
non-industry laboratories.5 The `industry list, on the other hand. included new
drugs that may not have embodied large steps forward but that are in frequent
use and thereby seem to have the confidence of the country's physicians.6 A large
majority of the drugs on `this list were dis~overed and developed within industry
laboratories. While :tl~ere was some controversy over the origins of specific
drugs. it seems likely that both sets of lists and their associated conclusions are
essentially correct.
To unravel the issues here, it is necessary to deal explicitly with the nature
and functions of industry research. In the succeeding sections of this paper, we
shall undertake this discussion in the light of the alternative pos'itions taken by
the committee and the industry.
Research and development in the pharmaceutical industry is carried on within
a context of strong product rivalry, `and this factor has affected the type of work
that is undertaken. Considerable attention is focused on the rapid introduction
2 In 1963. e.g., company-financed research and development totaled $282 million. This
sum equaled nearly 9 per cent of total industry sales (Pharmaceutical Manufacturers As-
sociation, "Pharmaceutical Industry Research and Development Activity, 1963-64" [Wash-
ington, D.C., 1964 (mimeographed)], p. 1, and Pharmaceutical Manufacturers Association,
"Ethical Pharmaceutical Sales, 1963" [Washington, D.C., 1964 (mimeographed)], p. 2).
Although N.S.F. data on company-financed H ,& D as a percent of net sales in manufac-
turing industries for 1963 is not yet available, 1961 data show that the highest percent-
age for an industry other than drugs and medicines was 4.4 per cent (National Science
Foundation, Research and Development in Industry, 1961 [February, 1964], p. 86).
This position was based on statements such as the following, which appeared through-
out the proceedings: "The question is what then is the goal of this admittedly large-scale
laboratory effort of our industry? Partly to exploit and market those foreign and non-
industrial advances and compounds that I have mentioned. Mostly, however, to modify
the original drugs-just enough to get a patentable derivative, but not to change it enough
to lose the original effect" (Testimony of Frederick H. Meyers, associate professor of
pharmacology, University of California, in U.S. Senate, Administered Prices in the Drug
Industry [hereinafter cited as "Administered Price Hearings"]. [Hearings before the
Subcommittee on Antitrust and Monopoly (Washington: 86th Congress, 2d Session, 1960)]
(Part XVIII, p. 10394).
Pharmaceutical Manufacturers Association, Time Pharmaceutical Industry, Key Pacts
at a alance (Washington, D.C., n.d.). The industry reported that of the ninety-four im-
portant drugs developed since 1945 and in frequent use in 1960, forty-nine, or over half,
were discovered by drug companies or by university scientists financed by drug companies.
The original committee list appears in' Administered Price Hearings, Part XIX,
pp. 10943-45, while the supporting comment is made in U.S. Senate Subcommittee on
Antitrust and Monopoly, Report of the ytudy'of Administered Prices in the Drug Industry
(Washington: 87th Congress, 1st Session, 1961), p. 115.
6 E.g., the original industry list included Achromycin as well as Terramycin, two anti-
biotics that have similar therapeutic effects. This list appears in Administered Price
Hearmngs, Part XIX, pp. 10840-54.
PAGENO="0538"
2088 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
of new products, and, as a result, the number of introductions exceeded 3,800
in the decade ending in 1960. Of these, the largest number by far consisted of
new combinations of existing drugs and of new dosage forms.8 The remainder,
or 432 in total, were new chemical entities, and these absorbed the largest share
of the research effort. It should be noted, however, that this category included
salts anti other derivatives of known drugs as well as entirely new substances.9
The nature of industry research may be inferred to some extent from informa-
tion on new products. As we might expect, there appears to be a large effort to
invent around existing patents, and this has led to the introduction of a consider-
able number of new drugs whose therapeutic effect is quite similar to that of
products already on the market. We should note, however, that rarely do different
chemical compounds have perfectly identical therapeutic properties and that small
differences may be important for some patients.
To a great extent, industry research activities build upon the scientific achieve-
ments of the past. They are concerned, primarily, with the smaller, albeit less
uncertain, steps forward. Although these activities may duplicate what has
already been clone and also frequently may lead to new drugs that represent at
best a very minor advance, it is still true that a number of important modifications
and variations have been discovered within industry laboratories.
The second major emphasis of industry research concerns the process by which
new scientific know-ledge is translated into marketable products. Product rivalry
is strong, and this has insured that new drugs will not lag far behind advances in
scientific knowledge, no matter whether these advances originate in university,
government, or foreign laboratories. This is an important area, because in most
cases substantial research and development problems remain even after the
original discovery has been made. Not only is there frequently a substantial gap
between laboratory synthesis and large-scale manufacturing methods, but also
considerable analysis and testing is required before a drug may be marketed. And
these activities, which may well be more costly than basic research, are under-
taken for the most part within industry laboratories.
Whether the impetus for a new drug comes from competitive successes, from
Jasic research carried on within the industry, or from new scientific knowledge
arising from non-industry sources, the problems of synthesis and testing remain,
and these absorb the major share of industry research expenditiures. Large
quantities of the promising compound are normally required before efficient pro-
duction techniques are available, and large sums are expended for this purpose.
It has been estimated, in fact, that approximately one-half of total expenditures
on research with the industry goes to support the work of synthesizing, purifying,
modifying, and preparing suitable substances for subsequent physiological tests.'°
Once this process is completed, the new substance is subjected to intensive bio-
logical investigation. The first concern is to determine the therapeutic properties
of the su,bstartce this encompasses the major purpose of the drug as well as side
effects. In addition, it is necessary to test for toxicity levels and to determine the
compound's potency in order to gain knowledge of appropriate dosages. For these
purposes, the pharmaceutical industry used in its research facilities some nine
million laboratory animals during l96l.~'
There is, moreover, a great deal of uncertainty concerning the therapeutic
properties of new substances. As would be expected, most of those tested in
laboratories are found to be without sufficient promise to justify clinical testing on
human beings. In 1958, for example, nearly 115,000 substances were subjected to
biological tests by pharmaceutical firms, while only 1,900 were considered worth
testing clinically.~ The Food and Drug Administration requires that manufac-
turers undertake efficacy studies on new drugs and that the results of these tests
Paul de Haen. "New Product Survey" (New York: privately printed by author. 1900).
Althou~h new chemical substances are not involved in these cases, considerable atten-
tion may still be required. E.g.. a good deal of work was done in an attempt to combine
Glucosamine with certain antibiotics in the hope that this would increase the absorption
of the drug into the blood stream (Administered Price Hearings, Part XVIII, p. 10257).
During 1959. sixty-three new chemical entities were introduced. Of these, eleven were
new salts of old products and twenty-three were derivatives of known drugs. The re-
mainder of twenty-nine were entirely original products (ITS. Senate, Drug Industry Anti-
trust Act [hereinafter cited as "Antitrust Act Hearings"} [Hearings before the Sub-
committee on Antitrust and Monopoly (871i Congress, 1st Session, 1961) 1, Part II, p. 888).
~° Statement by Austin Smith. president of the Pharmaceutical Manufncturers Asso-
ciation, in Administered Price Hearings, Part XIX, p. 10725.
1~ Ruported in New York Times, August 26, 1962. p. 60.
12 Administered Price Hearings, Part XIX, p. 10725.
PAGENO="0539"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 2089
be submitted to the administration before a new chemical entity may be
introduced.
Clinical studies are normally carried out with the help of physicians and medi-
cal institution:s throughout the country. New drugs are contracted out to clini-
clans in medical schools or hospitals for controlled testing. While there may be
a sizeable liaison group among the pharmaceutical research staff, the actual test-
ing is done, in most cases, by non-industry personnel. In 195~, only one major
company supported directly the maintenance of a clinical-research group having
hospital facilities; this was the Lilly Laboratory for Clinical Research located
within the Indianapolis City Hospital.13 Even when the testing was performed
outside of industry facilities, however, the pharmaceutical firm concerned nor-
mally supplied the compound that was to be tested.
While it is difficult to determine directly the cost of clinical trials to pharma-
ceutical firms, we can observe the division of industry research expenditures
spent within and outside of the companies as some indication of the relevant
magnitude. In this regard, less than 10 per cent of total R & D expenditures was
in the latter category.'4 To this should be added, however, the cost of preparing the
compounds used in the trials.
Although the laboratory and clinical-research activities described above have
a highly applied and developmental character, still they are necessary before
medical benefits can be realized from new scientific knowledge. With regard to the
major share of industry-research activities, these are best viewed as a comple-
ment rather than as an alternative to those undertaken outside the industry.
III
To evaluate industry research activities, it is necessary to place them within
the context of the total medical research effort. In this regard it should be
stressed that the industry provides only about 30 per cent of total medical-
research outlays within the United States,15 and the proportion would fall still
further if foreign research expenditures were included in the total. Given the
relative size and essential complementarity of industry and non-industry research,
it follows that the two cannot adequately be evaluated by the same set of
standards. Despite the fact that the same word is used to describe both sets of
activities, they are essentially different.
One of the major problems of the investigatory proceedings and the joust be-
tween the committee and the industry concerned the failure of either side to
recognize this difference. The committee criticized the industry on the apparent
failure of industry scientists to provide many dramatic new breakthroughs on
which major therapeutic advances are based and also on the lack of important
new drugs in recent years. Whatever the validity of these charges, they are not
founded on a realistic appraisal of the role and function of industry research. At
the same time, however, the industry's apparent claim that its research is respon-
sible in large measure for the great advances in medical science that have been
achieved in recent years is equally inappropriate. Both of these positions imply
that industry and non-industry research can be evaluated on a single set of
standards, and this neglects the considerable differences in purpose and approach
between the two areas of work.
A valid judgment of industry research and development must necessarily be
founded upon the nature of the complernentarities present within the larger
boundaries of medical and health-related research. These complementarities are
important, and it does appear that the differences that exist between the two
areas of research may well constitute a necessary and desirable division of
labor. The peculiar attributes of each probably create for it a comparative
advantage for the specific activities which it, in fact, undertakes. While it is
difficult to be definite in these matters, it does seem likely that basic or funda-
mental research is better carried on in a university or government laboratory
13 American Foundation, Medical Research, A Mid-Century ~S'urvey (2 vols.; Boston:
Little. Brown & Co., 1955), I, p. 592.
" In 1958 the amount spent outside of Industry facilities was $13.4 million out of a
total of $170.0 million, while in 1959 the comparable figures were $15.3 million and
$190.0 million (Administered Price Hearings, Part XIX, p. 10724).
15 In fiscal year 1960, the pharmaceutical industry allocated $215 million to research
and development activities out of a total national expenditure of $715 million. Govern-
silent provided $400 million, while other private support equaled $100 million (Antitrust
Act IIear~ngs, Part III, p. 1705).
PAGENO="0540"
2090 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
than in a commercial facility. Within a profit-motivated firm, there is a latent
conflict between an atmosphere of freedom, which is necessary for good scientific
endeavor, and the necessity of direction from those who are responsible for the
conduct of the firm.16 In addition, it appears that those firms which have had the
greatest success in basic research have moved furthest toward creating a uni-
versity environment within their facilities.17 While additional corroborating fac-
tors could be mentioned, it is sufficient to remark that we should not be surprised
if the major advancements in medical science are achieved in laboratories out-
side the pharmaceutical industry.
At the same time, moreover, there are a number of reasons w-hy highly applied
research and development activities are best carried on within the confines of the
industry. Not only do these tasks assume a highly routine character which tends,
in a large number of cases, to make them uninteresting to university scientists,
but also the relevant procedures are sufficiently well understood so that they may
be reasonably well directed. An atmosphere of completely free inquiry is not
crucial where both the goals and the procedures are relatively w-ell defined.
These considerations, moreover, may explain the fact that although the pharma-
ceutical industry provided approximately 30 per cent of the total funds for
medical research in 1960, it utilized only 12 percent of the total number of persons
doing medical research with the Ph.D. degree and only 5 per cent of the total
number doing medical research with the M.D. degree.18
Even more important, however, are the economies that are likely to result from
combining applied research with production in the same organization. Within a
technical industry, the road from laboratory to factory is not level or direct,
and large gains may result from institutional arrangements that provide the
smallest possible barriers to the flow of information. To insure a free and unre-
stricted flow- it is probably necessary that the latter stages of research and
development, as well as production, are undertaken w-ithin the firm.
It is interesting, in this regard. to note the conclusions of a recent study that
contrasted the American and Soviet pharmaceutical industries. Unlike the United
States, research and production are carried on by completely separate agencies
in the Soviet Union. The authors report that as a result of this arrangement,
`there are difficulties of communication and coordination between [research]
institute and factory which constitute a major bottleneck in getting pharma-
ceuticals into production." 19 While the divorcement of research from production
represents merely one additional factor out of many that distinguish the Soviet
industrial structure from that of the United States, still the apparent results
from this divorcement do seem to point to a major set of problems.
The division of labor that exists betw-een industry emphasis on product de-
velopment and non-industry emphasis on the more fundamental areas of re-
search provides a system that is rational in principle although we know little of
whether the optimal share is more or less than the 30 per cent presently occupied
by the industry. If, however, our concern is with the use and possible misuse of
such scarce and limiting factors as highly trained personnel rather than merely
with dollars, the quantitative pro~ilems appear in a different perspective. A large
drug-industry effort does not divert, to a substantial degree, scarce resources
from university and government laboratories to its own facilities but, rather,
uses, for the most part, a different and non-substitutable class of personnel. Thus,
the opportunity cost of industry research in terms of non-industry research is
likely to be relatively low.
When evaluated on its own terms, it seems clear that the industry accom-
plishes an important research task in a generally effective manner. It has been
responsible for a large number of pharmacological modifications and improve-
ments that have been introduced as well as for an acceleration in the process of
testing and developing new- drugs, and both of these gains provide a high degree
16 This "conflict" is discussed in John Jewkes, David Sawers, and Richard Stillerman,
The Sources of Invention (New York: St. Martin's Press, 1958), chaps. vi, vii.
~ See, e.g.. the statement by James B. Fisk. "Basic Research in Industrial Laboratories."
in Dael Wolfie (ed), Symposium on Basic Research (Washington, D.C. : American Asso-
ciation for the Advancement of Science, 1959), pp. 159-67.
18 National Institutes of Health, Manpower for Medical Research-Reqnireineists and
Resources, 1965-1970. (A Report for the Committee on Appropriations [Washington:
U.S. House of Representatives, February. 19621), p. 24.
~9 Raymond A. Bauer and Mark G. Field, "Ironic Contrast: U.S. and U.S.S.R. Drug
Industries," Harvard Business Review (September-October, 1962), p. 93.
PAGENO="0541"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2091
of social value.20 It should be recognized, however, that these gains are on a very
different scale from those related directly to significant advancements in medical
science and to dramatic improvements in public health.
