Urban Land Policy Issues and Opportunities Volume II SWP283 World Bank Staff Working Paper No. 283 May 1978 May 1978 Prepared by: Harold B. Dunkerley Alan A. Walters John M. Courtney William A. Doebele (consultant) Donald C. Shoup (consultant) Malcolm D. Rivkin (consultant) Urban Projects Department Copyright (D 1980 The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. The views and interpretations in this document are those of the authors and should not be attributed to the World Bank, to its affiliated : organizations, or to any individual acting in their behalf. The views and interpretations in this document are those of the authors and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting in their behalf. WORLD BANK World Bank Staff Working Paper No. 283 May 1978 URBAN LAND POLICY ISSLES AND OPPORTLTNITIES - VOLUME 2 This paper brings together a general overview of urban land issues affecting developing countries and 'ive supporting papers on individual aspects. They are the products of a program to review urban land issues directed by Harold B. Dunkerley, Senior Adviser of the Urban Projects Department. There are three papers in each volume. In Volume 1, the introductory overview paper provides a general context for consid- eration of urban land issues, with particular attention to those which most directly impinge on the preparation and implementation of projects and programs in which the World Bank is involved. The first of the supporting papers deals with problems encountered in calculating the economic valuation of land on the basis of opportunity cost. The second supporting paper provides a thorough discussion of different types of urban land tenure in relation to objectives of equity and efficiency. In Volume 2, the initial paper deals with measures to influence the allocation of surplus values created in the development of urban land, including various forms of land taxation and government acquisi- tion and development of land. The other two supporting papers in this volume deal with other forms of regulation of land use, the general limitations to which they are subject, and the characteristics of individual regulatory tools. Prepared by: Copyright c 1978 The World Bank Harold B. Dunkerley 1818 H Street N.W. Alan A. Walters Washington, D.C. 20433 John M. Courtney William A. Doebele (consultant) Donald C. Shoup (consultant) Malcolm D. Rivkin (consultant) Urban Projects Department PREFACE This and a companion volume bring together a general overview of some of the most important urban land issues now affecting developing -countries and a-series of supporting Dapers on individual aspects. The introductory overview paper has been designed to provide a context, or perspective, for consideration of those urban land issues which most clearly impinge on the preparation and mplementation of projects and pro- grams in which the World Bank is involved. Trhe analysis should, however, be of interest to a much wider audience than those directly involved ir. these urban projects and programs. At issue are a variety of aspects sh.ich, as the recent TN Habitat Conference showed, are of almost universal concern. These issues include the rapiditv of rises in urban land prices, the potential and problems of capturing publicly created land values, and various developmental problems inherent in ownership and land use rights. The first of the supporting papers deals with problems encoun- tered in calculating the economic valuation of land based on the opportunity cost of using landr for one purpose rather than another. The discussion in the overview paper of the underlying relationships behind shifts in the levels of urban land prices in response to growth in incomes, population and the provision of services is here given a more rigorous mathematical treatment. The second supporting paper provides perhaps the most thorough discussion available of different types of urban land tenure in relation to objectives of equity and efficiency. A third paper deals with measures to influence the allocation of surplus values created in the development of urban land, including various forms of land taxation and government acquisition and development of land. Finally, two papers deal with other forms of regulation of land use, the general limitations to which they are subject, and the characteristics of individual regulatory tools. The program as a whole was directed by Harold B. Dunkerley, Senior Adviser of the Urban Projects Department, who also wrote the over- view paper. He was assisted by Douglas H. Keare, Chief of the Urban and Regional Economics Division, Development Economics Department, and by Suzanne Henneman, who conducted a special survey of actual experience of land problems in urban project work. Alan A. Walters, William A. Doebele, Donald C. Shoup, Malcolm D. Rivkin and John M. Courtney contributed the supporting papers. Acknowledgment is also due to many colleagues in the Bank, particularly Orville F. Grimes, Johannes F. Linn, Callisto E. Madavo, Rakesh Mohan, Maurice Mould, Anthony J. Pellegrini and Bertrand M. Renaud, who helped revtiew earlier drafts of these papers, anc whose constructive suggestions have been largely incorporated in the present texts. Warren C. Baum Vice President Central Projects Staff - iil- TABLE OF CONTENTS VOLUME . Page No. Preface ii Urban Land Policy Issues and Opportunities - An Overview, Harold B. Dunkerley, Urban ProJects Department 1 The Value of Urban Land Alan A. Walters, Urban Projects Department 65 Selected Issues in Urban Land Tenure William A. Doebele, Harvard University 99 VOLUME 2 Preface Land Taxation and Governmert Participation in Urban Land Markets, Donald C. Shoup, University of California, Los Angeles 1 Some Perspectives on Urban Land Use Regulation and Control, Malcclm D. Rivkin, Rivkin Associates, Inc. 85 Urban Land Use Regulation John M. Courtney, Urban Projects Department 127 LAND TAXATION AND GOVERNMENT PARTICIPATION IN URBAN LAND MARKETS: POLICY ALTERNATIVES IN DEVELOPING COUNTRIES Professor Donald C. Shoup University of California at Los Angeles -2- Abstract The :a-dly growing cities of developing countries exhn4 it i3 severest for= the coaflicts izvolved in attemvts to achieve both efficient land use and 'airzess in the distribution of =icome or other benefits frog land. 1= th-ls paper several forms of taxation are evaluated as means of resolving thlis conflict. It is argued that in many countries 4nc-eased taxation of land value is desirable, not only because land value can be taxed without decreasing-and perhaps even increasing--the incentives to allocace land efficiantly, but also because 4t is a way to finance the costs of local sublic services that may benefit landowners. Because land value .5 an i ortant ftaccion of total national wealth in developing countries, land taxation PoCicy may also be onre of the most. effective means to i=lement general poli4cies toward redistribution of income and wealch. nc'reased oaricipaticn. of the government as an owner in the lant r=aket is another option for achieving the goals of land policy, although man-y local pubLic authortites have a very limited capacity for effective management of land on a large scale, and some curreat for-as of public intervention, such aS slum clearance projects, may do more harm than good. One promising form of gover-ment intervention now used in some countrias is land readjustment, wnereby a public authority assembles land for conversion f-om raral to urban use, installs all public services, and finances the cost of :he cperation from the increase in land value caused by the new infra- structure; the government sells enough of the serviced land to pay all its costs, and the remaining sites are returned to the original rural land owners _1z traoortion to their ilnitial land contribution. This paper argues that the land readjustment process could be more -widely appl ied in many ci ties exper4encing rapid growth, and that a larger share of the resulting betterment could be retained by government agencies responsible for carrying out tne readjustment. Other forms of government participation in the land market, as well as methods of i3roviag the private land subdivision process, are also discussed. 3 TABLE OF CONTENTS Page No. I. INTRODUCTION 6 II. OBJECTIVES OF LAND POLICY 9 Efficient Use of Land and Other Resources in Urban Areas 9 Equity in the Distribution of Benefits from Urban Land 10 Government Revenue 11 III. STATEMENT OF THE PROBLEM 12 Bases for Public Intervention in the Land Market 12 Public Sector Land Demand 12 Land Market Imperfections 13 Distribution of Betterment 15 Urban Planning and the Land Market 15 Land as Wealth 16 The Shortage of Serviced Land 18 Land Values and the Distribution of Public Service Benefits 19 IV. LAND TAX POLICIES 28 Objectives of and Constraints for Land Tax Policies 28 To Provide General Revenue 28 To Provide Revenue to Finance Expenditures on Specific Public Services 28 To Provide Incentives for Efficient Allocation of Resources 28 To Reduce Inequities in the Distribution of Land Ownership, Land Income, or Benefits of Land Use 29 Adminstrative Feasibility 30 Problems of Transition 31 Political Acceptability 31 Timing of Tax Payment 32 -4- Land Tax Systems 32 Site Value Taxation Versus General Property Taxation 33 Revenue Productivity 33 Resource Allocation Effects 33 Distribution Effects 38 Administrative Feasibility 42 Problems of Transition 43 Political Feasibility 43 Taxation of Land Value Increases 43 Revenue Productivity 44 Resource Allocation Effects 45 Distributional Effects 46 Administrative Feasibility 48 Timing of Tax Payments 50 Problems of Transition 50 V. DIRECT PUBLIC PARTICIPATION IN URBAN LAND MARKETS 52 The Objectives of Public Participation 52 To Improve the Land Use Planning of Newly Urbanized Land 53 To Provide Government Revenue to Finance P.ublic Infrastructure 53 To Allocate the Serviced Land to New Residents, Firms, and Government Use 53 To Encourage Efficient Use of Land After it is Converted 54 Land Readjustment 54 Land Banking 59 Land Banking of Sites for Future Government Use 60 Land Banking to Reduce the Rate of Land Price Inflation and Improve the Pattern of Urban Growth 63 Private Development with Public Permission 67 VI. RECOMMENDATIONS 73 Tax Administration 73 Increased Taxation of Urban Land Value 74 -5- Page No. Taxation of Betterment 74 User Charges as an Alternative to Taxes 75 Land Readjustment 76 Subdivision Regulations and Service Level Standards 77 Land Transactions in Site and Service Projects 77 BIBLIOGRAPHY 79 -6- I. INTRODUCTION 1. it may seem obvious that the urban land market cannot operate well without a high degree of public intervention. But it is also apparent that many local public authori-ties-have a very limited capacity for effec- tive intervention, and that some forms of intervention, such as slum clear- ance, now do more harm than good. When both views are held simultaneously, the challenge is both to discover what sorts of public intervention in the land market are most (and least) effective, and also to understand how the private land market can itself be better used to achieve public objectives. The objectives and rationale for government intervention in the land market are the subjects of Chapters II and III of this paper. 2. The large variety of possible public interventions in the land market can be classified into four broad categories: (1) Land Tenure Institutions, (2) Land Use Regulation, (3) Land Taxation, and (4) Direct Public Participation in the Land Market. The third and fourth of these are the subject matter of this paper, but they cannot be treated in isolation from the first two. 3. A country's land tenure institutions are sometimes taken as part of the given framework within which land policy must operate, or can themselves be considered as susceptible to change to achieve policy objec- tives. The comparative advantages and disadvantages of a variety of possible land tenure arrangements (squatting or de facto tenure, private freehold, private leasehold, public freehold, public leasehold, communal ownership) have been fully described and evaluated by William Doebele (Selected Issues in Urban Land Tenure), and the subject is not covered here except to note below where the type of tenure system bears on other types of land policy. For instance, a policy of leasing publicly owned land for private housing can be a substitute for land taxation as a revenue source, and can also be a means of regulating land use. Conversely, if there is an effective system of land taxation and regulation, there may be few advantanges to public leasing of land for development. 4. Land use regulation includes zoning, subdivision requirements, building codes, rent control, and other forms of administrative control involving government permission or prohibition of private land uses. The topic of regulation is treated in the companion papers by Malcolm Rivkin and John Courtney (Some Perspectives on Land Use Regulation and Control and Urban Land Use Regulation), and, as with land tenure, is not dealt with here except in its relation to other methods of intervention, especially taxation. For example, the process of granting permission for the use of land commonly involves gains in land value which may be a suitable tax base; conversly, where planning regulations are strict such gains associated with changes in permitted uses are not taxed, there is a stronger incentive for individual landowners to seek changes in public plans by legal or illegal means. _ 7 -- 5. The third general category of public intervention in the land market is by taxation of land or improvements to land. LT.he wide variety of possible taxes that involve explicit intervention in Lrta land market includes general property taxation, site value taxation, betterment levies, land value increment taxes, public infrasructure charges, and the vacant land tax. In- addition, other taxes may have important land market effects, most notably the almost universal exemption of imputed income of owner occupied homes from income taxation. Each of the varieties of land taxes, and the problems of introducing them, are evaluated in Chapter IV. 6. A clear distinction between taxation and other types of inter- vention is that because the revenue goal is so important, taxes are not always seen also as tools for influencing either land use or the distri- bution of the benefits from land among members of society. The chief excep- tions to this neglect of non-revenue goals of tax instruments are the tax on vacant land to stimulate development in certain zones (as in Taiwan and Chile), the use of property tax rates that vary according to whether land is used in accordance to its zoning (as in Jakarta), and taxes on speculative gains in land value (as in Korea). However, even when unintended, the effects of taxes on land use can be significant. Thus, in the design of land related taxes even more care than usual should be directed to ensuring that they either be neutral in their effects on incentives to use land, or else intentionally alter incentives in a desired way. In practice, many forms of property taxation can be shown to have effects on land use which are directly contrary to the objectives of other land policies. The chief example is the general property tax, which in one view is a tax on (and disincentive to) the production of housing. 7. In addition to concern for the revenue productivity and in- centive effects of land related taxes, another desirable feature of a tax system is that the burden of taxation be fair. Fairness of the dis- tribution of tax burdens can be measured either by the degree to which individuals are taxed in proportion to public service benefits received (a major justification for betterment levies), or the degree to which indivi- duals are taxed in proportion to their ability to pay. 8. Other desirable features of a land or property tax system are horizontal equity (e.g., equal tax burdens on properties of equal value, which for land related taxes requires the ability to assess land values with reasonable accuracy), low administrative and compliance costs, and political acceptability. 9. The fourth general category of public intervention in the land market is direct government participation in the land market, as a buyer, a seller, a landlord, or a user of land. This is the subject of Chapter V. Governments inevitably acquire land for their own direct use in providing public services, but the method of acquisition can be an important element of land policy. For instance, advance acquisition of the sites for key public facilities can represent public commitment to a particular land use plan, and thereby stimulate private investment related to the facility. -8- 10. Although the governmmt can exert considerable influence over land markets through acquisition for its own use, it can have an addlitional impact by preparing and servicing land for private use. Many cities in developing countries have more than doubled in area during the past decade. In this context it is crucial for the conversion of non-urban land to urban uses- -to .ae- haudle4 efficieatly, for.-it1is. at- that. point that the physical and social characterisities of the new urban pattern are essentially deter- mined, for better or for worse. The public role is particularly important here because public utility extensions to previously unserved land determine land use possibilities and land value. The government is thus a major I participant in the process of converting raw land to urban uses. The issue of whether and how the government should increase its role in the conversion process to include purchase and later resale or leasing of land for subse- quent private use is treated in Chapter V. Land banking and land readjustment are discussed, as well as the alternative policy of improving the process of private subdivision. II. OBJECTIVES OF LAND POLICY 11. It seems clear that no statement of land policy objectives can suit all circumstances, expecially since the economic, social, and politi- _calzcontexts vary- tr-emendously. among and-even within developing c-ountries. Overly general statements of goals usually result when everyone wants a set of objectives yet no one agrees what the specific objectives should be. Nevertheless, a discussion of possible objectives can serve to sharpen the issues (and conflicts) involved, and can help to provide a framework for decisions regarding the usefulness of specific kinds of public intervention. Also, as a method of organizing thinking about policies, the derivation of policies from objectives can be a useful procedure. At the least, it helps to uncover inconsistencies among policies, even if it does not reflect the ways in which conflicts and compromises among interest groups actually -determine policies. In any case, it is necessary to specify some set of objectives of public land policy, because what is considered to be a land problem depends on what the objectives are. 12. Three general objectives of urban land policy are proposed here: efficiency, equity, and government revenue. These three are clearly not the articulated objectives of land policy in every country, but nevertheless serve as a useful organizing framework for a discussion of specific policy instruments. Each objective is discussed below. Efficient Use of Land and Other Resources in Urban Areas 13. In a market economy, individuals' objectives with regard to the use of land are seen as many and varied. In this context, the objective of public policy is not to achieve some collectively determined land use pattern, but rather to create a framework so that an efficient pattern will result when all members of society make their own land use choices. 14. In a static sense, an efficient allocation of land is one where each parcel of land is assigned to its highest valued use, with value understood to include not only the private value in that use but also the social value of net external benefits or costs imposed by that use. In a dynamic sense, the objective is not only to encourage an efficient allo- cation of land at any one time, but also to encourage flexibility in the transition from one land use to another in response to changes in demand. Because of the durability of most improvements built on land, there will often be conflicts between the desire for full utilization of land at any one time and the desire for flexibility in changing land use over time; some land will usually be kept temporarily vacant to preserve its avail- ability for a planned future use. This competition between current and future use can become especially severe when urban growth is rapid. 15. Another objective of land policy can be to promote efficient allocation of other resources as well as of urban land; thus land use regulations are commonly used in an attempt to reduce traffic congestion, - 10 - air and water pollution, and similar resource misallocations that occur in other markets. However, a risk here is that land policies may be employed to affect resource allocation indirectly where other more direct policies would be more appropriate, or that in pursuing too many other competing objectives, the objective of efficient land use itself may be unduly sacrificed. 