PAGENO="0001" 90th Congress } JOINT COMMITTEE PRINT IMPACT OF THE PROPERTY TAX: ITS ECONOMIC IMPLICATIONS FOR URBAN PROBLEMS SUPPLIED BY THE NATIONAL COMMISSION ON URBAN PROBLEMS TO THE JOINT ECONOMIC COMMITTEE CONGRESS OF THE UNITED STATES ~ ~ ~ 0 ~ ~ COLLFCE ~ ~ ~ IrrsJ~ ~ L!Bp~ ~ / ~ ~`4~' ~1. 08102 V ~JUNi ~ MAY 1968 8 ~OV. DOC. Printed for the use of the Joint Economic Committee U.S. GOVERNMENT PRINTING OFFICE 94-0540 WASHINGTON : 1968 sale by the Superintendent of Documents, U.S. Government Printing Office Cf Lj~ ~ Washington, D.C. 20402 - Price 20 cents PAGENO="0002" SENATE JOHN SPARKMAN, Alabama I. W. FULBRIGHT, Arkansas HERMAN E. TALMADGE, Georgia STUART SYMINGTON, Missouri ABRAHAM RIBICOFF, Connecticut JACOB K. JAVITS, New York JACK MILLER, Iowa LEN B. JORDAN, Idaho CHARLES H. PERCY, Illinois WILLIAM H. MOORE HOUSE OF REPRESENTATIVES RICHARD BOLLING, Missouri HALE BOGGS, Louisiana HENRY S. REUSS, Wisconsin MARTHA W. GRIFFITHS, Michigan WILLIAM S. MOORHEAD, Pennsylvania THOMAS B. CURTIS, Missouri WILLIAM B. WIDNALL, New Jersey DONALD RUMSFELD, flhinois W. E. BROCK 3o, Tennessee JOHN B. HENDERSON DONALD A. WEBSTER (Minority) (H) GEORGE R. bEN JOINT ECONOMIC COMMITTEE (Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.) WILLIAM PROXMIRE, Wisconsin, Chairman WRIGHT PATMAN, Texas, Vice Chairman JOHN R. STARE, Executive Director JAMES W. KNOWLES, Director of Research EcoNo~IIsTs PAGENO="0003" LETTER OF TRANSMITTAL MAY 20, 1968. To the Members of the Joint Economic Committee: Transmitted herewith for the use of the members of the Joint Economic Committee and other Members of Congress is a research report on the "Impact of the Property Tax" prepared by Prof. Dick Netzer, of New York University, for the National Commission on Urban Problems, chaired by former Senator Paul H. Douglas. This study has been reproduced as a joint committee print because of its special relevance to the committee's hearings on selected aspects of the Report of the National Advisory Commission on Civil Disorders. This paper is the first in a series of studies initiated by the Douglas commission as a complement to its own staff studies and extensive public hearings held throughout the country. In this connection, I would like to commend Chairman Paul H. Douglas for the outstanding work his commission has undertaken. His leadership continues to set the same high standards of excellence which he maintained as a U.S. Senator and provided to the Joint Economic Committee during his years as chairman. While the committee is pleased to have this study available for our own inquiry, it is not at this time expressing either approval or disapproval. WILLIAM PR0xMIRE, Chairman, JOint Economic (lommittee. (III) PAGENO="0004" PAGENO="0005" CONTENTS Page Letters of transmittal III Impact of the Property Tax: study prepared for the National Commission on Urban Problems VII (V) PAGENO="0006" PAGENO="0007" IMPACT OF THE PROPERTY TAX EFFECT ON HOUSING, URBAN LAND USE, LOCAL GOVERNMENT FINANCE By DICK NETZER Professor of Economics, New York University Prepared for the consideration of THE NATIONAL COMMISSION ON URBAN PROBLEMS (VII) PAGENO="0008" COMMISSION MEMBERS PAUL H. DOUGLAS, Chairman Washington, D.C. DAVID L. BAKER Garden Grove, Calif. HUGO BLACK, Jr. Miami, Fla. LEWIS DAVIS New York, N.Y. JOHN DEGROVE Boca Raton, Fla. ANTHONY DOWNS Chicago, Ill. EZRA EHRENKRANTZ San Francisco, Calif. ALEX FEINBERG Camden, N.J. JEH V. JOHNSON Poughkeepsie, N .Y. JOHN LYONS St. Louis, Mo. RICHARD W. O'NEILL New York, N.Y. BICH~RD RAVITCH New York, N.Y. CARL E. SANDERS Atlanta, Ga. Mrs. CHLOETHIEL WOODARD SMITH Washington, D.C. To~i J. VANDERGRIFF Arlington, Tex. COLEMAN WOODBURY Madison, Wis. (VIII) PAGENO="0009" NATIONAL COMMISSION ON URBAN PROBLEMS, Washington, D.C. Hon. PAUL H. DOUGLAS, Chairman, National Commission on Urban Problems, T'Vashington, D.C. DEAR MR. CHAIRMAN: We are forwarding to you and the CoIn- mission this first in a series of background study reports on a number of key issues and problems which the President and the Congress have asked this Commission to examine. The views expressed in this and other study reports are those of the authors and do not necessarily represent those of the Commission or of its individual members. This report entitled "Impact of the Property Tax" was prepared by Dr. Dick Netzer, who is professor of economics and the head of the All-University Department of Economics at New York University. No one is more qualified to examine this problem than Dr. Netzer whose work in this field is widely known. He was assisted in his research by Mr. Stanley Natkins, of the Graduate School of Public Adminis- tration at New York University~ The study was prepared under the guidelines devised by Mr. Allen Manvel, associate director of the staff of the Commission. Sincerely, HOWARD E. SHUMAN, Executive Director. (IX) 94-054 O-68----2 PAGENO="0010" PAGENO="0011" FOREWORD This study is one of many undertaken by the staff and consultants to the National Commission on Urban Problems. It is part of the research effort being undertaken by the Commission prior to its own recommendations to the President, to Congress, and to the Secretary of Housing and Urban Development. The findings and conclusions reached in this property tax study are those of Dr. Dick Netzer, and do not necessarily reflect the views of the Commission àr the staff. The report is being released in advance of the Commission's findings and recommendations because it deals with one of America's vital urban issues. If the report stirs public discussion and wider understanding of this issue, its publication \~ill serve a useful purpose. This report focuses on one of the specific charges to the Commis- sion-the study of "local tax policies with respect to their effect on land and property cost and on incentives to build housing and make improvements in existing structures" (sec. 301, Housing and Urban Development Act of 1965). The author, Dr. Netzer, is an authority on these questions. He finds that the property tax as now constituted is a deterrent to new housing and to the maintenance of existing housing, and that it places a particular burden on low-income renters. A tax ha'ving these characteristics-at a time when some estimates indicate that about one out of every 10 American dwelling iunits is substandard-requires urgent attention. The property tax is the biggest single source of local government financing, and increased revenues are needed to maintain and upgrade essential local services. In this context, suggestions for improving and strengthening the tax, as well as for finding alternatives, are high-priority matters. In addition to problems of property taxation, the Commission is charged with studying a number of other matters; namely, building codes, zoning and land use, housing codes, Federal taxation as it affects housing, development standards, and how to provide an ade- quate supply of housing for low-income families. It is hoped that this and forthcoming background reports, along with testimony in the Commission's extensive public hearings, will give insights to the public which are pertinent to the solution of current city problems. PAUL H. DOUGLAS, Chairman. (XI) PAGENO="0012" PAGENO="0013" CONTENTS Page URBAN FISCAL PROBLEMS AND PROSPECTS 1 THE URBAN PROPERTY TAX 4 Nature of "the" property tax 4 State use of the tax 5 Legal coverage 5 Assessment organization 6 Role of the tax in urban public finance 6 Wide variations 8 Sources of property tax revenue 12 Property tax considered as a sales tax 13 EFFECTS OF THE PROPERTY TAX IN URBAN AREAS 16 Housing and the tax burden 16 Share paid through rents 17 Special burden on the poor 18 Property tax deters improvements 21 Urban development and land use 22 Central city economics 22 Intrasuburban variations 26 AN APPRAISAL OF THE PROPERTY TAX 29 Effects on housing and urban development 29 Effects on other resource allocation 30 Effects on income distribution 30 Quality of administration 32 Inherent and remediable defects 35 POSSIBLE REMEDIES 36 Increase State-Federal responsibilities 36 Other local revenue sources 39 User charges for public services 39 Local income taxes 40 Land value taxation 41 Taxation of land value increments 43 Improvement of the existing institution 43 Tax base consolidation 44 Fiscalfederation 44 Better administration 45 Hardship adjustments 45 Housing tax incentives 46 Deemphasis SCOPE OF COMMISSION RESEARCH 47 (XIII) PAGENO="0014" PAGENO="0015" URBAN FISCAL PROBLEMS AND PROSPECTS In recent years, as throughout the 20 years following World War II, local and State government public expenditures have been increasing more rapidly than has the Nation's total output and income. (See table 1.) Public expenditures in urban areas have always been signifi- cantly higher, in relative terms, than those in nonurban areas, and re- cently have been increasing slightly faster, in dollar terms, within the urban areas. This is to be expected, since nearly all the Nation's popu- lation growth has been occurring in urban areas. But urban population growth does iiot explain the rate of increase in public spending. Indeed, the increase in per capita local government expenditures in metro- politan areas was more rapid than the increase in aggregate gross national product between 1957 and 1962. Perhaps most striking, public expenditures in the larger central cities have been climbing steeply, despite their losses or slow growth in population. In the most recent 4-year period for which data are avail- able, expenditures of municipal governments in the larger cities rose by 27 percent, as shown in table 1-about three-fourths as rapidly as exl)enditureS of all other local governments combined. To be sure, substantially more external aid to central cities in the provision of public services has been forthcoming in recent years. State and Federal aid to central city governments has risen considerably more rapidly than have central city expenditures. Also, the direct role of State governments in the provision of public services in and for the central cities has expanded considerably. Since the passage of the In- terstate Highway Act in 1956, the States have been far more active in the construction of central city highways than previously. In a growing number of States, the State government is directly involved in urban mass transportation, in park and open space activities, and in housing programs. In some States in the Northeast, expansion of State higher education programs has had an important effect on central city popu- lations. Desipte all this, the taxes imposed by central city governments, collected from static populatioiis and slowly growing central city economies, continue to rise sharply. The explanation for rising public expenditures in urban areas is not hard to find. In the central cities, local-tax-financed outlays for services directly linked to poverty (in the health and welfare fields) have not been static; the central cities of the 12 largest metropolitan areas ac- count for one-eighth of the country's population, but nearly two-fifths of health and welfare outlays financed from local taxes. For central city governments, the problems associated with poverty and race are by far the most urgent of public problems. (1~ PAGENO="0016" 2 TABLE 1.-Percentage increases in non-Federal public expenditures, 1957-62 and 1962-65/66 Perce ntage increase 1957-62 1062-65/66 Gross national product 1 Total expenditures: All State and local governments 27 48 33 35 All local governments 46 35 Local governments in metropolitan areas- all SMSA's 2 47 (3) 38 largest areas Central city governments in large cities ~. - - Per capita expenditures: All State and local governments (3) 31 36 . 34 27 28 All local governments 34 28 Local governments in metropolitan areas- all SMSA's 30 (3) 38 largest areas (3) 24 1 For calendar years 1957-62 and 1962-66. 2 For identical collections of metropolitan areas in 1957 and 1962. Not available. 4 rncludes only the municipal government (excludes separate overlapping county, school district, and special district governments); for the 42 cities with a 1960 population of more than 300,000 excluding Honolulu. 5 Based on estimated 1957, 1962, and 1966 populations. Source: Adapted from various publications of the U.S. Census Bureau, Governments Division. Neither poverty nor racial disabilities can be eliminated solely by governmental action, and still less by action by local or State and local governments combined (that is, governments other than the Federal Government). But local governments do have a major respon- sibility to grapple with these problems and can make a major contri- bution toward their alleviation. In the American system of govern- ment, it is local governments which are responsible for providing edu- cational services that over time will have a major bearing on the chances the poor and racially disadvantaged have to overcome their disadvantages. Local governments are also responsible for a. wide range of health and welfare services, which are almost entirely oriented toward the poor in American cities. They have had, since the. late forties, major responsibilities in connection with the housing of the poor. And, as far as the poor are concerned, local government recrea- tional facilities are about the only recreational facilities available. A second major set of problems confronting the older central cities lies in the fact that they have a huge legacy of obsolescence. Their stock of housing and other social capital-that. is, public and quasi- public facilities of all kinds-is old, often physically deteriorated, and generally far from competitive with the new parts of the same urban areas. It may be, as some have argued, that the best national policy would be to allow this obsolescence to continue, and allow fur- ther deterioration of the older parts of the older cities. In this case, population would decline in these sections and, Presumably at some stage, values would be so low that private renew-al of such areas would become possible. Or, if desirable, public renewal could be undertaken, but on the basis of exceedingly low values. Developments in recent years suggest that this is hardly a likely course of action. For one thing, there is the plight of those who, be- PAGENO="0017" 3 cause they are poor, or Negro, or both, have little chance to escape the deteriorating areas. Amelioration for these hundreds of thousands of people is both politically and morally necessary. Quite apart from moral isues, most cities and the Federal Government appear to have decided that it is necessary to replace obsolete social capital and to compete for residents and businesses in an atmosphere of rising ex- pectations. That is, the cities feel they must offer an environment of pTlbhc facilities and services which, together with other attractions that the central locations may have, offset the blandishments of the newer and presumably more modern sections of the metropolitan areas where standards of public services and amenity are high indeed. In the newer sections of metropolitan areas-the new portions of central cities as well as the urbanizing fringes of the metropolitan area-the main governmental problem is the provision of the new social capital needed by a rising population, and a population which has peculiarly heavy demands for public services and facilities, notably schools. In the aggregate, these urban problems have caused a diversion of resources from private to public use, via tax increases. This relative expansion of the public sector is costly in another way. If local governments are to command resources, they must pay prices for these resources which are competitive with those prevailing in the economy, notably salaries of public employees. If they are to expand more Tapicily than the private sector, they must bid away resources by paying even more. This they are doing, as is shown by the rapid increase in urban government salary levels, especially for occupational groups whose talents are in heavy demand in the private sector. As table 1 suggests, the rate of increase in the expenditure of urban governments is not tapering off; if anything, it is increasing. This is consistent with our observations of the urban fiscal scene (with almost continual fiscal crises), and our observations of the urban social scene, with the huge unmet needs for new and improved public services. But these trends do conflict with some of the recent projections of the outlook for State and local finances in the decade ahead. These projec- tions are generally optimistic, in that they foresee no great fiscal strains, largely because of an expected tapering off of the rate ~f the increase in expenditure. The projections may be right, but there is room for skepticism. For example, the projections have not allowed for the recent surge in expenditure for public assistance programs. Between 1962 and 1965-66, local government public welfare expenditures rose by nearly 50 per- cent, a rate of increase nearly double that found in the more compre- hensive sets of projections; in New York State, public welfare expenditures will be more than double those implied by the projections sponsored by the Council of State Governments.' `For a recent evaluation of the projections, see Revenue Sharing and Its Alter- natives, hearings before the Subcommittee on Fiscal Policy of the Joint Economic Committee, U.S. Congress, July 31-Aug. 3, 1967, especially pp. 65-106. 