Iv
During the investigations, the Kefauver Committee proposed a major revision
in economic policy toward the drug industry. This proposal, however, was re-
jected by the parent Committee on the Judiciary and was not included in the
19G2 Drug Act. The change would have instituted compulsory licensing at reason-
able royalty rates three years after a patent had been granted. It is useful here
to note the implications of this change in the light of the analysis presented
above.
Although this proposal was intended to produce more competitive levels of
prices and profits, the industry maintained, that it would also lead to much re-
duced research expenditures. This position seems to be generally correct. The
primary motive for large research efforts in this industry has been the drive to
achieve effective product differentiation.21 With compulsory licensing, there
would be a sharp decline in the extent of differentiation based currently on
chemical differences among products.
Nevertheless, it is not at all certain that compulsory licensing would signifi-
cantly lower the rate of introduction of the most important new products. The
largest proportion of these come originally from non-industry laboratories. It is
true that pharmaceutical companies would have less incentive to undertake
projects of a long-term nature, such as many of those in basic research, because
of the diminished prospect for large gains over a prolonged period of time. These
firms account, however, for only a minor share of the work which currently is
done, and they also appear to lack a comparative advantage in pursuing basic
research in an extensive manner.
Even with compulsory licensing, research and development would still com-
prise an important element of industry behavior. There would still be gains
from achieving product differentiation. New drugs that embody a large element
of therapeutic improvement are also likely to provide a high degree of product
differentiation. Having developed a differentiated product, not only would firms
benefit substantially from a head start in promotion and selling, but also this
advantage would last for the initial, prelicense period of patent protection, dur-
ing which time monopoly gains could still be attained. In addition, patent royal-
ties might become an important element of the rewards resulting from successful
research.
Compulsory licensing appears on balance to be a useful and desirable policy to
adopt. While the magnitude of research would decline, it does seem probable
that projects of limited medical value and of lowest industry comparative advan-
tage would be eliminated first. There are substantial social gains to be derived
from industry research, but the marginal social productivity of research may
well decline rapidly after a certain level has been reached. We should be wary of
believing that much is to be gained from ever higher levels of research and
development, and it is quite possible that the present effort may exceed that
required to fulfill the major research functions and responsibilities of the
industry. So long as the decline in research expenditures was not of over-
whelming proportions, it may well be a small "price" to pay for a more competi-
tive determination of pharmaceutical prices and profits.21
~ This is in contrast to the Soviet pharmaceutical industry. Bauer and Field state that
"the testimony of the well-informed Soviet sources is that the separation of research
from production tends to produce substantial delays in the availability of drugs to physi-
cian and patient" (Bauer and Field, op. cit., p. 94). At the same time, however, excessively
rapid process leads to new drugs that have not been adequately tested, and this charge has
been laid on the doorsteps of the American pharmaceutical industry. This point was com-
mented on by a medical educator who noted sharply: "There is no short cut from chemical
laboratory to clinic, except one that passes too close to the morgue" (Administered Price
hearings, Part XVIII, p. 10418).
21 The role and function of product differentiation and its relation to research and de-
velopment activities is discussed in William S. Comanor, "Research and Competitive
Product Differentiation in the Pharmaceutical Industry in the United `States," Economica
(November, 1964), pp. 372-84.
Smaller research facilities may also lead to increased efficiency in pharmaceutical re-
search and development. When research output is measured in terms of private rather than
social gains, there appear to be substantial diseconomies of scale in research within the
larger firms in this industry. See William S. Comanor, "Research and Technical Change in
the Pharmaceutical Industry," Revieiv of Economics and Statistics (May, 1965), pp. 182-
PAGENO="0542"
2092 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
[From The Review of Economics and Statistics, Volume XLIX, November 1967, Number 4,
PP. 423-440]
ADVERTISING MARKET STRTTCTURE AND PERFORMANCE
(By William S. Comanor and Thomas A. Wilson*)
This paper presents an empirical analysis of the role of advertising in con-
sumer goods industries. The primary finding is that advertising has a statis-
tically significant and quantitatively important impact upon profit rates which
provide a measure of market performance as well as indicate the existence of
market power. This result is robust, and the estimated multivariate equations
account for half of the inter-industry Variance of profit rates.
This finding has implications which are precisely the opposite of the conclu-
sions reached by Telser in a recent important article.1 This contradiction is a
reflection primarily of differences in the conceptual and statistical approaches
adopted rather than differences in data or sample, for with minor exceptions,
we used the same set of industries, and drew upon the same basic data for
advertising outlays.
We shall therefore proceed as follows. First, we shall describe the conceptual
framework used. Then we shall examine the relationships which are likely to
exist between product differentiation, advertising and entry barriers. Finally,
w-e shall present the empirical results which are the core of this paper.
FRAMEWORK OF ANALYSIS
The analytical approach is to examine the joint effect of various dimensions
of market structure upon profit rates. Not only do profit rates provide some
indication of market performance in terms of the normal criteria of allocative
efficiency, but also high returns signal the possible existence of market power.2
If exercise in the direction of profit maximization, market power should lead
to rates of return which exceed those in competitive industries that are
comparable in terms of risk and growth of demand.
In this framework, concentration is simply one dimension of market structure
and is not of itself a measure of monopoly or market power. Another major
dimension is the height of entry barriers, which is determined in part by tech-
nical factors such as the extent of production economies of scale relative to the
size of the market, the absolute amount of capital required to operate a plant
of minimum efficient scale, and other absolute production cost disadvantages of
new entrants.
Product differentiation, a third major dimension of market structure, plays
a dual role. Not only does it directly influence the character of competition
among established firms, but it also raises the height of entry barriers.3 In this
study, however, w-e do not deal directly with product differentiation, hut focus
instead upon advertising expenditures, which are both a sympton and a source
of differentiation. Not only are advertising budgets influenced by product and
market characteristics, but also they depend on the policies pursued by indi-
vidual firms. In addition, past advertising outlays appear to be important deter-
minant of the extent of product differentiation. Differences in advertising,
therefore, reflect both structural and behavioral differences between industries.
On these grounds, the empirical analysis which follows takes the form of
multi-variate regression equations which explain the inter-industry variation
in profit rates as a function of different combinations of the following variables:
Seller concentration,
The rate of growth of demand,
*A preliminary version of this paper was presented at the December 1965 meetings of
time Econometric Society. The authors thank Frank Edwards for his discussion of the
paper at that time, and William Hughes and Lester Taylor for their comments on a sub-
sequent draft. Special thanks are due Richard Caves and R. B. Heflebower, both of whom
provided us with extensive and constructively critical comments.
The authors are grateful for financial assistance provided by the small grant program
of the Ford Foundation and a grant from the Alfred P. Sloan Foundation.
1 Lester Telser, "Advertising and Competition," Journal of Poltical Economy (Dec.
1964).
Low or average profit rates (10 not necessarily indicate that market power is absent.
Firms may become lax in minimizing costs when the discipline of competition is weak.
For a discussion of such behavior, see Carl Kaysen. U.S. vs. United Shoe Machinery Cor-
poration (Cambridge: Harvard University Press, 1956) 114-116.
Joe Bain, Barriers to New Competition (Cambridge: Harvard University Press, 1956)
21,
PAGENO="0543"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2093
Economies of scale in production in relation to the size of the market,
Absolute capital requirements for a plant of minimum efficient scale,4 and
Advertising.
The specific variables used, the alternative functional forms, and other specifica-
tions of the estimated equations are des'cribe~ below. The conceptual relationship
between advertising, product differentiation, and the height of entry harriers is
discussed in the next two sections.
Before proceeding, however, it is useful to contrast the framework adopted
here with that used by Telser. One of his major empirical findings is that the
simple correlation between advertising outlays as a percentage of sales and the
level of seller concentration is statistically insignificant. In each of the years
studied, he finds that thi.~ coefficient is about 0.16, and from this, concludes
that "There is little empirical support for an inverse association between
advertising and competition."
This approach raises the problem of whether concentration ratios are an
adequate measure of the extent of competition. Telser justifies their use by
stating that "Concentration of sales among the four leading firms is a widely
accepted measure of monopoly." 6 While this statement is unfortunately correct,
it ignores the fact that the concentration ratio measures only one dimension
of market structure, and is therefore an inadequate indicator of market power,
which depends on additional structural variables as well as on established be-
havior patterns. The significance of advertising expenditures depends on whether
they represent an additional factor affecting the achievement of market power.
The weak correlation between concentration and advertising simply indicates
that these are independent rather `than collinear variables.
ADVERTISING AND PRODUCT DIFFERENTIATION
The relationship between advertising outlays and product differentiation is
important for an evaluation of the competitive effects of advertising because
the former reflects the policies adopted by individual firms, while the latter is
a dimension of market structure.
The degree of product differentiation in a market is measured by the cross
elasticities of demand and supply which exist among competing products. Low
cross elasticities of demand between these products indicate that buyers prefer
the products or brands of particular sellers and w-ill not switch in significant
num!bers in response to small differences in price. Low cross elasticities of supply,
on the other hand, signify that firms are unable to imitate the products of their
rivals sufficiently well to eliminate these consumer preferences. While cross
elasticities between the products of existing producers affect the character of
the rivalry which exists between them, cross elasticities between the products
of established firms and potential entrants influence the height of entry barriers
posed by product differentiation.7
Product differentiation reflects two sets of factors: the basic characteristics of
products within the market, and the present and past policies of established
firms with respect to advertising, product design, servicing, and distribution.
On the demand side, products are more likely to be differentiable when buyers
are relatively uninformed about the relative merits of existing products. This
is particularly important for differentiation achieved via advertising. On the
supply side, differentiation is more likely where the products of rivals cannot
easily be imitated and where new entrants have difficulties in producing products
which are sirnular to those sold by successfully established firms. In producer
goods industries, successful imitation requires investment in product design
and adequate service facilities. In consumer goods industries, successful imita-
tion may require investment in advertising as well.
No attempt was made to measure any other absolute cost disadvantages of new
entrants. Bain found that only in those industries in which established firms controlled
scarce natural resources were these important. Bain, op. cit., 155-150. In addition, no
attempt was made to measure risk. The sources of the data and various technical ad-
justments are described in the appendix.
Telser, op. cit., 544 and 55S.
°Ibid., 542.
It is important to distinguish product differentiation from product variety. The steel
industry, for example, aroduces a great variety of products which are sold to knowledgeable
buyers. but product differentiation is minimal. In contrast, the cigarette industry offem's
a smaller variety of nroducts, but product differentiation-based largely on extensive
advertising-is great. Bain, op. cit., 127-129.
PAGENO="0544"
2094 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
It is noteworthy that Bath, in his authoritative examination of product
differentiation in 20 manufacturing industries, found advertising to be the most
important source of product differentiation in the consumer good industries
in his sample. Distribution policies are also important where forward integration
in prevalent, while customer services and product design play contributing but
relatively minor roles.
For typical consumer goods industries, then, a persistently high level of
advertising expenditures can be viewed in two ways:
a) If firms behave reasonably, high levels of advertising indicate that the
product is differentiable. In this sense, advertising is a symptom of differentiation.
b) The high level of advertising is itself an important determinant of the
level of differentiation which is realized by established firms vis-~-vis potential
entrants. In this sense, advertising is a source of product differentiation.
Provided that firms act reasonably, observed advertising expenditures provide
a useful measure of the extent of product differentiation. We write reasonably
rather than rationally since, in an oligopolistic market, rational policies are not
unambiguous. What is rational policy for the group acting in concert is not
rational policy for the individual firm expecting to gain a march on its rivals.
It is quite possible, moreover, that rivalry via advertising among established
firms is carried to the point of diminishing returns in terms of group profit
rates. However, even in this case, the result of extensive advertising rivalry may
be to permit the achievement of higher future profits for the group by raising
entry barriers.
THE EFFECT OF ADVERTISING ON ENTRY BARRIERS
Although advertising is only one source of product differentiation, it is espe-
cially important in a number of consumer goods industries where it has a strong
direct impact on entry barriers.8 In these industries, new entrants generally are
forced to sell at a price below the established brands or else incur heavy selling
costs. This explains the phenomenon of unbranded products selling at prices
substantially below those of highly advertised products even where there is
little "real" difference between them. On this account, established firms can
set prices above existing cost levels, including advertising and other selling
expenses, without inducing entry.
Product differentiation via advertising affects entry barriers in three ways,
each of which is analogous to the other determinants of overall entry barriers.
First, high prevailing levels of advertising create additional costs for new
entrants which exist at all levels of output. Because of buyer inertia and loyalty,
more advertising messages per prospective customer must be supplied to induce
brand switching as compared with repeat buying. Since the market which pros-
pective entrants must penetrate is made up largely of consumers who have
purchased existing products, advertising costs per customer for new entrants
will be higher than those of existing firms who are maintaining existing market
positions. Moreover, the costs of penetration are likely to increase as output
expands and customers more inert or loyal need to be reached.° This effect of
advertising creates an absolute cost advantage for established producers, since
they need not incur penetration costs.
In addition, the effect of advertising on firm revenues is subject to economies
of scale which result from the increasing effectiveness of advertising messages
per unit of output as well as from decreasing costs for each advertising message
purchased. The first source of economies will exist whenever the effect of
advertising on consumer decisions is sufficiently important that a threshold
level of advertising is required for a firm to stay in the market and maintain
its current market share. In such a situation, larger firms have the advantage of
being able to spread this cost over more units of output and thereby spend less
per unit sold. This advantage creates economies of scale at the firm level, since
an established firm does not have to spend twice as much on advertising to
8lbid 114-143.
These penetration costs depend on past as well as current advertising outlays by
established firms. The importance of past outlays is examined by Kristian S. Palda who
concludes that "distributed lag models both give a better fit to the Plnkham data and
forecast better than he models which do not incorporate lagged effects." The Measurement
of Cumulative Advertising Effects (Prentice-Hall, 1964), 94.
PAGENO="0545"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2095
maintain a market share which is twice that of a rival. Higher output levels are
associated with lower unit costs.'° As a result, smaller firms, including most
entrants, are placed at a strong disadvantage."
Economies of scale in advertising also result when the cost per advertising
message declines as the number of messages supplied increases. An increased use
of some forms of advertising leads to a lower most per message, and available
evidence suggests that this is very important for advertising on national television
and in national magazines.'2
If advertising in a particular industry is characterized by economies of scale
for either of these reasons, an entrant will suffer an additional cost disadvan-
tage if he enters at a relatively small scale. If he enters at a scale sufficient to
realize available economies of scale in advertising, however, his actions are
likely to influence the price or advertising policies of the established firms. The
possible reactions of established firms increase the costs and risks of entry.