1! Equity in the Distribution of Benefits from Urban Land 16. Simply because land value is an important fraction of total wealth in developing countries, general policies toward redistribution of income and wealth will involve some redistribution of land rent, of land value increases, of land ownership, or of access to land (as by tolerance of squatter settlements). 17. - In addition-to using land policy to pursue the general goal of equity in the distribution of income, a more direct land policy objective is to identify and attempt to reduce any distributional inequities that are caused by the special characteristics of the urban land market itself. 2/ For instance, some or all of the benefits of many local public services can be shifted from renters to landowners by means of increased rents and land values at benefitted sites. Thus, a distributional objective of land policy can be that land value increases resulting from public service benefits are either targeted on intended beneficiary groups or are recouped by the government. The distribution of other land value changes, such as those caused by public regulatory decisions or by general comminity growth, may also seem inequitable because they are unearned, and an objective of policy may be that these too be captured for government use. _/ The difficulty of employing land use policies to improve transportation resource allocation is suggested by the fact that some planners recom- mend off-street parking space requirements in all new buildings (in order to get parked cars off congested streets) while others recommend prohibition of off-street parking in new buildings (on the expectation that this would reduce congestion by encouraging use of mass transit). Either policy has defects in comparison to a policy that directly brings transportation prices into line with marginal social cost, as is now being tried in Singapore (Watson and Holland, 1976) . This general issue is discussed by Edwin Mills (1974). 2/ The distributional intent of some government intervention may in reality be to favor politically influential individuals or groups in society. The continuing debate over whether local governments employ land use regulations, such as "exclusionary" zoning, to keep low income and minority populations from certain residential areas or even from whole jurisdictions is only one example of the difficulty of inter- preting government actions in terms of the pursuit of equity. - 11 - 18. In devising land policy to reduce distributional inequities, an important objective should be to avoid introducing significant new inequities that would be caused by the transition from the previous policy to a new one. Changes in land policy can generate significant windfall gains and losses of their own, and these should also be considered in zevaluating.the ..eq.uity .consequences- of a-prapLased. policy. Government Revenue 19. A third objective of land policy is to provide resources to public authorities responsible for providing urban services, either as general revenue for providing area-wide services or as specific revenue to finance individual projects. This objective is linked to the first two by the considerations that (a) revenue should be raised without creating incentives for inefficient private land use, (b) the revenue instrument can be designed to raise revenue from landowners in proportion to public service benefits received by them, and (c) the progressivity of the revenue instrument is a means of redistributing income from land. - 12 - III. STATEMENT OF THE PROBLEM 20. A land problem is a divergence between the desired objectives and the actual performance of the land market. Thus, it is necessary not only to set -obj ectives, but- also-to identify- those- characteristics of the urban land market which cause its performance to fall short of the objec- tives that have been set. It is a long list. However, in cataloging the sources of urban land market failure, there is a risk of creating un- realistic expectations about the improvements that will be wrought bv government intervention to correct for these failures. Some "problems" may be created by-unattainable or inconsistent objectives, and some forms of intervention may not only fail to contribute to ultimate land policy objec- tives, but may actually prevent accomplishment of the stated objectives or of other more important objectives. Thus a goal of the discussion is not only to- identify -where incre-ased public intervention can improve market performance, but also to suggest how current forms of public intervention can be improved (which in some cases simply means terminated). 21. The first section outlines the theoretical bases for public intervention in the urban land market (without, however, assuming that intervention will in practice always bring society closer to its objectives). The second and third contain a more detailed discussion of selected problems that are particularly important in developing nations. Bases for Public Intervention in the Land Market 22. The following five grounds are suggested for public intervention in the urban land market. 23. Public Sector Land Demand. The public sector is itself a major user of land in its provision of such public services as roads, schools, parks, etc. It must therefore have a policy for selecting and acquiring the sites it requires to satisfy its own land demand, which is constantly expanding as urban growth occurs. Elements of public land policy regarding acquisition of land for public use include: (a) how the government selects land for future use, and how this land is reserved from alternative private development until public use occurs; (b) how land is valued for purposes of project design; (c) how private owners are compensated for land taken for 24. A particularly difficult issue involving government land acqui- sition is whether the price paid for land should include any development value created by the expectation of the public project itself, or whether the price should reflect only the value the land would have in the absence - 13 - of the project. 1/ Another issue is that of when land should be acquired-- long before actual use when more land is available in an undeveloped state and at a lower price, or closer to the time of development, when public site requirements are better known. If acquisiton is delayed until the time when land is needed for use, the best sites may already have been developed ftr -lower valued pr-ivate uses. 25. Policies related to public acquisition of land during its conver- sion from rural to urban use are discussed in Section V below. 26. Land Market Imperfections. There are several important features of the urban land market that can impede goals of achieving efficiency and equity in the allocation of land. In many developing countries the land--registration systems are-defective. This deficiency in land registra- tion systems may simply mirror the -more-fundamental uncertainties of who has precisely which rights to what land. Such ambiguities derive from historical patterns, in which one set of vaguely defined rights is layered on another. The resulting insecurity of tenure for many families is widely suspected to be a deterrent to investment in housing and other improvements to land. Other results are that the cost of transacting land can be very high, and it can be difficult to tax land value or secure payment for improvements when ownership and use rights are not clearly identified. The important and complex issues of government policy toward property rights, land registra- tion, and tenure are analyzed by William Doebele (Selected Issues in Urban Land Tenure) and are not explicitly treated here, except to the extent that other land policies discussed below must often be adapted to circumstances where insecurity of tenure is a major problem. 27. As in any market, monopoly of land ownership can result in higher prices and a reduction in supply. Because of the spatial uniqueness of individual sites, there can be some monopoly power in land submarkets even where overall concentration of ownership is low. There is little information on the degree to which private monopoly in urban land ownership is a problem in developing countries. 28. Almost the opposite problem of monopoly is excessive fragmenta- tion of land ownership so that the high costs of obtaining agreement among many different owners may preclude the benefits that coordinated action can 1/ To some extent, the same issue could be said to arise in the compensa- tion of specialized labor or capital whose market price may rise as a result of increased public demand for these factors in the execution of public projects. Given sufficient factor mobility, such price rises are limited by the degree to which factors elsewhere in the economy are unemployed. The very nature of urban land--particularly its spatial uniqueness--severely constrains this mobility and confers on the landowner some monopoly power. However, the government's power of compuLsory purchase of land does at least introduce the possibility of purchasing land at the price it would have in the absence of the project. - 14 - bring. Where there are advantages of large-scale development, the existence of many small contiguous but separately owned sites can create a hold-out problem, with each small owner having an incentive to hold out for a price higher than he would require if the sale were not to someone planning a land assembly. The generally favorable results of land readjustment as a public policy for dealing with fragmentation of ownership at the urban fringe-are discussed- in-Chapter V. 29. A particularly difficult land market problem is that the highest valued use for any one site depends not only on the characteristics of that site, but also on the uses of many other sites. Given these externalities or interdevendencies among land uses, it can be shown that it is impossible for a decentralized competitive land market to produce a stable equilibrium set of land uses (Koopmans and Beckmann, 1957). 30. Quite aside-from problems of instability, inefficiencies and inequities result when land users ignore the external costs and benefits they impose on neighbors. The extent to which externalities among land uses are perceived to be a problem undoubtedly varies greatly with cultural traditions and income level. In developed countries, intervention to deal with externalities among land uses often takes the form of exclusive-use zoning of large areas. In developing countries, however, it may be more important to encourage propinequity rather than separation of different types of land use in order to increase employment opportunities within neighborhoods and to reduce transportation costs. The advantages of large-scale private development as an alternative to detailed government land use regulations in dealing with land use externaliries are discussed in Chapter V. 31. There are certain indivisibilities associated with choice of an overall urban form. Even if each parcel of land has by marginal adjust- ments been allocated to its highest valued usa, the aggregate result may not be a global optimum; for instance, each owner may individually choose a lower density land use adapted to an individual waste disposal system, while all acting together might have chosen higher densities and a collec- tive waste disposal system. Such land use problems are often described in terms of the "prisoner's dilemma," (Davis and Whinston, 1961) and are an important theoretical justification for public intervention in the land market. However, given the great uncertainty as to what constitutes a desired urban form, it is not clear that concentrating land use decisions in the hands of government employees will necessarily lead to an improved urban form. 32. Externalities are not only imposed from one private land use to another, but some private land uses also imoose uncomDensated costs on the public sector in general. If taxes or charges for roads, sewers, water supply, etc., do not recover the full costs of extending those services to new locations, there can be both an inefficient pattern of land develop- ment and high net costs to the community at large. This problem is not - 15 - fundamentally a land market imperfection, because if all land developers or users were charged the full cost of providing public se ces at each location, no new development would impose any net costs for The public sector. With full cost recovery practices the higher costs of public services in difficult-to-serve locations would instead be borne by those who choose to develop or live on such sites. This would also provide a built-in tendency to guide development to locations where the public service costs (and therefore charges and taxes) are lowest. However, it is often theoretically difficult to calculate marginal cost, technically difficult to meter service use, and politically difficult to charge different prices at different locations for what appear to be the same service. Thus other methods may be needed to guide land development toward locations where the public cost of providing services is low. Land policies designed to relate taxes and charges more closely to public service costs are discussed in Chapter IV. 33. Distribution of Betterment. An important question is how the Dattern of land ownership affects the distribution of the benefits of public services, and of the benefits of urban development in general. The benefits of public services can be shifted from initial low income beneficiaries (i.e., those who receive publicly supplied electricity, water, transporta- tion, etc.) to higher income landowners if increases in land rents and land values occur at the sites where the services are provided. Therefore, in order to achieve objectives regarding the distribution of the benefits of public services, it may be necessary to intervene in the urban land market to affect the pattern of land ownership, to recoup the betterment caused by public sector activities, or to control land prices directly. CThis problem is discussed in the next section.) 34. Going beyond the benefits of public services,- a more general distributional question is the extent to which the benefits of urban growth and development are capitalized into higher land values, and whether public intervention in the land market should redistribute the land value increases. Even if it is agreed that increases in land rents do fulfill the important function of rationing land use, it may also be felt that as a result landowners receive increases in income they have not earned. This contrast between the two functions of land rent--to allocate land use and to distribute the income from land--is one of the recurring motivations of land policy. Methods by which some of the increases in land value can be redis- tributed to reduce inequities without dulling the incentives for efficient land use are discussed in Chapter IV. 35. Urban Planning and the Land Market. In one view, urban planning is simply land use planning. An alternative view is that urban planning is a far more comprehensive activity, concerned with meeting a diverse array of social and economic objectives, and that intervention in the land market is but one of the means of implementing urban planning. In either view, an important task in designing a package of land policies is to ensure that _ 16 - they contribute to, rather, than conflict with, the implementation of urban planning measures. 36. A frequent problem in urban planning is that the public plan appears to prevent the most profitable private use of individual sites, in the sense that if the site were exempt from planning restrictions, the owner would receive a higher price for it. In some cases this is a valid market signal that the planned use is inappropriate, and that some other use would better satisfy demand. In other cases, holding the land to the planned lower value use is justified by the fact that permitting the higher value (as indicated by the market) use of the site would entail greater net external costs elsewhere in the system. Thus, a dilemma in policy coordi- nation is to devise land policies (such as taxes on land value increases) that (a) will reduce the private market incentives to violate a public land use plan, but which (b) will not at the same time eliminate the private market demand signals that tell when planned uses are not what people want., This dilemma is particularly strong in developing countries, where rapid rural-to-urban migration has in many cases totally overwhelmed formal land use controls such as zoning, building codes, or subdivision regulations, with new residents obtaining dwellings only in defiance of, rather than in accordance with, legal requirements. Both squatting and illegal subdi- visions are important sources of land for housing in the cities of develop- ing countries, in many cases because the public authorities have not respon- ded to demand quickly enough. Estimates of the number of residents living in "uncontrolled settlements" are rarely less than one quarter of a city's total population (IBRD, Urbanization, Sector Working Paper, June 1972, p. 82), 37. If urban planning standards for legal subdivision of land for - housing are so high as to prohibit affordable housing for the rapidly growing low income urban population, it is arguably preferable not to coordinate all land policies in support of implementing an urban plan. Quite obviously, what is needed is both urban planning that recognizes the great variability of individual desires and circumstances, and also a better coordination of land policy tools to implement the planning. The general subject of urban planning is not treated here, but the effect of land policy, especially the varieties of land taxation, on the incentives to USE land in conformance with urban land use plans is treated in Chapter IV. Land as Wealth 38. Quite aside from the previously discussed special characterisitcs of the land market which justify government intervention, the important role of land as wealth is an independent source of public concern regarding land. General policies toward the taxation and distribution of income and wealth will inevitably have some impact on the pattern of land ownership, and one method of redistributing wealth is by intervention in the land market. Also where land value is a large share of total wealth, ownership of land may be a more reliable indication of ability to pay taxes than are other measures available to tax authorities. 39. There are almost no data on the share of land value in total wealth or on the distribution of land ownership among income groups. 1/ Research on this question would be very useful, because as discussed in Chapter IV, there are persuasive theoretical reasons to believe that land value may be taxed with little or no distcrtion of incentives to allocate land efficiently. Information on the magnitude of land valu' is important to indicate possible revenue yield; information on the distribution of ownership of land is necesssary to indicate where the burden of tax would fall. 40. A particularly sensitive factor of importance in some multi-ethnic countries is the disproportionate ownership of land by some ethnic groups. Imbalances in the pattern of land ownership among ethnic groups may of it- self create social conflict, and thus be a motive for public intervention.2/ Public intervention can, however, also cause ethnic imbalances in ownership patterns. Laws permitting land ownership only to members of certain ethnic groups are found, for instance, in Fiji and Malaysia. 41. A related problem is the question of ownership by non-citizens. Each owner may individually want the option of selling his land to the highest bidder, but yet at the same time approve of laws prohibiting sale of land to foreigners. This is most likely to be an issue in nations in the early stages of development, but is of some concern in industrialized nations as well (as suggested by recent Canadian taxes on sales of land to foreigners). 1/ Allan Manvel (1968a, Table 1) estimated that in 1966 the market value of land in the United States represented approximately 41 percent of the total market value of all (rural and urban) taxable locally assessed real estate, while the comparable figure for all urban pro- perty was 31 percent. Raymond Goldsmith (1962) estimated that for the United States in 1956, land value was 29 percent of the total value of privately owned land and structures. Mason Gaffney (1970) estimated that land value in the United States is approximately half of total property value. Risden (1976) gives an estimate for Jamaica that land value was J$ 1.869 billion compared to a total capital value of all real estate of about J$ 3 billion, or about 62 percent. 