94-054 O-68--3 PAGENO="0018" THE URBAN PROPERTY TAX All of this is, of course, relevant to an appraisal of the property tax, since the property tax is by far the largest single source of funds to finance public services in urban areas. If, as seems likely, urban public expenditure continues to rise sharply, and if there are ~io major changes in fiscal institutions, the property tax will bear a large share of the burden; the various deficiencies of the tax will be aggravated by its use at continually increasing rates. NATnRE OF "THE" PROPERTY TAX A century ago, the property tax in nearly all States was one which, in law, applied at a uniform rate to all forms of ~rivately owned wealth (with exemptions for charitable and similar organizations) within a taxing jurisdiction-to all types of real property, to tangible personal property, and to wealth in intangible forms, notably bank deposits and securities. Even then, the property tax differed greatly among the thousands of local units employing the tax. The level and composition of legally taxable wealth varied widely; the effectiveness with which local assessors discovered and valued this wealth also varied widely; and the ratios of public expenditures financing by the tax to assessed values had a wide range, producing major variation in tax rates, whether measured as a fraction of assessed values or of the underlying market value of the taxable forms of property. All these sources of variation continue to exist., to a pronounced extent, among the 70,000 or so local govermnent units which are au- thorized to levy property taxes. But., in addition, the generality of the property tax has been substantially reduced by successive State stat- utory and constitutional enactments exempting various classes of personal property from the ordinary general property tax. A further complication is that some of these exemptions have been accompanied by new forms of property taxation on the exempted classes of prop- erty, most commonly on intangibles, motor vehicles and public utility property; these "special property taxes" have rates which are not those applied in general property taxation, but usually a uniform statewide rate. Still another complication is that the State governments, all of which relied heavily on the property tax in the 19th century, have withdrawn from property taxation, though in varying degrees. The result has been to create distinctive statewide property tax systems in each of the 50 States (and the District of Columbia), im- posing 51 different legal patterns on top of the wide intrastate varia- tion in taxable wealth, administrative performance, and fiscal require- ments. *Thus, there are really 70,000 or so different property taxes in the United States, grouped into 51 systems with common legal settings. As of 1962, nearly 96 percent of all property tax revenue came from local government general property taxation; for the purposes of this report, this major component of the tax will be treated as if it. were (4) PAGENO="0019" 5 the entire property tax. Another 1 percent of property tax revenue came from local special property taxes, and about 3 percent of the revenue accrued to State governments, mostly in the form of special property taxes.2 STATE USE OF THE TAX Although the State share of the tax is small in the aggregate, it is not uniformly so. Consider the 47 States (including the District of Columbia) containing standard metropolitan statistical areas (SMSA's). In 21 of them, State property tax revenue amounted to 1 percent or less of total State-local property tax revenue in 1965-66. In another 14, the percentage ranged between two and five. At the other extreme, there were four States in which the State government per- centage was more than 15.~ However, in nearly all the large urban concentrations, the State share of property tax revenue is very low. In all except five of the States containing very large SMSA's-the 38 largest with populations over 700,000-the State share was 5 percent or less.4 LEGAL COVERAGE5 The process of exemption has gone furthest with respect to intangi- ble personal property, for two rather obvious reasons. First, some forms of intangibles (like securities) are easily concealed and all forms are highly mobile; this makes assessment of intangibles extra- ordinarily difficult and comes close to converting the tax into one on honesty. Second, even if the assets could be discovered readily, a gen- eral property tax at the rates common in this country-5 percent and more-is a confiscatory tax when applied to the conventional low-yield liquid assets; consider, for example, a 5-percent tax on demand de- posits in commercial banks, which bear no interest at all. In any event, there are no local property taxes on intangibles in 20 of the 47 SMSA States and in 11 of the 23 States with the largest SMSA's-over 700,000 population. In 14 of the SMSA States (and, eight of the large SMSA States), there are local special property taxes, usually applied at very low rates, on some forms of intangibles. In 13 of the SMSA States, some or all types of intangibles are subject to local general property taxation, but even in these cases, intangibles generally amount to only trivial proportions of total assessed values. In Illinois, for example, intangibles valuations amount to only about 1 percent of total assessed values, and barely equal 3 percent of total privately owned bank deposits, not to mention the vast array of other types of intangibles which are also legally subject to tax. Some forms of tangible personal property are subject to local gen- eral property taxation in all except foui of the SMSA States. The From U.S. Census Bureau, Property Taxation in 1962 (State and Local Gov- ernment Special Studies No. 47, November 1964). From TJ.S. Census Bureau, Governmental Finances in 1965-66 (1967). This grouping of SMSA's is related to data availability and will be used fre- quently in subsequent paragraphs. The Census Bureau now publishes annual data on the finances of local governments in these areas. The information in this section is for 1961, from U.S. Census Bureau, 1962 Census of Governments, vol. II, Taxable Property T7alues (1963, revised August 1964), pp. 5-6. PAGENO="0020" 6 great bulk of taxable personal property consists of business machinery and equipment and inventories. Household goods are generally taxed and fully taxable in only nine of the SMSA States. Motor vehicles comprise a significant part of the tax base in some, but by no means all, metropolitan areas, as the following tabulation shows: Number of S tates containing SMSA's Large SMSA's Taxed under local general property tax 21 7 Taxed under local special property tax 5 2 No local property taxes 21 14 ASSESSMENT ORGANIZATION Another dimension of interstate property tax system differentials is the geographic organization for assessment. This is of significance from the standpoint of concern about fiscal disparities within metro- politan areas, or at least from the standpoint of obtaining information about these disparities. The smaller the geographic unit at which the assessment is done, the more difficult it is to utilize available informa- tion on taxable capacity and tax rates for local government units within a single metropolitan area~, since individual assessing districts may assess taxable wealth at very different proportions of its market \Taliie In 12 of the 47 SMSA States (and 7 of the 23 large SMSA States), assessment is done by city, village, and town officers, rather than by countywide assessment organizations. In several other States, the pat- tern is a mixed one, with some countywide assessment. In two-thirds of the SMSA States (and 13 of the 23 large SMSA States), county- wide assessment is the general rule, for metropolitan areas. ROLE OF THE TAX IN URBAN PUBLIC FINANCE Prior to the 1930's, the property tax provided three-fourths or more of the general revenue of American local governments aiid more than four-fifths of their locally raised general revenue (that is, excluding payments from other levels of government), as table 2 shows. Largely because of the greatly increased grants-rn-aid for welfare and school purposes from State and Federal governments, the property tax de- clined sharply as a proportion of total general revenue in the 1930's and 1940's. From the early 1950's to the early 1960's, the property tax stabilized at just under half of total general revenue. However, in the past few years, a slow decline in its relative role apparently has resumed (see Figure 1), again largely due to increased grants-in-aid. In a sense, table 2 understates the long-term decline in the relative role of the property tax in the financing of urban public services. State governments are important direct providers of public services, through State government agencies, in addition to their indirect role in financing local government~. The more important direct State gov- ernment services include highways, public assistance (in a majority of the States), higher education, mental hospitals and correction. The relative importance of direct State government expenditure has risen PAGENO="0021" 7 H,gi CONTRIBUTION OF PROPERTY TAXES TO LOCAL GOVERNMENT REVENUE (selected years, 1927 to 1966) Percent ~ 40 __ 7 __ .111 1927 1946 1956 1965-66 source: table 2 considerably since the 1920's. But in the 1920's, the States still relied to an important extent on the property tax, which is no longer true. Taking this into account, the property tax probably finances less than one-third of the costs of State-local urban services today, compared to over two-thirds in the 1920's. TABLE 2.-Trends in the role of the property tax in local government finances, 1927 to 1965-66' [In percent] 1961-66 1962 1956 1946 1927 Property tax revenue as percent of total local government general revenue: All local governments Local governments in 38 largest SMSA's2 45 47 48 50 49 58 74 Other areas 43 46 Property tax revenue as percent of locally raised general revenue: Alllocalgovcrnments Local governments in 38 largest SMSA's 2 67 66 69 68 69 78 82 Other areas 69 70 on u.s. census Bureau, 1962 Census ef Gevernments, vol. vi, No. 4, Historical Statistics en Gerern- snental Finances and Employment (1964) and Leral Gerernment Finances in Selected A1etrapelitan Areas in 1965-66 (1967). 2 Those with pcpulations over 700,000. 3 Total general revenue less revenue from other governments. PAGENO="0022" 8 If intergovernmental payments are ignored and attention is focused on general revenue raised from local sources, the decline in the rela- tive importance of the property tax was concentrated in the late 1940's and early 1950's, to consequence of increased utilization of local non- property tax sources. On a nationwide basis, however, the property tax continues to provide about two-thirds of locally raised general revenue, both within and outside the major urban concentrations. Moreover, the relative decline merely reflects a n~ore rapid increase in other revenue-in absolute terms, property tax revenue continues to rise sharply. In the period of less than 4 years, from 1962 to 1965-66, property tax revenue rose by nearly 30 percent-$5.4 billion on a nationwide basis and $2.7 billion in the 38 largest SMSA's. Because State aid looms relatively larger in the finances of local governments in the less urbanized parts of the country, the property tax accounts for a larger proportion of local government general revenue in the largest SMSA's than elsewhere, but a smaller propor- tion of locally raised general revenue, because nonproperty taxes are largely an urban phenomenon (table 3). Also, as is to be expected, property tax revenue is substantially higher on a per capita basis in the larger urban areas. WIDE VARIATIONS There is considerable variation among the major urban areas in the extent of their dependence on the property tax. In part, this is a con- sequence of State aid arrangements, but this is by no means the whole story. By far the most important differential in State aid provisions relates to assignment of the responsibility for administration of public assistance. Where the State government handles public assistance, State grants to local government for welfare purposes are, quite naturally, very small; where the local governments have the adminis- trative responsibility, it is accompanied by sizable grants from the States. Thus, in the 21 large SMSA's where public assistance is pri- marily a local responsibility, intergovernmental revenue per capita in 1965-66 was $120; in the other 17 large SMSA's, with primary State government responsibility for public assistance, the corresponding figure was only $66. However, as table 3 shows, relative dependence on the property tax differed little in the two types of areas. This suggests that the prop- erty tax (much higher on a per capita basis in the local welfare respon- sibility SMSA's) finances a not inconsiderable share of the locally raised expenditure for public assistance. It also suggests that differ- ences among areas in reliance on the property tax cannot be explained mainly by State aid difference. In fact, the major source of the differ- ence appears to be the extent to which local nonproperty taxes are employed. These differences are large ones. In SMSA's located in States where local nonproperty taxes are virtually nonexistent-in New England, New Jersey, and Indiana, for example-the property tax accounts for 80 to 90 percent of locally raised general revenue. By contrast, in SMSA's where income taxes are widely used by local governments, such as Cincinnati and Pittsburgh, the property tax percentage is only about 60. In Illinois and California SMSA's, where sales taxes are used widely by local governments, the property tax percentage is in the 65 to 70 range. PAGENO="0023" 9 TABLE 3.-The role of the pro pery tax in local government finances, 1965-66 1 38 largest SMSA's2 Areas where welfare All local governments Outside 38 largest SMSA's Total 38 SMSA's primary responsi- bility of- State Local govern- govern- ment ment Property tax revenue as percent of- Total general revenue - Locally raised general 45 43 47 48 46 revenue4 Property tax revnue: In millions Percapita5 67 69 66 64 67 $23, 836 123 $11, 573 101 $12, 263 154 $3, 954 $8, 309 121 177 1 Based on U.S. Bureau of the Census, Local Government Finances in Selected Metropolilan Areas in 1965-66 (1967 and Governmental Finances in 1965-66 (1967). 2 Those with populations over 700,000. 3 The classification is based on statewide ratios of State government direct expenditure for public welfare and of local government direct expenditure for public welfare to total public welfare expenditure. 15 of the areas are in States in which State direct expenditure in 1965-66 was 75 percent or more of the total and another 2 in States with State/State-local ratios of 50-75 percent. 17 are in States with local/State-local ratios of 75 percent or more and another 4 in States with local/State-local ratios of 50-75 percent. 4 Total general revenue less revenue from other governments. 5 1965-66 revenue divided by estimated July 1, 1965, population. Even lower percentages can be found in some areas dominated by large central cities which rely very heavily on nonproperty taxes. (The instances cited in the preceding paragraph concern SMSA's in which both central city and suburban governments employ such taxes.) As table 4 shows, the property tax looms much larger in local government finances in the large SMSA's for county governments, school districts, and small municipalities and townships than it does for the central city governments, which on the average obtain barely over half-51 percent-of their locally raised general revenue from the property tax. TABLE 4.-Dependence on the property tax by local governments in the 38 largest SMSA's, 1965-66, by type of Government 1 Type of government Property tax reve nue as percent of- Total local government general revenue Locally raised general revenue 2 All local governments in 38 largest SMSA's Counties Municipalities (34 central cities with 1960 populations of 300,000 or more) (Other municipalities) Townships School districts- Special districts 47 46 39 (37) (47) 65 55 27 66 75 53 (51) (58) 80 88 32 1 Based on U.S. Bureau of the Census, Local Government Finances in Selected Metropolitan Areas in 1965-66 (1967) and City Government Finances in 1965-66 (1967). 2 Total general revenue less revenue from other governments. PAGENO="0024" 10 In general, the property tax finances a significantly smaller share of the costs of public services in central cities than in outlying parts of the larger metropolitan areas, because of use of nonproperty taxes by the municipal governments in most of the large central cities. This is illustrated by the first two columns in table 5, for 11 large SMSA's.° In nine of the 11 areas, the relatively smaller role of the property tax in the central cities is evident. In the Baltimore area the differences are small, and in New Orleans the property tax is relatively insignificant in both central city and suburbs. But despite the smaller relative im- pOrtance of central city property taxes, on a per capital basis property tax collections in the central cities are equal to or greater than such tax collections in the suburbs, except in the Washington, New York, and Philadelphia areas, as shown in the last two columns of table 5 and in figure 2. And in these three cases, the disparity in the relative role of the property tax is greater than the disparity in the per capita prop- erty tax revenue; that is, the very heavy use of nonproperty taxes more than offsets the lower per capita property tax payments. This is just another way of making the common observation that tax burdens are, in most cases, greater in central cities than in suburbs; usually this is reflected in the property tax data, but elsewhere the explanation lies in nonproperty taxes. ~ SMSA's selected for table 5 are, with two exceptions, those in which the central city is an independent city not in any county, or has a combined city- county government structure, thus simplifying the comparison. The exceptions are Chicago and Detroit. TABLE 5.