Finally, if economies of scale exist either in production or in advertising, the
need to obtain funds for advertising will give rise to capital requirements over
and above those needed for physical plant~ and equipment. Furthermore, this
investment in market penetration will involve a particularly risky use of funds
since it does not generally create tangible assets which can be resold in the
event of failure. The required rate of return on such capital will therefore be
high.
These various effects are illustrated diagramatically in figure 1. Curve APC
represents average production costs for established and prospective firms, and
l\1ESP is minimum efficient scale in production. Curve AAC describes average
advertising costs for existing firms as well as for new entrants after they
have become established. It denotes unit advertising outlays which are required
in order to maintain a firm's market position and to preserve a given volume
of sales once it has been established. This will depend on both the total level
of advertising outlays and their distribution among established firms, and
therefore, it describes prospective advertising costs for entrants only if existing
firms do not react to any loss of market share. To the extent that they do respond,
required advertising outlays will be higher. Curve ATC, the vertical sum of
these two curves, represents average total costs for established firms." MES
denotes the minimum efficient scale in both production and advertising for an
established firm with a given market share.
In addition, curve AMPC describes average market penetration costs for
new entrants. Penetration costs represent an investment in establishing a market
position and therefore depend on the opportunity cost of capital as well as on
total penetration expenditures.'5
10 In the automobile industry, for example, the two smaller firms during the 1950's were
forced to spend more than twice as much on advertising per car sold as did either Ford or
General Motors. Between 1954 and 1957 Studebaker and American Motors spent annually
on national advertising approximately $64.04 and $57.89, respectively, per automobile sold
while G.M. spent $26.56 per unit and Ford spent $27.22 per unit. Chrysler was in an
intermediate position, spending $47.76 per unit. Leonard W. Weiss, Economics and
American Industry (New York: John Wiley and Sons, 1961) 342.
11 This result occurs within the relevant market. When a firm in a regionally segmented
market expands its national market share by moving into new geographic areas, unit
advertising costs do not decline.
"The extent of discounts given to large advertisers is documented In Federal Trade Com-
mission vs. The Procter 1 Gamble Company, Brief for the Federal Trade Commission in
the Supreme Court of the United States (Dec. 1966), 12-13.
13 For simplicity, we assume here that advertising constitutes the only form of selling
expense.
14 Penetration costs include extra advertising outlays which are required for entry.
These outlays will represent total penetration costs if the price charged by the entrant
is the same as that set by established producers. If the entrant is forced to set a price
below that of existing firms, there are additional penetration costs which equal the price
differential times the amount of output sold by the entrant at the lower price.
81-280-68-pt. 5-35
PAGENO="0546"
2096 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
FIGURE 1. - ADVERTISING AND ENTRY BARRIERS
Unit
Costs
0
This schedule therefore denotes the required rate of return on capital invested
in market penetration times the total expenditure required to establish a given
volume of sales, all divided by the number of units sold. The figure illustrates the
case where average penetration costs rise throughout the relevant range of output.
This assumes that the growing difficulty involved in winning over customers
with stronger preferences for the products of established firms, reinforced by
rising required rates of return as the absolute amount of capital required for
penetration increases with the scale of entry, is not fully offset by economies of
scale in advertising or by bandwagon effects for the new entrant's products.
Curve ACN represents average costs, including peneration costs, for new en-
trants, and MESN is the most efficient scale for entry if the reactions of estab-
lished producers are neglected. From this, it follows that EP is the miniinvm
price at which entry will occur. If MESN is a negligible fraction of the market,
EP is the entry-inducing price. If, however, MESN is a significant fraction of the
market, entry is unlikely to occur even at price EP because the entrant will
expect established producers to contest the encroachment of their market position
through an increase in advertising outlays or by a reduction in price. The gal)
between EP and ATC' represents, therefore, the minimum price-cost margin
which may induce entry.
This figure demonstrates, moreover, that the interaction between rising pene-
tration costs and economies of scale at the firm level is important even if no
allowance is made for the reactions of existing producers. If economies of scale
in both production and in advertising w-ere absent, the relevant price-cost margin
would be simply M, w-hich is less than EP-ATC'.
VARIABLES USED IN THE MULTIVARIATE ANALYSIS
In this section, we brefly define and discuss the rationale for the selection of
each of the specific measures.
Profit Rates
The profit rate variable used is profits after taxes as a percentage of stock-
holders' equity,15 averaged within each industry for firms with assets exceeding
15 This profit rate variable was used originally in Joe S. Bain, "Relation of Profit Rates
to Industry Concentration: American Manufacturing, 1936-4940," Quarterly Journal of
Economics (Aug. 1951), 296-297, and Bain, Barriers to New Competition, 192.
- .- ACN
ATC
MESP M~SN MES Output
PAGENO="0547"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2097
$500,000. This procedure avoids the difficulty, noted by Stigler, of profit with-
drawals in the form of executive salaries in small and closely held corporations.'0
Profit rates are also averaged for the period 1954-1907, which covers a complete
business cycle.
Although the profit rate on stockholders' equity is viewed as a more appropri-
ate variable than the rate of return (including interest) on total assets,'7 we
examined whether the empirical results would be sensitive to this decision. The
simple correlation between the profit rate on stockholders' equity and the rate of
return on total assets is 0.93. In addition, the correlation coefficients between each
of these variables and the 1954 four-firm concentration ratio'8 are, respectively,
0.30 and 0.33. These results suggest that our empirical findings are unlikely to be.
sensitive to the choice of a specific profit rate valuable.
Advertising
In light of the discussion in the preceding section, it is useful to examine the
absolute volume of advertising expenditures by existing firms as well as~ the
advertising-sales ratio. The latter variable probably provides a good indication
of the absolute cost disadvantage of the new entrant at small scales of entry, but
is likely to be a less accurate index of the economies of scale and absolute capital
requirements effects of advertising.
We have, therefore, calculated two measures of advertising intensity: adver-
tising outlays per dollar of sales for firms with assets greater than ~50O,OOO, and
average advertising expenditures per firm among firms which account for 50 per
cent of industry output.'9 Both advertising variables are averages for the years
1954 through 1957.
Seller Concentration
Concentration is a sufficiently promin~nt variable in the literature to warrant.
introducing it in three alternative formulations. First, a trichotomous classifica-
Lion based on Kaysen and Turner's classification of market groups is used.2°
Second, the average four-firm concentration ratios published by Stigler are intro-
duced.2' Finally, a dichotomous classification is constructed on the basis of Bain's
finding that a critical point is reached when the eight-firm concentration ratiG
exceeds 70 per cent.22
Economies of Scale in ~Production
Economies of scale in production presumably exist primarily at the level of the
plant rather than the firm. In the absence of better estimates for most of the
industries in the sample, a measure is derived from the size distribution of
plants within the relevant industries. Since cost minimization is an element of
profit maximization, large multi-plant firms should operate plants which are
sufficiently large to realize available scale economies. Where demand is not a
limiting factor, moreover, competition among firms should lead directly to plants.
which equal or exceed minimum efficient scale.2' At the same time, however, small
plants may exist. These may have been built in an earlier period, before demand
had expanded or a technology which required large scale had been developed, or
`~ George J. Stigler, Capital and Rates of Return in Manufacturing Industries (Princeton
University Press) 1.25-127.
17 This is because firms presumably maximize profits, rather than the sum of profits
plus interest payments. The rate of return on stockholders' equity will therefore be a.
more sensitive indicator of the extent of freedom from competitive constraints.
18 In two cases out of 41, our industry classifications differed from those presented by
Stigler. In both of these cases, "Screens and Venetian Blinds," and "Radio, T.V., and
Phonograph," it appeared that Stigler had combined these industries with smaller, mis-
cellaneous industries. In these calculations, therefore, we used data for the more aggregated
industry to stand for its major component.
`~ The procedure used was to select successive asset size classes of firms until 50 percent
or more of industry sales was covered. The proportion of sales in the boundary size class
required to reach this degree of coverage was used to determine the amount of advertising
and the number of firms from that size class included in both the numerator and denomina-
tor of the measure of advertising per firm. For some industries, the largest size class
accounted for more than 50 per cent of sales. In such cases, the measure is simply
advertising per firm in the largest size class.
20 Carl Kaysen and Donald F. Turner, Antitrust Policy, 27.
21 Stigler, op. cit., 214-215.
22 Bain, "Relation of Profit Rates to Industry ConcentratIon," 314.
23 Since the bulk of the evidence suggests that cost curves in manufacturing are "L
shaped" rather than "U shaped," plants which exceed minimal efficient scale will typically
be efficient plants. See J. Johnston, $tatistical Cost Analysis (New York: McGra,v-}Tfl1,
1960) 44-168.
PAGENO="0548"
2098 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
they may result from the entry of small firms. They may also exist in pockets of
the market which are geographically segmented or may specialize in narrow
product lines which are not representative of the industry generally. It is there-
fore important to select a measure which is insensitive to the entry of single-
plant firms of sub-optimal scale.
The measure used is based on average plant size among the largest plants
accounting for 50 per cent of industry output. This average plant size is divided
by total output in the relevant market to obtain the scale economies variable used
in the regression analyses.21
A test of the reliability of this variable can be made by comparing minimum
efficient scale as a percentage of industry output with Bain's estimates. Not only
did Bain concentrate on a smaller number of industries, but also he used varied
forms of information. Therefore, his estimates can be considered a benchmark
against which to appraise various methods of estimating the extent of scale eco-
nomies. We examine both the method described above and an alternative method,
the Survival Technique, as used by Weiss.25
Of the 20 industries examined by Baiu, data on the size distribution of plants
are available for 19. Across those industries, the correlation coefficient between
Bain's estimates and those derived from the method proposed above is 0.89.26
Estimates b~sed on the Survival Technique are available for 13 of the industries~-~--~.
studied :~y Bain, and the correlation between these estimates and those presented ~
by Bain is 0.66. Whan the comparison is limited to these same 13 industries, the
correlation coefficient between estimates derived from the size distribution of
establishments and those published by Bain is 0.86. The method propose'd above is
more consistent with Bain's estimates than those computed from the Survival
Technique,~ and it will therefore be used in the succeeding analysis.tm
Absolute Capital Requirements
This amount of capital required for entry at the scale of a single efficient plant
is based upon the above estimates of economies of scale. The average output level
of plants at estimated minimum efficient scale is multiplied by the ratio of total
assets to gross sales for the industry.25
The Rate of Growth of Demand
The rate of growth of demand is measured by the rate of growth of sales be-
tween 1947 and 1957.~° A period of this length was chosen iu order to emphasize
the long-run effects of the growth of demand, and the terminal years selected
were both years of nearly full employment for the economy as a whole.
Composite Variables Representing Technical Entry Barriers and Advertising
Industries were classified into three groups on the basis of the two variables
which measure technical entryl barriers (economies of scale and absolute capital
requirements). Dummy variables identifying industries with high and moderate
technical entry harriers were used in some regression models.3'
21 An alternative measure based on average plant size among the largest plants ac-
counting for 70 per cent of output was also constructed. This variable was highly corre-
lated with the variable used.
25 See Leonard Weiss, "The Survival Technique and the Extent of Suboptimal Capacity,"
Journal of Political Economy, (June 1964), 246-261.
~° In these calculations, Bain's estimates of minimum efficient scale for the Steel In-
dustry refer to "Steel Works and Rolling Mills," while for the Copper Industry, they refer
to `Primary Copper."
27 In addition, the survivor technique has the great disadvantage of frequently yielding
indeterminate results. See Weiss, op. cit., 258-259.
28 Ideally, some account should be taken of the slope of the average cost function below
minimal efficient scale. The variable used in this paper, in contrast to the estimates de-
veloped by Bain, ignores this aspect of economies of scale.
`~ This measure is likely on the average to understate capital requirements. The book
value of total assets will normally be less than their replacement cost, as a result of in-
flation in preceding years. In addition, a new firm is likely to have higher input costs
while it is learning the production and distribution techniques required in the market.
tm The ideal measure, of course, would be the rate at which the demand curve shifted
over time. The rate of growth of sales is an exact measure only if the price elasticity
of demand is unity or if prices did not change over the period.
~` High technical entry barriers are assumed to exist if either the scale of an efficient
plant exceeded ten per cent of the market or the capital required for efficient entry
equalled at least $50 million. In addition, if scale economies fell between six and 9.9
per cent and if capital requirements amounted to more than $25 million, the same
classification is designated. Low technical barriers exist when either scale economies
amounted to less than three per cent and capital requirements are less than $25 million
or when the former was less than six per cent but the latter sum was under $10 million.
Moderate technical barriers are assumed in all other cases.
PAGENO="0549"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2099
Similarly, industries were classified into three groups on the basis of the two
advertising variables. Dummy variables identifying industries with high and mod-
erate advertising expenditures were also used as an alternative measure of ad-
vertising intensity in some equations.32
Local Market Dummy T7ariable
In some of the regression equations, a dummy variable was introduced to iden-
tify the `three local market industries.
THE SAMPLE OF thDIJSTRIES
For the reasons presented above, the empirical analysis is confined to con-
sumer goods industries.33 Of the 41 industries included in the analysis, 29 produced
non-durable consumer goods, and the remaining 12 produced consumer durables.
In size, the industries ranged from motor vehicles with average sales of over $20
billion per year to hats wi'th average sales of only $122 million per year. Spuri-
ous size effects are `absent, however, since the dependent variable is expressed
in ratio form. The core of the analysis is based upon unweighted interi'ndustry
regressions. However, as su:bsequent tests indicate that lieteroscedasticity is pres-
ent, weighted regressions are also estimated.
Three o'f t'he industries-soft drinks, bakery products, and dairy products-sell
in local markets, and this factor influences the appropriate measures of both con-
centration and scale economies. Two `techniques `are used to handle this problem:
a) Economies of scale are estimated in relation to the `typical local market, and
the Kaysen and Turner concentration classifications which take into account the
local market character of these industries, are used.
b) In some equations, national concentration ratios are introduced along
with a dummy variable which identifies the ideal market character of these three
industries.
One interesting characteristic of the underlying data is that the distribution of
advertising-Sales ratios across the 41 industries is high'ly skewed. Twenty-five
of the industries have ratios below three per cent while eight have ratios between
three and six per cent and only six have ratios w-hich exceed six per cent. In the
latter group, perfumes have an advertising-sales ratio of 15 per cent, cereals and
drugs ten per cent each, soap nine per cent, malt liquor seven per cent, and soft
drinks slightly over six per cent. Notable industries in the intermediate group
include cigarettes and wines with ratios of about five per cent each.
Although advertising per firm is positively correlated with advertising per dol-
lar of sales, the positions o'f two important industries change radically depend-
ing on which variable is used to measure advertising intensity. While tires and
tubes and motor vehicles have quite low advertising-sales ratios, both are `among
the small group of industries with high or very high average advertising outlays
per firm among the leading firms. As a result, tires and tubes and motor vehicles
are classified respectively as industries with moderate and high overall adver-
tisi'ng barriers.