2/ For instance, Kimani (1972) found that in Nairobi, Africans, who comprise over 70% of the population, owned only 9% of the area, 16% of the parcels, and 7% of the assessed value of all individually owned land. In terms of man/land ratios, there was 1 acre of African-owned land for every 800 Africans, I acre of Asian-owned land for every 32 Asians, and 1 acre of European-owned land for every 7 Europeans. In a study of land ownership in central Georgetown, Penang, Lee (1975) found that of the total land owned by the major ethnic groups in Malaysia, the Chinese own about 88 percent, the Indians about 8 percent, and the Malays about 4 percent. Even among the urban Chinese, however, the majority were renters rather than owners of land. - 18 - The Shortage of Serviced Land 42. A, or perhaps, the, major land problem in many rapidly growing cities is the shortage of land supplied with public services such as water supply, sewage disposal, electricity, and public transportation. However, the meaning of "shortage" must be carefully pianed down. Shortage in the sense of a difference between what now exists and what could be achieved with available resources is one thing. Shortage in the sense of a differ- ence between what now exists and what would be required to provide services to all occupied urban land at a required "standard" is very different. As mentioned at the outset, some problems may result from unattainable objec- tives rather than from inherent defects of the land market. 43. In many cases municipal governments supply services, or require them to be supplied by private land developers, only at a standard above that for which low income families have an effective demand. The services must therefore be supplied either with a subsidy, or only to higher income families who can pay the cost. Available subsidies are rarely sufficient to bridge the gap between the cost of the standard level of services and price that low income families are able to pay; an inevitable result is that many families will end up living on land without any public services even though they do have an effective demand for some services which could be provided without a subsidy. In this situation, the shortage stems from the imposed standard rather than from the lack of available resources. 44. It is hard to estimate the magnitude of the problems caused by mandating standards of public services above that for which there is an effective demand. However, in internal studies in the World Bank, it has been estimated that the percentage increase in capital costs from a "minimum" to a "standard" level of services is quite large for some services, though not for others: For example, there is virtually no difference for electricity and street lighting, whereas that for circulation and storm drainage exceed a factor of two. 45. Government sponsored site and service projects generally provide plot sizes and service standards well below what would be required for a legal private subdivision. Thus, in a sampling of site and service projects financed by the World Bank between 1972 and 1974, the average cost per serviced plot appears to be about $1000, with land cost accounting for between 10% and 25% of total cost. 46. -Since site and service projects can accommodate so few of the eligible population, the policy of legally prohibiting similar private land subdivisions of the same standards is clearly questionable, especially since the inevitable result seems to be illegal private subdivisions or squatting on unserviced land. 1/ This alternative policy of encouraging _/ The arguments against minimum standards for housing as well as for land subdivision are made forecefully by J.F.C. Turner (1972, 1976). In regard to standards for the size of urban housing sites, an argument could even be made for enforcing maximum rather than minimum size limits. - 19 - or at least permitting private land subdivision at standards consistent with effective demand is explored below in Chapter V. 47. Even where there is an effective demand for the standard of services that a municipality has decided to supply, political or technical problems- car irev the-govenment- from -chargi-ng- a-price- to service reci- pients which will repay the government's cost. For instance, if others have not in the past been required to pay for the installation of water mains on their land, it may be difficult to charge for installation in new areas, even though the owners would be willing to pay the full cost of the service rather than do without it. The result of failing to recover from the users the cost of the service is that the quantity of serviced land is constrained by the subsidy available for supplying services rather than by effective demand, and some sort of rationing must result. With rationing, horizontal equity is violated, and the poor almost always lose out on what is provided. 48. The general point of this discussion about the meaning of a shortage of serviced land is that, for whatever reason (too high standards of service supply, institutional constraints on recovery of costs for services that are supplied, or both) there can be an effective demand for some, perhaps low, level of services at a price that would recover the cost of supply, and yet the services are not supplied. This is the sort of "shortage" of serviced land that can be remedied with available resources. Unfortunately the pursuit of higher standards or the failure to charge for what is provided can be the real barrier to eliminating the shortage. 49. An absence of publicly supplied water or other services to land does not, of course, mean that these services do not exist. Instead, there results the substitution of higher cost private alternative means of supply, such as hand-carried rather than piped-in water. Where these privately supplied services are sold at a price above what more efficient public provision would cost, this is a clear indication that there is an effective demand for public services that could be but are not provided on a cost recovery basis. Of course, one technological impediment to charging for services may be the high cost of metering and billing for use. Where lack of an appropriate mechanism for charging for services is the chief impedi- ment to public supply on a cost recovery basis, the alternative of better- ment taxation of finance public services may be not only an important method to reduce the shortage of serviced land, but also a method of reducing inequities in the distribution of the benefits of public services. The relation of betterment to the supply of public services is discussed next, and betterment levies in Chapter IV. Land Values and the Distribution of Public Service Benefits 50. Many urban projects are undertaken with the explicit intention of targeting benefits on certain subgroups of the population who are deemed especially needy. The subgroups of intended beneficiaries are - 20 - usually categorized in a number of ways: the poor, squatters, minorities, migrants, renters, farmers, etc. When a project's target population has been identified, the criterion for evaluating the project is not limited to a simple comparison of the project's aggregate costs and benefits tc find the way of achieving the greatest possible excess of benefits over coEts. The distribution of the project benefits and costs among recipients is another criterion for evaluating the project, with some sacrifice in net benefit traded off against a more approved distribution of benefits if necessary. 51. One method of explicitly considering distributional effects in project evaluation is to weigh the value of a marginal increase in consumption to different recipients differently, with the weight given to marginal increases in consumption inversely related to the existing level of consumtion (Squire and van der Tak, 1975). To do so it is necessary to know not only the distribution weights, but also whose income will be increased by the project. For instance, in site and service projects the income group of the target population is usually specified, not only so that the project can be designed for their circumstances but also so that the project can be evaluated on the basis of who its intended beneficiaries are. 52. A difficulty in using distributional weights in project evalua- tion is that the incidence of benefits may be shifted from an initial low income beneficiary to some higher income beneficiary by increases in land rents at the project location. If so, the benefits are still there, but it would be incorrect to weigh the benefits using distributional weights appro- priate for the initial low income beneficiaries. In order to weigh benefits appropriately, the extent of benefit shifting must be explicitly considered, and the possibility that the benefits of urban projects may be shifted is strongly illustrated by recent theoretical and empirical research on the use of land value changes to measure the net benefits of projects. The use of land value changes as the measure of project benefits does not rest on the assumption that increases in land value are the project benefits, or that a project's objective is to increase land values. But, if receipt of the project's benefits (such as water or sewage) is contingent on being located at particular sites, completion for those sites will raise their rental value. If the rent of sites served by the project increases by an amount equal to the net benefit enjoyed by the site's residents, the present discounted value of the rent increases (which would be the increase in land value) is equal to the present discounted value of the net benefits of the project. 53. The use of land value increases as the measure of benefits is convenient when the benefits themselves (such as improved health or comfort) are difficult to measure in dollar terms. If the benefits could be measured directly, there would be no need to look at land value increases as well, and to include them as additional project benefits would be double counting. - 21 - 54. William Wheaton (1977, p. 142) has recently shown that the bene- fits of highway investments can be accurately measured by the "marginal change in consumer surplus under the derived demand function for travel. All of the changes in -the honsing-and-h-and market-tha-t accompany-highway investment can be completely ignored in benefit calculations..." Even if this is true in regard to the measurement of benefits, changes in housing and land values are highly relevant in regard to the distribution of benefits, for they indicate that some benefits may be shifted from highway users to landlords. 55. If there are instances where land rents do rise by an amount equal to the value of net benefits to a site-specific project, then all the localized benefits of the project are shifted to landowners. For instance, if the net benefits (over annual user charges) of a new sewer system are valued by each household at $100 per year, and if the rental price of houses in the newly served area is bid up by $100 per year as a result,-the net benefit of the sewer is thereby shifted from the residents to the owners (if they are different persons), and the values of the pro- perties would presumably rise by the capitalized value of the $100 per year rent increase. 1/ The use of the change in property value to measure the present value of the stream of future benefits logically implies that the owner of property is the ultimate beneficiary; this can have a large impact on the target efficiency of a sewer project. 2/ 56. Whether or not the changes in property values accurately mea- sure benefits is not a settled question, but there is no disagreement that public service levels do affect rents and property values, and thus that some shifting of benefits through the urban land market does take place. When this possibility of shifting is neglected in project formula- tion, later evaluation may show that the target efficiency of programs w~ith spatially localized benefits designed to help specialized groups will turn out to be unexpectedly low, if the residents are not also land- owners. The unfortunate result can be projects that sacrifice efficiency in the name of equity without actually achieving the stated distributional goals. 1/ Tenants with long fixed-rent leases would share some of the net benefit that would otherwise accrue to the owner. 2/ If the government expenditure significantly increases the quantity of serviced land, the market price of all serviced land may decline, so that all residents, and not only those in the newly serviced area, will receive a small reduction in rent. - 22 - 57. An example of the possibility of benefit shifting is illustrated by the Bustee Improvement Organization of the Calcutta Metropolitan Develop- ment Authority. Bustees in Calcutta, which house about one-third of Cal- cutta's population, are areas of small and rather impermanent huts, with an interesting teuure arrangement. Much of the land is rented by its owners to housing etr epreawrs., -alL-ed._thkar-tenants, who-buiLd the huts and in turn rent them to the occupants. Bustee clearance is of course resisted by the thika-tenants, and the government has instead attempted to upgrade the bustees by installing public services such as water and electricity supply, drainage and sanitary facilities (Grimes, 1976, p. 26). The intended beneficiaries are, presumably, the bustee dwellers, but if the improvements lead to increased demand for huts in the serviced areas and thus increased rents, some and perhaps most of the benefits will accrue to the landowners and to the thika tenants who own the huts, with the distribution of benefits depending on how quickly the housing and land rents adjust. 1/ This is not to say that a bustee improvement scheme is unwise, but it is meant to illustrate the idea that the final incidence of the benefits of localized public programs depends greatly on land ownership and tenure in the serviced area. This, of course, also has implications for the way in which tax revenues to finance such programs should be raised, a subject explored in the next chapter. 58. A topic related to the capitalization of benefits into land values is the incidence of user charges for urban services such as water and electricity, the use of which can be metered. User charges are dis- tinguished from benefit taxes or betterment levies according to (1) the degree of volunteerism or compulsion in the citizen's decision to consume and pay for the service, and (2) the degree to which the benefits of the service are concentrated on specific users or are available to everyone if produced at all. User charges are generally specified charges per unit of consumption when the public service can be provided in purchasable units (such as gallons of water, kilowatts of electricity, bus-miles of travel), where the user is able to choose the quantity to be consumed, and where the benefits of the service accrue largely to the user. Benefit taxes are more appropriate where service beneficiaries are identifiable but where there is a greater difficulty in defining or metering the units of service consumed; the service can be provided to users at zero marginal charge, but with a tax for the availability of the service, whether used or not. The betterment levy is a benefit tax in this sense, although these taxes are sometimes advocated where user charges could be employed instead. In some cases both finance measures are used for the same service: a tax or connection charge on those to whom the service is extended, plus a charge per unit of service consumed. An example would be a special assessment 1/ The elasticity of supply of huts on the serviced land will limit the size of the rent increase. However, the supply is presumably quite inelastic in an already densely built area that is being up- graded. If there were rent control, this would limit the shifting of benefits from bustee dwellers. Rent control, however, brings other problems of its own. - 23 - on all properties in an area served by water mains, with a metered charge for the amount of water actually used. 59. The extension of metered urban services, such as piped in water, to previously unserviced areas may provide a large consumer surplus to residents who value the services (in an all-or-nothing consumption choice) more highly than the total user charge they pay for the service, even if the service is totally self-financed by user charges. If there are properties identical in all respects except that one has water services available and the other does not, residents would presumably bid up the rent of the site with water toward the value of the consumer surplus derived from the water supply. This suggests that the benefit of low user charges will in part be shifted to land owners through rent increases. The final distributional impact of low service charges may thus be quite different from the intial impact unless most residents are owner occupiers. 60. One would not expect all the consumer surplus from the consump- tion of every public service to be capitalized into land values. Certain public services may enhance some residents' willingness to pay for their sites by more than they enhance the market value of the sites. The value of a public service to those residents who have a relatively strong preference for the services is thus not fully capitalized into the value of their land, and land value changes would underestimate the benefits of supplying the service to them. Conversely, some residents who place a low value on the newly provided service may find that their rents increase by more than does their willingness to pay, and their consumer surplus can be reduced by the service provision. 61. The situation can be illustrated by considering two consumers, A and B, who have different demand curves for piped-in-water, which is newly supplied at price P . A consumes QA water and receives a consumer surplus CHE. B consumes and receives a consumer surplus CFG. Both have identical sites and both pay the same market-clearing per square G C ~~ H F \ A DB QA QB Qwater - 24 - meter of land. If the equilibrium land rent is bid up by an amount that just absorbs A's consumer surplus, B would continue to enjoy a consumer surDlus, FHGE (triangle CFG minus triangle CHE), which is not reflected in increased land rent. 1/ 6Z2. f_ L ilbriuz hangt inpxlne. depends on the pattern of indi- vidual demands for water. If only a few persons have a large consumer surplus while most others have a small consumer surplus, the price of land would presumably change very little, because the demand by the marginal bidders for land would change very little. If, however, all residents receive a substantial consumer surplus, the price of land may rise considerably.' The two situations in the land market are pictured below, with the quantity of land newly provided with water assumed to be fixed as SL , and with DL and D'L representing the demand for land before and after the provision of piped water. Figure 1 represents the situation where the-preferences are such that the consumer surplus from water con- sumption is significant for all residents. Figure 2 represents the situa- tion where land values are not bid up because not all residents receive a consumer surplus from the opportunity to consume the new service. 1/ This argument would hold only if the size and number of sites in the newly serviced area remain constant after introduction of the water supply. However, if piped water is available to each resident regardless of the size of the site he occupies, then presumably there would be a reduction in land consumption per capita, as new residents bid for land in order to enjoy the consumer surplus associated with water supply. - 25' - \~~~~~~~~~~ I land latid~~~~Si Piand \ | .. Qland Figure 1 'N i-and | land land 1 1,. Dland Figure 2 - 26 - 63. For instance, the situation in Figure 2 could arise if the price per gallon of the piped water is set so high that some residents choose not to connect to the mains, and thus receive no consumer's surplus, or if the fare on a new public transport service were sufficiently high that some residents could not afford to ride it. This suggests that if net benefits, _ --i; tt~F2aa- conaumet .surp.lus,accrua-to some- but -eot all consumers of a new public service, this net benefit will not be reflected in increased land values. Thus, changes in land values from before to after a service is provided can underestimate the net benefits of public services and in this case landowners do not receive all the net benefit. 64. There is a further conceptual and empirical problem with the use of land value changes to measure the net benefit of providing a service. The benefits of future public services, to the extent that they are fore- seen, will be discounted into the present value of land before the services are actually-provided. Therefore, property values will rise even before the service is provided, and a simple comparison of land values just before and just after the service is provided will underestimate the full rise in value. 65. To avoid the problem of anticipated benefits in measuring before- and-after changes in land value, an alternative measure is the difference in land values between sites with and without a public service. However, by a similar argument relating to consumer surplus, the difference in land values between two areas identical in all respects except that one is provided with a public service and the other is not will generally overestimate the net benefit of extending the service into the second area. Before the service is extended into the second area, those who receive the greatest consumer surplus from the service would bid most for land in the served area. A land value differential between the two areas would be established at the level representing the present discounted value of the consumer surplus of those who in equilibrium are indifferent between ldcating in the two areas. All those whose consumer surplus is less than the land value differential will locate in the unserved area. Therefore, the observed land value difference between the served area and unserved areas would overestimate the benefit of extending the service to the previously unserved area, precisely because all those who initially located in the unserved area did so because they valued the availability of the service at less than the associated land value difference. 