-The property tax in local government finances: Central cities versus out- lying parts of selected large SMSA's, 1965-66 1 Area 2 Property tax revenue as percent of locally raised general revenue Central city Outlying Property tax revenue per capita Central city Outlying Baltimore 73 71 $136 $108 Boston3 82 87 258 177 Chicago4 71 79 156 157 Denver 65 80 164 143 Detroit5 69 74 156 130 New Orleans 39 34 48 37 New York Philadelphia St. Louis 48 88 46 70 44 76 175 88 111 234 121 114 San Francisco6 68 72 253 238 Washington, D.C 29 70 110 131 1 Based on U.S. Bureau of the Census, Local Government Finances in Selected Metropolitan Areas, 1965-66 (1967). 2 Except where otherwise indicated, the central city is the city named in the stub and the outlying parts of the area comprise the remainder of the SMSA. The "central city" is Suffolk County; the SMSA is Massachusetts State economic area C. The "central city" is the whole of Cook County. 2 The "central city" is the whole of Wayne County. The outlying parts exclude Alameda County, because Oakland has central city characteristics. PAGENO="0025" $300 11 Fig.2 PROPERTY TAXES IN SELECTED LARGE METROPOLITAN AREAS, 1965-6 6 (Per capita amounts) SMSA $100 $200 BALTIMORE BOSTON CHICAGO 1/ DENVER DETROIT 2/ NEW ORLEANS NEW YORK PHILADELPHIA ST. LOUIS SAN FRANCISCO WASHINGTON, D.C. OUTLYING AREA E~ CENTRAL CITY ~~;; source: table 5 94-054 O-68----4 1/ `Central city' is whole of Cook County. 2/ `Central city' is whole of Wayne County. 3/ `Outlying area' excludes Alameda County. PAGENO="0026" 12 SOURCES OF PROPERTY TAX REVENUE Cutting across the geographic variations, there is an important eco- nomic distinction in property taxation. The tax can be viewed as comprising two distinct forms of taxation, with quite different eco- nomic characteristics. The first is the tax on residential property and on consumer-owned personal property, where the latter is subject to tax, largely consisting of motor vehicles. This tax is one of selected types of consumer ex- penditure, with economic effects similar to those of other types of consumption taxes. The second is the tax on real and personal property used in business and agriculture. Its economic effects are in some ways similar to those of other forms of business taxation, but are dissimilar in other ways. The dissimilarities take three major forms. First, this is a tax on business inputs-land and capital-rather than on profits or gross receipts, the base of the predominant nonproperty types of business taxation. Second, the impact of the tax therefore varies among indus- tries on the basis of the structure of inputs, falling most heavily on those which employ large amounts of land, buildings, equipment and inventories relative to output. Third, few other forms of business taxation vary so much among small geographic areas-they are usually statewide or nationwide in scope. Table 6 indicates the estimated distribution of assessed values sub- ject to general property taxation in SMSA's in 1961. Based upon the not unreasonable assumption that the sources of SMSA local govern- ment property tax revenue closely reflect this distribution of assessed values, a crude estimate of 1962 property tax revenue by type of 1)roperty is presented in table 7. Roughly half of the revenue was derived from taxes on housing and another 42 percent from taxes on nonfarm business property. The remaining 8 percent includes taxes on farm property (as expected, this is minor in urban areas), taxes on consumer-owned tangible personal property, and taxes on property which cannot reasonably be allocated between businesses and house- holds, like vacant lots and intangibles. TABLE 6.-Estimated distribution of assessed values subject to taxation in SMSA's, by type of property, 1961 1 general property Type of property In billions Percent distribution Total Single-family houses Other housing Acreage and farms $244.6 100.0 97. 1 25. 1 6.8 39. 7 10.2 2.8 \Tacant lots 5.3 2.2 Railroad and public utility property 2 Commercial realty 16. 6 36.2 6. 8 14.8 Industrial realty 2 3 Other real property ~ 19. 9 1. 0 8. 2 . 4 Personal property 2 36. 7 15. 0 1 Adapted from U.S. Census Bureau, 1962 Census of Governments, vol. II, Taxable Property Values (1963; revised August 1964). Involves the following estimates by the author: distribution by property type of partial exemptions; distribution by property type of State-assessed property; public utility property included in reported data on induitrial realty (locally assoncd). 2 Includes both State-assessed and locally assessed property. 3 Excluding railroad and public utility property. About $400,000,000 is separately assessed mineral rights. Note: Because of rounding, detail may not add to totals. PAGENO="0027" 13 TABLE 7.-Crude estimate of sources of local government property tax revenue in SMSA's, by type of property, 1962 1 Type of property Percent distribution Housing2 Railroad and public utility property Other real and personal property used in nonfarm business `.~_ Other 50 7 35 8 Total 100 1 Based on distribution of assessed values in table 6, adjusted for estimated revenue from local government special property taxes and its distribution by property type. 2 Includes estimated housing property listed as acreage and farms. 3 Includes separately assessed mineral rights; 80 percent of all personal property; and commercial and industrial realty. 4 Includes agricultural property; vacant lots; consumer-owned motor vehicles; household goods; intangi- bles subject to general property taxes; and other and unallocable real property. PROPERTY TAX CONSIDERED AS A SALES TAX The half of the property tax revenue which comes from taxes on housing can be compared, if they are considered to be consumption taxes, to consumer expenditure for housing. In 1962, cash outlays for housing by SMSA homeowners and rental payments by SMSA renters amounted to an estimated $34.9 billion.7 Housing property taxes amounted to 19 percent of this total; they amounted to 24 percent of housing expenditure excluding th~ tax paymei~ts, which is the conven- tional American way of looking at other forms of consumption taxa- tion.8 Business property taxes, on the other hand, can best be com- pared to the value of business output-in this case, estimated national income originating in nonf arm private business enterprises in SMSA's. Business property taxes amounted to about 2.4 percent of this total in 1962, in SMSA's.° There are significant geographic differences in the relative roles of housing and business property as sources of property tax revenue. Quantification of these differences for entire metropolitan areas is complicated by the existence of so many submetropolitan assessment organizations, with differing levels of assessment. In table 8, data are presented for 19 areas, mostly single counties in multicounty SMSA's, for which the assessment-level problem can be overcome and for which there are available central city data. Housing comprises only 20 per- cent of the property tax base in the Atlanta area,1° but nearly 70 per- cent of the tax base in the San Antonio area. However, most of the areas cluster around the 50-percent level. This is an updating to 1962 of 1960 housing census data, based on changes in the housing stock and in the housing components of the Consumer Price Index between 1960 and 1962. This analysis is amplified below-. Corporate profits tax liability in 1962 amounted to 9.4 percent of national income originating in corporate business, on a nationwide basis. This includes State as well as Federal corporate profits taxes. 10 This is because homestead exceptions amount to roughly 20 percent of the gross assessed value of taxable property in Georgia; these exemptions, by defini- tion, apply almost entirely to housing in urban areas. PAGENO="0028" 14 TABLE 8.-Housing as a percent of assessed values subject to general property taxation ire selected metropolitan county areas, 1961 1 Area Entire designated Central city Outlying portions area Bexar County, Tex. (San Antonio) 68. 6 61. 9 93. 4 Washington, D.C., plus Montgomery and Prince Georges Counties, Md Shelby County, Tenn. (Memphis) Maricopa County, Ariz. (Phoenix) 1 San Diego County, Calif. (San Diego) 1 Tarrant County, Tex. (Fort Worth) Franklin County, Ohio (Columbus) Los Angeles County, Calif. (Los Angeles) 1 59. 4 55. 7 53. 9 53. 6 52. 7 51. 7 51. 0 52. 3 55. 0 52. 4 53. 5 51. 3 48. 1 51. 1 68. 0 58. 1 49. 4 53. 8 55. 9 58. 6 50. 9 Cook County, Ill. (Chicago) Hamilton County, Ohio (Cincinnati) Multnomah County, Oreg. (Portland) 1 King County, Wash. (Seattle) Alameda County, Calif. (Oakland) 1 Jefferson County, Ky. (Louisville) Lucas County, Ohio (Toledo) Cuyahoga County, Ohio (Cleveland) Jackson County, Mo. (Kansas City) Jefferson County, Ala. (Birmingham) Fulton County, Ga. (Atlanta) 1 50. 4 48. 1 46. 4 46. 3 45. 5 44. 7 42. 5 39. 8 39. 6 37. 7 20. 6 44.4 43. 1 44. 0 43. 8 39. 6 41. 3 42. 9 25. 7 35. 7 37. 3 21. 5 61. 4 53. 8 54. 1 50. 3 49. 9 47. 8 41. 9 54. 8 57. 6 38. 2 14. 8 1 Adapted from U.S. Census Bureau, 1962 Census of Government, vol. II, Taxable Property Values (1963; revised, August 1964). The areas selected met 2 criteria: (1) assessed vlue data for the central city were shown in the census volume; and (2) countywide assessment prevails, so that the census county-area data are not distorted by differing assessment levels. The 1 exception is Washington, D.C.; the 2 Maryland counties appear to have assessment ratios similar to those in the central city. In the referenced (1) cases, partia ex- emptions were allocated between housing and other property types by the author; except in the case of Atlanta, these were of minor consequence and the estimates do not affect the results significantly. Despite the increasing decentralization of economic activity in met- ropolitan areas, most central cities continue to have substantially heavier concentrations of business activity within their boundaries than do the outlying parts of their metropolitan areas. At the same time, the central cities have relatively more low-quality housing than do the suburbs, and substantially more multifamily housing (which is generally less valuable, per unit, than single-family housing, even if of equivalent quality). Therefore, housing usually comprises a lower proportion of the tax base in central cities than in suburban areas. This is true for 13 of the 19 areas in table 8, especially for Chicago, Cleveland, Kansas City, Washington, D.C., and Oakland. The housing share is roughly similar in central city and suburbs in five cases; it is significantly higher only for Atlanta, where the outlying sections are small in population relative to the central city." There do not appear to be systematic regiona~ differences in the housing share of the metropolitan area tax base; the areawide differ- ences appear to be related more to the legal coverage of the tax (a function of State laws), and to specific economic characteristics of individual metropolitan areas. However, table 8, as well as data for other areas not shown in the table, suggest that the central city-sub- urban disparities in the housing share are sharpest in the Northeast 111t should be observed that table 8 deals primarily with individual major counties rather than with entire metropolitan areas. However, of the 19 areas listed, seven consist of entire single-county SMSA's as defined in 1961. The seven are Bexar, Shelby, Maricopa, San Diego, Franklin, Lucas, and Jefferson. PAGENO="0029" 15 and Midwest. This seems to be related to two factors. First, the large older central cities in these regions developed in an era in which nearly all economic activity was concentrated in the central cities, and they still retain much of this activity, particularly commercial activity. In contrast, the newer parts of the country have developed with a more dispersed pattern of activity from the start. Second, the large old cities do have very large blocks to old low-quality and low-value housing and some of them have very large proportions of multifamily housing. This suggests two rather distinct groups of metropolitan areas, insofar as the effects of the property tax are concerned. One additional partial explanation for the disparities in the housing share of the tax base lies in the frequent favorable treatment of housing by central city assessors, which is discussed at greater length in a subsequent section. For example, were housing and business prop- erty assessed at similar fractions of market value in cities like Chicago, a substantial fraction of the apparent central city-suburban disparity in tax base composition would disappear, but by no means all of the disparity. PAGENO="0030" EFFECTS OF THE PROPERTY TAX IN URBAN AREAS HOUSING AND THE TAX BURDEN As suggested previously, the property tax on housing is analogous to taxes commonly known as consumption taxes; that is, general sales and selective excise taxes. Like the ordinary consumption taxes, the great bulk of the burden of the housing property tax appears to rest upon housing consumers, whether they are owner-occupants or ten- ants. There are some exceptions; the chief one is that owners of rental property canot shift the burden of that portion of the tax which falls on the land underlying their buildings. But, for the country as a whole, probably well over 90 percent of all property taxes on housing are borne by housing occupants. Since this is so, it is useful to view property taxes in relation to con- sumer expenditures for housing. This expresses the relationship in a form similar to that with which we are familiar in connection with other consumption taxes-a sales tax of 4 percent of taxable purchases, for example. There are two sets of data available to illustrate this re- lationship. The first expresses property taxes as percentages of the esti- mated rental value of housing (as found in the national income ac- counts). As table 9 shows, property taxes average about 19 percent of the rental value of nonfarm housing in the United States currently, equiv- alent to an excise tax of nearly 24 percent on rental value, excluding property taxes (parallel to the way in which sales tax rates are stated, as percentages of sales before sales tax is added). The second set of relationships, also shown in table 9, is perhaps a more realistic one, from the standpoint of housing consumers them- selves. This is the relationship noted in the previous section, and esti- mated for all SMSA's as of 1962. (The table 9 data are for 1960.) It expresses taxes as percentages of actual cash outlays for housing- expenses of owner-occupants or rental payments of tenants. The per- centages are very high expressed in this way, too, especially outside the South. Converted to a before-tax form, they range-excluding the South-from sales-tax-equivalent rates of 18 percent for large apart- ment houses outside New York City to 30 percent or more for single- family houses in the Northeast, and multifamily properties in New York City. In general, the upper end of this range applies to most of the Nation's large cities outside the South. (10) PAGENO="0031" 17 TABLE 9.-Property taxes on housing as percent of housing expenditure 1960 1065 I. In relation to rental value of nonfarm housing (national income data): 1 Owner-occupied housing 18. 1 18. 9 Rental housing 19. 3 19.4 All nonfarm housing 18. 5 19. 1 II. In relation to actual housing expenditure or rents (census data) : 2 Owner-occupied single-family houses- All United States 17 In standard metropolitan statistical areas 19 Northeast region 24 North central~~~~_ 20 South 10 West 18 Rental properties- 1-4 unit properties 17 5-49 unit properties 17 New York City 23 Elsewhere 16 50-or-more unit properties 20 New York City 23 Elsewhere 15 1 Includes the imputed rental value of owner-occupied houses. Based on data in U.S. Department of Commerce. The National Income and Product Accounts of the United States, 1929-65, A Supplement to the Survey of Current Business (1966). Taxes on rental property estimated by the author. 2 Based on 1960 Census of Housing data, including special tabulations for New York City. The data, except for the New York figures, appear in Netzer, Economics of the Property Tax (Brookings Institution, 1966), tables 2-8 and 5-6. For owner-occupied houses, taxes are expressed as percentages of annual cash housing outlays by owners; for rental properties, taxes are expressed as percentages of rental receipts for mortgaged properties. SHARE PAID THROUGH RENTS What this means is that very large numbers of urban families pay, via their rents, or directly if owner-occupants, taxes which amount to very sizable increments to their housing costs. This is shown more directly in table 10, which contains a distribution of housing units in multifamily rental housing subject to property taxes amounting to a sales tax equivalent of 20 percent or more. New York City is broken out, because it accounts for so large a fraction of the multifamily housing stock (about one-fourth), because its property tax rates are high (though no higher than in most other cities in tile Northeast and a few elsewhere), and because its sales-tax-equivalent tax rates are so high (in part because rent control keeps rents, the denominator of the fraction, down). For the country as a whole, as of 1960, 3.6 million households-more than half the total in this type of housing-were subject to rates of 20 percent or more, and 1.2 million were subject to rates in excess of 33.3 percent. PAGENO="0032" 18 TABLE 10.