MAJOR EMPIRICAL RESULT5
The simple correlation coefficients betweer~ profit rates and each explanatory
variable are presented in table 1. All of the coefficients have the expected sign
and all are statistically significant in at least one functional form. Moreover, the
lograrithmic relationship appears appropriate' in the case of the growth of demand
variable, while the opposite is the case with regard to the advertising-sales ratio.
As the latter variable is already expressed as a percentage, it is measured in units
comparable to the dependent variable.
32 High barriers are assumed either when the advertising-sales ratio exceeded eight
per cent or when advertising expenditures among leading firms averaged more than $20
million. The same classification is given to industries where the average ratio fell be-
tween four and eight per cent and average annual expenditures amounted to between
$5 and $20 million. Low barriers exist when the advertising-sales ratio was less than two
per cent and average expenditures less than $5 million or when the ratio lay between two
and four per cent and expenditures did not exceed $1 million per firm. In other cases,
moderate advertising barriers are assumed to exist.
~ Time Petroleum industry is excluded because of the difficulty of obtaining profit data
comparable to other industries in view of the special tax treatment of that industry.
PAGENO="0550"
2100 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE 1.-Simple correlation coefficients-Profit rates and various dimensions of
market structure
Market structure variables
Correlation wi
th profit rates
Natural units 1
Logarithm 2
Growth of demand
0. 17
0.42
Capital requirements
Economies of scale
3 0.43
0.25
0. 57
20.37
Advertising-sales ratio
Advertisingperfirm
Concentration ratio (4 firms)
0.42
3Q43
0.36
0.27
3Q5Q
4Q35
I The units of measurement are described in the text above.
2 In computing these coefficients, the structural dimension is measured in logarithms although the profit
rate is measured in natural units.
3 Indicates coefficient is statistically significant at the 99 percent level.
4 Indicates coefficient is statistically significant at the 95 percent level.
NoTE-Tests of significance are made on the basis of 1-tailed t tests.
The correlation between profit rates and concentration in table 1 is based on a
continuous four-firm concentration ratio. The relationship between these tw-o
variables was also examined in terms of discrete groupings. Industries were
divided according to both the three-way classification scheme proposed by Kaysen
and Turner, and a two-way classification depending on whether the eight-firm
concentration ratio exceeded or was less than 70 percent. The following results
are obtained:
Number of
industries
Average profit
rates
Kaysen and Turner trichotomy:
Type I oligopolies
Type II oligopolies
TJnconcentrated
13
14
14
8.4
9.2
6.3
Dichotomy based on 8-firm concentration ratio at 70 percent:
Concentrated
8
10.0
Unconcentrated
33
7.5
All industries
41
7.9
While the distinction between Type I and Type II Oligopolies seems, to the
average, unimportant, there do appear to be substantial differences in profits be-
tween concentrated and unconcentrated industries. These differences are impor-
tant in both classification schemes.
The core of the empirical work is the multiple regression equations which
relate profit rates to various combinations of the explanatory variables. A.
set of linear equations is presented in table 2. As may be ojserved, the
advertising-sales ratio and the measure of capital requirements appear to
be the most important explanatory factors. Their regression coefficients are
generally significant even w-hen all other variables are included. The vari-
able describing economies of scale seems from these results to be quite weak,
although it has the expected sign in all cases. The advertising per firm coefficient
is significant if the advertising-sales ratio is not included. Where both are in-
cluded, it tends to be insignificant. Advertising outlays per firms are correlated
w-ith absolute capital requirements (the simple correlation coefficient between
these variables is 0.40) and this variable is not as statistically important in con-
junction with the latter as is the advertising-sales ratio.
PAGENO="0551"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2101
TABL
E 2.-~-Multiple regression
equations expla
ining pr
ofit rates-Linear results
Adver-
Econ- Absolute
Growth
Concentration
Intercept tising- Advertising
sales per firm
ratio
omies capital
of scale require-
ments
of de-
mand
classsesa
Type I Type R2
II
Cor-
rected
R2
(1)
(2)
(3) -
(4) -
(5) -
0. 049 *0. 424 0. 00000059
(2.4) (1.1)
0. 052 0. 296 *0. 00000114
(1.6) (2.0)
0. 051 ~0. 437 0. 00000050
(2.1) (1.1)
0.048 **0 499 -
(3.1)
0. 058 *0. 00000112
(2.1)
0. 113 ~0. 000281
(0.8) (3.0)
0. 113
(0.7)
~0. 000282
(3.0)
0.116 **0 009315
(0.8) (3.5)
0. 145 *0. 000227
(0.9) (2.3)
0. 0014
(0.6)
0. 0018
(0.7)
0. 0012
(0.5)
0.0016
(0.7)
0. 0012
(0.5)
-0. 0158 0. 0084 **Ø 47
(1.0) (0.6)
-0. 0058 0. 0115 *_0 32
(0.3) (0.8)
-0. 0091 0. 0126 **0 46
(0.7) (1.0)
-0.0117 0.0085 **045
(0.7) (0.6)
-0. 0147 0. 0146 **0 38
(0.9) (1.0)
**0 34
*0. 19
**0 35
**034
~ 25
a Based on Kaysen and Turner groupings.
NoTE-Figures in parentheses are t values. The statistical significance of the regression coefficients is
tested by means of 1-tailed t test and of the multiple correlation coefficients by means of the F-ratio test.
*Indicates coefficient is statistically significant at the 95-percent level.
**Indicates coefficient is statistically significant at the 99-percent level.
In none of the equations do the estimated coefficients of the concentration
dummy variables exceed their standard errors. In addition, the coefficient for
Type I Oligopolies has a negative sign throughout, which does not coincide with
a priori expectations. While the impact of cOncentration is examined at greater
length below, the linear results suggest that the partial effect of this variable
may be relatively unimportant when it is intrbduced in conjunction with variables
reflecting product differentiation, the height of technical entry barriers, and the
rate of growth of demand.
In an alternative formulation, four dummy varaibles were defined to represent
high and moderate technical entry barriers and high and moderate advertising
intensities, and these were introduced in place of the advertising, economies of
scale, and absolute capital requirements variables. The results are presented in
table 3. The dummy variables designed to measure the influence of technical
barriers are not statistically significant, and in the second equation, the estimated
coefficient for industries with high technical barriers is less than the coefficient
for industries with moderate barriers. This result reflects in part the correlation
between the high technical barrier dummy variable and concentration. The cor-
relation coefficient between the dummy variable for high technical barriers and
the Type I concentration dummy variable is 0.53. When the four-firm concentra-
tion ratio is used, this coefficient rises to 0.68. This collinearity obscures the
separate effects of concentration and technical entry barriers.
TABLE 3.-Multiple regression equations containing composite variables
Advertising Technical entry Concentration a Growth Cor-
Inter- barriers barriers of de- R2 rected
cept mand R2
High Moderate High Moderate Type I Type II (logs)
(1) 0. 056 **0 0552 0. 0080 0. 0154 0. 0090 -0. 0132 -0. 00085 *0. 020 ~ 46 **0 33
(3.3) (0.6) (0.9) (0.5) (0.7) (0. 05) (1.9)
Advertising Technical entry
Inter- barriers barriers
cept -
High Moderate High Moderate
Concen- Regional Growth Cor-
tration industry of de- R2 rected
ratio b dummy mand R2
variable (logs)
(2) 0.044 *0.0403 -0.0049 -0.0150 -0.0013 0.000596 0.0311 *0.024 5*048 ~~036
(2.3) (0.3) (0.8) (0. 1) (1.4) (1.4) (2.4)
Based on Kaysen and Turner groupings.
Four-firm concentration ratio.
NoTE-Figures in parentheses are t values. The statistical significance of the regression coefficients is
tested by means of one-tailed t test and of tl1e multiple correlation coefficients by means of the F-ratio test.
5lndicates coefficient is statistically significant at the 95-percent level.
*allsclicates coefficient is statistically significant at the 99-percent level.
PAGENO="0552"
2102 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
To consider further the impact of concentration, the continuous concentration
ratio was introduced into the analysis. As this variable is available only on a
national basis, a dummy variable was also introduced to identify industries which
sell in local rather than in national markets.34 It is interesting that the con-
tinuous variable appears to have a stronger impact on profits than do the concen-
tration dummy variables. Both the coefficient for concentration and for the local
market dummy variable are significant at the 90 per level.
In contrast to the weak effect of composite technical barriers, the composite
advertising variable has a strong effect. In both equations, the high advertising
variable is statistically significant and the coefficient appears relatively stable.
In addition, the growth of demand variable becomes significant when introduced
in logarithmic form.
Additional sets of regression equations are presented in tables 4 and 5. In both
sets, capital requirements are introduced in logarithmic form and the estimated
coefficients are generally significant. However, some degree of collinearity exists
between the measures of capital requirements and scale economies. When both
variables are introduced into the equations, the variance of the estimates in-
creases, and in some cases, the capital requirements coefficient becomes statisti-
cally insignificant.
In table 4, the advertising-sales ratio is used to measure the entry barriers
created by high advertising expenditures, while in table 5, the dummy variable
representing high advertising barriers is introduced. Both advertising variables
are statistically significant in all of the equations presented, and the estimated
coefficients of both variables are very stable. The dummy variable for high adver-
tising barriers appears to be somew-hat stronger than the advertising-sales ratio.
However, the coefficients of the other structural variables included in these
equations are not sensitive to the particular advertising variable used.
TABLE 4.-Multiple regression equations explaining profits rates- Major findings
with advertising-sales ratio
Inter-
cept
Adver-
tising-
sales
ratio
Capital Econo-
require- mies of
ments scale
(logs) (logs)
Growth Concen- Regional Cor-
of tration industry 122 rected
demand class dummy 122
(logs) variable
(1)
(2)
(3)
0.042
0.042
0. 039
*0.362
(2.4)
*0.362
(2.3)
*0.343
(2.3)
**0 0097
(3.2)
**0 0096 0.000067
(2.5) (0.01)
**0 0105
(2.8)
0.016 *4046 **040
(1.6)
0.016 *4046 **038
(1.6)
0. 015 0. 0043 0. 0278 4*0. 49 *40 40
(1.4) (0.3) (1.5)
(4)
0.038
*0.341
(2.3)
0111
(3.6)
0.014 0.0280 **049 4*042
(1.4) (1.6)
An industry is concentrated if the 8-firm concentration ratio equals or exceeds 70 percent; otherwise it is
unconcentrated.
4lndicates coefficient is statistically significant at the 95-percent level.
**Indicates coefficient is statistically significant at the 99-percent level.
NoTE-Figures in parentheses are t values. The statistical significance of the regression coefficients is
tested by means of one-tailed t test and of the multiple correlation coefficients by means of the F-ratio test
~ These market characteristics had already been accounted for in the Kaysen and
Turner groupings.
PAGENO="0553"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2103
TABLE 5.-Multiple
regression equations explaining profit rates-Major findings
with high advertising barrier
Intercept
High Capital Econo- Growth Concen- Regional
adver- require- mies of tration industry R2 Corrected
tising ments of scale demand class dummy R2
barrier (logs) (logs) (logs) variable
(1)
0.053
**0 0379 *00066 *0.018 **048 **042
(2)
0. 069
(2.8) (2.0) (1.9)
**0 0388 0. 0047 0. 0038 ~0. 019 - - **0 49 **0 42
(3)
0. 048
(2.8) (1.1) (0.7) (2.0)
**0 0395 *0 0089 0. 015 -0. 0063 *0. 0318 **0 53 **0 45
(4)
0.048
(2.9) (2.4) (1.5) (0.5) (1.8)
**0 0379 **0 008~ *0.016 *0.0316 **052 **046
(2.9) (2.5) (1.7) (1.8)
An industry is concentrated if the 8-firm concentration~ratio equals or exceeds 70 percent; otherwise it
is unconcentrated.
NoTE-Figures in parentheses are t values. The statistical significance of the regression coefficients is
tested by means of 1-tailed t test and of the multiple correlation coefficients by means of the F-ratio test.
*Indicates coefficient is statistically significant at the 95 percent level.
**Indicates coefficient is statistically significant at the 99 percent level.
In both tables 4 and 5, a dummy variable identifying industries with eight-
firm concentration ratios which exceed 70 per cent was introduced in order to
examine the effect of this aspect of market structure in yet another specification.
The coefficients, however, remain smaller than their standard errors.
The regional industry dummy variable was also introduced into the equations
in both tables. In all cases, its estimated parameters are significant at the 90
per cent level, and reach the 95 per cent level in table 5. While this variable was
used originally to correct for the use of concentration ratios calculated on a
national basis, it appears to have an independent effect which does not depend
on the presence of the other variable. It is useful, therefore, to compare the
structural features of the three local market industries included in our sample
w-ith `the others. Relevant data are presen'ted in table 6. The scale economy vari-
able is the most sensitive to this industry characteristic. As wo~ild be expected,
these industries, on the average, have much higher estimates of the ratio of
minimum efficient scale to market than do the national industries. We should
expect, th'erefore, that the local market dummy variable represents the increased
importance of economies of scale as well as the higher concentration levels in
local markets.
The multiple correlation coefficients for these equations are aways statistically
significant. The included variables typically' account for about half of the total
variation in industry profit rates.
TABLE 6.-Local and regional industry characteristics
Industry Concentra-
tion class 1
Advertis-
ing-sales
ratio
(percent)
Advertis- `
ing per `
firm (mu-
lions)
Capital
require-
ments
(millions)
Economies
of scale
(percent)
Growth of
demand
(ratio)
Profit rate
(percent)
Soft drinks Type I~ - -
Dairy Type IL~ -
Bakery Type IL~ -
Average of 3 indus-
tries
6. 2
2. 2
2. 9
3. 8
80. 26
15. 12
1. 97
5. 78
$0. 75
2. 09
2. 57
1. 80
8. 2
14. 2
8.3
10. 2
1. 98
1. 16
1. 69
1. 61
10. 0
7. 9
9. 3
9. 1
All industries
3. 3
6. 03 `
24. 32
4. 7
1. 83
7. 9
1 Kaysen and Turner groupings.
PAGENO="0554"
2104 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
THE PROBLEM OF MULTICOLLINEARITY
As was noted above, a number of the explanatory variables included in the
analysis are collinear to some extent. While the simple correlations between the
advertising-sales ratio and the other independent variables are typically low,35
it is useful to examine the sensitivity of the estimated coefficients for this variable
to changes in the specifications of the regression equations. The results are pre-
sented in table 7. The estimated coefficients are reasonably stable, ranging from
0.30 to 0.52. In addition, seven of the 15 coefficients presented are significant
at the 99 per cent level, seven at the 95 per cent level, and the remaining coefficient
at the 90 per cent level.