66. The finance of services by user charges would tend to reduce the capitalization of project benefits into land values. For instance, if water is supplied to all houses without charge, the consumer surplus which may be capitalized into land values is larger than when the use of water is metered and charged for. 1/ Without user charges, everyone resident in the 1/ A rate system sometimes recommended consists of a free initial allow- ance to provide water necessary for minimum sanitary needs, with a charge for all consumption above this amount. The value to residents of this initial block of water could, however, also be bid away in rent increases. - 27 - served area pays for the water surply thrcugh land rents, while w-th user charges the heaviest users are directlv charged and the land rent would rise by a smaller amount, thus imposing a smallet burden on those who use little water. 1/ The incidence of user cbharges may be very simiJar to the incidence of betterment levies waien tal !apitaJfiztim of c nsumEr surpluses in land values is consicered. Given the well noin.D e.ficiency advantages of user charges, tte added attraction of .inding thb-:m also favored on both the benefit and oa the ab.i;y- -ray rrincipt s of equijy is a strong argument for user charges. 1/ the analogy can be made to the distinction between apartment buildings in whilch all utilities are paid by the landlord and those in which all utilities are paid by the tenant. Tenants would presumably bid the rent up higher in the buildings with "free" utilities, though some tenants would prefer to pay a lower rent and pay for their own utili- ties. - 28 - IV. LAND TAX POLICIES After all the proper subjects of taxation have been exhausted, if the exigencies of the state still continue- to. requirap new taxes, they must be imposed on improper ones. -The Wealth of Nations 67. Many land problems have been perceived either as ones caused by improper tax policy, or as ones that could be cured by a proper tax policy. For instance, apparent underuse of urban land held for speculation has been attributed to undertaxation of capital gains (whether by low tax rates, undue postponement of tax payments, or by outright evasion); the large numbers of people living on unserviced land has been attributed to an inability or unwillingness to finance such services by taxation of the betterment which they provide; and it is felt that inequalities in the distribution of income are increased by the unearned and untaxed increases in land value that accompany growth and change. There are clearly a numbier of problems for which some form of land taxation can be advocated as a solution; since the efficacy of taxes on land must be considered in light of the specific ob,ectives they are intended to achieve, it is important to specify just what is wanted of a land tax policy, and what the constraints are. ObJectives of and Constraints for Land Tax Policies 68. From the discussion of Chapter I, possible objectives of tax policy in the land market are eight in number. (1) To Provide General Revenue, especially for local governments. Performance can be gauged by the land tax revenue per capita, or by land tax revenue as a percentage of total tax revenue or of GNP. The income elasticity of the land tax revenue is a relevant indication of future revenue productivity. (2) To Provide Revenue to Finance Expenditure on Specific Public Services, especially where the services raise the market value of benefitted sites. Performance can be measured by the same revenue indices used to measure general revenue performance, or by the share of specific project costs which is financed by taxes or other charges on benefitted land. (3) To Provide Incentives for Efficient Allocation of Resources in the urban land market. Taxes should either be neutral in their resource allocation effects - 29 - in otherwise well functioning markets, or provide corrective incentives in imperfect markets. An important goal here is that land taxes should rein- force urban planning measures, for example by levying higher taxes on land not used in conformance with zoning regulations, or by levying taxes on land value gains associated with changes in public land use plans (and thus reducing the private rewards for and private resources devoted to getting plans changed). 69. There are no obvious measures of how well a tax succeeds in promoting (or preventing) efficient land use. Accurate or even approximate estimates of the welfare gains or losses associated with various forms of land taxes in developing countries are probably out of the question. Perhaps the best that can be done is simply to point out how the tax affects -private- incentives to allocate land to its highest valued use (the usual criticism is, of course, that most existing forms of property taxation discourage rather than encourage efficient use of land, especially by discouraging investment in improvements). 70. Because of the difficulty of measuring the overall efficiency of land market operation, the resource allocation objectives of taxation are usually stated in terms of intermediate objectives, such as to lower land prices, to promote early development of vacant land, or to curb land speculation. A problem with all these intermediate objectives is that achieving them may not in all circumstances produce an efficient alloca- tion of land. For instance, it is not axiomatic that a reduction in land prices, if any policy can achieve this, will mean that resources will be more efficiently allocated; the most obvious impact on resource allocation is that everyone will want to use more land. Similarly, it is not clear that early development is always a more efficient use of land than later development; not all land can be developed at once, and some must be re- served for future use. In practice, some of these intermediate objectives may be related more to distributional concerns than to their contribution to an efficient use of land. While the questions of efficiency in allocation and equity in distribution are never independent, it is useful, at least for purposes of analysis to keep them separate. (4) To Reduce Inequities in the Distribution of Land Ownership, Land Income, or Benefits of Land Use. This can refer either to inequities in the existing distri- bution of land and land income, or more specifically to inequities in the distribution of changes in land values, especially where land value changes are directly caused by public expenditure or other public actions. 71. To estimate the distributional impact of any tax, one must first estimate its incidence; for most taxes it is probable that at least part of the burden will be shifted to someone other than the individual legally obligated to pay the tax. Even with the long existing general property - 30 - tax there are wide differences of opinion regarding its incidence, and whether it is progressive or regressive. 1/ In addition, the incidence of the same tax can be different in different countries because of great variations in the effectiveness and uniformity with which the tax is administered. For example, the frequency and accuracy of assessment, the- pattern .of d.e I. and de- La exempeios- -ttm- enforcement of col ections, and other factors can greatly affect the incidence of "the" property tax in developing countries. 2/ 72. Insofar as the incidence of taxes can be determined at all, a frequently discussed distributional objective is to tax some of the increases in land value caused by government activities. A similar objective is to tax increases in value associated with re-use (development) of land, but to let increases in the existing-use value accrue to owners. Encompassing both of these objectives is the taxation of all increases in land value, on the grounds that-increases in land value are not earned but are instead caused by the actions of others. However, even an apparently straightforward tax on windfall gains may be shifted, and if so it may not achieve the distributional objectives of recapturing the windfalls.3/ 73. Quite aside from using taxes to capture changes in land value, an objective may be to redistribute existing land or land income. For instance, a progressive tax rate on land value can both provide an incentive to deconcentrate land holdings and also (if borne by the landowner) serve to redistribute the income from land that remains in concentrated holdings. (5) Administrative Fea sibility. Theoretically appealing tax proposals must survive through legislation, admini- stration by government officials, and compliance by taxpayers. Many slippages can occur between the orderly theoretical requirements of the system and its performance in practice. Perhaps the most important question here is whether the tax base can be clearly 1/ For instance, McClure (1976) has surveyed the fluidity of his own views on the incidence of the property tax. In a 1968 study he estimated that in Colombia two-thirds of the property tax was borne in proportion to non-food expenditures, and one-third was borne by upper income owners of capital; in another study done in 1972 he assumed that the tax was borne entirely by owners of the taxed property. 2/ Bahl (1975), Lent (1974), and Lion (1976b, 1976c) survey the widely varying property tax practices in developing countries. 3/ For instance, if a tax on realized land value gains reduces the sale of land from farmers to land developers at the urban fringe, this may raise the price of land for housing; if so, part of the tax could be shifted from owners of raw land to users of urban land. - 31 - defined and then accurately assessed, either by govern- ment officials or by self-assessment. 1/ The collec- tibility of the tax is also an issue, because it varies widely among different kinds of taxes. 74. Another question is whether the successful operation of the tax system requires a cadre of incorruptible and technically competent public officials; one of the chief constraints in designing a tax system may be the average level of experience, competence, or motivation of the avail- able government tax officials, although this can certainly change over time. Important measures of the administrative feasibility of a tax system are the cost of operating the system in relation to the tax revenue collected, and the percentage of tax obligations that are collected. (6) Problems of Transition. This refers to short-run problems involved in a switch from the current tax system to a system that is considered better in the long run. An especially important consideration is that if future tax obligations are discounted into the present value of land, current landowners bear all of the burden of tax increases or receive the windfall gains caused by tax reductions. Another transition problem is that when land taxes are changed, other land policies, such as zoning, may have to be altered to make them consistent with the new tax. (7) Political Acceptability. Unless there is widespread political acceptance of the land tax system, landowners may expect that a change of political party in power will lead to repeal of the tax. If a new tax policy (or any other policy) is expected to fail and be repealed, owners can often delay undertaking the taxed events (such as land sale or development) in hope I/ There is no fixed standard for accuracy in assessment. The usual measure of assessment accuracy is the coefficient of dispersion of assessment ratios. This measures the average deviation of the assessed value from market value, after the effect of general underassessment of all property has been accounted for. In the United States in 1971, fewer than half of taxing jurisdictions achieved less than a 20% average deviation of assessed value from market value for nonfarm homes (Maxwell and Aronson, 1977, p. 151). Holland and Oldman (1972, p. 18) found that for a sample of properties in Jakarta the deviation of assessed value from market value exceeded 50% percent for more than two-fifths of the properties, even after the effect of underassessment of all properties had been accounted for. - 3 of the repeal. This will lead to low revenue collec- tions and depressed market activity while the tax is in effect, and thus increase the demand for repeal. Widespread political acceptability of a tax can create an atmosphere in which landowners will be convinced that the tax will persist. (8) TiminR of Tax Payment. The operation of the tax system should avoid large increases in tax payments except when gains are realized or when public service benefits are received. Land Tax Systems 75. There are many variations of land tax systems that are con- ceivable, and most seem to have been tried somewhere. For purposes of comparison, the different tax systems can be classified according to four main features: the definition of the tax base, the method of assessment, the structure of tax rates, and ease of collection. In regard to urban land, important tax bases are (1) land value alone (site value taxation); (2) total value of land and buildings (geaeral property taxation); (3) changes in land value or in total property values, either when land is transacted or periodically assessed on unrealized gains (land value incre- ment taxation, betterment levies, and special assessments); (4) the total value of land or total property value when it is transacted (real estate transfer taxes); and (5) land area, front footage, or some other physical rather than value measure of land (special assessments). The first three of these tax bases, along with their methods of assessment and tax rate struc- tures are discussed below and the relative advantages and disadvantages of each are evaluated with reference to the objectives and constraints pre- viously listed. 76. To evaluate a tax on any of these bases, it is first necessary to estimate (or assume) the incidence of the tax. Two approaches to inci- dence estimation are possible. The first is the differential incidence approach, in which government expenditures are assumed to be fixed; an increase in one tax must thus be matched by a decrease in some other tax, so the problem really becomes one of estimating the incidence of two taxes. This approach seems most appropriate in discussing large structural changes in the tax system, such as a change from a general property tax on both land and building value to a site value tax on land alone. 77. The second incidence approach is to assume that a tax increase on any of these land bases is used to finance a specific expenditure in- crease, and examine the results of both the tax and expenditure together. This second approach seems more appropriate for estimating the effects of taxes, such as betterment levies, which are designed to finance local public services by recouping land value increases that are caused by the public services. - 33 - 78. - - In dilcussing the .tax.bases, we begin by.comparing systems of site value and general property taxation in a differential incidence frame- work, and then proceed to discuss taxation of land value increases to finance expenditures on land improvement projects. .Si:te. >Value Taxation Yex-s, _qengrI_ Proper Taxation 79. Site value or land value taxation is an alternative to general property taxation which many economists have long advocated on theoretical grounds; there is, however, no decisive empirical evidence to demonstrate that a change to site value taxation would produce a large change in land use, land prices, housing prices, the frequency of renewal or in other land market phenomena. The direction of such changes seems logically clear in theory, but the actual magnitude of the changes that would occur is not. 80. - --Although many of the important political and economic questions -regarding land va-lue versus general propert-y taxation cannot be answered here, the two systems can at least be compared on the basis of the taxation objectives previously outlined. 81. Revenue Productivity. The relative revenue potential of a site value tax depends in part on the ratio of land value to total property value. There are few estimates either of aggregate land value or of the share of land value in total property value, and there is general agreement that existing assessed values for tax purposes are poor indicators of true values. 82. The actual revenue performance of land value taxes in developing countries that employ them has not been impressive. For instance, Jamaica, which in 1974 completed an 18-year process of changing from a general property tax to a site value tax system, collected only 5% of its total tax revenue from land taxes in 1975-76; the land tax revenues were US$11.54 per capita (Risden, 1976). In Taiwan, where there is a longstanding commit- ment to the principles of land value taxation, the urban land value tax revenues were 1.5% of total tax revenues, and rural land taxes contributed another 2.2%, compared to 3% contributed by the "house tax" assessed on the value of improvements (Lent, 1976). Land value tax payments are deductible from gross income in calculating taxable income for personal income taxation. At the subnational level, land taxes are a more important source of revenue; they contributed between 10 and 20 percent of provincial tax collections in Taiwan between 1967 and 1974 (Grimes, 1974). 83. Resource Allocation Effects. There are two basic threads in the argument that shifting the property tax base from improvement value to land value will improve resource allocation efficiency. One aspect of the argument is that taxation of buildings, or of any capital improvements to land, is a disincentive to improvements which reduces the amounts of both new construction and maintenance of property, and increases the price of housing and other real estate. If the argument that a tax on vacant land is - 34 - a disincentive to holding land idle is accepted, the logical corollary is that a tax on buildings is a similar disincentive to construction of housing and other real estate, a disincentive that can hardly be afforded in cities with rapid population growth. A property tax on land value alone, which does not vary with the individual's decisions to build, presents no such _ disincentiver to cons truc-tin. n-'-Thus!,-onea-rgument-iir-favor ^of chamging the general property tax to a site value tax is that it reduces the disincentive to investment in the production and maintenance of real estate, while at the same time it can maintain property tax revenues. 84. A second argument for changing the tax base from total property value to site value alone is that there are benefits that derive from high taxes on land, quite aside from the benefits that derive from the accompanying reduction of taxes on improvements. Since few taxes are thought of as having any advantages other than the revenue they produce, the claim that site value taxation can actually increase the efficiency of the urban land market certainly sets it apart from most other taxes. The argument is that if land is taxed according to its capacity to produce revenue (regardless of whether the owner has developed the land to its full revenue producing capacity), there is a stimulus to develop the land to its full capacity in order to pay the taxes. It should be noted that even without the tax liabilities, a wealth maximizing landowner always has the incentive to develop land to its maximum revenue producing capacity. However, in imperfect capital and land markets, the stimulus of the neces- sity to meet the cash flow requirements of a land value tax may be a strong encouragement to allocate land to more intensive uses (thus providing employment in construction). 85. Since the development capacity of land serviced with urban infra- structure is higher than that for raw land, site value taxation would put greater pressure on the development of already serviced land, relieving some of the political pressure to extend new services to raw land. This should result in some infilling of the vacant or presently underutilized areas of the city, as well as promote the early use of newly serviced land. Since the development capacity also depends on a site's zoning, taxation according to land value should also promote the use of land for the purposes for which it is zoned rather than for some less intensive purpose. 