-Estimated number of households living in rental housing subject to high property tax rates, 1960 1 [In thousands] Real estate tax relative to rental receipts, stated as a sales tax equivalent 2 New York City Elsewhere in United States U.S. total 33.3 percent or more 541 676 1,217 25 to 33.3 percent 20 to 25 percent Total, 20 percent or more 568 293 513 1, 021 1,081 1, 314 1, 402 2, 210 3, 612 1 Adapted from U.S. Census Bureau, 1960 Census of Housing, vol. V, Residential Finance, pt. 2, Rental and Vacant Properties and from special tabulations for New York City. The census data for this purpose cover only mortgaged properties acquired before 1919; the estimates shown here have been adjusted to cover all rental properties. These data cover only properties with 5 units or more. 2 Real estate tax as a percent of rental receipts less real estate tax. These very high tax rates are greatly in excess of the rates applicable to other forms of consumer expenditure, with the exception of taxes on liquor, tobacco, arid gasoline. The highest retail sales tax rate currently is 5 percent, and this rate applies only in places which exempt food from the tax; the typical sales tax rate is closer to 4 percent. If we exclude liquor, tobacco, and gasoline, all the indirect taxes which fall upon consumers-including shifted business taxes as well as ordinary sales and excise taxes-probably amount to less than 10 percent of nonhousing consumer expenditure-less than half the level of housing taxation. It is simply inconceivable that, if we were starting to develop a tax system from scratch, we would single out housing for extraordinarily high levels of consumption taxation. More likely, we would exempt housing entirely from taxation, just as many States exempt food from the sales tax. SPECIAL BURDEN ON THE POOR The situation is even worse than the preceding discussion implies with respect to the low-income population. We have not discussed income heretofore. But the housing property tax is heavily regressive, absorbing a much higher fraction of the incomes of the poor than of the rich. This is largely because housing expenditure looms so large in the budgets of poorer families; in addition, the poor tend to be concentrated in the high property tax rate central cities. Table 11 indicates how heavily the property tax does burden poorer renters in the country as a whole, and for New York City, the only individual city for which good evidence exists.12 The burden on the poor in the latter case is even heavier than in the country as a whole. However, because rent control in New York City moderately reduces ratios of rent to income for poorer families in the city, the housing property tax is probably more severe for poor families in other large northern cities than it is in New York, on balance. `~ The New York City figures apply to homeowners as well as renters, but are dominated by the latter, who occupy 79 percent of the city's housing units. PAGENO="0033" 19 TABLE 11.-Estimates of housing property taxes as a percent of income, by income class Income class All renters, United States, 1959-60' New York City 1960-61 2 Less than $2,000 $2,000 to $3,000 $3,000 to $4,000 $4,000 to $5,000 $5,000 to $7,000 $7,000 to $10,000 $10,000 to $15,000 Over $15,000 8. 5 ~* 9 3. 0 2. 5 2. 1 1. 8 1. 6 1. 4 8. 6 5. 6 4. 1 3. 4 2. 8 2.4 2. 2 2. 7 1 From Netzer, op. cit., table 3-8. 2 Adapted from Alan D. Donheiser, "The Incidence of the New York City Tax System," in Graduate School of Public Administration, New York University, Financing Government in New York City (1966), p. 177. This regressivity, illustrat~d in figure 3, is of concern, and not just from the standpoint of equity pure and simple. High taxes on housing, relative to taxes on other uses of the consumer's dollar, are likely to discourage expenditure for housing. But they are likely to be an espe- cially severe deterrent to the poor, since they have so little leeway in family budgeting-there is little else that they can forego in order to rent better housing. This amounts to saying that the poor probably are more sensitive to price (rent) differentials than are the rich, out of necessity rather than choice. The general effect of these very high property taxes on the supply of housing to the low-income population has two components. First, taxes raise the cost of housing to the occupants and simply put a sig- nificant part of the existing housing stock beyond the reach of many in the low-income population. For example, take a large urban family with an income of $4,000. The maximum tolerable rent for such a f am- ily (if the rent is not paid by public assistance) might be said to be 30 percent of income, or $100 a month. Assume that the prevailing level of property taxes is 25 percent of rents in that city. Were property taxes on housing eliminated, this family could afford to rent an apartment now renting for $133 a month, without exceeding the 30 percent rent/ income ratio, an apartment which is now out of reach. Second, high property taxes discourage consumption of and invest- ment in housing in general, by the entire population.'3 There is ample evidence that consumers will buy more and better housing if its price is lower, just as they do with regard to most other objects of consump- tion. It can be and has been argued persuasively that one of the most effective ways of helping the low-income population (with respect to housing) is to rapidly increase the total supply of housing in a par- ticular city and metropolitan area; a decrease in prices (rents), while having no immediate effect on total housing supply, does create a larger effective housing market for those who now suddenly can afford 13 It should be noted here that any tax tends to reduce income available for expenditure by consumers, and that a reduction in disposable income will re- duce demand for housing as well as demand for other consumption goods and services. But in general, this reduced-income effect will depress demand for housing much less than will the increased price effect of a tax which, like the property tax, specifically applies to housing as such. PAGENO="0034" 20 Fig.3 HOUSING TAXES PAID ON RENTED NONFARM HOUSING AS A PERCENT OF FAMILY INCOME OF RENTERS BY INCOME CLASS, 1959-60 Percent ~, 4~7 -~---~_~ ~/~` vç~ 7'~/~' 7~7 . I under 2.000 3,000 4,000 5,000 7,000 10,000 15,000 $2,000 to 3,000 to 4,000 to 5,000 to 7,000 to 10,000 to 15,000 and over source: table II Income class of renters more of the existing housing. The evidence from the 1950's strongly suggests that the housing conditions of the poor improved most radi- cally in those areas in which the total suppiy of housing rose i~'iost rapidly. The process by which this occurs is related to the rate of turn- over of housing; this argument has been carefully developed in a re- cent article, using New York City evidence.14 There is yet another aspect to all this. Most low-income families do spend large proportions of their incomes for housing, but not all of them do so. In the 1960 census, there were 3.8 million renter households with incomes below $3,000 for whom rent/income ratios were coin- puted. Roughly 600,000 of these spent less than 25 percent of their incomes for housing. Now it is a worthy objective to attempt to per- suade some of these families to moderately increase the proportions of income spent for housing, by offering them greatly improved housing 14 See Frank S. Kristof, "Housing Policy Goals and the Turnover of Hous- ing," Journal of the American Institute of Planners, August 1965, pp. 232-245. PAGENO="0035" 21 conditions at small increases in rent. A reduction in the property tax burden could do just that, by reducing the prices of housing across- the-board. To return to the $4,000 urban family, assume a present rent/ income ratio of 20 percent, which means a monthly rent of $67. A 50 percent reduction in property taxes, combined with an increase in the rent/income ratio to 25 percent, would permit this family to rent an apartment which now rents for $97 a month, a 45 percent improvement in housing conditions (presumably) with a one-fourth increase in actual rental payments. PROPERTY TAX DETERS IMPROVEMENTS This line of argument applies not only to the central city poor. It can be generalized to apply to the broader problem of improving the future prospects of the large, old central cities as attractive resi- dential locations for families of all income levels. An essential in- gredient of such an improvement program must include offering a stock of housing which affords a reasonable alternative to the housing available in suburban locations. This means modernizing of that por- tion of the central city housing stock which is amenable to moderniza- tion,'5 replacing other parts of the old housing stock with new hous- ing, and building housing on the limited amount of available central city land. High property taxes effectively shrink the market for all these types of improvement in the central city housing suppiy. This is perhaps most obvious for owner-occupants of unmodernized older housing considering rehabilitation; the rehabilitation is not only costly and difficult to finance, but in most cities, it also will result in some in- creases in assessment. In some cases, the assessor will heavily discount the improvements, but in other places where the old housing is assessed at levels which are low for that city, the rehabilitation may trigger a very large increase in the assessment and in tax liability. In any event, one of the few case studies of this subject suggests that fear of poten- tial property tax increases can be a potent deterrent to improvements of central city residential properties.'° The effects of high property taxes in shrinking the market for new central city housing seem to be obvious to policymakers, for they are increasingly attracted to tax exemption or abatement schemes for specified types of new housing, of types similar to the New York arrangements under which 70,000 middle income housing units have been built. It should be noted that the issue here is not the height of central city tax rates compared to those in the suburbs, for middle income families. Indeed, frequently the central city tax bills, for the housing which could be an alternative to suburban housing, will be lower than the taxes in the suburbs. In suburban communities, particularly bedroom suburbs, the public services that a family receives or has access to are very closely tied to the local taxes that the same family pays. Therefore, in a sense, the property tax in maily suburbs is analogous to a general charge for `~ Experience in a number of cities suggests that there is very little old hous- ing which is not amenable to modernization, although some of it is very costly. `~ George Sternlieb, The Tenement Landlord (Urban Studies Center, Rutgers- The State University, 1966), Chap. 11. PAGENO="0036" 22 the use of public services, or perhaps even to a local income tax. It is unlikely to be a deterrent to consumption of housing; that is, to the expenditure of consumer income for housing. For the central cities, this is not the case. Central cities provide a wide variety of services and tax a wide variety of property types. Individuals cannot reasonably assume that the prices of housing con- fronting them include an identifiable tax component which is in effect a charge for a preferred package of public services. WThat they do observe is that housing is expensive in the central city. It may not be any more expensive in the central city than in the suburbs. But an effective city-rebuilding strategy requires that the central cities en- courage more private expenditure for housing, and this may in turn require that housing be much cheaper in the central city than in the suburbs. Moreover, the suburban nexus between taxes and public services is likely not to be present in the central city for yet another reason- inevitable more of the new central city housing will be rental housing, rather than owner occupied. Therefore, the taxes per se will not be apparent to the central city housing consumer-only the rentals, which reflect a tax component. And, making the comparison still worse, prop- erty taxes paid through rents are not deductible in computing Federal income tax liability, as are taxes on owner-occupied properties. URBAN DEVELOPMENT AND LAND USE The argument thus far is that the property tax does have one impor- tant class of effects on urban development and land use: by reducing consumer demand for housing in central cities, it tends to retard cen- tral city rebuilding and may very well make suburban areas appear relatively more attractive for households in a position to choose sub- urban residential locations as an alternative to central city locations. There are two other classes of effects: (1) central cities versus suburbs in respect to nonresidential land use; and (2) the pattern of land use in outlying parts of metropolitan areas. CENTRAL CITY ECONOMICS There has been much discussion of the effects of taxes on the loca- tion of economic activity in recent years. Most analysts have come up with negative findings, to the effect that State-local tax differentials have little impact on location, largely because tax differentials are so small, relative to the differentials in other costs of doing business, especially labor and transport costs. However, most such studies have been done on a statewide or inter- regional basis, using States or regions as the units of observation. Nontax costs do differ greatly among widely separated locations, so the usual findings are not surprising. But within a single metropolitan area, nontax cost factors are likely to differ only slightly; indeed, in some cases, local taxes may be the only costs of doing business which differ among alternative locations. An analysis of tax differentials for manufacturing in the New York area in the late 1950's indicated very PAGENO="0037" 23 large differentials in State-local tax burdens, with the older central cities having markedly unfavorable positions.'7 A more recent study of New York City's finances provides fairly clear evidence that these major business tax differentials actually have stimulated decen- tralization of economic activity away from the central city.'~ Moreover, if the tax differentials are unfavorable to central city locations, this reinforces rather than works against the powerful eco- nomic forces making for decentralization of economic activity from central cities to suburbs. The central cities have been losing out rela- tively (and in some cases absolutely) with respect to manufacturing and wholesale distribution of goods and with respect to population- serving activities (which follow population out to the suburbs). Tax differentials may trigger decisions to move away from the central city which might have occurred in any case, but at a later date. But this acceleration of locational shifts is a loss to the central cities. Of course, not all economic activities are equally susceptible to tax differentials. For one thing, some kinds of activities have heavy re- quirements for special types of public services, which higher central city taxes may buy. Then, too, there will be no effect on location if the tax disadvantages do not exceed the special advantages of central city locations-for example, for corporate headquarters in large-city cen- tral business districts-or if there is no choice at all, typically the case for banks, public utilities and newspapers. Are actual property taxes on business property enough higher in central cities so that they do in fact stimulate or contribute to migra- tion of economic activity from central cities to the outlying parts of metropolitan areas? It is difficult to assemble conclusive evidence on this, but there is some fragmentary evidence which supports a positive answer to the question. Consider the data in table 12. The last column of the table presents estimates of the differentials in effective tax rates for all types of tax- able property for 24 large central cities and their immediately adjacent suburban hinterlands. There are only four cases in which suburban effective tax rates are clearly above those in ceiitral cities (5 percent or more higher) and another three cases in which effective rate levels are roughly the same in central cities and suburban territory. In the remaining 17 cases, central city tax rates are substantially higher; they are nearly twice as high as in suburban areas in such older central cities as Newark, Cincinnati, and Baltimore. `~ Alan K. Campbell, "Taxes and Industrial Location in the New York Metro- politan Region," National Tax Journal, vol. II (September 1958), pp. 195-218. 18 See the papers by Leslie E. Carbert, James A. Papke, William Hainovitch, and Henry M. Levin, in New York University Graduate School of Public Admin- istration, Financing Government in New Yor1~ City (1966). In this case, non- property taxes appear to be the real culprits. PAGENO="0038" TABLE 12.-Property tax differentials, central cities versus suburbs-selected large metropolitan areas, selected years between 1957 and 1961 1 [Values show outlying portions of area as percentage of central city] Region Central city Suburban area Percent ratio of suburb an area to central city Approximate effective tax rate relationship 2 Per capita property taxable values , Per capita property tax revenue (1) (2) (3) (4) (5) 4 (6) Rest of: Northeast New York City Philadelphia Buffalo SMSA SMSA Erie County 131 146 112 134 94 96 102(100) 64 86(92) North-central Newark Rochester Chicago Detroit Cleveland St. Louis Milwaukee Essex County Monroe County~. Cook County Wayne County_~. Cuyahoga County. SMSA Milwaukee 158 100 123 102 106 96 138 91 58 93 87 88 105 91 58(51) 58(75) 76 85 83(97) 109 66(81) West S Cincinnati Kansas City Columbus Los Angeles San Francisco San Diego Seattle Denver Portland County. Hamilton County - Jackson County~. Franklin County~ Los Angeles County. SMSA San Diego County. King County SMSA Multnomah 122 52 117 102 85 100 91 90 77 66 62 137 68 89 90 66 58 47 54 119 117 67 105 90 73 64 61 South Oakland Baltimore Washington, D.C Atlanta Louisville County. Alameda County~. SMSA SMSA Fulton County_. - - Jefferson County - 90 110 95 82 145 78 58 96 71 98 87 53(55) 101 87 68 I From Dica Netzer, op. cit., table 5-7, p. 118, and app. 0. The presentation hero omits Central city equals 100. data for SMSA's which are not among the 38 largest. Col. S divided by col. 4. 