TABLE 7.-Sensitivity of regression coefficients for advertising variables to changes in
specification of regression equations
A. ADVERTISING-SALES RATIO
[N-variable in natural units; L-variable in logarithms; I and II denote Kaysen and Turner concentra-
tion classes; H and M denote high and moderate dummy variables]
Estimated
coefficient
Value of t Concentra- Regional Capital Scale Growth of Advertis-
tion industry require- economies demand ing per
dummy ments firm
0.42
0.30
0.42
0.43
2.27 1,11 N
1.61 1,11 N
2.40 1,11 N N N
2.50 1,11 N
0.48
2.75 N N
N
0.52
3.24 N N N
0.50
2.92 N N N
0.34
2.21 I L N L
0.35
2.29 I L N L
0.49
3.28 N N L
0.41
2.41 N N L N
0.40
2.54 N N L
0.46
3.02 N N L
0.36
2.43 L
0.33
2.19 N N L
B. DUMMY VARIABLES FOR HIGH ADVERTISING BARRIERS
Estimated
coefficient
Other variables included
-
Value of I Concen- Regional Capital Economies Growth of Technical Moderate
tration industry require- of scale demand barriers adver-
dunimy ments tising
barrier
0.010
0. 049
0.050
2.28 N N L H,M M
3. 67 I, II L
3.25 1,11 N L H,M M
0.045
3.39 N N L
0.038
2.80 L
L
0.039
2.83 L L L
0.038
2.89 N N L L
The stability of the estimated coefficient for the dummy variable denoting high
advertising barriers is also tabulated. These coefficients appear to be insensitive
to changes in the specifications of the equations. They lie between 0.038 and 0.050.
Six of the seven coefficients presented are significant at the 99 per cent level and
the remaining one at the 95 per cent level. The estimated effect of either the ad-
vertising-sales ratio or the high advertising dummy therefore does not appear
to be affected by which of the other variables are included in the equations.
~ The simple correlation coefficients between the advertising-sales ratio and other struc-
tural variables are as follows: the log of economies of scale, 0.27; the log of capital re-
quirements, 0.21; the log of growth of demand, 0.40; and the four-firm concentration
ratio, 0.10.
PAGENO="0555"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2105
While there may be other variables which affect the estimated relationships
between advertising and profits, the importance of both advertising variables is
relatively insensitive to changes in specification of the variables and the models
examined in this paper.3°
Significant correlations also exist between capital requirements, economies
of scale, and concentration. This is not surprising, for one should expect the first
two variabes to have some effect on the latter. At the same time, however, con-
centration is influenced by other factors, such as the past record of merger ac-
tivity in the industry. To examine the extent to which concentration is explained
by scale economies and capital requirements,~ two multiple regression equations
were fitted. The results, which are striking, appear in table 8.
TABLE 8.-Multiple regression analysis-Concentration and technical entry
barriers
Concentration a
Intercept
Capital
requirements
(logs)
Economies
of scale
(logs)
~
Regional
industry
dummy
variable
R2
Corrected
R2
(1) Natural units
(2) Logarithms
49*9
3. 85
**7 08
(3.9)
**0. 244
(5.1)
**6 91
(2.6)
**O. 238
(3.4)
-11.2
(1.2)
-0. 294
(1.2)
**0 71
**Ø 81
**0 68
**0. 79
a Four-firm concentration ratios.
Figures in parentheses are t values. The statistical significance of the regression coefficients is tested by
means of one-tailed I test and of the multiple correlation coefficients by means of the F-ratio test.
a*Indicates coefficient is statistically significant at the 99-percent level.
Absolute capital requirements, scale economies, and the local market dummy
variable ~ together account for a substantial share in the variation in national
concentration ratios. In logarithmic form, over 80 per cent of the variation is ex-
plained by these variables. What is surprising is the small share of variation
left to be accounted for by other factors. With this high a degree of inter-corre-
lation, it is understandable that the estimated coefficients for concentration are
not statistically significant. The role of cencenration appears closely linked to
that of technical entry barriers and there is little remaining influence which is
evident.E
HETEROSCEDASTICITY AND WEIGHTED REGRESSIONS
An examination of the residuals from a leading equation (number 4 in table 5)
revealed that heteroscedasticity is present, as small industries typically have
large residuals. There are two possible reasons for this phenomenon. The smaller
industries may tend to have fewer firms, so that the variance of average profit
rates is larger. The smaller Industries may also have smaller firms. Previous
studies have indicated that the variance of profit rates among small firms is
greater than among larger firms,30 and this would also account for a larger vari-
ance for smaller industries.
~ This result, however, does not apply to average advertising expenditures per firm,
which is more strongly correlated with the other explanatory factors. As a result, its
statistical significance in regression analysis appears to depend on which of the other
variables are included in the estimating equation.
37 The regional industry dummy variable was included because the concentration ratios
are constructed on a national basis. The negative sign on the coefficient represents simply
the downward bias of the national ratios in those industries.
iS One should be wary of drawing any policy conclusion on the basis of this equation.
Merger activity may be highly correlated with entry barriers. Furthermore, there is some
element of spurious correlation between the scale economies measure and concentration.
The scale economies measure used here is 0.5 times the reciprocal of the number of the
largest plants required to account for one-half of industry output. It is, therefore, related
to plant concentration. Since plant concentration and firm concentration may be expected
to be correlated even in the absence of variations in relative scale economies, some
spurious correlation exists between concentration and relative scale economies. (The
authors are indebted to Joe S. Bain for the elaboration of this point.)
~9 Sydney S. Alexander, "The Effect of Size of Manufacturing Corporation on the Dim-
trihution of the Rate of Return, this REVIEW, XXXI (Aug. 1949), 22.9-235.
PAGENO="0556"
2106 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
To determine an appropriate weighting scheme, an empirical approach was
adopted. The variance of the residuals was calculated for successive quartiles in
the distribution of industry sales. From this tabulation, it was clear that the use
of industry sales is inappropriate as a weighting variable as it would give too
much emphasis to the largest industries. The square root of sales, however, is
nearly proportional to the variance of the residuals, and was therefore chosen as
the weighting variable.
The weighted regressions were fitted both for all industries, and for all in-
dustries except motor vehicles, as this industry is an outlying observation with
respect to some of the variables, including the weighting variable. The results
appear in tables 9 and 10. As is clear, the ft2 of each of the weighted regressions
is considerably higher than the W of its unweighted counterpart. This is to be
expected, since the weighting procedure deliberately emphasizes industries with
smaller residuals and the ft2 measures the proportion of the weighted variance of
the dependent variable explained by the regression equation. Another way of look-
ing at this is that weighting essentially involves multiplying the equation by the
root of the weights (in this case by the fourth root of sales) and proceeding by
ordinary least squares. The ft2 indicates the success at predicting profit rates
multiplied by the fourth root of sales.4°
The results are impressive. About 75 percent of the weighted variance across
all industries is accounted for by these equations and about 65 percent of the
w-eighted variance is explained when the outlying auto industry is excluded.
TABLE 9.-Weighted regressions with advertising sales ratio
Inter-
cept
High ad- Capital
vertising- require-
sales meats
ratio (logs)
Econo-
mies of
scale
(logs)
Growth
of de- Concentra-
mand tion ratio
(logs)
Regional
industry 132
dummy
variable
Cor-
rected
132
(1) a. All indus-
tries
0. 040
*0. 29 **0 013
(1.9) (4.8)
0. 0084
(1.0)
028 **9 76
(1.9)
**0 72
b. Motor ye-
hices cx-
cluded~__
0. 045
**0 44 *0 00~7
(3.1) (2.4)
0. 0096
(1.3)
0. 020 **0 67
(1.4)
62
(2) a. All indus-
tries
0.056
*0.28 **0 010
(1.8) (3.1)
0. 0046
(0.9)
0. 0096
(1.1)
**075
**071
b. Motor ve-
hicles ex-
cluded
0. 074
*w 42 0. 0010
(2.8) (1.2)
0. 0052
(1.2)
0. 011
(1.4)
67
62
(3) a. All indus-
tries
0.040
*0.29 *0. 014
(1.9) (2.4)
0.0081 -0. 00003
(0.9) (0.08)
*0.028 **076
(1.8)
**072
b. Motor ve-
hicles ex-
cluded -
0. 017
**0 43 0. 0090
(2.9) (1.6)
0. 0088 -0. 00011
(1.1) (0.3)
0. 021 **Q 67
(1.4)
**0 61
* Indicates coefficient is statistically significant at the 95-percent level.
** Indicates coefficient is statistically significant at the 99-percent level.
NoTE-Figures in parentheses are t values.
The high advertising barrier dummy variable and the advertising-sales ratio
variable are introduced alternatively. The former is significant at the 09 per cent
level in all equations, the latter at the 95 per cent level when the auto industry
is included, at the 99 per cent level otherwise. The collinearity between capital
requirements, economices of scale and concentration is again evident, but in
contrast to the unweighted regressions, the economies of scale variable is some-
times significant when introduced alongside capital requirements (and the latter
variable is sometimes insignificant).
°° It is important to note that the increase in 132 is no indication that the weighting
used is the correct one. Indeed, a very lsigh 132 cnn be obtained by weighting with industry
sales, which is clearly inappropriate. A subsequent test, moreover, was made on the extent
of heteroscedasticity in the weighted regressions. The residuals from equation is in
table 10 were calculated and the successive variances of these residuals were compared
with the mean root sales in the relevant quartile. The fact that the two varinbies were
nearly proportional provides some confirmation of the use of the square root of sales as
the weighting variable.
PAGENO="0557"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2107
These results, moreover, provide additional evidence of the stability of the co-
efficients for these two advertising variables. They also point to the joint signifi-
cance of technical barriers to entry and of concentration, but the collinearity
among these variables prevents precise measurement of the separate effects of
each of these variables.
THE PROBLEM OF CAUSALITY
We have found that the inter-industry variation in profit rates can be ex-
plained quite well by a model incorporating the rate of growth of demand, some
measure of advertising intensity, and variables reflecting the importance of con-
centration and technical barriers to entry. The relationship between profits and
either of the advertising variables introduced into the equations is quite robusL
Throughout this paper, we have assumed that the direction of causality is from
the independent variables to profit rates. Could the reverse be the case?
A plausible case can be made that a significant feedback exists from profits to
advertising expenditures, since advertising reflects the discretionary behavior
of firms as well as the extent of product differentiation. Indeed, we should not be
surprised if a time-series analysis, which emphasizes short-run effects, revealed
that changes in profits preceded, rather than followed, changes in advertising
expenditures.
There are a number of factors, however, which suggest that the causality of
the observed relationships runs largely from advertising expenditures to profits.
A cross-sectional study tends to emphasize the long-run differences between in-
dustries, and this in turn is more likely to reflect the structural rather than the
behavioral aspects of advertising. Profit levels cannot influence those market
and product characteristics which permit product differentiation via advertising.
Firms with high profit rates will not have higher optimum advertising expendi-
tures than firms with low profit rates in the same market situation. The pursuit
of profits will hence limit the extent to which profits will be "spent" on advertis-
ing, especially over a period of several years.
TABLE 10.-Weighted regressions with high advertising barrier
Inter-
High
adver-
Capital Econ- Growth Concen- Regional
require- omies of tration industry 112
Cor-
rected
cept
tising
barrier
ments of scale demand ratio dummy
(logs) (logs) (logs) variable
112
(1) a. All industries_ -
b. Motor vehicles
0. 052
**0 035
(2.9)
**0 0080 0. 012 *0. 027 ~0. 78
(2.5) (1.5) (1.9)
**Ø 75
excluded
0.055
**0 032
*0 0064 0. 012 0. 025 **Ø 65
**Ø 59
(2) a. All industries - -
0. 090
(2.7)
**0 037
(1.8) 1 (1.5) (1.7)
0. 0034 *0 0073 0. 013 **Ø 78
**0 75
b. Motor vehicles
(3.1)
(1.0) (1.7) (1.6)
~
excluded
0.097
**Ø 032
0. 0010 *0. 0080* *0. 014 **0 65
**0 60
(3) a. All industries - -
b. Motor vehicles
0. 053
(2.7)
*20 035
(2.9)
(0.3) (1.9) (1.8)
0. 0093 - -- 0. 011 -0. 00010 *0. 028 **0 78
(1.6) (1.3) (0.3) (1.9)
~
**0 75
excluded
0. 057
**Ø 031
(2.4)
0. 0085 0. 011 -0. 00016 *0. 027 2*0 65
(1.5) (1.3) (0.4) (1.8)
*20 58
*Indicates coefficient is statistically significant at the 95-percent level.
**Indicates coefficient is statistically significant at the 99-percent level.
Figures in parentheses are t values.
In addition, if high profits lead to high advertising expenditures, we should
expect that industries which have high profits for reasons other than product
differentiation (e.g., concentration or technical entry barriers) would tend to
have high advertising expenditures as well. Yet, as we noted above, advertising
is only weakly correlated with the other dimensions of market structure.
CONCLUDING COMMENTS
On the basis of these empirical findings, it is evident that for industries where
products are differentiable, investment in advertising is a highly profitable ac-
tivity. Industries with high advertising outlays earn, on average at a profit rate
which exceeds that of other industries by nearly four percentage points. This
PAGENO="0558"
2108 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
differential represents a 50 percent increase in profit rates. It is likely, moreover~
that much of this proftt rate differential is accounted for by the entry barriers
nreated by advertising expenditures and by the resulting achievement of market
power.
We note also the significant joint impact on profit rates of concentration and
4he entry barriers created by scale economies and high capital requirements. Al-
4hough the composite effect of these factors is clearly important, a more precise
indication of the distinct effect `of any of these variables is hazardous because
of the high degree of collinearity. As would be expected, the rate of growth of
demand has an important positive impact on profits. Models which incorporate
~these variables fit the underlying data reasonably well, accounting for approxi-
matey 50 per cent of the variation in industry profit rates.
These empirical results suggest that factors which promote product differentia-
flion may be as important as those which influence the size distribution of firms
:in terms of their effect upon the achievement of market power. Current policies
-which tend to emphasize the role played by concentration may need to be supple-
:mented by those concerned directly with the nature and extent of product dif-
ferentiation. Policies dealing with these matters would be an important compon-
sent in a general policy designed to promote competition.
APPENDIX
DATA SOURCES AND TECHNICAL ADJUSTMENTS
The industry data used are reported at or aggregated to the level of I.R.S.
"minor industries," which are roughly comparable to S.I.C. three-digit industry
groups. The source for each variable is listed in table Al.