86. Despite the generally accepted theoretical arguments that site value taxation will lead to more capital intensive use of land, most of the empirical studies of the resulting urban patterns where site value taxation is used have not shown that land use patterns are markedly differ- ent from what happeas where general property taxation is used.l/ Richman (1965) found no effect in Pittsburgh where land is taxed more heavily than 1/ This may partially be explained by the fact that the land tax rate is not very high in those cities where buildings are not taxed. - 35 - buildings, and Woodruff and Ecker-Racz (1969) found no differences between suburbs in Australia that use different-property tax systems. One of the most interesting investigations was by W.A.V. Clark (1974) who examined the land use patterns in Auckland, New Zealand, where within one metropolitan area each jurisdiction has the local option of using any of three property tax bases: total value of land and buildings, unimproved land value, or annual rental value of the property. Despite the fact that all three systems were operating side by side in one metropolitan area, no significant differences between jurisdictions were found in either land use or changes in land use. 1/ This was so despite the fact that the property tax revenues from each system were comparable to that in most U.S. cities. Archer (1972) found that site value taxation stimulated redevelopment in the central business district of Sydney, but also found that site value taxation had little effect on land use in Brisbane, Australia, because of the long time lags between reassessments. 87. Grieson (1974) using a general equilibrium model, estimated that if the property tax rate is 3% and the interest rate is 6%, elimination of taxes on improvements would increase the supply of improvements by 23% and reduce their price by 23% (if the elasticity of demand for real estate were unity). Grieson also estimated that for a property tax rate of 3%, the deadweight loss is 12% of the value of real estate and 25% of property tax revenue. 2/ 88. Pollock and Shoup (1977) used a partial equilibrium framework to analyze the effect of property taxes on a sample of resort hotels in Hawaii, and estimated that elimination of the tax on structures could lead to an 1/ It is interesting that the "old view" of the incidence of the general property tax seems appropriate in this case. When only one part of a metropolitan area untaxes improvements, the elasticity of supply of capital to the untaxed area should be greater than if all jurisdictions untaxed capital. Thus, the investment response to untaxing buildings should be greater in this situation than if a whole city or nation changed to a site value tax. In addition, when only part rather than all of an area untaxes buildings, the elasticity of supply of construc- tion materials, labor, and operating inputs for buildings would be greater than if the whole area untaxed buildings. Similarly, the demand for real estate services are more elastic in one jurisdiction than in an entire metropolitan area. All those considerations suggest that if changing to site value taxation would produce a large effect, it would be detected in this situation. 2/ Even though he used a general equilibrium model, Grieson implicitly adopted the "old view" of property tax incidence in which the interest. rate is unaffected by the property tax rate. increase in investment in construction by up to 25%, depending on the assumptions regarding the elasticity of supply of capital and other inputs, and the elasticity of demand for hotel space. 89. This brief survey of empirical attempts to estimate the resource allocation impact of a change from general property taxation to site value taxation suggests that there is no clear-cut answer to the question of whether site value taxation will produce a markedly different or more efficient land use pattern. 90. Even the theoretical argument that a general property tax reduces investment in improvements has been challenged by the "new view" of property tax incidence'in which capital is assumed to be in fixed supply to the nation, so that the effect of the tax is to reduce the rate of return to suppliers of capital, but not to reduce capital formation. If so, and if the property tax rate is uniform on all property in the country, there would be no inefficiency introduced by the tax. 1/ 91. One final resource allocation effect that may be a specific obJec- tive of site value taxation is to reduce the price of land. As explored by Walters (The Value of Land), the market price of land depends on the appropriation ratio, which is the proportion of the true rental value of the land which is received by the landowners: net rent received by landowner appropriation rate - total economic rent of land When a site value tax diverts to the government some portion of the rental value, the appropriation ratio declines, and the market price of the land should also decline. The economic cost of land to a new buyer does not decline, however, because the cost to the buyer is the sum of the market price of the land and the stream of future tax payments. Thus, the reduction in market price which results from the tax is not a benefit to the new buyer; 1/ Linn (1976a) has further challenged the old view of the property tax. He points out that the assumption that the supply of urban land is perfectly inelastic can be questioned on the grounds that the quantity of urbanized land in most cities does expand over time. If the supply of urban land is price elastic, then a local tax on land value may be shifted. Whether this is true depends on how the tax is imposed. If a fringe land owner has some say in whether his land will or will not be included in the urban tax district, a land tax will decrease the willingness of rural landowners to be included in the urban dis- trict, and thus reduce the quantity and raise the price of urban land. On the other hand, if planners or tax authorities decide the extent of the tax district, the tax obligation is independent of the owner's and would not be shifted from the landowner. Thus, whether or not a land tax can be shifted depends very much on the way it is admini- stered. - 37 - the buyer is simply purchasing a lesser interest in the land. In effect, the owner is paying rent to the government for the use of his own land, as the government has appropriated a part of the ownership rights by the tax. 92. It seems clear on theoretical grounds that higher land taxes will reduce the market value of land, and there is some empirical verifi- cation for this effect. The effect of a site value tax on the rate of appreciation of land values must, however, be analyzed somewhat differently. As developed by Walters, the rate of return to land owners is the sum of the rate of return of the land in current use plus the rate of price apprecia- tion of the land. For vacant fringe land awaiting development, the rate of return in current use may be very low or even zero; if so, the rate of appreciation must cover both the supply price of capital (the interest rate) and the land tax. Thus, land prices must increase faster in the presence of an annual land value tax in order to provide the same after-tax rate of return earned on other assets. This seems intuitively reasonable if inves- tors are assumed to invest so that after tax rates of return on assets are equalized. Upon imposition of an annual tax on land value, the after-tax yield is returned to its former rate by a one-time fall in the value of land at the time the tax is imposed (if the tax is an unexpected event), with subsequently more rapid appreciation in value to provide an after-tax rate of return equal to the prior value. Thus, although a site value tax would be expected to lower the general level of land prices, it should not be expected to reduce the rate of appreciation of land prices. 93. This line of reasoning is consistent with Grimes' (1974) finding that "in none of the countries examined ... was a significant relation discovered between land taxes and the rate of increase of land prices. With a few exceptions, particularly Great Britain's on-and-off experiments with betterment levies, land values seem to have risen regardless of the land taxation policies in force." A speculative profits tax on raw land value gains would be expected to lower the value of raw land in the period prior to development, but to increase its rate of appreciation from its lower initial base. 94. One variety of land tax that is specifically designed to affect resource allocation rather than produce revenue is the vacant land tax. To reduce speculation and encourage use of land that is already provided with infrastructure, an annual tax can be placed on either vacant or under- utilized land, and at either a uniform rate or one that escalates with the length of the period over which land is held idle. An important argument in favor of a tax on vacant land is that it may be necessary to stimulate a critical mass of development to achieve the economies of agglomeration that compact, contiguous development can bring. On the other hand, it would not in general seem desirable to encourage the immediate development of all land that is serviced with infrastructure. Many sites need to be held for later expansion of facilities or for denser development not yet justified. - 38 Even in planned new commnities where governments control the investment, not all sites are developed at once. Also, in practice, the definition of what is or is not "development" of a vacant site may be an important issue. If the definition is too loose, premature and inappropriate buildings may be constructed merely to satisfy the tax requirement. If defined at a rigidly high standard, it may be an undue hardship for many low income families to meet the requirement. 95. In Taiwan a surtax of between 200% to 500% of the regular land tax is levied on private land that has been designated as a building site but which remains vacant, with vacant land including all sites on which the value of improvements is less than 10% of the land value (Lent, 1976). The tax is levied only in selected areas, which in 1974 included less than 1% of all land subject to the conventional land value tax. However, between 1968 and 1973, it has been estimated that the vacant land tax led about half of the owners subject to the tax to construct improvements on their land. 96. To conclude the discussion of the effect of land taxation on resource allocation, it should be noted that even if a tax on pure rent does not distort land use decisions, the macroeconomic effects of the tax can be important. Feldstein has recently argued that if a tax on land reduces land value, one result is that a larger amount of desired wealth must therefore be accumulated in the form of produced capital (Feldstein, 1977). Because land and capital are substitutes in investors' portfolios, a tax that reduces land value would increase the demand for physical capital as a store of wealth, and thus cause an increase in the equilibrium stock of physical capital. In countries where land value is an important means of holding-wealth, a land tax that reduces the market value of this component of wealth may be an important stimulus to accumulation of physical capital. The increase in the capital stock occurs in response to a decrease in the value of land because savers will seek to substitute addi- tional physical capital for the diminished value of land. If so, a land tax could be viewed as a way to stimulate investment and accelerate development. This argument is difficult to test empirically, but is of obvious relevance in countries where land is the traditional method of holding wealth. 97. Distributional Effects. A shift in the property tax base from general property value to land value would reduce some distributional inequities and produce others. If the land value tax is considered to be a levy on the pure economic rent of land and therefore unshiftable, the incidence of the tax change is on the owners of land at the time the tax change is "announced." 1/ In practice, no such change of tax base could 1/ There is a certain amount of circularity in the definition of what is the economic rent that is taxed. One definition of pure economic rent of land would be that part of the return to land that can be taken from the land owner without affecting the use of the land. - 39 - ever be suddenly announced. As one instance, in Jamaica it was eighteen years between enactment of legislation introducing site value taxation in 1956 and the completion of site value assessments in 1974. 98. If the burden of a change toward site value taxation is incident on land owners, the most important element in evaluating the equity con- sequences of the change is an ethical judgment about the distribution of land itself. If land ownership is highly concentrated in higher income groups, the redistribution accompanying the change in tax base would at least be popular with a large number of taxpayers. Indeed, one of the factors mentioned by Risden (1976) as a reason for introducting site value taxation in Jamaica was the very unequal distribution of landownership: in 1965, 3% of all farms were larger than 25 acres, yet they accounted for 63% of total farm acreage; 80% of all farms were smaller than 5 acres, and accounted for only 15% of total farm acreage. 1/ 99. Similarly, land ownership is highly concentrated in Hawaii, which in 1964 adopted a graded real estate tax with heavier taxes on land than on buildings. In 1960, approximately 84% of all privately owned land in Hawaii, and 70% of all privately owned land on the highly urbanized island of Oahu, was in holdings of 5,000 or more acres (Baker, 1961, p. 8). By 1962, more than two thirds of new residential development land on Oahu was available only with leasehold tenure, and seventeen landowners accounted for 99.8% of all residential leaseholds on Oahu (Vargha, 1964, pp. 10-12). Thus, if the shift in tax base does cause windfall losses to some landowners, in Hawaii the losers may be relatively few compared to the number whose taxes would be reduced (though in the short run land tenants with leases requiring the land tenant to pay all property taxes would also suffer). 100. In regard to the burden of a site value tax on low income families, it should be recognized that if the tax is capitalized into land values, only the initial owners of the land at the time the tax is imposed bear the burden of the tax, and not renters or subsequent owners. After the tax has been capitalized, it is no longer incident on any current owner; as noted earlier it is like a rental payment to the government for use of an asset co-owned by the government. Bird (1976) has pointed out, however, that 1/ This view corresponds to Woodruff's and Ecker-Racz's findings: "The fact that a number of Australian and New Zealand comm=nities have been voting to change to unimproved capital value basis for property taxa- tion has less implication for the inherent superiority of that system over others, as it generally reflects taxpayer desire to minimize tax bills. Communities vote to shift to unimproved capital value when they are on the outskirts of a developing metropolis and a majority of homeowners stand to benefit at the expense of a few. The reverse is true in older prosperous sections of metropolitan areas where the shift. would be in the opposite direction and store and factory owners would benefit at the expense of the more numerous homeowners. (Woodruff and Ecker-Racz, 1969, pp. 185-186). - 40 - current landowners can be said to bear the. burden of even a previously capitalized tax in the politically relevant sense that they would be the gainers if the tax were reduced. Insofar as the tax later appropriates some portion of subsequent unanticipated and therefore uncapitalized windfall gains in value (such as those created by public investments), which it is likely to do in the long run, subsequent owners will also bear some of the burden of the tax in that their share of the windfall gains (or losses) will be less. 101. A site value tax levied at high enough rates could also serve as a betterment capture device, although the low effective rates found in most countries indicate that such attempts are not yet important. When taxation is according to the market value of land, any increase in land value automaticaly increases the annual tax liability (usually with a lag that depends on assessment practice). This increase in taxes occurs regard- less of whether the increase in value was caused by public investment, public regulatory decisions, someone else's private investment, or general community growth, and regardless of when or whether the land is transacted by sale. Thus, site value taxes are broader in concept than either better- ment levies, which are generally used only to recoup the cost of public investments that benefit specific properties, or capital gains taxes, which are generally triggered only at the time land is transacted. Also, in contrast to betterment levies or capital gains taxes, which usually must be paid in a lump sum or by installments within a few years, site value tax increases that arise from land value increases are a permanent annual payment in proportion to the value increase. Thus, the site value tax appropriates as community tax revenue a proportion of all increases in the value of urban land; the tax increase is independent of any activity on the owner's part, and thus should not affect the allocation of land. Since the equal-yield tax rate on land value is higher than that on total property value, the site value tax captures a greater proportion of betterment than does the general property tax. 102. The distributional consequences of raising the tax rate on land must be considered along with the consequences of lowering the tax rate on improvements. Unfortunately, as two recent surveys (Bird, 1976; McClure, 1976) on the incidence of general property taxes point out, there seems to be increasing uncertainty about the incidence of the property tax, espec- ially in developing countries. Each new study seems to identify new areas of uncertainty about who pays the property tax. Not only is the theory unsettled, in most developing countries even the factual basis on which incidence estimates depend is scant. For instance, there is generally very little information on the distribution of land and capital ownership among income groups, although it is well suspected that the ownership of both is largely in the hands of the highest income group. 103. In this uncertain context, Linn (1976c) examined the differential incidence of a change from general property taxation to site value taxation in Colombia, and found that it would have either an approximately neutral - 41 - or somewhat progressive impact in the long run. The ratio of land to total property value is highest for the least valuable properties, declines toward the middle value categories, and is slightly higher for the high valued properties. Linn points out that the switch to site value taxation would be more progressive if the lower valued properties are exempted from taxation. The exemption could be set at a level that would exempt the very poorest from tax payment, so long as their wealth in land was small. This practice is followed for the general property tax in many states of the United States, and depending on its method of implementation could mitigate most significant hardship for the very poorest. 1/ In regard to rental property, since the tax is assumed to be incident on the landowner rather than the renter, exemptions for rented properties would not help the occu- pants and would have to be justified on the grounds that the landowner deserves the break. 104. Another modification of the site value tax which might be con- sidered is a progressive rate structure to crudely align land tax burdens with ability to pay. The land value tax rate in Taiwan is graduated from 1.5% to 7.0% of assessed value, and in 1974, about 43% of all land tax revenue was derived from land assessed at more than the 1.5% basic rate (Lent, 1976, p. 4). Owner-occupied land is taxed at a preferential rate of 0.7% and factory sites at 1.5%; in this way the rate structure can be employed to implement specific land use policies, such as to promote home- ownership and industry. The land tax rate in Jamaica is similarly graduated, from 1% of assessed value up to a maximum of 4.5% of the assessed value over J$5O,000 (US$44,000). Because of this progressive rate structure, parcels valued at more than J$50,000 were fewer than 1% of all parcels but yielded more than half the total revenue from the land tax (Risden, 1976, p. 18). 105. A progressive rate structure is not well suited to a general pro- perty tax because at the upper end of the rate scale a high tax rate on improvements would be a strong disincentive to investment in capital inten- sive land uses. By contrast, the high tax rates on land are no disincentive to investment in improvements, and may instead have the desirable effect of deconcentrating land holdings. 106. If a site value tax rate is to be progressive, it seems logical to consider the tax base to be total value of the land holdings, rather than the individual parcel, or else the rate structure simply encourages subdivision without ownership change. In Taiwan, the progressive tax rate is levied on the cumulated value of all land held by an owner within each prefect or municipality, rather than on the cumulated value of all the owner's land in Taiwan. This seems reasonable if the objective is to avoid monopoly power in the land market, because the relevant market is clearly the local one. 1/ An alternative to exempting low-valued properties is some form of 1"circuit-breakert that exempts only low income owners from property taxation. Most of the circuit-breaker formulas used in the United States are, however, so complex and require so much information that they are probably unsuitable for most developing nations. - 42 - 107. Administrative FeasibilitY. Since both site value and general property value tax systems are in use, both are clearly feasible. It is, however, quite difficult to compare how well each system does what it is supposed to do. In regard to relative costs, one study found that in New Zealand, which permits local governments to tax either unimproved land value or the capital value of land and improvements, the cost was $NZ0.60 per assessment for unimproved land value, and $NZ4.00 per assessment for land and improvement value I/ (Brown, 1968). This suggests that a pure site value tax will be cheaper to administer than a hybrid "graded" property tax that has a higher tax on land than on buildings, because so long as build- ings are taxed at all they must be assessed. 