2 Figures in parentheses are based on direct computations of effective rate differentials from various local sources. The other figures are derived by dividing the data in col. ~ NOTE: SMSA-Standard metropolitan statistical area. by those in col. 4. PAGENO="0039" 25 These figures do not apply solely to business property. However, some inferences can be drawn about business property on the basis of the limited evidence on differential assessment practices, by type of property, in some of the large central cities, from the 1962 Census of Governments.19 This evidence suggests that effecti ye tax rates 20 on business property areprobably significantly lower in Kansas City and St. Louis than in their suburbs; but in San Francisco, New York, and Cleveland, central city taxes on business property are probably higher. In the latter cases, the rough parity in overall rates appears to be a consequence of favorable central city treatment of housing in assess- ment practices. In Atlanta, San Diego, and Los Angeles, central city business prop- erty tax burdens are probably less unfavorable, relative to their suburbs, than table 12 indicates. But in Chicago and Seattle, the busi- ness tax differentials are even more unfavorable than the table would suggest. In the other cases, the table 12 figures are probably reasonably accurate reflections of the differentials for business property. All this suggests that there are a fair number of large central cities in which business tax differentials could be a real factor in accentuating the dispersal of economic activity away from the central cities. This conclusion is based on data which are several years old. How- ever, there is no real indication that the situation has changed signifi- cantly in the past 5 to 10 years.21 Indeed, the continued rapid rise in expenditure by the local governments serving central cities, combined with the much slower rate of increase in their taxable property values, may indicate that the situation has deteriorated. Table 13, for 10 selected central city areas, shows that nominal property tax rates have increased very steeply in recent years in a number of cases. If the ratios of assessed to market value have not declined substantially (which does not seem to be the general case), then there have been some substantial increases in central city effec- tive tax rates, including the rates on business property. For presentation and discussion of evidence, see Netzer, Economics of the Property Tax (Washington, D.C.: Brookings Institution, 1966). 20 Effective tax rates express the relationship of the tax charged to the market value of the property-an appropriate basis for comparing levels of property taxation in different localities. The nominal tax rate is that set by local tax authorities and applied to the officially assessed values of properties. Compari- sons of nominal rates are likely to be meaningless or misleading unless one also has information about valuations-i.e., the relation between assessments and full market value-in the particular localities involved. 21 In the case of the most extreme differential in table 12-Newark-the situ- ation may have worsened. For 1967, the effective property tax rate in the central city was 6.6 percent, compared to a 3.25 percent average for the remainder of the SMSA. See I~Iorris Beck, "Fiscal Disparities in Metropolitan Areas," paper pre- sented at the National Tax Association Conference, Atlanta, Ga., Oct. 23, 1967. PAGENO="0040" 26 TABLE 13.-Changes in assessed values versus changes in tax revenues, selected csntral city areas, 1957-65 1 Percent increase in per capita Central city areas in listed SMSA's Assessed Tax revenue Implied percent increase in nominal values property tax rates 2 Baltimore, Md 18 49 27 Boston, Mass 2 37 35 Cleveland, Ohio 17 46 25 Indianapolis, Ind 7 24 16 Milwaukee, Wis 31 57 20 Minneapolis-St. Paul, Minn.. 21 30 8 Newark, N.J 57 Paterson-Clifton-Passaic, N.J 6 54 46 Portland, Oreg 29 42 11 Providence, R.I 28 90 48 1 The first 2 columns are based upon computations made for the Advisory Commission on Intergovern- mental Relations in connection with its report, Fiscal Balance in the American Federal System. The ACIR data are for average annual increases, and have been converted back to 8-year increase figures. The ACIR tax revenue data are for total tax revenue, not just property tax revenue; the areas selected are those in which non-property-tax rove sue is relatively insignificant and in which major annexations or reassessments did not occur in the 8-year period. 2 Computed from data underlying the first 2 columns on the assumption that changes in the relatively minor non-property-tax revenue did not appreciably affect the growth rate of total tax revenue. INTRA-SUBTIRBAN VARIATIONS There are very large numbers of local government units in metro- politan areas with property taxing power-over 7,000 such units in the 38 largest SMSA's at present. The average large SMSA thus has about 200 taxing units. This implies, of course, that the size of most of the units is small-an average population of less than 15,000 per unit outside the central cities. Small size alone assures that there are wide disparities within a single SMSA in regard to the level and adequacy of the tax base and the height of tax rates. If the units are sufficiently small, the location of individual plants can have major fiscal effects on the individual taxing jurisdiction. The evidence indicates clearly that differentials in property tax base per capita (or per pupil, for school districts) within metropolitan areas are very large ones, even if the extreme cases, like industrial enclaves and resort communities, are ignored. Ranges of 10 :1 are not uncommon.°2 Such wide variations in taxable capacity have two kinds ofeffects, from the standpoint of urban development. First, they per- mit individual communities within an urban area to offer public serv- ices which differ greatly in scope and quality. To some extent, this need not exercise anyone; rich communities, like rich families, will enjoy a higher standard of living than poorer places in any society which places a high value on choice and diversity. However, it is not so easy to be complacent about wide variations in the quality of those public services which are basic to the future well-being of metropolitan areas (and to that of the Nation), like education, nor about wide variations in the quality of services which have effects on neighboring communi- ties, like disposal of wastes and control of air and water pollution. See Netzer, op. cii., pp. 124-125. PAGENO="0041" 27 Second, although rich communities generally spend more than poorer ones, in most cases they do not spend as much more as their superior tax bases would permit. The result generally is that the rich communities have lower tax rates than do their poorer neighbors.23 This is often observed in aggravated form in connection with small communities with extensive concentrations of business property. The public service requirements of the business property are low, but it call yield very large amounts of tax revenue. The consequence is an extremely low tax rate. This, in turn, has two effects. It encourages economic activity to locate in low-tax jurisdictions, which may or may not be the optimal locations for particular forms of economic activity. Equally important, it encourages communities to plan land use for fiscal advantage, rather than on the basis of broader considerations: The* examples of the industrial tax colonies and the high income/low tax rate enclaves have encouraged large numbers of communities to zone to attract tax base and repel consumers of public services. From a fiscal standpoint, the best of all possible worlds appears, to many suburban decisionmakers, to be develop- ment of the community's vacant land by campus-type offices and laboratories and by housing expensive enough to assure that there will be few schoolchildren (because of the anticipated age levels of the owners). The popularity of this "fiscal merchantilism"-efforts to export service costs and import tax base-has generated a good deal of argument about the extent to which the various forms of land development "pay for themselves." There are, of course, endless possibilities for local variations in cost-revenue situations for particular classes of property. The presence of factories and shop- ping centers can increase some costs disproportionately, under certain circumstances, and have adverse rather than favorable effects on property tax rates. In some places, more expensive residential property is associated more with larger numbers of schoolchildren per family than with less expensive residential development. But by and large, the intuitive judgments of local policymakers are likely to be right since, in suburban areas, the school levy is usually more than half the total property tax levy; business property, that is, produces a school tax "profit," and much residential property produces a school tax "deficit." And if the latter is tile case, the best way to minimize the "problem" is to have as few "deficit-producing" houses as possible, for example, by requiring very large minimum lot sizes for single-family houses. Thus, in the New York area as of 1960, nearly half the vacant land in the 22-county region was zoned to require single- family houses on lots of 1 acre or more. This is not the place to comment in depth on the adverse conse- quences of land use decisions made on the basis of property tax considerations. It should suffice to say that efficient patterns of land use in metropolitan areas, in the broadest sense, are not necessarily those that maximize the current fiscal position of 23Ibid pp. 125-430. PAGENO="0042" 28 governments in the area. There is more to urban society than the contemporary public sector and governments are seldom in busi- ness to maximize their own welfare. Even if maximization of the public fisc were the overriding criterion for land use planning, the property tax is far from the whole story; moreover, a multiplicity of taxing jurisdictions, each striving to maximize its own cost- revenue combination, is highely unlikely to maximize for the area as a whole.24 It may be that this problem will be of lesser importance in future years. One reason for asserting this is that, as the larger SMSA's grow and spread, more and more of the urbanization will occur in previously rural areas which are served by relatively large (in a geo- graphic sense) taxing units; much recent urbanization has occurred in closer-in territory already carved up into rather small units. The larger the geographic size of the jurisdiction, the less the temptation toward "fiscal zoning," since each individual land use decision will have a lesser impact on the overall fiscal results. However, this favor- able tendency will be realized only to the extent that large numbers of new taxing units are not created as urbanization arrives; this is a function of State laws governing the creation of new units of govern- ment, which differ greatly.25 In addition, it has been claimed by a number of observers that dif- ferentials in tax rates, expenditure levels, and tax bases have been de- clining within metropolitan areas, at least among the jurisdictions out- side the central cities.26 This observation appears to be almost entirely a consequence of the urbanization of previously rural jurisdictions which then become like the already urbanized ones. There does not seem to be any pronounced tendency toward uniformity when the corn- parison is confined to communities already nrbanized at the beginning of the comparison period; this part of the problem does not appear to be withering away. In short, the conclusion here is that the incentive to plan land use for fiscal ends is now, and has been, a real one, and will continue to distort urban development pattei~ns as long as these four conditions obtain in metropolitan areas: (1) The existence of a large number of separate and relatively small local government units. (2) Responsibility on the part of these governments for the provision and a major part of the financing of costly public services. (3) Reliance by these governments mainly on the property tax, the base of which is closely related to land uses within each jurisdiction. (4) Placement of extensive powers over land use patterns in the hands of these governmental units. 24Ibid., pp. 131-132. 23 There is some reason to believe that the generation of new units is slowing down. Between 1962 and 1967, the number of municipalities was virtually un- changed, while there was an increase of 800 (about 5 percent) between 1957 and 1962. Also, the rate of increase in the number of special districts (not all of them with property taxing power) slowed down somewhat in the latest 5-year period and the number of school districts continued to decline as rapidly as in the previous 5-year period. See U.S. Census Bureau, 1967 Census of Governments, "Governmental Units in 1967," preliminary report (October 19G7). 26Netzer, op. cit., pp. 132-135. PAGENO="0043" AN APPRAISAL OF THE PROPERTY TAX This appraisal consists, in the main, of a catalog of defects, some of them a summary of previously elaborated points. The property tax does have virtues, so the public interest is in doing what can be done to minimize the defects, rather than merely describing the defects of the institution. The principal virtues are pragmatic rather than philo- sophical ones. The tax exists; it produces very large revenues; and our society and economy have adjusted to and worked through many of the baleful effects of the tax, at least of present levels of property taxation. These are significant advantages, but they are not enough to compel us to reject reform out of hand, or to ignore methods of reducing reliance on the tax over time. EFFECTS ON HoUsING AND URBAN DEVELOPMENT To briefly recapitulate the appraisal developed in the previous sec- tion, the property tax, on balance, has the following kinds of sig- nificant effects on urban housing and development: 1. The tax amounts to a very high consumption tax on housing expenditure and thus tends to reduce consumer demand for hous- ing. This in turn tends to limit growth in the stock of urban hous- ing and to limit improvement in the quality of the existing housing stock. 2. These effects are not likely to be very evident in suburban communities, especially the better-off ones, for two reasons. First, the connection between property tax payments and local public services provided homeowners is a clear one in most suburbs. Sec- ond, the Federal income tax advantages of homeownership, for relatively well-off taxpayers, offset the property tax in large measure. 3. However, the deterrent effects of high taxes on consumption of, and investment in, housing in large central cities are serious. This is likely to be more true for tenants than for homeowners, and more so for lower income groups. For upper income groups, the outcome may be a marginal encouragement to the observed trends toward suburban residential locations; for the poor, the outcome will be less and poorer housing. 4. In many places, the property tax helps to make central city locations relatively unattractive for some types of business activ- ity, although the evidence on this is by no means unequivocal. 5. Outside the central cities, in metropolitan areas which oper- ate with large numbers of local taxing units (most SMSA's out- side the Southern States), there are very large differentials in property tax bases. One consequence is an unacceptably wide vari- ation in the scope and quality of the public services essential for urban living and for the achievement of a good urban environ- ment. (29) PAGENO="0044" 30 6. Another consequence is wide variation in effective property tax rates. This can promote a less-than-optimal pattern of the location of business establishments. Equally important, it often leads to emphasis on fiscal criteria in land use planning-efforts to maximize the fiscal position of each of the large number of taxing units-rather than concern with the broader social, economic, and esthetic considerations which should be the basis for planning decisions in urban areas. EFFECTS ON OTHER RESOURCE ALLOCATION The effects of the property tax on central city housing and on the location of economic activity in urban areas are by far the most im- portant of its resource allocation effects, for urban areas. There are, however, some other types of effects related to the fact that property taxes are usually much heavier burdens on some types of industries than on others, notably on particular types of public utility enter- prises. Some of these effects are of relatively minor importance for urban areas in general, although important for specific firms. For example, the characteristically high property taxes on gas utilities increase gas rates to consumers relative to the prices paid for fuel oil, thus marginally discouraging use of gas for heating.27 A more important effect is among competing forms of transporta- tion. Because railroads own their own rights-of-way, and because of frequent discriminatory assessment of railroad property, property taxes on railroads are very high relative to those on others types of transportation. It is estimated that, in 1957, property ta.xes amounted to 4.9 percent of national income originating in railroad transporta- tion but only 0.04 percent of national income originating in air trans- portation; 28 the corresponding figure for motor freight transportation was probably about 2 percent. Since these are regulated industries, the tax differentials do show up in rates to users, and encourage use of the nonrailroad modes. On balance, this facilitates decentraliza- tion of industrial activity away from central cities, since one of their major advantages in the past has been their superior rail freight facilities. High and rising taxes on agricultural property 29 reduce net income of farmers and thus, on balance, help speed the conversion of farmland on the rural-urban fringe to urban uses. The more general effect on agriculture probably is to encourage marginal farmers to leave the industry and concentrate farmland in fewer but more heavily capi- talized units of production. EFFECTS ON INCOME DISTRIBUTION One of the more frequent charges against the property tax is that it is more regressive 30 with respect to personal incomes than alterna- tive forms of taxation. Before examining the evidence on the distribu- ~` In New York City, natural gas rates to domestic consumers might be 10 per- cent lower, were property taxes no higher than those reflected in retail fuel oil prices. Since the prices of the two heating fuels are roughly equivalent now, the property tax appears to be a real deterrent to gas use. Netzer, op. cit., p. 26. ~ In 1957, these amounted to nearly 8 percent of national income originating in agriculture, versus 2 percent for nonfarm business. Ibid., pp. 24, 28. 