The sample was chosen originally to gain complete coverage of all consumer
goods industries. All "miscellaneous" industries were eliminated, however, be-
cause of the obvious conceptual problems. In addition, three other industries were
dropped from the sample: newspapers, while technically a manufacturing in-
dustry was considered to have sufficient "service" elements to make its inclusion
inappropriate; petroleum refining, because of the unusual statistical problems
which result from the tax treatment `of mineral depletion; and motor vehicle
parts, because of the lack of comparable Census data. Average profit rates and
advertising-sales ratios for the remaining 41 industries are presented in table A2.
The variables are defined and explained in the text. The calculation of the
technical entry barrier variables and the rate of growth of demand involved using
both Census and I.R.S. data. The various specific adjustments made to reconcile
data drawn from these two sources and reported at different levels of aggrega-
tion are described in the next two sections.
Technical Barriers to Entry
These variables are based on data from the 1954 Census of Mann factures. To
carry out `these computations, it is necessary to relate industries as defined by
the Census Bureau to those of the Internal Revenue Service. This is done on the
basis suggested by the Census Link Project.4'
Within 5.1.0. four-digit industries, average plant size among the largest plants
which account for 50 per cent of industry output is used as the estimate of mini-
mum efficient plant scale (MES). Data on shipments are used in all cases where
available. In the few remaining cases, the calculations are based on value added.
When the ratios of MES to industry output are obtained, the average percentage
among component four-digit industries within the relevant I.R.S. industry is
calculated, using shipments as weights where available and value added as
weights elsewhere.
In determining the capital requirements variable, the scale of an efficient plant
is measured in most instances by the value of the shipments but in a few by
value added. In the latter cases, these figures are multiplied by the ratio of ship-
ments to value added for the same four-digit industry but in a later year.
41Bureau of the Census, Enterprise Statistics (1958), Part 3.
PAGENO="0559"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2109
TABLE Al-SOURCES OF DATA
Variable Source
(1) Profit rate Internal Revenue Service Source Book of Statistics of IncomeS
Average values for 1954-57.
(2) Advertising Do.
(3) Concentration:
(a) Trichotomous and dichotomous classi- Carl Kaysen and Donald F. Turner, Antitrust Policy, statistical
fications. appendix.
(b) Continuous 4 firm ratio George J. Stigler, Capital and Rates of Return in Manufacturing
Industries, 206-215.
(4) Economies of scale relative to market 1954 Census of Manufactures.
(5) Absolute capital requirements 1954 Census of Manufactures and Internal Revenue Service Source
Book of Statistics of Income.
(6) Rate of growth of demand Internal RevenueServiceSource Book ofStatisticsof Income.
(7) Local market dummy variables Carl Kaysen and Donald F. Turner, Antitrust Policy, statistical
appendix.
When estimates of MES measured in shipments for all four-digit industries
are obtained, these are averaged, using value added as weights, to derive the
value in the larger I.R.S. industry. These averages are then multiplied by the
appropriate assets-sales ratio for the I.R.S. industry, and the resulting figures
used to represent the level of capital required for efficient entry.
In the case of the motor vehicle industry, Census data are unavailable. Bain's
estimates for this industry are therefore used for both the extent of scale econ-
omies and the level of capital requirements.
For the three regional industries-soft drinks, dairy products, and bakery
products-it is assumed that the appropriate market is the typical large metro-
politan area. Output data are not available for four-digit industries by standard
metropolitan area. Consequently, value added data for larger three-digit group-
ings are used. Average value added in large metropolitan areas for the three-digit
industry is multiplied by the ratio of total national shipments in the relevant
four-digit industry to total value added in the associated three-digit industry
to obtain the estimate of four-digit industry shipments within the typical local
market. This figure is then used as the denominator in the estimate of M.E.S. to
market for the four-digit industry.~ Where necessary, the resulting estimates are
aggregated to the I.R.S. minor industry level as described above.
Rate of Growth of Demand
This variable is the ratio of I.R.S. gross sales in 1957 to that in 1947. In a few
cases, however, LR.S. data are not available for both years and alternative pro-
cedures are used.
In 1947, the I.R.S. industries "Cigars" and "Cigarettes" are aggregated as
"Tobacco Manufactures." To disaggregate the reported figure for gross sales,
Census data on value of shipments (excise taxes excluded) for 1947 were ex-
amined. The ratio of shipments in "Cigars" to total tobacco manufacturing is
multiplied by I.R.S. gross sales for tobacco manufactures, and the resulting prod-
uct used to denote gross sales in "Cigars." Gross sales in "Cigarettes" is obtained
residually.
~ Output data were available for 56 standard metropolitan areas for the beverage in-
dustry (S.I.C. 208) ; 59 such areas for bakery products (S.I.C. 205) ; and 61 such areas
for dairy products (SIC. 202).
~ These corrections were made for the following subindustries: SIC. 2081-bottled
soft drinks: SIC. 2021-creamery butter; SIC. 2027-fluld milk and other products~
SIC. 2051-bread and related products.
PAGENO="0560"
2110 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE A2.-AVERAGE PROFIT RATES AND ADVERTISING-SALES RATIOS IN 41 CONSUMER GOODS INDUSTRIES,
1954-57
[In percent[
Profit rate Advertising sales ratio
(1) Soft drink 10. 0 6. 2
(2) Malt liquors 7. 2 6. 8
(3) Wines 7.3 5.2
(4) Distilled liquors 5. 0 2. 1
(5)Meat 4.6 .6
(6) Dairy 7.9 2.2
(7) Canning 6. 4 2. 9
(8) Grain mill products 7.0 1.9
(9) Cereals 14.8 10.3
(10) Bakery prodscts 9. 3 2. 9
(11) Sugar 5.8 .2
(12) Confectionary 10.6 3. 5
(13) Cigars 5.3 2.6
(14) Cigarettes 11.5 4.8
(15) Knit goads 3.8 1.3
(16) Carpets 4.5 2.0
(17) Hats 1.6 2.2
(18) Men'sclothing 5.9 1.2
(19) Women's coithing 6. 1 1. 8
(20) Millinery -1.3 .8
(21) Furs 5.7 1.0
(22) Furniture 9. 7 1. 5
(23) Screens and venetian blinds 9.3 1.6
(24) Periodicals 11. 7 - 2
(25) Banks 10. 1 2. 4
(26) Drugc 14. 0 9. 9
(27) Soaps 11.7 9.2
(28) Paints 9.9 1.5
(29) Perfumes 13. 5 15. 3
(30) Tiresand tubes 10.2 1.4
(31) Footwear 7.6 1.5
(32) Handtools 11.4 4.2
(33) Household and service machinery (not electrical) 7. 3 1. 9
(34) Electrical appliances 10. 3 3. 5
(35) Radio, TV, and phonograph 8.8 2.2
(36) Motorcycles and bicycles 5. 2 1. 1
(37) Motor vehicles 15. 5 - 6
(38) Instruments 12. 0 2. 0
(39) Clocks and watches 1. 9 5. 6
(40) Jewelry (precious metal) 5. 3 3. 2
(41) Costume jewelry 1.4 4.0
I.R.S. data for "Screens and Venetian Blinds," in 1947, are included in
"Miscellaneous Furniture." As a result, Census value added data for both 1947
and 1957 are used. Similar data problems exist in the case of "Drugs," "Per-
fumes," "Instruments," and "Costume Jewelry." In all of these instances Census
data in 1947 and 1957 are used. With regard to "Perfumes" and "Costume
Jewelry," information on value of shipments is available in both years and is
therefore used, while value added data are used in the other cases.
Concentration
The classification of industries into Kaysen and Turner concentration classes
is based on data for 1954, and their definitions are used in most cases. Where
I.R.S. minor industries include a number of Kaysen and Turner markets,
weighted averages based on value of shipments are used to determine the appro-
priate classifications.
Further adjustments are made to account for the local market character of
three industries: soft drinks, dairy products, and bakery products. In addition.
the soap industry falls just below the boundary between Type I and Type II
Oligopolies. Since the same major firms are dominant in the important product
lines, the industry is reclassified as a Type I Oligopoly. The radio and television
industry falls just above the same boundary. It is classified as a Type II Oligopoly
since the largest single subindustry falls into that category. For appliances, the
Kaysen and Turner definition differed substantially from that used by the In-
ternal Revenue Service. After an appropriate adjustment the concentration data
indicated that this industry is a Type II Oligopoly.
PAGENO="0561"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2111
The classification of industries according to whether the eight-firm concentra-
tion ratio exceeded or fell short of the 70 per cent level is based on the concen-
tration ratios presented by Kaysen and Turner.
(Whereupon, at 11 :55 a.m., the hearing was adjourned subject to call
of the Chair.)
S1-280-68-pt. 5-36
PAGENO="0562"
PAGENO="0563"
APPENDIXES
APPENDIX I-CORRESPONDENCE BETWEEN MR. BENJAMIN GORDON AND DR. P. M.
COSTELLO RE TECHNOLOGICAL PROGRESS IN THE PHARMACEUTICAL INDUSTRY
MARCH 6, 1968.
Dr. P. M. COSTELLO,
Department of Economics, Smith College,
Northampton, Ma$s.
DEAR DR. COSTELLO: It is my understanding that you have done considerable
research concerning technological progress in the pharmaceutical industry.
I should be very grateful if you would submit to our Monopoly Subcommittee
a brief summary of your findings in this field. Your contribution on this aspect
of the subject would be very useful in our study of this industry.
Your cooperation is appreciated.
Sincerely,
BENJAMIN GORDON,
Staff Economist.
SMITH COLLEGE,
No-rthampton, Mass., March 13, 1968.
Mr. BENJAMIN GORDON,
Staff Economist, Select Committee on Small Business,
U.S. Senate, Washington, D.C.
DEAR MR. GORDON: In reply to your letter of the 6th, I am enclosing a paper
which examines the pace of technological progress in the ethical drug industry.
This study was developed from research for my doctoral dissertation which con-
sidered the narrower question of patents in the antibiotic segment of the industry.
Sincerely,
P. M. CosTELLO.
TECHNOLOGICAL PROGRESS IN THE ETHICAL DRUG INDU5TRY*
(By P. M. Costello, Assistant Professor of Economics, Smith College,
Northampton, Mass.)
One critical test of an industry's market performance is its rate of tech-
nological progreSs. For the ethical drug~ industry this dimension of performance
is of added significance in view of the increasing role played by drugs in the
management of human diseases. This paper examines a sample of 528 new drugs
innovated between 1945 and 1965. The first section presents estimates, based upon
medical evaluation and opinion, of the proportion of these innovations that
represent significant advancement in the art of chemotherapy. The second section
considers the problems of constructing an adequate criterion `to judge the pace
of advancement from the antibiotic segment `of the industry.
I. THE RECORD OF TECHNOLOGICAL PROGRESS
The sample of new drugs introduced `to the market between 1945 and 1965 was
compiled mainly from the American Medical Association's New and Nonofficial
Drugs. In general the drugs selected for analysis were new chemical structures
excluding combinations of old drugs and new dosage forn~s. The exclusion of the
latter two categories was founded on the assumed lower order of scientific ability
required to produce such products. But even within the category `of new chemical
structures major differences in the effectiveness of these drugs were immediately
apparent.
*An earlier draft of this article received valuable comment from Professors Joel B.
Dirlam and Robert Britt.
2113
PAGENO="0564"
2114 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In an attempt to meet the qualitative problem two categories are defined. The
first is designed to include drugs which were in one degree or another an im-
provement over older products. These varied from drugs which were considered
medical breakthroughs to drugs which reduced the incidence of undesirable side
effects observed with established chemotherapeutic agents. Drugs which fell in
this category were counted as examples of technological progress.
The second category includes those drugs which from a technical point of view
were no advancement over older products. The range in this category is from
drugs which had the same degree of effectiveness and the incidence of side effects
as older products to drugs which were removed from the market or limited
in their application because of undesirable side effects. Also included in this
category are drugs where medical opinion was divided or clinical experience
too limited to form a statistically valid sample of the value of the drug. Drugs
falling in this category were counted as examples of product differentiation which
exhibit no significant technological progress.
To classify each of the drugs in `the sample the Medical Letter1 and the AMA's
New and Nonofficial Drugs were selected as authoritative publications that best
summarized clinical experience with each drug. Thus the classification of a
drug is directly dependent upon its technical efficiency relative to a particular
medical problem. This scheme gives no weight to sales volume as an index
of the relative value of a drug. This recognizes that physicians are responsive
to the promotional claims of the manufacturers and will prescribe a new drug
before sufficient clinical experience is amassed to establish the comparative
value of the new drug relative to older products. As a result a drug significant
in the management of a disease of low incidence in the population and a corre-
spondingly low volume of sales receives the same weight as a drug which cures
diseases of high incidence and high volume of sales.2
Biases in the sample are undoubtedly numerous and difficult to evaluate and
could operate in either direction. First, the source p~iblications do not clearly
indicate the basis of selecting a particular drug for discussion. Inclusions may
well be in response to the promotional claims of the manufacturer with the
recognition that the majority of practicing physicians have neither the facilities,
the time, the training, nor the range of patients necessary to conduct statistically
valid tests of such claims. Second, definitive judgment of the place of a particular
drug in the physician's tool kit may require several years of experience before
the population treated is sufficiently large to expose undesirable side effects.
Thus, a drug currently considered significant may, with further testing, be
found to have only a restricted application. Third, drugs now considered in-
significant may, through modifications in techniques of administration, or in
combinations with other drugs, be found valuable in the management of specific
diseases. Finally, there is a bias in restricting the sample to new chemical struc-
tures. It is possible that combinations of old drugs, for example, may increase
the number of diseases controllable through chemotherapy. But there is no
evidence `to date that this technique has produced anything of importance.
In Table I of the sample of 528 new chemical drugs is divided according to the
classification scheme outlined above. Column I contains estimates of the total
number of new chemical structures innovated yearly between 1945 and 1965.
1 The Medical Letter is a publication designed to acquaint the practicing physician with
the latest information on drug products.
2 For a different approach to the qualitative problem see W. S. Comanor, "Research and
Competitive Product Differentiation in the Pharmaceutical Industry in the United States.'
Economics, Vol. 31. November 1964, pp. 372-84; and "Research and Technical Change in
the Pharmaceutical Industry." Review of Economics and Statistics, Vol. 47, May 1965,
pp. 182-90. Comanor's investigation of the whole range of new drug Innovations weights
each innovation by its sales volume for the first two years following Its introduction. My
objection to following this approach In the present study is that the time required to
develop experience with any particular new drug to support a definitive opinion on its
technical worth seems to exceed the two-year period following introduction.