108. With a general property tax there is a greater opportunity to check the accuracy of assessment by reference to sale values in trans- actions. This is more difficult to do with a site value tax, because unless land is vacant, what is sold is not what is assessed. At the urban fringe, however, where land values can change very rapidly, there are more vacant land transactions to give evidence of current values. This is particularly important during rapid urban growth if the tax is to capture much of the betterment that occurs as land is urbanized. 109. In some countries the lack of defined site boundaries and ambiguity in ownership makes taxation of land value so difficult that the real estate tax base is improvements only. 2/ For instance, Ghana taxes only the repro- duction cost (less depreciation up to 25% of reproduction cost) of buildings (Lent, 1974). 110. Regardless of whether land or total property value is the tax base, in an inflationary environuent it would seem worthwhile to index assessments to a general price index or specific real estate price index between revaluations. Otherwise assessments can become badly out of date and thus make it difficult to fairly assess new properties entering the cadastre. With rapid inflation it also seems desirable to avoid "rolling" reassessments where some fraction of the assessments are brought up to date each year. This system, frequently practiced in the United States, results in periodic surges in tax payments for properties that have been reassessed. Each year the reassessed owners will tend to protest because in effect a larger share of the total tax burden is shifted to those who have just been reassessed according to market value, even though the relative value of their property compared to all other properties may not have changed. 1/ Technical problems involved in assessing site value for tax purposes are described in Vickrey (1970) and Hicks (1970). 2/ Adam Smith was perhaps overoptimistic in his belief that "The quantity and value of the land which any man possesses can never be a secret, and can always be ascertained with great exactness" (Wealth of Nations, Book 5, Chapter 2, Part 2). - 43 - Ill. Problems of Transition. The problems of transition toward heavier site taxation have been mentioned earlier. Thus, imposition of site value taxation amounts to the appropriation of some share of land's value to the government. Perhaps for this reason, some proposals for a movement toward site value taxation involve a slow and phased reduction in building taxes, with a correspondingly slow increase in land taxes, as in Hawaii. Since the increase in tax liability is deferred somewhat, the impact on current landowners is reduced. On the other hand, a phased introduction of land taxation also defers whatever benefits may accrue as a result of the reform. An increase in taxes is a risk that must be accepted by all landowners, and maintenance of the status quo rewards those who gambled that the tax system would not change. 1120 Political Feasibility. Much of the question regarding the poli- tical feasibility of shifting the basis of taxation from improvements to land value depends on the distributional aspects and transition problems previously mentioned. In New Zealand where there is local option regarding the tax base, 75% of the cities, boroughs, and towns employ site value rather than annual value or capital value of the improved property. 113. One important political problem with the use of land taxation as a tool to promote widespread ownership of land is that any progress toward the goal will increase the number of those opposed to the means employed. Taxation of Land Value Increases 114. Neither site value taxation nor general property taxation specifi- cally address the issue of capturing socially induced increases in land values. Site value taxes to perform this function in part, but a significant portion of the value increases are taxed only if the tax rate is high and if assessments are kept current. A more precise instrument for the purpose is a tax based on increases in land values, of which there are several important variants. 115. First, where the goal of the tax is to finance site-specific local public expenditures, an appropriate tax base is the increase in land values (betterment) caused by the expenditures. If assessed before the project is undertaken, the tax base is the increase in land values that is expected to occur as a result of the expenditure. In some cases this is an alternative to finance of services by user fees, especially where it is technically or politically difficult to charge in proportion to use of services. 116. A basic distinction in taxes on betterment caused by public expenditures is between charges that recoup only the project costs, and those that go beyond cost recovery to capture more (or even all) of the betterment generated by public improvements. In regard to betterment levies - 44 - that are limited to recoupment of project costs, Doebele (1975) and Linn (1976a) have recently described and analyzed in great depth this sort of tax in Bogota, and Macon and Manon (1975) have surveyed the use of betterment levies throughout Latin America. All three studies provide an excellent discussion of the important tasks of (a) determining the area benefitted by the public expenditure, (b) determining the total betterment to be recovered, and (c) division of the levy among benefitted properties. 117. Second, there can be taxes on land value increases at the time property is transferred, which is either a special capital gains tax on land or simply part of the general system of capital gains taxation. The tax is thus based on what has already happened to land values, and the revenue is generally not earmarked for any specific property-related use. This sort of tax is more commonly administered by a national than a local government, as is the case in Taiwan. Usually, either the tax rate is lower on capital gains for land that has been held for longer periods, or the tax base is adjusted to remove the effects of inflation over the holding period. 118. A third type of land value gains tax is on the gain in value associated with a change in land use, whether a tax on the development value of land at the time when development or redevelopment occurs, or at the time planning permission is given for development. For this sort of tax, the increases in the value of land in existing uses are not captured (unless by other capital gains taxes, betterment levies, or land value taxes), but the gains associated with transition to a higher use are subject to a special tax. Because of its intimate relationship to the land use planning process, this type of gains tax is probably best adapted to coun- tries where there are very strong land use regulations. However, even in Britain, which does have strong land use regulations, there has been con- siderable difficulty in administering the development gains type of tax, and the problems would probably be much greater in almost all developing coun- tries. 119. Revenue ProductivitI. Few countries appear to collect a large amount of revenue from taxes on increases in land values. Grimes concluded in his survey of social appropriation of betterment in a selected group of both advanced and developing countries that: In all countries examined ... receipts from betterment levies and land value increment taxes were low compared with receipts from property taxes and other revenue sources. In some cases, particularly Great Britain and South Korea, this disappointing performance was compounded by expectations when the taxes were introduced that the revenue obtained could not fail to be impressive. (Grimes, 1974, p. 71) 120. Most of the increment taxes examined by Grimes were national taxes designed to produce general revenue. Even in Taiwan where there is more of a tradition and commitment to taxation of land, the land value increment tax yielded only 2.7% of total tax revenue in 1975 (Lent, 1976). - 45 - 121. The revenue performance of local land value increment taxes designed to recover the cost of specific projects has been more impressive. In Bogota, "valorization" charges are used to finance public improvements of benefit to specific areas, and the charges are based on the increases in land value caused by the public improvement. In some years the revenues from these valorization charges have exceeded 50% of the total revenue collected by the general property tax (Linn, 1976a, p. 38). The revenue potential of land value increment taxes appears to be greatest when the tax is clearly linked to a specific expenditure. If it is understood that the expenditure will not occur unless it is financed by the increment tax, landowners' objections to the tax may be muted. Without the project there will be no increment tax, but neither will there be a land value increment. 122. Resource Allocation Effects. One important contribution of the betterment levy type of increment tax is that it introduces some discipline in the process of citizen demand for public expenditures. Where expendi- tures that benefit specific sites are financed from general revenues, there is considerable competition in the political arena for the valuable public decisions that confer land values gains without any compensating payment. As Walters points out (The Value of Urban Land) in such situations people are induced to spend real resources in legal (or illegal) procedures to secure public expenditures. If instead there is a tax on the land value gains conferred by public expenditures, this reduces the incentive to spend resources merely to influence the distribution of public investments, and the resources saved thereby could be put to more productive uses. 123. Another resource allocation effect of taxes on increases in value caused by site-specific public expenditures is that this may in some coun- tries be the only effective way to finance installation of public services on land being converted from rural to urban use. As discussed in Chapter III, and also by Walters (The Value of Urban Land), the "shortage" of serviced land in developing countries is due in part to an unwillingness or inability to charge a cost-recovery price for public services. For instance, if utilities are priced at a low historical average cost that is lower than the marginal cost of extending the utility network, extensions will require a subsidy; this will constrain the amount of services that can be provided, and cause the price of serviced land to rise relative to the price of unserviced land. The obvious remedy is to raise the price for extending utilities, but if this is not done, an alternative is to finance the service provision by taxation of the land value gain that arises when the services are extended. In this situation the tax effectively takes the place of a price for services, and permits a more efficient and extensive program of servicing raw land. 124. The rate structure of the increment tax may also have an effect on resource allocation. For instance, in Taiwan there is a progressive tax rate on land value increases which varies from 20% of the gain on land value increases of less than 100% up to 80% of the gain on increases over 300%. Since the gain is generally larger on land held for longer periods, - 46 - this rate structure presumably places a lower tax rate on land held for short periods (a practice usually attributed to speculators). A progressive rate on the absolute (rather than percentage) size of the gain would also have this effect, and both forms of progressive rate structure encourage "straw" transactions to make one large gain appear to be several small gains. Despite this incentive, in 1975 over 60% of the revenue from the Taiwan land value increment tax derived from transactions subject to the 801 tax rate (Lent, 1976). 125. Another possible resource allocation effect of an increment tax based on capital gains on transacted land is that at high tax rates it can produce a lock-in effect. 1/ If this leads to leasing rather than taxable sale of land, a result can be an actual decrease in owner occupancy; there is, however, almost no evidence as to how important this lock-in effect is in the land market. In Taiwan the tax on realized gains is supplemented by a tax on the unrealized increase in the assessed value for properties that are not transacted in any ten year period; this reduces the incentive to delay sale of land in order to avoid taxes on the increase in value. 126. Distributional Effects. No less than with other taxes, the inci- dence of betterment levies or other land value increment taxes is not ne- cessarily on those legally obligated to pay them. Neutze took the position that it is "virtually impossible to devise a tax that will directly reduce in substantial degree, or directly capture a significant part of, the increment in the value of land resulting from its conversion to urban use. Almost any kind of tax will mostly be passed on to the final consumer of the developed land" (Neutze, 1970). Archer (1976) cites evidence that the Sydney betterment levy of 30Z of gains due to rezoning was passed on to land users. Many land sale contracts even included an explicit provision that the buyer should pay the Sydney levy, and this quickly generated a popular belief that the levy was passed on. 127. If the tax is not passed on to users but is instead incident on land owners, the presumption is that the future tax obligation would, like the site value tax, be incident on those who own the land at the time the tax is introduced, rather than on those future owners of land at the time the tax is payable. To the extent that the tax can be anticipated, all future buyers and sellers would expect the tax and then capitalize it into 1/ Very recently there has been an active debate on this issue in Korea. The Minister of Finance favors high taxes on land value gains in order to divert revenue from the real estate sector to other sectors of the economy. The Minister of Construction has proposed lowering the taxes on transaction of land in readjustment projects (discussed in the next chapter) in order to facilitate the flow of resources within the real estate sector. - 47 - the transaction price of the land. Land prices would be lower, but the subsequent rate of return would be unaffected. 1/ 128. Although the expected future betterment levies will be discounted into the present value of land when the tax is introduced, the exact amount of future betterment and therefore future levies is never known with cer- tainty for every site. Especially in developing countries where there is such a shortage of serviced sites, there is a large number of possible locations for future public infrastructure investments. Therefore even if the average rate of return on land holding were no greater than on other assets, some owners can receive large land value increases from the public investments while others receive little. In this case, a betterment levy would tend to capture the large imperfectly anticipated gains at the time they occur, and thus reduce the variation in after-tax gains in land income among owners. 129. Although landowners may not oppose imposition of an increment tax if the revenue is used to provide the services that cause the better- ment, there is an important equity aspect of betterment taxation in low income areas which should not be neglected. Some lower income owner occupants or renters may have a very low willingness to pay for the bene- fit of public investment, such as paved roads. However, these public investments may make the neighborhood more attractive to higher income groups, who will bid up rents and land values in the area. Thus, the original owner occupants may find that the new public investment may cause their property values to rise by more than the betterment levies, but they will prefer to sell their properties to new higher income occupants rather than to remain and pay the tax out of their current income. In one sense, they are better off because their net wealth has increased (by an amount equal to the difference between the betterment and the tax); at the same time they may find that unless they move from their now more valuable property and take their capital gain, their unchanged cash income would 1/ Linn found that "one of the reasons for the relatively slow growth rate in land values experienced in Bogota is the fact that a signi- ficant proportion of the betterment value resulting from access to publicly provided services such as roads and streets, and public utili- ties, is taxed and priced away through the application of the valoriza- tion system and relatively high connection fees for public utilities. By implication compared with the typical LDC cities, only relatively small windfall gains from publicly induced betterment are capitalized into land values, thus resulting in a re-latively slow real property value growth." (Linn, 1976b, pp. 45-46). Because valorization taxes are levied at the time infrastructure is installed, capitalization of these taxes into land Xvalues would reduce the rate of increase of land values, when infrastructure is installed, but would not reduce the subsequent rate of appreciation. - 48 - be insufficient to pay the betterment tax. This can be said to be an efficient outcome, because the land will be allocated to those who value it most, given that the public facility is going to be built. However, if costs of moving are high, some of the initial residents might consider their own welfare reduced by the process, and might prefer a situation without the public facility and without its accompanying betterment and betterment tax. 130. Finance of public works by betterment levy is equivalent to requiring residents of an area to receive a public service and pay for it (through the levy), or else move out, realizing whatever net capital gain remains after payment of the levy. 1/ Depending on the size of the capital gain, and the amount of consumer surplus received at the old site, their welfare may be either increased or reduced. When this analogy is extended to proposals for recapture of all the betterment provided by public facilities, the residents are in the situation of having to receive the new service and pay the charge, or else move out without any capital gain (which is fully taxed away by the betterment levy). All those who value the new service at less than the charge suffer a net loss whether they move or not (unless they were initially indifferent between moving and staying, which seems unlikely). Insofar as the relatively poorer residents are those with less willingness to pay for new public services, it may be that attempts to capture 100Z of betterment may reduce the welfare of low income owner occupants (if there are any), and-some mitigation of the charges in terms of ability to pay should be incorporated. 131. In regard to rental rather than owner occupied properties, the circumstances are somewhat different. If a new public service causes rents to be bid up in the benefitted area, some former renters may find that they prefer to move rather than pay the rent increases, in which c-ase they suffer a welfare loss equal to the costs of relocation. Even some who stay may value the new service less than the accompanying rent increase and thus would be made worse off. But if the tax is incident on the land- owner, there is no reason not to levy betterment tax on the increased values of rental properties; for income property, the increase in rents provides the means to pay the betterment levy, and the landowner's welfare is not reduced if the betterment levy is less than or even equal to the increase in rental value of the property. 132. Administrative Feasibility. Betterment levies and other increment taxes are clearly feasible because in some cities they appear to be already quite well administered. The technical sophistication of the valorization system in Bogota is very impressive. One problem is that the type of levy 1/ Information on the pattern of subsequent land transactions in areas where public projects are financed by betterment levies (as in Bogota) would be necessary to establish whether this form of project finance induces poorer residents to move out. - 49 - that seems to promise most in terms of financing infrastructure--the levy on land value gains that occur at the time of public service provision but which are not realized in a market transaction --is the most difficult to assess. The problems in assessing land value for site value tax purposes have been noted, but it is even more of a problem to assess land value changes. For example, suppose a new road increases the market value of a site from $100 to $150. A 10Z error in the assessment of both the "before" and the "after" value could produce an assessment of the change in value anywhere from $25 (- $135-$110) to $75 (- $165-$90), or up to a 50% error in assessment of the true change of $50. Thus, accuracy of assessment of changes in land values can be considerably less than the accuracy of assess- ing land values themselves. 