30 Regressive taxes exact a higher percent of income as income decreases; progressive taxes exact a higher percent of income as income rises. PAGENO="0045" 31 tion of property tax payments by income class, it is worth noting the probable income distribution effects of other forms of State-local taxation: 1. General sales taxes.-In those jurisdictions where food is exempt, these taxes are regressive for the very lowest income groups, roughly proportional in the middle-income ranges, and regressive again for upper-income groups. Where food is included in sales taxes, they are much more regressive, and throughout all income categories. 2. Selective sales taxes are usually quite regressive. 3. Personal income taxes covering all forms of personal income are very progressive in those States with graduated rates, and mildly progressive where there are flat rates but personal exemp- tions or credits. The typical flat-rate local income tax without exemptions and applying only to earned incomes is proportional for lower and middle-income households, but regressive for the rich, who have substantial property income. 4. Business taxes.-Taxes on gross receipts, value added and profits (if it is assumed in the latter case that the bulk of the tax is shifted forward to consumers) have an incidence pattern something like the general sales tax with food exempt, that is, moderately regressive. If it is assumed that a substantial part of taxes on corporate profits is borne by stockholders, then these taxes become progressive among upper-income groups. In the aggregate-fOr the country as a whole-the incidence of the property tax is roughly proportional to income, except among the lowest income groups, where the tax amounts to very high fractions of incorne.~' The property tax therefore appears at least as good on income distribution grounds as most other State-local tax forms, except for personal income taxes. However, the aggregate conceals important variations. The important distinctions are, first, between nonresidential and housing property taxes; and second, within the housing sector, among geographic areas. The nonresidential property tax is rather like a general consump- tion tax, because much of the tax, notably that on utilities and most trade and service activities, is shifted forward to consumers. The tax is therefore regressive up to incomes of about $10,000. However, because some of the tax is not shifted-including that part on the land underlying nonresidential structures-the tax tends to be progressive above that income level. As for property taxes on housing: Rather good evidence on incidence by income class of property taxes on owner-occupied houses strongly indicates that this com- ponent of the tax is even more regressive than the nonresidential component. Somewhat less direct evidence indicates that the tax on rented housing is still more regressive. However, because renters tend to be both poorer and decidedly smaller consumers of housing (and hence pay less property tax, via rents), when the two series are combined, the picture is less clear. Residential property taxpayments decline sharply as a percentage of aggre- gate income as income rises in each class in the lower half of the income distribution but the percentage climbs again in the middle- 31 For amplification, see Netzer, op. cit., chap. III. PAGENO="0046" 32 income range up to the $10,000-$15,000 level. For the income class over $15,000, the percentage again drops.32 The combination of homeowners and renters has an important geo- graphic characteristic-it combines central city residents and subur- banities. Within each of these geographic groupings, housing property taxes are probably markedly regressive. This is surely the case with respect to large central cities, as a number of studies of individual areas indicate; see, for example, the evidence for New York City in table 11 above. For such cities, the property tax is decidedly inferior to other forms of local taxation on income distribution grounds. There is one important consequence of the regressivity of the prop- erty tax in central cities. As indicated earlier, the property tax in central cities finances substantial expenditure which, in effect, redis- tributes income in kind-health and welfare services, education for children in low-income families, et cetera. Thus, to a considerable ex- tent, central cities find themselves in the position of taxing the poor to provide services to the poor, which is surely nonsense. Equity in taxation refers not only to the burden of taxation on households in different income groups; it also includes the extent to which tax burdens differ among the households within a particular income group. Since consumption of housing is a major determinant of property tax burdens, and since this differs widely within income groups, property tax burdens do in fact differ considerably within in- come groups. There is some degree of this "horizontal inequity" even under an income tax, since exemptions, deductions, and the statutory definitions of income subject to tax result in differences in tax liability among taxpayers with a given total income. Under the Federal income tax law, in most income classes, the typical taxpayers will have tax liability which is roughly 20 to 25 percent more or less than the mean tax liability in their income classes. But taxes based on consumption expenditure or property values give rise to much larger degrees of horizontal inequity, since tax liability depends upon consumption decisions as well as inconie status. For New York City, for example, it is estimated that the residential real estate tax has a coefficient of variation of roughly 65 percent in all income classes. That is, the typical family's housing tax burden is almost two- thirds greater or less than the mean burden for its income class.33 In summary, the regressivity among income classes of the housing component of the property, and its horizontal inequity within income classes, are serious defects in the property tax, relative to other con- ceivable forms of local taxation. QUALITY OF ADMINISTRATION Ordinarily, the principal factors in an appraisal of any tax are its resources allocation effects (including effects on the location of eco- nomic activity, for taxes imposed on less than a nationwide basis) and its income distribution effects. We usually assume that at least the larger units of government can administer any major form of taxa- tion tolerably well. This assumption is not valid for the property tax, `~Ibid., p. 40. 33Melvin and Anne White, "A Per~ona1 Income Tax for New York City: Equity and Economic Effects," in Graduate School of Public Administration, New ~!ork University, "Financing Government in New York City" (1966), pp. 460-465, 482-488. PAGENO="0047" 33 however, even if we confine the analysis to relatively large units of government. It is generally agreed that, according to the principal measure of the quality of administrative performance in the property tax-uniformly in the ratios of assessed to market value for individual properties-the existing quality of administration is very bad indeed. Moreover, some observers argue that the potential for high-quality performance at reasonable administrative costs is very limited, espe- cially within large urban areas.34 There are two principal elements in the indictment of assessment ad- ministration. First, the evidence is clear that, within most assessing jurisdictions, even the most common, least heterogeneous properties- single-family nonfarm houses-are assessed at widely varying frac- tions of market value. Second, owners of different types of property in a single area are treated differently by assessors, sometimes systemat- ically but often erratically. The best evidence on the lack of uniformity within property types is that of the Census of Governments-comparisons of the selling prices of houses actually sold with their assessed values, and expressed as a "coefficient of dispersion," that is, the average deviation from the median assessment/sales ratio as a percent of the median ratio. As table 14 shows, in 1961; 21 percent of the areas located within SMSA's had coefficients of 15 percent or less and another 24 percent were in the 15 to 20-percent range. But a sizable number of areas, including six of the country's very large cities, had coefficients of 30 percent or more. The 20-percent level is often taken to indicate good assessment qual- ity. But it appears to be a very modest standard of administration when compared to sales or income tax administration. A 20-percent coefficient means that the typical homeowner can expect to be faced with a tax bill which is 20 percent more or less than it should be, given a legal requirement for uniformity. It is hardly conceivable that sales or income tax administrators would be satisfied with an average error in tax payments of 20 percent-S percent would be a more likely stand- ard, and one which is realized in Federal income tax administration. Not one place in SMSA's satisfied this more rigorous standard in 1961. TABLE 14.-Dispersion of assessment ratios for nonfarm houses in selected local areas in SMSA'S, 1961 1 Number of areas Percentage of total number Total number of selected areas with co- efficients in intraarea dispersion (in percent)2 of 483 100 Less than 15 102 21 15 to 19.9 115 24 20 to 24.9 102 21 25 to 29.9 71 15 30 to 39.9 62 13 40 or more 31 6 1 From TJ.S. Bureau of the census, 19C~ Census of Governments, Vol. II, Taxable Property Values, Table 16. 2 Average deviation of ratio of assessment to sales price (for houses sold in a 6-month period) from median ratio in the area, as a percent of the median ratio. 34Netzer, op. cit., pp. 173-183. PAGENO="0048" 34 There is some evidence that it is by no means inexpensive to realize a higher degree of uniformity. Typically the low-dispersion places spend more for administration, relative to revenue, than do the high- dispersion places. But although more costly administration may im- prove things somewhat for the relatively easy-to-assess single-family houses, it is by no means clear that a greater administrative effort can produce reasonable standards of uniformity for other types of prop- erty, which are often complex and, in many cases, infrequently sold. Thus both assessors and independent appraisers must rely on some proxy for market value (which is the conventional legal basis for taxation), and any proxy variable is likely to have a coefficient of correlation which significantly differs from ± 1.0 in a world with all sorts of market imperfections. The most coin- monly used proxy for buildings is depreciated reproduction costs, with both depreciation and reproduction costs determined for individual properties by applying detailed formulas and tables, rules which cannot help departing from the market's evaluation in individual cases. Another common proxy used for income- producing properties is capitalized earnings. The earnings data are likely to be objectively determinable, but the capitalization rate may not coincide with that of prospective purchasers. In their search for "true value," assessors combine the varieties of evidence: sales of similar (but not identical) properties, re- production costs, an evaluation of the observed condition of the subject property and of its environment, and income. However objective the basic evidence, its weighting is entirely subjective and varies from case to case. * * * It would appear that the tax on business property in general can be regarded as a collection of taxes nominally meas- ured by market value but actually measured by a varied and changing set of evidence more or less related to market value. Nor is the situation much better in a large jurisdiction with regard to residential realty which is varied in age, type, style, and loca- tion; only where the stock of housing is homogeneous in all major respects (as in a suburban subdivision built within a brief span of time) is there any assurance that market value and its proxies will be closely correlated.35 All this refers to uniformity within classes of property. The second count in the indictment refers to differences among types of property. In a fair number of large cities, the assessment ratio for single-family houses differs significantly from that for other types of property; more often than not, the ratios for single-family houses are lower than for the other types.3° A frequent pattern is that found in New York City-very high assessment ratios for office buildings, large re- tail stores and hotels; only slightly lower ratios for multifamily hous- ing and for industrial structures; considerably lower ratios for one- and `two-family houses; and extremely low ratios for vacant lots. To some extent, such differences among property types represent an accommodation to economic realities, albeit an extralegal and erratic accommodation. For example, relatively high tax burdens for central ~ Ibid., pp. 180-182. 20Ibid., table 4-1, p. 76. PAGENO="0049" 35 bu~iness office structures probably reflect an awareness of the central city's strong competitive position in this field, as contrasted with its weak competitive position with regard to industrial activity. But one cannot be sanguine about all the differentiation. The fre- quently high assessment ratios for multifamily housing may very well conflict with central city housing policy goals; few large old central cities can expect substantial expansion in their supplies of housing to result from a policy of favoring low-density single-family housing. Similarly, underassessment of vacant land is hardly consistent with any rational urban development program. And, more generally, lack of uniformity in assessments, if widely appreciated by taxpayers, has a real, if intangible, effect on a city's ability to obtain more revenue from the property tax. Voters are far more likely to resist tax increases if they believe the tax to be unfairly administered than if they be- lieve it to be reasonably equitable. Bad administration is, in this sense, an Achilles' heel, and is so considered even by those whose appraisal of the tax on other grounds is far more favorable than the appraisal presented in this paper. INHERENT AND REMEDIABLE DEFECTS It should be noted that this catalog of defects in the property tax can be subdivided in another fashion. Some of the defects appear to be fundamental to the nature of the property tax; others, notably poor assessment of single-family houses, are much more remediable; and still others largely relate to the fact that the tax is levied by so many jurisdictions operating within a single urban area. The major defects which seem to be inherent in the property tax are these: 1. Its adverse effects on the central city housing stock; 2. The difficulty in uniformly assessing business property; 3. The horizontal inequity of housing taxes within income classes; 4. The regressivity of housing taxes among tenants and among homeowners; and 5. The lack of neutrality among types of economic activity, particularly in connection with taxes on transportation and public utility property. The defects which relate to local government fragmentation, cor- rectable at least in part were the tax to be levied over broader geographic areas, are these: 1. The adverse effects of high central city business property taxes; 2. The effects on urban development patterns outside the central cities; and 3. Some part of the regressivity of the tax, in particular that part which results in taxing the central city poor to provide public services designed to alleviate or overcome poverty. This classification, then, suggests some possible policies to overcome the major failings in the property tax. PAGENO="0050" POSSIBLE REMEDIES To the extent that the defects of the property tax are inherent ones, the principal remedies must take the form of some reduction in the reliance on the property tax* for the financing of urban public services.37 The alternative financing is by higher levels of government which do not use the property tax, or by other local government revenue sources. Reduced reliance on the property tax will also diminish the importance of the other types of defects, mainly related to the fragmented local government patterns. But these latter defects can also be remedied in part by reforms within the institution of the property tax itself. INCREASE STATE-FEDERAL RESPONSIBILITIES The increase of State and Federal Government responsibilities is an obvious route toward reduced reliance on the