PAGENO="0565"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2115
TABLE I-ESTIMATE OF TECHNOLOGICAL PROGRESS FOR NEW CHEMICAL DRUGS, 1945-65
Year
Estimate of the oumber Estimate of the number
Total new chemical representing product representing techno-
drugs differentiation logical progress
Col. III as a percent
of col. I
(I)
(II) (III)
(IV)
1945
3
2
1
0.33
1946
16
14
2
.31
1947
16
11
5
1948
14
10
4
.29
1949
28
21
7
.08
1950
25
23
2
1951
21
17
4
1952
22
21
1
.13
1953
38
33
5
1954
35
33
2
1955
27
24
3
.11
1958.
27
26
1
1957
38
35
3
1958
41
37
4
.10
1959
52
46
6
.12
1960
34
28
6
1961
18
16
2
.10
1962
20
19
1
1963
16
14
2
.13
1964
16
15
1
1965
21
20
1
.05
Total
528
465 63
.12
Industry figures for new chemical drugs are available for the period 1951-61.°
Comparing these figures with those in Column I indicates that the industry
data exceeds the sample data by a yearly, average of 10.8 drugs. In part this dis-
crepancy is due to differences in the year of introduction reported for certain
drugs and perhaps due in part to differences in the definition of new chemical
drugs used by the industry and the AMA. It seems likely however, that if any
of these innovations reported by the industry were important they would have
been reported to the medical profession through the AMA's publications. The
effect of including the industry data here would be to reduce the proportion of
total innovation that represented technological advancement.
Beginning in the late 1940's there is a persistent increase in the number of
drugs marketed each year until 1959, followed by a marked decline up through
1964. The series representing technological progress, Column III, however fails
to increase appreciably. For the period 1945-50 the average ratio of innovations
to advances is 5 to 1, while for the period after 1950 this ratio increases to an
average of 10 to 1. Both series exhibit ,a decided break appearing in 1960 and
1961 respectively.
These broad movements can in part be explained by a combination of structural
factors internal to the industry and external institutional factors. For the
1945-50 period the industry engaged in a process of forward vertical inte-
gration. The objective of this integration was better control over prices through
the elimination of the firms solely engaged in packaging. The major example was
set by Parke Davis and Lederle, the innovators of the first broad spectrum drugs.
Both of these firms were fully integrated into the retail pharmacy and hospital
markets and held strong patents on their innovations. In refusing domestic li-
censes on these products the traditional marketing pattern in the industry was
altered. With the elimination of the packagers, direct control over price became
possible. This example of the use of the patent and forward integration was un-
doubtedly a factor in Merck's merger with Sharp and Dohne as well as Pfizer's
entry by internal expansion.
The structural change in the industry suppressed one form of competition-
imitation from packagers or from integrated firms under license agreements.
The alternative avenue open wa's to invent around the patent to produce a close
substitute. Thus the increase in the proportion of innovations which represented
product differentiation was partly in response to the closing of one method of
competition.
~ Pharmaceutical Manufacturers Association, Prescripting Drug Industry Fact Book
(Washington, PMA), 1963, pp. 2-5.
PAGENO="0566"
2116 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The decline in both innovation and technological progress observed in the
early 1960's can be explained as the industry's reaction to criticism generated by
the Kefauver investigation and the stricter controls imposed by the FDA during
this period. Table II indicates that the rate of increase in expenditures for
research and development fell measurably in 1960, leading to a decline in inno-
vations as well as in the number that represent technological progress.
TABLE Il-RESEARCH AND DEVELOPMENT EXPENDITURES AND EMPLOYMENT FOR ETHICAL DRUGS
Year
Expenditures Percentage R. & D. Percentage
(millions) change employment change
(thousands)
1956
$94
1957
104 0.11 4.7
1958
128 .23 5.1 0.08
1959
154 .20 5.9 1.6
1960
163 .06 6.0 0.2
1961
181 .11 6.2 0.3
1962
198 .09 6.7 0.8
1963
215 .09 6.8 .01
1964
238 . 06 7. 2 . 06
1965
268 .13 7.6 .06
Source: National Science Foundation, Review of Data on Science Resources, Washington, various issues.
In general, if technological progress is measured as the percentage of ismo-
vations w-hich find general medical acceptance, it is clear that over time the pro-
portion of drugs that can be classified as representing meaningful progress de-
clines. Or to state the case somewhat differently, the series on the number of
new drugs of lasting value appears rather constant over time in spite of the
marked increase in the number of innovations. These data then tend to support
the observation that while the industry has increased its expenditures on re-
search and development, much of the added effort appears to have resulted in
product differentiation.
IT. CRITERION FOR JUDGING PERFORMANCE
Attempts to judge the "fairness" of prices or profits are generally made on
the basis of comparisons of price to costs of production, or profits of one firm in
comparison with those of another. But with technological progress as the per-
formance variable no such objective criterion exists. In the present case we do
not know whether the proportion of innovations representing technological prog-
ress w-ould have been higher or low-er if the industry did not, for example, have
access to patent protection. If the research resources which were directed to
patent circumvention had been used to attack more significant problems, it is pos-
sible that the research output would have been higher. But it is also possible
that given the quality of these resources, product modification was the best they
were capable of in their employment in the industry. The problem is of course
a familiar one, in that w-e cannot rerun the industry's reactions under different
structural characteristics.
An alternative approach is to compare progress under different market struc-
tures. But even here competing hypotheses of the effect of structure on perform-
ance are in existence and debated. The Schumpeter version that only firms with
some degree of market power have the resources to innovate is opposed by the
hypothesis that innovation offers competitive firms escape from the rigors of
competition. The antibiotic segment of the drug industry does shed some light
on these competing hypothe~s and can aid in the judgment of whether prog-
ress has been adequate.
Table III presents data on the innovation record for the ant.ibiotic segment
of the industry. Total innovations are divided between the so-called narrow- and
broad spectrum drugs corresponding to a competitive segment and a monopo-
lized segment of the market. In the narrow- spectrum market approximately 30
percent of these innovations represented some degree of technological progress,
w-hile in the broad spectrum market one-half of the innovations were signifi-
cant in this respect.4 In light of different market structures associated w'ith these
The four broad spectrum drugs and their classification are as follows: Both
chioramphenicol and chlortetracycline are counted as evidence of technological progress.
Oxytetracycline and tetracycline are counted as evidence of product differentiation as both
are very close substitutes for chiortetracycline.
PAGENO="0567"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2117
two markets, the differences in the proportion of innovations representing an
advance lends initial support to the hypothesis that monopoly is a necessary
condition for progress. But examination of the development of the broad spec-
trum market and its operation weakens this conclusion.
TABLE Ill-PRODUCT INNOVATION IN THE NARROW AND BROAD
SPECTRUM ANTIBIOTIC MARKETS
Narrow spectrum Broad spectrum
Year innovations innovations
1945 1
1946 1
1947 2
1948 1 1
1949 1 1
1950 3 1
1951 1
1952 1
1953 2 1
1954 2
1955 3
1956
1957 3
1958 3
1959
1960 3
1961
1962 2
1963 1
1964 2
1965 2
Total 34 4
The origins of the antibiotics are found in Alexander Fleming's discovery of
penicillin in 1928 and in the development research undertaken by a group headed
by H. W. Florey in England in the late 1930's. England's involvement in the war
and problems of mass producing penicillin caused Florey to seek commercial
sources of supply in the United States. The majority of firms contacted by
Florey expressed little interest in pursuing the research necessary for large
scale production. But he was successful in his contact with the Department of
Agriculture's Northern Regional Research Laboratory (NRRL). With the entry
of the United States into the war a cooperative program was instituted by the
Committee on Medical Research (CMR). The program included a number of
drug firms, university and government laboratories, as well as laboratories in
England. The commercial laboratories of the drug firms initially concentrated
their research on the synthetic production of penicillin,5 while the noncommercial
laboratories concentrated on the problem of mass production. The development
of improved mold stains by the university laboratories and the deep fermenta-
tion' production process by the NRRL essentially solved the key problems of
mass production.°
By the mid-1943 the CMR shifted the emphasis in the penicillin program from
research to production. To overcome the reluctance of the firms to abandon
their synthetic research and adopt the available technology, the program was
expanded to include firms not initially~ under contract with the government
It has been argued that the drug firms concentrated on the synthete approach at the
expense of the production technology suggested by the NRRL. The reason offered for this
approach was that product patents were unavailable on penicillin, and the Department of
Agriculture held the key process patents. Richard Harris, Annals of Legislation, the
Real Voice, The New Yorker, for March 14th, 21st and 28th, 1964. Especially page 69
of the March 14th issue, ". . . the firms were too busy trying to corner patents on the
various processes in the production of penicillin to produce much of it, and the government
began to press them to work together." A similar conclusion was reached in Federal Trade
Commission, Economic Report on Antibiotic Manufacture (Washington, Govt. Printing
Office, 1958), pp. 37-38. In defense of the firms involved it should be noted that his-
torically, the synthetic production of drugs had made the `natural technology' obsolete. For
a description of Merck's experience and others see, Tom Mahoney, The Merchants of Life
(New York, Harpers, 1959), Chaps. 11, 13, 16.
Until the mid-1940's penicillin was produced by two methods. The less efficient `sur-
face technique' involved the growth of penicillin on the surface of the culture medium.
With the `deep fermentation technique' the medium was aerated allowing growth through-
out the medium.
PAGENO="0568"
2118 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
through the CMR and pressure on those under contract to begin production. In
addition, government funds were provided for plant construction and modifi-
cation,7 and as a result, the War Production Board permitted the sale of peni-
cillin through normal trade channels in March of 1945.
After the removal of government restriction on the safe of penicillin, market
forces operated to reduce prices. Tariff Commission data indicate that between
1945 and 1950 the realized price of penicillin fell from $3,055 to $282 per pound,
while production increasing from 11,746 to 329,746 pounds.8
The behavior of penicillin prices can be explained by two factors-conditions
of entry and the existence of a bulk market. A product patent on penicillin was
unavailable, and the process patents were held by the Department of Agriculture,
w-hich follow-ed a policy of freely licensing applicants; entry could not be re-
stricted in manufacturing or selling based upon patents. Available evidence indi-
cates that there were 16 penicillin suppliers in 1945, and 13 in 1950.~
The second factor was that several of the major drug producers were not
integrated forward into direct selling to hospitals and retail pharmacies. These
firms generally sold a significant proportion of their output through the bulk
market to packagers who resold under their own brands. The FTC study re-
ported 27 packagers in 1950, increasing to 45 by 1956, concluding that entry
at this level was relatively easy.1° Thus packagers purchased their supplies in
the open market and sold in competition with the integrated producers.
The emphasis on price competition in the penicillin market was not at the
expense of product improvement. Various modificat~ions of penicillin were
introduced. For example, Squibb introduced the first orally effective penicillin
in 1945, Wyeth introduced benzathine penicillin which has a slower rate of
absorption in 1951, and in 1955 Lilly marketed penicillin V claiming higher
potency. Product patents were generally obtained on these specialties but usually
only after interference proceedings had been declared by the Patent Office.
Settlement took the form of withdrawal of a competitive patent application in
return for a license to produce and sell under any patents that be issued. While
data on realized prices of these products are not reported separately by the
Tariff Commission, these prices generally drifted downward following the price
of the basic drug although at a somewhat higher level.
Thus in the penicillin market the underlying condition of freedom of entry
and the bulk market produced a market competitive in price and with product
improvement.
The second major antibiotic following penicillin was streptomycin, discovered
by Selman Waksman at Rutgers University in 1943. Under the terms of the
research contract between Waksman and Merck, the latter received exclusive
commercial rights to this drug. To exploit its position, however, Merck had to
first convince the Patent Office that streptomycin was sufficiently important to
warrant patent protection,11 and second, to integrate forward into the retail
market to avoid dependence upon the bulk market. Both steps were necessary.
For with the sale of a patent product to packagers, the patent monopoly was
exhausted and price cutting could legally begin.12
The demonstration of the potential importance of streptomycin resulted in
Waksman's request to Merck that right of commercial exploitation be given
up. The sequent agreements provided for the assignment of the patent applica-
tion to a nonprofit foundation at Rutgers, and royalty abatement of $500,000 to
cover Merck's development costs. In 1948 the patent was issued to the founda-
tion which followed an unrestrictive licensing policy.13 Seven firms in addition to
Merck were licensed to manufacture streptomycin. Four of these firms, including
Merck,14 were not integrated forward, selling primarily in the bulk market to
FTC, op. cit.. pp. 47-56.
U.S. Tariff Commisoion. PVntlzetzc Organic Chemicals, United Production. and ,S'ale,
Reports No. 159 and 167 (Washington. Govt. Printing Office), various issues.
° The decline in the number of manufacturers was largely due to their adoption of
inefficient production techniques. With the end of military purchases and the onset of
price competition these firms wculd not cover costs and transferred their resources to
other markets.
10 FTC. on. cit., p. 66.
11 Prior to streptomycin the Patent Office took the position that the isolation of products
of naturo did not rise to the level of invenion and hence were unpatentabble. Merck
successfully arrued that antibiotic products were transitory in neture and that their
isolation constituted invention within tile purpose of the statutes. F. Cacciapaglia, Jr.. nail
H. B. Rroakman. The Proposed Drug Industry Antitrust Act-Patents, Pricing and the
PubPc, The George Washington Law- Review, Vol. 30. June 1962, p. 890.
12The doctrine that the first sale of a patented product exhausted the patent monopoly
was set out in Boston Stores Co. v. American Gramophone, 246 US 8 (1917).
11 FTC. op. cit.. p. 229.
14 Merck did not enter the retail market until its 1952 merger with Sharp and Dohne.
PAGENO="0569"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2119
packagers. The market results were similar to those in the case of penicillin.
Tariff Commission data indicate that the, realized price of streptomycin fell
from $2,866 per pound in 1946 to $160 in 1950. Competitive market conditions
did not prevent Merck from further research on streptomycin. In 1948 dihydro-
streptomysin was marketed as an improved version of streptomycin, and the
product patent was issued to Merck in 1950. Six firms were licensed, three of
which were bulk sellers. The price of dihydrostreptomycin has closely followed
that of streptomycin.
The marketing of the broad spectrum ailtibiotics in the late 1940's and early
1950's ended the competitive characteristic of the postwar antibiotic market.
In late 1948 Lederle Laboratories, a division of American Cyanamid, introduced
the first broad spectrum antibiotic, chiortetracycline, closely followed by Parke
Davis with chioramphenicol. As both firms were fully integrated into direct
selling the bulk market was bypassed. With the issuance of the product patents
competition could be legally excluded. Evidence through 1960 indicated that
neither company made bulk sales to packagers and all requests for domestic
licenses were refused. The prices of these drugs were reduced from their intro-
ductory level of $15 per bottle of 16-250-milligram tablets, reaching a floor in
1951 at $5.10 and remaining constant until the opening of the Kefauver investi-
gation in 1960.