1/ 133. An alternative method of assessing betterment caused by a specific project is to assume that all benefitted properties increase by the same percentage amount. The problem remains, however, to delineate the area within which properties are benefitted, and at the boundaries the un- reasonableness of such an assumption will often be apparent. 134. The difficulty of assessing changes in land value caused by public projects is emphasized by the theoretical and empirical difficulties that have been encountered in research on the relationship between public services and land values. Bahl, Coelen, & Warford (1973) found that it was difficult to disentangle betterment caused by public expenditures from betterment that is caused by other simultaneous factors, such as increase in population and development of other sites. By comparing project areas with control areas they estimated that the increases in property value associated with sewage installation ranged between 20% and 48% in Nairobi, while there was a 30% drop in property values associated with a sewage project in Kuala Lumpur; the latter result they attributed to other influences not adequately accoun- ted for in the comparison of the control area to the project area. 135. The difficulty encountered in assessing betterment caused by public investments suggests that it would not be feasible to collect all of the betterment caused by public expenditures; rather, the inaccuracy of the assessment process may require considerably less than full re- capture or else many owners would end up paying taxes greater than the betterment received. 1/ In regard to the cost of a system of taxes on unrealized gains, the administration costs for the Sydney betterment levy were 12% of reven- ues for 1969-73 (Archer, 1976). The land value increment tax in Taiwan, which is based on both realized and unrealized gains, had collection costs of between 1.8 and 4.0 percent of revenue in the years 1967-1971 (Grimes, 174). - 50 - 136. There is considerable variation in the governmental level at which taxes on land value increases are administered and to what level of government the revenues accrue. For betterment taxes that are intended to recoup land value increases caused by specific projects, the appropriate level of government to administer the tax would seem to be the one under- taking the expenditure. In some cases (for example, the Urban Development Institute in Bogota), the tax is administered by the local public authority that undertakes the value-increasing projects. In general, local govern- ments are more experienced in assessing and collecting property taxes, but national taxing authorities are more accustomed to administering taxes on transactions involving realized capital gains, whether on land or other assets. Thus, most of the taxes on realized land value increments are probably best administered at the national level, as in Taiwan. 137. Timin2 of Tax Pavment. With betterment levies to finance site- specific public expenditures, the tax payment is generally due at the time that exDenditures are made. In Bogota there is a procedure for stretchirng out payments for low income owners. Macon and Manon (1975) point out that if the government permits owners to pay betterment levies over an extended period, the effects of inflation can greatly reduce the real value of the payments over time, and they suggest that the payments be indexed to the general price level. 138. Macon and Manon concluded from their survey of betterment levies in Latin America that it is important to require payment of the levy before the project is completed. They attribute the failure of some betterment levies to a failure to consider the tax as a price that owners must pay for the project; once the project has been completed, there is much greater resistance to pay taxes on subsequent land value increases. 139. Problems of Transition. One of the chief problems in introducing a system of taxes on land value increases is to create the impression that the tax will persist. Otherwise, owners will have an incentive to delay undertaking the taxable action in expectatioan of avoiding the tax in the future. This impression of permanence requires that the existence of the tax not be an issue in every election, and thus all political parties should be in substantial agreement concerning the wisdom of the tax. 140. Given the difficulty that most countries have in administering general property tax systems that are conceptually much simpler than a system of betterment levies, the prospects of early or easy introduction of taxes on land value increases are not bright. In Colombia, the effec- tiveness of the betterment levy system is a result of several decades of experience. 1i/ No country could initiate such a system from scratch without / The political acceptability of the Colombian system of betterment levies can also be attributed to the fact that it is confined to recoupment of project costs, in full realization that in so doing, significant amounts of untaxed betterment will accrue to landowners. - 51 - a considerable investment in training of personnel. Because of this, there is much to be said for starting with a very simplified system and also for starting with projects where the land value gains are large in relation to project costs, and where not all project costs are recovered by betterment levies. - 52 - V. DIRECT PUBLIC PARTICIPATION IN URBAN LAND MARKETS The Obiectives of Public Participation 141. The government is a major participant in the process of converting raw land to urban uses because public infrastructure extensions play a large role in determining land use possibilities and land values. Given the rapid growth of many cities, it is crucial for the conversion of non-urban land to urban uses to be handled efficiently, for it is at this point that the physical and social characterisitcs of the new urban pattern are essen- tially determined, for better or for worse. To manage this conversion process some land policy objectives may best be served by government owner- ship--either temporary or permanent--of land. To orient the discussion of public land ownership it is important to settle on the institutional or ideological framework from which land is viewed. 142. In one view the private market is, and ought to be the dominant mechanism in allocating land use and the income from land, and the role of public land policy is to tamper with the private market to increase its efficiency and achieve certain other public objectives, such as a fairer distribution of income or increased government revenue. Stressing imperfec- tions of the private market as the rationale for public intervention implies that policy makers have an image of what a "perfect" private land market would be, and that one goal of intervention is to approximate its outcome. 143. A second view is that urban planning and public action are (or should be) the dominant means of allocating land among uses, and land income among members of society. Strong public control of land is seen as indispensable in achieving a socially desirable urban environment and an equitable distribution of income. The role of the private market is to operate in the interstices where planning and public control leave some room for choice. The role of public policy is to set appropriate objectives and to see that public managers acting on behalf of the commnity pursue those objectives. 144. It follows from the first view that the important issues regarding public intervention are (a) what objectives are better served by public intervention than by private action, and (b) what methods of public inter- vention are best suited to meet these objectives. Thus, each type of public intervention requires justification that it would improve the operation of the private market or achieve an end that the private market cannot. 145. Similarly, it follows from the second view that the issues are centered around the search for (a) the best methods of planning the urban environment and (b) the most effective means of implementing public plans. In this framework, private market activity can be justified for some pur- poses, but an important problem is to ensure that private market activity does not thwart public planning and the attainment of public objectives. - 53 - 146. Thus, one important issue related to government land ownership is the perspective from which other issues are defined, and this probably depends more on one's political philosophy than on economic analysis. As stated at the outset, the focus of this paper has been public policies to remedy the failures of the private land market rather than on how to replace it with some other methods of allocating land. From that standpoint and from the general objectives outlined in Chapter I, the following speci- fic objectives of public financial intervention in the conversion process seem important: (1) To Improve the Land Use Planning of Newly Urbanized Land This includes not only the physical layout of the land, but also the relationship of the new urban land to the existing urban area. An important aspect of this is to secure the benefits of site planning before land is occupied at urban densities, rather than to let it be settled and later upgraded at higher cost. (2) To Provide Government Revenue to Finance Public Infrastructure Because provision of infrastructure can cause increases in land value, a goal of public dealing in land can be to generate the revenue to finance the infrastructure by capturing appreciation in the value of publicly owned land. (3) To Allocate the Serviced Land to New Residents, Firms, and Government Uses In allocating publicly owned land, the goals of horizontal as well as vertical equity should be pursued. A frequent observation in government land distribution schemes is that both equity criteria are violated. Because government land development projects generally serve only a small portion of the eligible population, equally deserving families are often treated quite unequally; the few families in the project can receive a substantial subsidy while many similar families get nothing. 1/ Also families that do receive the subsidies are not necessarily those most in need. 147. Measures of the extent to which the government intervention achieves its equity objectives are the horizontal target efficiency (the degree to which all eligible members of the target group are aided) and the vertical target efficiency (the degree to which program benefits are actually received by members of the target group, rather than by higher income groups). 1/ This horizontal inequity can occur even if all families included in the project are treated equally. - 54 - (4) To Encourage Efficient Use of Land After It is Converted Public ownership can also serve as a method of enforcing public land use control, with land passing through public cwnership during the conversion to urban uses, or perhaps remaining in public ownership even after conversion. In serving this goal, public ownership is an altermative to land use regulations. 148. The proliferation of illegal subdivisions and squatments on the periphery of many cities in developing countries suggests that the conversion process has not been handled well, though in some cases, given the very low income of the residents, it may not be possible for public financial intervention in the land market to significantly improve condi- tions in these "unregulated" settlements without a large subsidy. There are, however, good reasons to believe the conversion process can be signi- ficantly improved, and several proposals--land readjustment, land banking, and private subdivision with public permission -- are discussed below. The land readjustment model appears particularly promising, because of its demonstrated successes in practice in Korea and other countries. Land Readlustment 149. Despite the fact that urbanization usually results in greatly iacreased land values, it is paradoxically very difficult to finance the public capital investments that should accompany or precede conversion of land to urban use. One existing model that has demonstrated great success in terms of both planning land uses and financing public investments is the process called land readjustment, which is extensively practiced in Korea, Taiwan, Australia, and Japan. I/ 150. In brief, land readjustment is a process whereby a public author- ity assembles numerous small parcels of raw land without paying monetary compensation to the owners, services and subdivides the land for urban use, returns most of the resulting building sites to the original owners in proportion to the value of their land contributions, and sells the remaining sites to recover all public costs. Land readjustment is a temporary form of public ownership to achieve unified control over large areas and a means of financing public service installation during the crucial land development stage of urban growth. 151. In Korea, city governments extensively participate in and direct the conversion of raw land into serviced building sites through land re- adjustment. There are many complex practical details, which are fully described by Doebele (1976a), but it is conceptually a simple process. When conversion of rural land to urban use is iminent, the area is declared a land readjustment project, either upon initiation by the government or by 1/ The land readjustment process in Korea is described in Doebele (1976a, 1976b). Archer describes a similar process used in Australia, where it is called land pooling. - 55 - petition from a qualified percentage (in Korea, 80Z) of the landowners in the readjustment area. The city government or readjustment authority then prepares a site plan for the entire area, replotting the land to accommodate both private building sites and public uses such as streets, schools and markets. The city also installs all the public infrastructure, such as paved roads, sewers, water supply and electricity. The resulting market value of the newly created building sites is then estimated, and the govern- ment retains a sufficient number of building sites for subsequent auction to pay for all the public costs incurred in the planning and installation of infrastructure. The remaining building sites are then returned to the original owners in proportion to the value of each's initial contribution of land to the total project area, and if possible the building sites returned to each original owner are within or near the specific parcel(s) he contri- buted. Farming of the original parcels can generally be carried on right up to the time when urban infrastructure is actually installed at each site, and a conscious attempt is made to minimize the time needed for transfor- mation from farm land to urban lots. 152. Land readjustment addresses some of the serious problems for which public land acquisition can be a solution, but in a way that involves no government purchase of land. The readjustment practice in Korea may well deserve imitation in other countries, though some of the methods of operation raise interesting questions. Doebele (1976a, p. 51) notes that the land readjustment program in Korea is accompanied by an "almost complete suppression of competitive private land subdivision in peripheral areas," and that the program is a "beneficient monopoly" which amounts to a "soci- alization of the land subdivision and development business." However, this monopoly of the subdivision process has not been used to divert from land- owners any betterment above that required to repay public expenditures for the readjustment. Since the practice is for the government to recoup only its own project cost, all of the increase in land value net of the infrastructure costs goes to landowners. The government is interested in making sure that the readjusted land has a high value, so that it can recover its investment in infrastructure. A result is very high net gains in value associated with readjustment (rarely less than a tripling in the market value of owners' holding, and often more); these large gains imply that the market is in great disequilibrium, with the value added by re- adjustment greatly exeeding its cost, signalling that more subdivision would be profitable (and probably socially desirable). 153. The quantity of serviced sites that the land readjustment agency must retain for sale to finance the project cost (the "cost equivalent area") depends on the market value of serviced land. In Korea the public authorities pursue a policy of keeping land prices high by making sure that "the lots they supply do not outrun demand at any time" (Doebele, 1976a, p. 51). However, the policy of keeping the prices of serviced land at levels high enough to repay project costs by sale of sites retained by the land readjustment agency stems from a simultaneous policy concerning the amount of land to be retained. If the fraction of land reserved for sale by government were greater, the project could be self financing at a lower price for the subdivided land, though with smaller profits to the initial owners. Variations of policy could be: - 56 - 1. "capturing" some of the gains from the land readjustment process by taking more of the finished sites for resale than is necessary to recoup only the government's costs; or 2. increasing the quantity of land readjusted every year, thus putting downward pressure on the price of serviced land. 154. In the first case land prices would stay high but the original owners would receive fewer building sites in return for their raw land contribution, and in the second case both the proportion of sites returned and the value of each would be reduced. In either case, there may still be a need to provide what appear to be large "windfalls" to participating owners in order to induce them to go through the process in the first place. The loss of agricultural income during the readjustment, the necessity of moving, the displacement from the time raw land is taken until bulding sites are returned, the loss of ovner occupants' consumer surplus, etc., all involve costs that are compensated for only by subsequent capital gains in. land value, and so not all land value gains represent increases in welfare for the owners. However, readjustment that returns building sites that are sometimes worth more than 20 times as much as the prior use value of the surrendered land would indicate that the land readjustment process as prac- ticed in Korea is indeed a "beneficient monpoly" that accomplishes a diffi- cult task well, but with a quite haphazard distribution of the profits. I/ 155. The use of land readjustment, with land contributions as the means of financing the provision of public facilities that provide better- ment, is particularly suited to countries where imperfections in the finan- cial sector make it difficult for landowners to finance betterment levies paid in cash. In addition, land readjustment has the advantage of almost completely ensuring compliance by landowners in payment of project costs, an important factor in countries where property tax collection is difficult. Since agreement on land contributions is necessary before the project begins, land readjustment as a method of financing urban infrastructure eliminates the delinquencies which can and do plague a betterment levy system where payment is in cash. 156. A related feature of the land readjustment process as a method of converting land to urban use is that the government does not have to pay any "holding costs" for land while the project is underway; these costs are voluntarily borne by the initial landowners who look forward to the higher valued readjusted sites in return. Thus, land readjustment 1/ One would expect that some of the expected returns from a readjustment would be foreseen and discounted into the value of land prior to readjustment. Thus agricultural land might well be transacted at a price above its value in current use prior to readjustment, and the measured increases in land value during readjustment would understate the full increase in value caused by the readjustment process. - 57 - is in some ways an alternative way to achieve the goals of public land banking, but with the raw land kept in private rather than public ownership until ready for development. 