property tax, and one which is in keeping with developments in fiscal federalism since the 1920's. These developments include both the transfer of direct responsibility for the actual performance of some functions from the local to State and Federal levels of government, and increased State and Federal financing of functions still performed at the local level. Consider four of the major functional classes of civilian public expenditure: education; highways; public welfare; health; and hos- pitals. In each case, the local government share of total direct public expenditure by all three levels of government declined appreciably from 1927 to 1965-66; for example, from 71 to 32 percent for highways and from 69 to 52 percent for welfare. But there were even larger declines in the proportions of local government expenditure financed from local, rather than State and Federal, revenue sources. As a result of this (and of another much less important factor, the expansion of local nonproperty revenue sources), the property tax now finances half or less of local expenditure for these functions, compared to 75 percent or more in 1927. The combined effect of functional transfers, increased grants-in-aid, and increased use of local nonproperty revenue sources is indicated by the following estimates of the percentages of total public expendi- ture of all levels of government financed by the property tax: 38 ~ This does not necessarily mean a rollback in property tax levies or rates. In practice it will mean a reduction in the relative role of the property tax; that is, financing increased expenditures in future years from revenue sources other than the property tax. are very rough estimates based on Census Bureau data; they some- what overstate the decline for education. (36) PAGENO="0051" 37 [In percenti 1965-66 1927 Education 37 Highways Public welfare 13 8 56 61 Health and hospitals 15 32 Suppose the 1927 percentages prevailed today, and expenditure levels were no different than they are. To finance these four functions, local governments would have needed, in 1965-66, an additional $23 billion in property tax revenue, above and beyond the $23.8 billion they actually received ($12.5 billion more for education; $5.5 billion more for highways; $5.1 billion more for health and welfare services). Thus, we can say that during this period, there has been an upward shift of nearly half the potential property tax burden. But a strong argument can be made for further upward shifts, in two functional areas. One is education; very small portions of the eventual benefits from education are recaptured within the con- fines of individual school districts, since our population is so mobile- perhaps no more than 20 percent on the average. This argues for a much increased role for external financing, especially at the Federal level. Since education now absorbs slightly over half of current prop- erty tax revenues, such shifts could greatly reduce reliance on the property tax. A second area is that of poverty-linked services, notably welfare and health services, which now absorb roughly 10 percent of total property tax revenues, but substantially more for the large central cities. A good case can be made for relieving the property tax of the job of financing all public services linked to the existence of poverty. Since this burden if concentrated in central cities, it would alleviate the central city-suburban disparities and the property tax problems these create; it would also alleviate the regressivity problem in the sense of taxing the poor for services to the poor. This is very much in keeping with the historic trends. During the last 30 years, each of the important institutional changes which re- duced pressures on the property tax has been associated with redis- tributive services. These include the Federal and State assumption of most public welfare costs in the thirties, via grants-in-aid, transfers of functional responsibilities, or direct Federal social insurance pro- grams; the steady expansion of the State government role in financing education in the past 20 years; the gradual increase in Federal fi- nancing of health services (either directly or through grants-in-aid), culminating in medicare and the 1965 social security amendments; the Federal role in the provision of housing for low-income people; and most recently, the new Federal participation in antipoverty programs and in the costs of education where there are extensive pockets of poverty. All of these Federal-State aids combined have not been suf- ficient to keep effective property tax rates from rising at a fairly rapid rate. But without external aid to urban-area local governments, the rise might have been far more rapid. PAGENO="0052" 38 Some evidence on the extent to which complete external financing of poverty-linked services would have alleviated central-city-suburban tax effort (tax revenue divided by personal income) disparities, had this been in effect in 1962 in 22 of the largest SMSA's, is presented in table 15. The estimates are those of Prof. `Woo Sik Kee of West Vir- ginia University. On the average, the tax effort disparities would have been cut in half: central city tax burdens would have been nearly 25 percent lower, while suburban tax burdens would have been 15 per- cent lower. Converted to effective property tax rate terms, this would have reduced central city tax rates to levels equal to or less than those in suburban areas in a number of cases. TABLE 15.-Measures of tax effort in central cities and suburbs in 22 largest S1~1SA's, 1962 1 (Per capita tax revenue, 1962, divided by per capita income, 1960) SMSA Actual ta x revenue Adjusted ta x revenue 2 113 Cities Suburbs Cities Suburbs Cities Suburbs New York 9. 5 7. 5 7. 8 7. 0 7. 3 6. 8 Chicago Los Angeles Philadelphia Detroit 7.4 8. 4 7. 4 7. 5 6. 1 7. 0 4. 9 5. 7 6. 6 7. 3 6. 6 6. 2 5. 8 6. 0 4. 6 4. 9 6. 2 6. 8 6. 1 5. 5 5. 6 5. 6 4. 4 4. 6 Baltimore 6. 9 4. 4 6. 0 4. 3 5. 3 3. 9 Houston 5. 9 5. 6 5. 4 5. 4 4. 7 4. 9 Cleveland 7. 4 5. 2 6. 1 4. 4 5. 5 4. 2 St. Louis 7. 6 5. 1 5. 9 4. 8 5. 2 4. 4 Milwaukee 8. 4 6. 5 6. 8 5. 4 6. 3 5. 2 San Francisco 7. 4 7. 2 6. 1 6. 0 5. 6 5. 6 Boston 11. 2 7. 4 8. 9 6. 8 8. 3 6.4 Dallas 5. 7 3. 7 5. 2 3. 1 4. 8 2. 7 Pittsburgh San Diego Seattle 7. 2 6. 3 5. 0 4. 9 6. 7 3. 6 6. 8 5. 3 4. 5 4. 7 5. 6 3. 2 6. 3 4. 7 4. 2 4. 5 4. 9 2. 9 Buffalo 7. 5 7. 0 6. 2 6. 2 5. 7 5. 9 Cincinnati 8. 2 4. 5 6. 5 4. 2 5. 7 3. 8 Atlanta 6. 3 3. 7 5. 1 2. 8 4. 5 2. 4 Minneapolis Kansas City Newark Mean 7. 0 6. 0 12. 3 6. 5 5. 4 7. 0 5. 3 5. 1 9. 5 5. 6 5. 0 6. 5 4. S 4. 5 8. 9 5. 3 4. 6 6. 2 7. 6 5. 7 6. 3 5. 1 5. 8 4. 8 1 Data computed by and presented in Woo Sik Kee, "City-Suburban Differentials in Local Government Fiscal Effort" (mimeo., Regional Research Institute, West Virginia University, October 1967). 2 Total tax revenue minus the estimated locally financed portion of expenditure for public welfare, health, and hospitals. 3 Total tax revenue minus the estimated locally financed portion of expenditure for public welfare, health, hospitals, and for education of children in families with incomes less than $3,000. To some extent, Professor Kee's estimates understate the reduction in the disparity. He is looking at tax effort as if all local taxes were charges against personal incomes received by residents of the communi- ties levying the taxes. But some part of the burden of taxes paid by businesses is "exported" to other communities, via higher product prices or reductions mt he profits accruing to nonresident owners of PAGENO="0053" 39 businesses.~9 And generally speaking, the export percentage is higher for central cities than for suburban areas, if for no other reason than the lesser role of housing in the central city tax base. Estimates made for New York City suggest that 25 percent of its tax burden was exported in the early 1960's.4° Similar estimates for suburban counties in the New York SMSA suggest that only about 15 percent of their tax burden was exported. Applying this correction to Kee's estimates, actual tax burdens on residents can be said to have been 7.1 percent of central city residents' personal incomes in 1962 and 6.3 percent for suburbanites. Had the poverty-linked services been entirely financed from external sources, the percentages would have been 5.5 for the city and 5.7 for the suburbs. Since effective prop- erty tax rates in central city and suburbs in this SMSA are roughly equivalent, the upward shift would have resulted in effective suburban rates nearly one-fifth above those in the central city, if the entire effect of the transfer fell on the property tax. OTHER LOCAL REVENUE SOURCES Changes in intergovernmental fiscal arrangements of the types pro- posed here could have a potent impact. Kee and other observers have asserted that much of the remaining disparity in tax burdens-where any does remain, after correction for tax exports-can be associated with the extent to which the central city accommodates the surrounding suburban population with places of employment, shopping and cul- tural facilities and, presumably, the public services supportive of these activities. To the extent that this is in fact the case, it strengthens the argument for greater reliance on other local government revenue sources, notably direct charges for the use of public services and facilities (paid by actual users wherever they may live) and local income taxes paid by residents and commuters alike. USER-CHARGES FOR PUBLIC SERVICES Local governments do currently employ user-charges; they obtain roughly 18 percent of their locally raised general revenue from charges for services (other than utility services) and from special assessments. About half of this amount comes from school lunch and similar charges, hospital charges and public housing rental payments, but charges apply to a wide range of other services. Despite this, there is considerable potential for greater exploitation of user-charges, in con- nection with activities which do not have significant income-redistri- bution objectives. The case for this has been put as follows: ~ * * many of the public services provided by local govern- ments are in many ways like those provided by public utility companies. That is, they are not provided uniformly to the entire population, but rather in distinguishable quantities and qualities to individual families in the population, who consume them in For comprehensive estimates of this, see Charles E. McLure, Jr., "Tax Ex- porting in the United States: Estimates for 1962," National Taa~ Journal, vol. 20 (March 1967), pp. 49-77. 4° See Alan D. Donheiser, "The Incidence of the New York City Tax System," op. cit. Table II, p. 162. PAGENO="0054" 40 accord with their personal preferences. For example, not all f am- ilies use the same amount of water, not all use the same amount of highway transportation, and so on. There is a strong case for financing such services in the same way public utility services are financed-that is, via user charges which are like prices, rather than through general taxes. If the purpose of providing the public service is to offer differ- ent consumers the services they want, and place some value on, then they ought to pay for such services in proportion to the costs. Otherwise, governments will be called upon to provide a great deal more of the service than people would be willing to consume if they did have to pay for it, which is a wasteful use of resources; or the service will be in such short supply that a form of nonprice rationing will be employed to allocate the service among con- sumers. The outstanding example of this is street congestion in cities: users pay for highways in the aggregate but not for specific individual uses of the streets, and therefore, not surprisingly, treat highways as a free good. The only deterrent to use of the streets at the most crowdec[ times and in the most crowded places is the value one places on time; the rationing in effect then results in those who place a low value on time pre-empting the street space from those who place a high value on time. Ordinarily, in our society, rationing is on the basis of price. Somebody who val- ues a service highly bids it away from someone who places a lower value on that service and would rather use his income for alterna- tive kinds of consumption.4' The most striking opportunities for greater utilization of user- charges, as this would suggest, are in connection with financing of urban highway and parking facilities and services, waste collection and disposal, and recreational activities; the potential revenue in these areas alone equals roughly one-tenth of property tax revenue on a nationwide basis. The potential is relatively larger in urban areas, especially the larger ones, which provide more of these services and are generally less effective exploiters of user-charges. LOCAL INCOME TAXES The case for local income taxation does not rest on the argument that it is a good device for central city taxation of commuters. To be sure, the potential here is a large one. The municipal governments of the 43 largest cities collected $3.3 billion in property tax revenue in 1965-66. Excluding Washington, D.C., their local income tax revenue was $242 million, largely made up of the collections in Detroit, Louisville, and the larger cities in Pennsylvania, Ohio, and Missouri. But if all 43 cities had a flat 1-percent tax on income earned within the central cities, revenue would have amounted to perhaps $1.3 billion, probably $300 million of this from commuters. This is a substantial fraction of property tax revenue for these units of government. The more general case for income taxation is that it does not have an especially adverse effect on housing, as does the property tax. It ~` Netzer, "Financing Urban Government," in James Q. Wilson, editor, The Metropolitan Enigma (Chamber of Commerce of the United States, 1967), p. PAGENO="0055" 41 escapes the regressivity charge. Moreover, for central cities, it is su- perior to local sales taxes and other loca' business taxes, since unlike the latter, it is highly unlikely to encourage migration of economic activity away from the central cities. Local income taxation heretofore has been largely a large city affair, except in Pennsylvania. However, in time, it may become more like the widespread use of local sales taxes, which in States like California and Illinois, are virtually universally used supplements to the State sales tax, collected by the State. As such, local nonproperty taxes have the general character of the State tax, since, if universally applied, they have no impact on the intrastate location of economic activity. They may be thought of as a susbtitute for outright grants-in-aid from State tax revenue, with the distribution of the funds based on the origin of the tax collections rather than on the traditional kinds of State-aid formulas. The latter can permit much more adequate reflec- tion of differences in needs among communities (although State-aid formulas frequently do not realize this potential). On the other hand, the local tax supplements permit much more local government discre- tion in the use of the funds. LAND VALTTE TAXATION An entirely different type of alternative revenue source would be a heavy tax on land values in urban areas as a partial substitute for currently collected property tax revenues. As the name implies, land value taxation is a tax on the value of land alone, irrespective of the value of buildings or the lack of build- ings on a site. A step in this direction is the graded or differential tax-the application of a higher tax rate to the land portion than to the improvement portion of property valuations. The argument for exclusive taxation of site values, or for substan- tially heavier taxation of land than of buildings, is an old one, and differential site value taxation is widely practiced-in western Canada, Australia, New Zealand, and South Africa, for example. The merits of the case have been submerged for many years by the extravagant claims of the proponents of site value taxation. Moreover, skepticism has been bolstered by the apparent absence of discernible effects in the places where site valuation is utilized. However, the case for~ site value taxation is a good one. The argu- ment, on equity grounds, is that most of the value of land is a conse- quence, not of actions by individual owners, but of collective invest- ment, community development, and population growth. Individual landowners therefore can realize large "unearned increments" over time. It is entirely appropriate for the community to recapture these unearned increments by taxation, and use them for community pur- poses. There are complications in this equity argument, related to the fact that most landowners have already paid, in their purchase prices, for at least some of the unearned increment, but by and large the equity argument makes sense. The economic argument is even more compelling. A tax on site value which is independent of the improvements on the site will not affect entrepreneurial decisions as to the use of the site; the best (most profit- able) use before tax remains the best use after the tax is imposed. In PAGENO="0056" 42 other words, the tax is neutral with regard to land use decisions. Since the present property tax, on both land and improvements, is not neu- tral but tends to discourage investment in buildings, a switch from the present tax to exclusive site value taxation (or to a tax heavily weighted on the land portion) would tend to have strong land use effects. Provided that demand permits, it would encourage owners to de- velop their sites more intensively, in an effort to minimize tax liability as a percentage of current receipts, since additional investment in buildings would not increase tax liability. Within individual urban