In 1950 the Chas. Pfizer Company introduced the third broad spectrum drug
oxytetracycline. While waiting for FDA clearance a sales force was organized,
thereby avoiding sales in the bulk market. With the issuance of the product
patent all requests for licenses were rejected. The marketing of oxytetracycline
caused Lederle to reduce the price of its entry, with Parke Davis meeting the
price cut. The last major broad spectrum drug was tetracycline, introduced in
1953 by Lederle, but also produced by Pfizer and Bristol Laboratories and sold
by Squibb and Upjohn as a result of patent interference claims. The introduction
of tetracycline did not disturb the broad spectrum price level nor were domestic
licenses granted to other producers or sellers.
Thus an examination of the early narrow spectrum market indicates a com-
petitive market structure at both the manufacturing and packaging levels. Com-
petitive behavior can be observed in price as well as in product innovation. The
availability of patents and forward vertical integration provided a means by
which the broad spectrum innovators could escape the rigors of competition.
The examination of the antibiotic segment of the industry points up quite
clearly the major flaw iii the measurement of technological progress employed
earlier. If the market is subject to some degree or form of monopoly control where
firms can reach agreement not to compete on the basis of price, and prices can be
maintained at relatively high levels, there is a tendency not to disturb these
agreements through product innovations. This has the effect of decreasing the
base and overstating the proportion of innovations which constitute technological
progress. Having achieved control over price and profits the major stimulant to
further research is removed. There is evidence from antibiotics that once market
control is established further research in the therapeutic area ceases. First, it
seems doubtful that tetracycline, introduced in 1953, is the ultimate broad spec-
trum drug. The strong patent positions' of the manufacturers and Pfizer's ag-
gressive defense of its patent on tetracycline undoubtedly reduce the potential
profitability of this segment of the market for firms contemplating entry through
research. The broad claims allowed in antibiotic patents would be a major factor
in discouraging entry directed research. The potential entrant to be free of
patent infringement suits would be forced to discover antibiotic producing micro-
organism unclaimed by the established manufacturers. Surmounting this prob-
lem there remains the threat of infringement suits designed to harass. And unless
the innovating firm is of equal size with its competitor the evidence indicates
that a policy of harassment will be successful in forcing him from the market.
Here one can cite the fact that Pfizer' drove 33 smaller competitors from the
tetracycline market between 1960 and 1965 by this method.
Second, there is evidence that Lederle, after the discovery and innovation of
chiortetracycline abandoned research in this area until faced with serious com-
petition from Pfizer in late 1952. In 1948 Lederle carried out experiments with
chlortetracycline that produced a substance with antibiotic properties. In the
patent dispute with Pfizer, Lederle claims that this substance was in fact tetra-
cycline. But in 1948, Lederle dominated the broad spectrum market with chiorte-
tracycline and evidently did not feel compelled to complete its experiment. But
in 1952, with Pfizer's discovery of tetracycline Lederle ". . . resumed the 1948
PAGENO="0570"
2120 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
work, and in December 1952 or January 1953 . . . produced tetracycline . . .""
This lends support to the argument that once a dominant position is established
and protected by patents, interest in further innovation declines until the basis
of the dominance is threatened.
Third, Parke Davis, one of the original innovators in the broad spectrum mar-
ket, has never made any additional contributions.
The observations from the narrow spectrum market suggests that where com-
petition remains an active force innovation continues. And in comparing the
approximate 30 percent of these innovations which were found to be significant
with the 12 percent for the industry as a whole indicates that competition is a
major determinate of technological progress. In the broad spectrum market where
competitive forces were suppressed one finds a high proportion of the innovations
significant. But to conclude from this observation that monopoly promotes prog-
ress would be in error. First, the number of innovations were considerably fewer
than in the narrow spectrum market. This in turn suggests that the agreement
not to compete on price included an agreement not to compete in terms of product
innovation as well. Second, an examination of the discovery of the broad spec-
trum drugs indicates that associated with the establishment of monopoly is a
decline in research within the therapeutic area. And only when the basis of the
monopoly is threatened by competition does research resume.
APPENDIX IT-PAPER, "RIsK AND CORPORATE RATES OF RETURN," BY I. N. FISHER
AND G. R. HALL,' THE RAND CORPORATION, SANTA MONICA, CALIF.
(Cited to text, p. 1820, supra.)
I. INTRODUCTION
Although economists have great interest in the correlation between risk and
profits, few studies have attempted to quantify the relationship.2 Consequently,
this paper considers the concept of risk differentials in corporate profit and pro-
poses a model for measuring them. Using this model, the risk-rate of return re-
lationship was estimated for a sample of firms in various industry groups. For
each industry group, average risk-adjusted rates of return were also obtained.3
Risk is defined as the inability to predict the outcome of a forthcoming event
with complete certainty. Entrepreneurs are viewed as making decisions in the
face of uncertaintyly on the basis of probabilistic expectations about future out-
comes.4 If certaintly is a situation where the entrepreneur's anticipation will as-
suredly be fulfilled, then uncertainty can be measured by the likelihood that
the actual outcome will differ from the anticipated outcome.
The foregoing definition accords with economic models of risk (Refs. 6, 23,
24), and it suggests studying risk by examining distributions of corporate rates
of return. Specifically, this approach intimates that the risk-rate of return rela-
tionship can be analyzed statistically in terms of the relationship between the
mean rate of return and higher moments of the distribution.5
15 American Cyanamid Company, Proposed Findings of Fact, Before the Federal Trade
Commission, Docket No. 7211, 1960. Vol. I, p. 16.
1 Any views expressed in this paper are those of the authors. They should not be in-
terpreted as reflecting the views of The RAND Corporation or the official opinion or policy
of any of its governmental or private research sponsors. Papers are reproduced by The
RAND Corporation as a courtesy to members of its staff.
This paper was prepared for presentation at the 1967 Annual Meeting of the Econometric
Society, December 27-30, Washington, D.C. Helpful comments and suggestions were
provided by F. D. Arditti, B. A. Greenberg, R. E. Johnson, L. J. Kleiger, M. H. Kosters,
A. Phillips and I. H. Plotkin.
2 Examples of the wide and diverse theoretical literature on risk and profits are Ref s. 6,
8, 11, 14, 17, 18, 20, 27, 32, and 34. Two important empirical studies of risk and rates of
return are Refs. 4 and 30. In this paper "risk" will not be distinguished from "uncertainty."
Compare Refs. 4, 6. 9. 11, 18, and 34.
The 88 firms included in the sample were selected from Fortune magazine's list of the
500 largest industrial firms. Eleven industry groups were selected, comprising a variety
of different types of industries. Firms in each group were chosen so that both middle-
sized, as well as large, firms would be included.
Economic theory contains two approaches to this problem: in one approach, the decision-
maker balances the various moments of the probability distribution of potential outcomes
on the basis of his utility function (Refs. 21 and 28), while in the other, the decision-
maker chooses among a set of dated financial claims defined over all future states-of-the-
world (Refs. 13, 14). We adopted the first, since the data do not justify the more elegant
approach.
It is assumed that the present and future sets of profit-generating opportunities for
each firm are determined exogenously. That is, a firm may select opportunities but cannot
influence the composition of any set of potential investments. Without this assumption,
the concept of risk becomes more complex. If firms can influence the investment-choice set,
however, presumably the observed variance of profits would decrease over time. Thus,
concentration on fairly long periods of time and a large sample of firms should lessen
the likelihood of this possible effect biasing the statistical results.
PAGENO="0571"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 2121
II. THE MODEL
Assume that firms maximize not profits, but expected utility, and let U (P+W)
be the firm's utility function.6 Utility is a function only of earnings, P (a random
variable), and net worth, W. The risk premium, R(P,W), is that amount required
to make the entrepreneur indifferent between the expected value of the uncertain
earnings, E (P+W), and the certain amount E (P+W) -R(P.W), corresponding
to the expected utility of the uncertain earnings (Ref s. 8, 26).
Earnings distributions and utility functions are not important per se; it is their
interaction that determines the risk component of profits. Suppose that both the
probability distribution of potential earnings and the firm's utility function are
known (illustrated in Fig 1(b) for a riskHaverse firm).7 Assume the probablity
distribution is curve (1). Both the probability distribution of utilty, shown as
(1) in (a), and its expected value, E(TJ1) are easily derived. Note that, although
the probability distribution of earnings is symmetric about the expected value,
E (P), the distribution of utilities is skewed to the left. This occurs because the
utility function is concave, resulting in a non-linear transformation from earn-
ings into utility. The expected value of the utility distribution, E (U1), is less
than the utility of expected earnings, U (E (P+W)), and the difference, trans-
lated into monetary terms, is the risk premium E (P) -P~.
Now ~uppose that the probability distribution is not curve (1) but curve (2).
This distribution is also symmetric about the mean, but the variance is larger.
The distribution of utilities is curve (2) in (a) and, as before, it is not symmetric
6 Here P refers to profits in the sense of increments to net worth, rather than the
profit rate.
The question of "whose" utility function is moot. There are various candidates, e.g.,
managers, stockholders, the chief executive officer, as well as others (Ref. 10). It is assumed
that each entrepreneur (management) is interested in maximizing the expected' utility
of the net worth of the firm on the basis of his judgments about stockholders' preferences.
Such an assumption permits us to explore the relationship of uncertainty to earnings
without having to deal with the complexities of adding stockholders' utility functions,
or the Modigliani-Miller view that stockholders can lever portfolios to offset corporate
management decisions about risk (Refs. 22, 24. See also Ref. 6).
If the firm is averse toward risk, the utility function is concave. This requires that
U' >0 and U' `<0, or that utility increase with earnings and net worth, but at a decreasing
rate.
Fig. 1-Effect of dispersion and skewness on risk premium
PAGENO="0572"
2122 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
about its mean, E (U2). The important point, however, is that the expected
utility has decreased as a result of the increased dispersion of the earnings
distribution. As a result, the risk premium, (E (P) -P) > (E(P) -P°). Con-
sequently, greater variance in the distribution of earnings implies greater risk
and, for risk-averse firms, leads to larger risk premiums. This implies that earn-
ings should be larger, on average, for firms with greater variation in their earn-
ings than for firms with little earnings variability.
Skewness may also have an important effect on the risk premium (Refs. 3,
12, 31). The entrepreneur may prefer positively-skewed earnings distributions
because the likelihood of extremely low earnings is smaller. This, also, is illus-
trated in Fig. 1. Curve (3) in (b) has the same expected value as (1) and (2),
but is skew-ed to the right. This function has been constructed so that the result-
ing distribution of utilities is symmetric about its expected value. In this ex-
ample, skewness offsets variance and the risk premium is zero, i.e., E (Us)
U (E (P +W)). Thus, positive skewness results in smaller risk exposure, w-hile
negative skewness leads .to greater risk exposure, implying that earnings should
be smaller, on average, for firms with earnings distributions positively skewed but
larger, on average, for firms with negatively-skewed distributions.
The results of Fig. 1 suggest that once the form of the utility function is
specified, risk exposure can be measured by characteristics of the probability dis-
tribution of earnings. The required risk premium becomes larger as the spread
of the earnings distribution increases, but the premium decreases as the dis-
tribution becomes positively skewed.8 This illustrates that risk exposure, as de-
fined here, can be measured by characteristics of the firm's earnings distribution.
Before testing this hypothesis, one link in the discussion of the relationship
between risk and earnings remains to be completed-that of the mechanism by
which entrepreneurial preferences for risk and profits are translated into in-
dustry profit differentials or risk premiums and discounts.
Conventional economic theory indicates that with well-functioning capital
markets the equilibrium rate of return will be identical among all activities. En-
trepreneurs theoretically seek those investments yielding the largest rates of
return. As capital is w-ithdrawn from less profitable activities, the rates of re-
turn in such activities rise. Similarly, the inflow of capital into higher-yield
investments forces the rates of return in these activities downward. Equilibrium
occurs when the rates of return of investment are identical among all activities.
When risk is considered, the adjustment process is more complex. Because
differences in risk exposure exist among alternative investments, entrepreneurs
balance risk against expected rates of return. Capital, therefore, is transferred
from low-return, high risk activities to high-return, low risk investments until
an equilibrium, characterized by a set of risk premiums reflecting differences
in risk exposure, is achieved. In this equilibrium, risk-compensated rates of re-
turn are equal among alternative investments, but observed or actual rates of
return will differ by the amount of the risk premiums.
8 This can be demonstrated formally in the following manner: Expand U(P+JV) in a Taylor series
A
about the point (P± W) =-E(P± W),
A A A A A fT" A A
U(P+W)= tJ(P+JF)+U'(P+JV)(P-p)±-1 (P± W)(P_P)+-i-~.~ (P+ W) (P-P+
A
Taking expected values and holding W, P constant,
A U" A U" A
E(U(P+JV))= U(P±JV)±~ -~- (P+JV)±~ -~- (P+JV)+
Rearranging terms, the difference between expected utility and utility of expected earnings Is
A / U" A U'/A
~ (P±W)±o~ -~- (P+JV)+
Equation (3) is the risk premium, R(P, JV), and it becomes apparent that the second, third, and higher
moments may affect the magnitude of the risk premium. Since U"0 and the risk premium becomes smaller as skewness increasesJligher
moments add little information about the characteristics of the distribution and are ignored. (See references
3, 26).
PAGENO="0573"
COMPETITIVE PROBLEMS IN ~FHE DRUG INDUSTRY 2123
In short, we posit that capital markets respond to risk as they respond to
expected rates of return. We should, therefore, expect to find a structure of risk-
compensated rates of return that motivate or discourage investment. Part of the
earnings differentials observed among alternative investments can be attributed
to risk; these are the risk premiums that compensate for differences in risk
exposure.
III. EMPIRICAL RESULTS
To test the hypothesis that profits are larger for firms with greater risk ex-
posure, it is necessary to translate the theoretical definition of risk into statis-
tical terms.° We can do this by assuming managers' anticipations, on average,
are correct, thereby permitting the observed mean rate of return to be used as
a proxy (Ref. 4). Risk exposure, as defined here, can then be measured by mo-
ments of the distribution of earnings.10
The risk variables were calculated from
r 2lh/2
I ~ ~r~-r~,) I
o1L6T1 ] (1)
and
(r~i-~~,)~
~ no-13 (2)
where
r~t= observed rate of return for firm i in year t;
~13=predicted rate of return from trend for firm i, year t;
oi standard deviation of rates of return about trend, firm i;
81=skewness about trend for firm i;
and n is the number of years included in the sample.
The model can now be stated explicitly as
~1=ro+bio1+b2S1 (3)
where
~= average rate of return on net worth for firm i;
r0= intercept; and b1, b2 are the coefficients of the standard deviation and
skewness, respectively-the risk coefficients.
The signs of these coefficients are expected to be
b1>O,
b2