jurisdictions, taxes on vacant land would tend to rise, thereby increas- ing the holding costs of vacant land and making the speculative with- holding of land from development a less attractive proposition. Thus, a switch to site value taxation is likely to have its maximum impact in two parts of a metropolitan area-in the central areas, where it would encourage more investment in buildings, and in the outlying sections, where it. would tend to discourage land speculation and the resulting patchy patterns of land development (less "leapfrogging" over sites withheld from the market). In theory, there are few if any legitimate economic arguments against site value taxation. On an operational level, there are grounds for hesitation. First of all, one may doubt the actual strength of the positive tend- encies associated with a switch to site value taxation. It is, after all, a major institutional change, and major institutional changes should not be pressed unless their positive effects are also expected to be major in extent. However, it should be noted that effective property tax rates in most American metropolitan areas are high and rising. The nega- tive land use effects of the present tax are likely to become increas- ingly apparent in time, and the likely benefits from a change in the basis of taxation will correspondingly increase. Second, there is some question about the revenue adequacy of site value taxation. Some calculations suggest that the present yield of property taxes on nonf arm realty substantially exceeds the total rental value of privately owned nonfarm land. Thus, even a 100 percent site value tax might not yield enough to fully replace the existing property tax (on real property, exclusive of personalty). This suggests that only a partial, rather than a complete, shift is possible, diluting the possible advantageous land use effects. Third, there are administrative problems if both land and build- ings are taxed, but at differential rates-the "graded tax" concept applying in Pittsburgh, Hawaii, and western Canada, for example. This makes it very important to accurately value land and buildings separately. Under a pure system of site value taxation, the building value is irrelevant. Under the conventional property tax, the distinc- tion between land and building for any individual site is also irrele- vant, although the statutes may require the assessor to make some statement about the notional separation. It seems likely that joint ad- ministration of the two different types of taxes will produce bad ad- ministration of the site value tax, in that assessors will tend to relate land and building valuations as they often do at present. Therefore, the proposal here is for a separate system of land value taxation, levied and administered, if possible, over a wide geographic area- a whole State or SMSA. PAGENO="0057" 43 TAXATION OF LAND VALtTE INCREMENTS The equity argument for taxation of increases in land values is at least as strong as that for annual taxes on total land values. This kind of tax is aimed at recapturing for the government a higher propor- tion of what economists call the unearned increment-the rise of land value that occurs, not through efforts of an owner but through gov- ernmental action (new highways, subway lines, zoning changes, etc.) and through growth of the population and industry of the community. Land value increment taxes strike directly at the unearned incre- ments realized by specific individual owners, and do not penalize present owners who have not realized substantial land value incre- ments. The claim in this case is not that community improvements tend to enrich landowners in general; tax liability occurs only when enrichment is demonstrated by the realization of capital gains on land. Variants of this approach to taxation are used in a number of coun- tries. In some Hispanic countries, specific public improvements are financed from taxes on the estimated (although not necessarily realized by sale) increase in the value of adjacent properties; this is done on a very large scale in some of the major cities in Colombia. In our own country, special assessments to finance street and sewer proj- ects have a similar rationale. In a number of European countries, this approach shows up in the income tax treatment of capital gains on land, which is much less favorable than that of other types of capital gains. Thus, land value increments could be taxes in a number of ways and at different levels of government. Perhaps the easiest method might be through State income taxation, by a special supplemental rate on capital gains on land or by including a larger portion of the gain as ordinary income (rather than the 50 percent now included in most States, following the Federal practice). This seems easiest, because the gains are already reported and legal definitions do exist and are applied. Like annual land value taxation, this form of taxation would be largely neutral with respect to the use of land and would not dis- courage new construction. However, its economic impact and revenue potential would be somewhat less. Presumably, the land value incre- ment tax would apply only to gains actually realized (including con- structive realization at the death of the owner). Very high tax rates would tend to postpone realization of gains, although closing the transfer-by-death loophole would reduce this. Nevertheless, the eco- nomic impact would be in the right direction and the equity effects appropriate. Therefore, this seems good policy, especially if straight- forward annual land value taxation does not prove acceptable. IMPROVEMENT OF THE EXISTING INSTITUTION Since some of the major defects in the existing institution relate to the fragmented structure of local government in urban areas, an ob- vious direction for reform is application of the property tax over wider geographic areas, thereby reducing tax rate disparities by evening out the differences in tax base per capita or per pupil. There are two approaches to this. PAGENO="0058" 44 TAX BASE CONSOLIDATION One would be to regionalize a segment of the tax base-eliminate local taxes on some types of property and levy property taxes on these types over a broader area, with either use of the proceeds for region- wide (or statewide) functions or distribution of the proceeds to local government units on the basis of some measure of need. A frequent suggestion along these lines is for regional or statewide taxation of business property, to eliminate local competition for economic ac- tivity and attendant pressures on land use planning.42 FISCAL FEDERATION A second approach is to regionalize the financing (and perhaps administration) of part or all of selected local government functions, but still utilize the property tax to the extent it is 110w used. The most dramatic proposal along these lines is to employ a statewide property tax for the great bulk of (non-Federal) school funds.4' There also have been proposals for metropolitan areawide school financing in, perhaps, some kind of fiscal federation. Under this scheme, State school aid would continue as at present, but paid to an areawide authority. The area authority would then levy a uniform areawide property tax and distribute this revenue and the State aid on a per pupil basis. For most school districts, this would provide adequate program levels. However, they would be free to supplement the levels with local levies, but pre- sumably only few districts would do so. Since property taxes for schools amount to roughly half of total property taxes, there would be an appreciable reduction in property tax disparities. Smaller, but real, reductions in disparties would result from metro- politan areawide financing of functions with a fundamentally regional character, like transportation and waste disposal. Within single-county SMSA's, this could be done on a county basis. In fact, even in multi- county SMSA's, significant reductions in disparities could be achieved by increased countywide financing. An indication of the potential can be found in aggregate data for SMSA's from the 1962 Census of Governments. Consider a few functions with a regional character, or for which countywide administration is -common in some parts of the country. Expenditure for these functions can be derived into that handled by county-scope units of governments (counties, combined city-county governments like New York, Philadelphia, and Baltimore, and large special districts) and that handled by subcounty units. The following is the approximate proportion of SMSA local government expenditure in 1962 handled by subcounty units: Percent Highways 55 Waste disposal 60 Parks and recreation 60 Health and hospitals 30 Public welfare 15 ~` It should be noted that the planning difficulties could be accommodated in another way-by regionalizing land iesc controls rather than taxation. There is much to be said for this course of action; indeed, there is hardly anything to be said in defense of land use planning by a huge number of small jurisdictions. But this is an entirely seperate subject. ~ has been proposed by Lynn A. Stiles, of the Federal Reserve Bank of Chicago. PAGENO="0059" 45 Now, had these functions been financed entirely on a countywide basis, roughly $1.7 billion of property taxes collected on a subcounty basis in 1962 would have been collected county-wide. This is roughly half the total property tax revenue of the subcounty units (excluding school districts) in that year. BETTER ADMINISTRATION The position in this report is that improved assessment of most types of complex business property is a utopian goal, but that it is possible to do a much better job with respect to housing, vacant lots and the simpler, more common types of business property, like small store buildings. The basis for this argument is that there are fairly frequent sales of such property to provide a basis for assessment. The requirements for realization of such improvements as are achievable are professionalization and adoption of truly systematic procedures- indeed, full computerization of the primary assessment process.44 These in turn imply large-scale assessment organizations. Except in the very largest States, this may very well imply state- wide assessment; it surely is not consistent with assessment districts having populations of very much less than 500,000. The general rule seems to be, if full use is to be made of the possibilities for computeri- zation, the bigger, the better. This, then, is no less radical a proposal than the others advanced in this report, since this country has been firmly wedded over many decades to the notion that small local assess- ment districts are an essential component of local self-government. HARDSHIP ADJUSTMENTS One way in which the burdensomeness of the property tax, including its regressivity, has been attacked has been through the device of special exemptions and abatements for various kinds of "hardship" cases. The homestead exemptions which became popular in the 1930's were one manifestation of this. More recently, there have been adop- tions of devices to relieve property tax burdens for older people. Almost without exception, exemptions and abatements have proven to be clumsy and inefficient methods of relieving hardship. If a partial tax exemption is offered to a whole class of property owners-such as the aged-it is likely to relieve the real hardship cases only if it is very generous, and hence very costly in foregone tax revenue. Mean- while, many property owners who are not hardship cases, benefit. If the generosity of the provision is tempered by revenue-loss considera- tions, it may be of trivial value for those really hard hit, and ad- ministratively complex as well. Moreover, tax relief for homeowners, whether aged or not, tends to ignore the frequently worse-off cases among renters. One way out of this is to offer carefully tailored credits for very burdensome property tax payments under State income tax laws. Such credits can be restricted to those whose income status makes it clear that the burdens are real ones. An example of this is the provision in Wisconsin, adopted in 1963, for income tax credits for the aged, both homeowners and tenants, who have both low incomes and pay ~ has been done on an experimental basis with extraordinarily good results in California. PAGENO="0060" 46 high proportions of their incomes in property taxes, directly or through rents.45 This kind of provision parallels the spreading use of income tax credits to offset the regressivity of State sales taxes, and could be usefully employed on a widespread basis. HOUSING TAX INCENTIVES The obvious deterrent effects of high property taxes on housing has led to the use of tax exemptions and abatements for specific kinds of new housing construction and rehabilitation, most notably in New York State. The New York programs have had some success in stirnu- lating construction of middle-income housing, with real though in- direct effects on the housing status of poorer families. Even from the standpoint of an exclusive emphasis on increasing the supply of housing, the tax abatement programs now being used have a serious drawback-they are administratively very cumbersome, and therefore slow-moving. When governments single out particular groups in the population for extraordinarily favorable treatment, they are likely to try very hard to insure that the benefits do in fact accrue to the worthy target population, rather than to unintended freeloaders with no special claim on the public purse. The attendant restrictions can greatly complicate matters. New York's most generous tax abatement scheme is a good example of this. Under this plan, designed to encourage rehabilitation of older housing occupied by lower income people, an owner can recover 75 percent of the cost of the improvements through tax reductions over 9 years. However, the improved property is subject to rent control, and the owner typically must forego rent increases he might otherwise be entitled to. As a result of this and other complications, the program is little used. The administrative difficulties could be overcome by a general tax abatement or exemption for all new housing investment, not just that which satisfies complex administrative requirements. The selective tax abatement programs are moderate in size,~~ and therefore the reduc- tion in taxes has a negligible effect in increasing property taxes on other types of housing. But an effective across-the-board exemption program for all new investment would be a different matter. It could result in a significant shift in the tax burden to older properties, in- cluding older properties occupied by relatively low-income households. Therefore, such a program is a questionable one unless it is linked to steps to reduce the reliance on the property tax in general, such as those suggested in the preceding section of this study. And if there is a substantial reduction in reliance on the property tax in general, the need for special housing exemptions will be greatly reduced. ~ Billy Dee Cook, Kenneth E. Quindry, and Harold M. Groves, "Old Aged Homestead Relief-The Wisconsin Experience," National Tax Journal, Vol. 19 (September 1966), pp. 319-324. ~ The various New York City programs as of the end of 1965 covered 75,000 units, 2.5 percent of the city's housing inventory. City of New York, Committee on Housing Statistics, Housing $tatistics Handbook (August 1966), table 1-3. PAGENO="0061" 47 DE-EMPHASIS In summary, the highest priority would seem to attach to dc-empha- sis of the property tax per se. It is a generally inferior tax instrument, although not the worst of all possible taxes. But an inferior tax be- comes a monstrous one if applied at high enough rates. There are alternatives to ever-increasing property tax rates in urban areas, alternatives which require a willingness to acept real change in that most conservative of all institutions, local government. SCOPE OF COMMISSION RESEARCH' Zoning and land use: Problems of land assembly. Regionalism in land use controls. Land value trends. New techniques in land use and development controls. Zoning in the suburbs. Zoning in the central city. Zoning case studies. Building codes, housing costs, and technology: Impact of local building codes, regulations and practices on housing costs. Structure of building codes. Building code administration. Analysis of the building industry. Urban technology. Development standards and urban design. Development standards and the development process. Labor practices and housing. Housing costs. Housing programs: Housing needs. Programs for expanding low- and moderate-income housing. Evaluation of social objectives of low-income housing programs. Evaluation of types of home ownership. Financing. Community development and renewal. Processing time and procedures for housing programs. Housing construction goals: implications. Housing codes: Goals and administration. Legal remedies for housing code violations. State of housing code enforcement. Code enforcement: Costs and effects. Housing code standards. Taxation and government finance: Demographic developments and prospects. Impact of the property tax. Federal taxes and housing. School districts: Lessons on governmental unit consolidations. Land taxation. `Listing does not imply all research performed will lend itself to publication. PAGENO="0062" 48 Housing and social problems: Housing and the large poor family. Creative neighborhoods. Racial and economic integration: Factors and problems. Statistical surveys: Canvas of 3,100 local governments on substance and administra- tion of zoning, planning, and building regulations (codes and standards). State study of land values, improvements, and assessments. Land use patterns in 130 major cities. Correlation of housing conditions and poverty areas. 0