157. A central feature of the readjustment process is the bartering of raw land for serviced land, and this may be greatly preferable to the money exchange system when a lack of financial institutions makes trans- action of durable assets difficult. However, one inevitable aspect is that many participants become owners of several serviced building sites in exchange for their raw land contributions (probably more than they them- selves can use), and if the unused sites are not immediately sold the new owners autoi4atically become "speculators" in serviced land. Although the main goal of land readjustment is to provide serviced lots for use, it may be that this feature of providing several building lots to each participant will actually delay use of some of the serviced land, and the information provided by Doebele (1976a, pp. 54, E-36, and E-40) suggests that develop- ment on the serviced land does take place slowly. The lack of a well functioning mortgage market compounds the problem of mismatch between those who own sites and those who are eager to build. If the financial markets worked smoothly, those who want to build their own houses could secure long-term finance to purchase land from those who own several plots; without a mortgage market such asset transactions are much more difficult, and this will impede construction on the serviced sites. If development on the serviced sites does not occur promptly, the large infrastructure investment yields little or no immediate return, and this casts doubt on statements that a shortage of building sites is the chief impediment to orderly urban expansion. An effort to strengthen the mortgage market in readjusted areas may improve the effectiveness of the readjustment process in achieving its ultimate goal of providing housing in well planned and serviced extensions to the urban area. Finance for land transactions need not involve mortgage subsidies, but rather the simple mechanism for extending payment for sites over a suitably long period. 158. An alternative modification to the land readjustment process to speed utilization of serviced sites would be to return only one or two building sites to each landowner, and to auction off all the rest, with the land sale proceeds, net of government planning and infrastructure costs, divided among participating landowners in proportions to their initial land contributions. 1/ This would serve the dual purposes of ensuring that the initial landowners would have sufficient cash to begin construction on their own returned building sites, and also of ensuring that few owners would have more sites than they could themselves use. The auction process could be further modified by "antispeculation" requirements in the sale agreement 1/ This is equivalent to forcing owners to sell their excess sites even if they don't want to, and is a departure from one of the basic principles of the readjustment process, i.e., the barter system. - 58 - which would specify that development on lots be completed or at least be underway before specified dates; this sort of requirement may have to be tempered in specific cases, to take into consideration unpredictable fluctuations in family incomes which would make planned construction difficult. 1/ 159. It may be expected that the initial landowners will object to being given cash instead of serviced sites in the exchange process. One predictable disadvantage for landowners is that, in disposing of sites, the land readjustment agency would not necessarily attempt to get the highest possible price for the sites but would instead sell them at less than market value to achieve other goals (such as providing sites for low income families). The leeway for introducing such redistributive goals at the expense of the initial landowners depends on whether the size of the remaining capital gain from readjustment is still great enough to induce landowners to participate. 160. Additional questions concern the operation of the land market in a readjustment area after readjustment. For instance, do recipients of several plots sell some to finance construction on others? Is the land leased to those who wish to build? How are subsequent transactions financed? If owner occupancy is a desired goal, how can the transfer of ownership of land from multiple plot recipients to those who wish to build for themselves be facilitated? How are capital gains on readjusted land taxed, and are there any lock-in effects of the taxes? 161. In broader contexts, one must look at the use of potential "monopoly profits" of the land readjustment process. The most pressing problem in some developing countries is the need to produce serviced sites for the multitude of impoverished families, though there is then the problem of deciding who among the great number of eligible families will receive the sites. There is also the problem of knowing how much of the increase in value can be diverted from the original owners while still securing their voluntary cooperation (or at least avoiding violent oppositon). One impor- tant factor in this decision concerns the alternatives to readjustment which landowners have. If private subdivisions are not suppressed, owners will certainly consider this alternative source of capital gain from conversion of rural land to urban use if all the gains from legal land readjustment are siphoned off by the government. In Korea there is close monitoring of deve- lopment and strict prohibition of illegal subdivision, so that almost all urbanization of rural land is accomplished through government-sponsored readjustment._2/ I/ Still another way to promote use of the serviced sites would be to impose a special vacant land tax on developed sites in the readjustment area. 2/ Between one-fourth and one-third of the existing developed area in Seoul is on readjusted land (Doebele, 1976a, p. 13). The prohibition of illegal subdivision is enforced by frequent aerial photography and demolition of unauthorized construction. - 59 - 162. Government-sponsored land readjustment appears to promise both an efficient way to convert rural land to urban use, and also a way for the government to capture some of the land value increases caused by the supply of government services. There are dangers, however, in a government monopoly of land conversion. Private (often illegal) subdivision is in- evitably an important method of providing housing for low income families where the government is slow to provide serviced sites, or provides them at too high a cost. In many cases, the reality is not that the government acts to counter problems caused by the private market, but rather that the private market functions to reduce problems caused by improper government activity or inactivity. Thus, it is not axiomatic (or in some cases even likely) that greater public control over the land market will improve the welfare of poor migrants. The introduction of land readjustment as a means of converting land to urban use should be encouraged only if it increases rather than decreases the supply of urban land. The results will depend greatly on the ways in which political power is exercised, and on the nature of the planning organization. Land Banking 163. If there is no greater evil than a problem misstated, discussion of land banking would seem to be in great danger. Various policies with .varying objectives have been proposed under the general name of land bank- ing. First, small or medium size parcels of land can be acquired in advance of need for such ubiquitous public services as schools, fire stations, parks, or government offices. Local governments of any size require a large number of these sites. The need for new sites arises chiefly because the services are expected to require-expansion for any of several reasons--a growing population, a desire to provide new or upgraded services to an existing population, changing technology, and so on. Since it is likely that forces of these sorts tend to cause frequent needs for further services, the related land acquisition tends to constitute an ongoing program with many broadly repetitive operations. 164. Second, much larger areas can be acquired in advance for community facilities that require extensive tracts of land. Facilities of this sort would be major recreational areas, reservoirs, airports, industrial parks, and the like. Most communities require no more than a few of these very large sites, and such acquisitions may take place infrequently. Acqui- sition in advance to meet these needs presents quite a different problem from the frequently recurring advance acquisitions of land for individually smaller scale needs, and it is not likely that it could be carried out in any routine program. The investment is so large that a mistake can be politically disastrous. Public debate takes place. In short, each acqui- sition must be considered in the context of a unique event. 165. Third, land can be acquired in advance primarily for the purpose of shaping a desired land use pattern or to influence land prices, rather than to provide the sites required to produce specific public services. To be effective, this requires control over a very substantial quantity of the - 60 - undeveloped land surrounding the existing urban area. Control does not necessarily have to take the form of outright ownership of all land; there is a wide variety of less comprehensive means, such as zoning, mapping, and subdivision control. But insofar as purchase in full fee is the method used, the program is of a completely different order of magnitude from acquisition of sites for specific public facilities. Both the purpose and methods of this kind of acquisition are very different from those of the first two types. In many ways they constitute a different subject. 166. First we consider land banking as the acquisition of sites for public use of the first two types, and then turn to public dealing in the land market on a grander scale to achieve broader objectives. 167. Land Banking of Sites for Future Government Use. When land prices are rising, government agencies, like everyone else, have a strong incentive to buy land in advance of need, especially in locations where population growth requires the provision of new land-using public facilities. However, if a government wishes to capture appreciation by acquring land in advance of need for its own uses or for later resale to developers, it may experience some of the same costs that private landowners or speculators experience. 168. First, if land is purchased early to get it at a lower market price, the government may have to borrow to finance the acquisition, and the interest costs on that borrowing would be a cost of "capturing" the subsequent appreciation in land value. It may, of course, turn out that the interest cost of capturing the appreciation may exceed the appreciation captured, which is one reason why many organizations, public and private, choose not to buy land long in advance of need. One advantage that govern- ments may, however, possess is an interest cost lower than for other specu- lators; if the government discounts future values at an interest rate lower than in the private market, the discounted present value of any future development value will be higher to the government than to a private specu- lator. The source of the lower interest rate may, however, stem only from the government guarantee of taxing power to repay borrowed funds rather than from any genuinely lower cost associated with government borrowing. Also, the proposition that a government has a lower interest cost than does a private borrower is an argument for turning all investment over to the public sector, not just investment in land. 169. Second, in most situations governments face capital constraints such that the marginal rate of return on investment can be higher than the government borrowing cost. In that case, the higher opportunity cost of capital would be the relevant discount rate, rather than the low govern- ment borrowing rate. 170. Third, the government would undoubtedly incur some costs of administering an advance acquisition program. For private speculators, the rate of appreciation of raw land value must include a premium to cover - 61 - both management expenses and taxes; to measure the net benefit of government advance acquisition, these costs of management and foregone taxes would also have to be included in calculating the public costs of advance acqui- sition. 171. Fourth, the rate of appreciation on privately held vacant land includes a premium to cover the probability of loss to squatters, and government owned land may be subject to the same or perhaps even greater risk, depending on the public authorities' attitude towards squatting. If the government is holding land for use as future housing sites when the land can be serviced, it may be politically difficult to displace any residents who decide tp occupy the land earlier (though not, as experience is showing, to "incorporate" them into the project). On the other hand, the government clearly has the policy power to protect its land if it wishes, as evidenced by large vacant parcels already in public ownership in many cities. Without debating the merits and demerits of squatting as a means of allocating land to low income migrants, if property in a public land bank is permanently occupied by squatters, the appreciation (if any) in land value is captured by them rather than by the government. The government itself would appear to make a loss on such transactions. 172. Fifth, there is always the chance of buying land that later turns out to be inappropriate for an intended public use. For instance, after a site has been acquired in advance of need, it may turn out that an alter- native site may be a preferable location for the planned facility (prefer- able in terms of the site characteristics and current market values). If, in this situation, the originally planned site is used anyway, whether because of inertia, a failure to consider alternatives, or fear of appearing to have bought the wrong site in the first place, the full value of the appreciation is not really a net gain to the government. 1/ This will not happen if governme;nt officials always correctly consider the price of land as an opportunity cost. But in some cases, the available inventory of publicly owned sites may be the only ones considered for a new facility, and if so the benefit of having "captured" appreciation may be far outweighed by the poor location of the completed facility. Unless future project locations can be accurately predicted, or unless the government is in practice flexible in selling or exchanging inappropriate sites to obtain better ones, land banking of sites to capture appreciation may worsen rather than improve the site selection process. 1/ There is some evidence that governments do not carefully evaluate the opportunity cost of land acquired in advance of need. In 1966 the U.S. National League of Cities conducted a survey of its membership regarding public advance land acquisiton. Of the cities that had advance acquisition programs only a quarter made any attempt to calculate the carrying cost for properties acquired in advance of need; there was virtually no attempt by any of the respondents to evaluate the effectiveness of advance acquisition in terms of costs or benefits (Shoup and Mack, 1968). - 62 - 173. Since there is always uncertainty concerning both the timing of future development and the value of land when it is developed, most vacant land awaiting development will not increase in value at a constant rate of appreciation. Rather, as events change expectations, land values can change very quickly at some times, either up or down. Insofar as government possesses advance information concerning its own future actions that will affect land values, advance acquisition has potential for cap- turing those sudden changes in value that would otherwise accrue to private owners. To be successful, however, the government has to act before infcr- mation becomes public, and government land acquisition is often a slow process. 174. Also, if the proceedings are kept secret, there is a great temp- tation and opportunity for trading by those with inside knowledge of the government's plans. In any case, the government purchases would soon become a signal of intentions, and land values would tend to rise in response even in the absence of any public announcements. Public projects of any real consequence are rarely unexpected events, because much planning and debate usually does or should occur before final decisions are made. If compre- hensive land use planning is done for the urban area, this too will indicate roughly where services will be provided, and expectations regarding whether or not the plans will be carried out will affect land values long before services are provided. 175. This brief list of the costs that governments should calculate when considering steps to form a land bank to capture appreciation on sites needed for future public use is not exhaustive, nor is capturing appreciation in the value of acquired land the only benefit of land banking. In practice, the chief benefit of advance acquisition may not be to capture price appreciation, but rather to preserve sites that will later be needed for a public purpose by preventing premature commitment of the land to some other permanent use in the private market. If the sites for future public projects can be identified long enough in advance, early acquisition may be necessary to prevent private construction that would later have to be cleared at great expense or which would preclude use of the site for the intended public purpose. In two case studies of advance land acquisition by local governments, avoiding the cost of demolishing private construction on land needed for public use was found to be more important than the benef it of capturing price appreciation on the acquired land (Shoup, 1969). In addition, advance acquisition that avoids later demolition costs provides a net gain for society, while capturing appreciation is a transfer gain from the private sector to the government. 176. Land banking of sites acquired for future public use is compatible with some interim private uses, such as agriculture, or any other existing use at the time of accuisition, but there are potentially important problems with some other types of interim use. It will often appear that anv signi- cant amount of land awaiting a future public use is "idle" and should at - 63 - least be used during the interim for housing or another important social purpose. This can be appropriate if the users agree to vacate the land when the time for its permanent public use arrives, because in some cases the shelter constructed by very low income families is of sufficiently temporary construction methods to make good use of a site for periods of only a few years. However, once it is agreed to use the land for housing in the "interim," pressures can understandably build up to make the "in- terim" use permanent. The spectacle of a government agency dispossessing a large number of households, despite earlier agreement that this would happen, may deter the government from going through with its earlier plans for a subsequent public use. This is even more likely if the residents construct more permanent improvements in order to solidify their claims to the land and to magnify the apparent injustice of reclaiming the land. And, of course, if there is any sizeable number of voters who would be affected, political pressure to support their claims would be expected. These dif- ficulties of making clearly enforceable contracts for short term uses is one of the reasons for keeping land "idle" while awaiting a later public use. 177. Land Banking to Reduce the Rate of Land Price Inflation and Improve the Pattern of Urban Growth. The objectives of advance acquisition of the sites for government use are to capture land price appreciation and to provide public facilities in an efficient response to urban growth. By contrast, the objectives usually mentioned for larger scale land banking schemes are to reduce land price appreciation, and to improve the pattern of urban growth. 178. Perhaps the most important role urban governments perform in affecting land prices is in providing the public services that permit intensive development. This usually has the effect of greatly increasing the value of the individual parcels receiving services, but it also tends to hold down the rate of increase of value of other already serviced build- ing sites. If land banking is coordinated with the extension of urban services to raw land, the effectiveness of the land servicing process may be increased so that the resulting price of serviced sites will be reduced. The desired result will ultimately have to come through an increase in the quantity and utilization of serviced sites, with the government land trans- actions being a method of facilitating the necessary expansion of urban land area during rapid population growth. 179. Before analyzing how land banking might reduce the price of urban land or hold down the rate of increase, it is necessary to specify carefully the meaning of land prices and their rate of appreciation. In a market economy, the conventional concept of the price of land is the price that would be struck in a voluntary bargain between a buyer and a seller, neither of whom was under coercion. In this sense, a land banking operation could sell land to certain users below the market price, and thus reduce the pric.e of land to that user, but this could not be said to reduce the market price of land. Rather than reducing the market price of land by reducing its scarcity, the result is more akin to a subsidy in kind (which may, of - 64 _ course, by well justified but which introduces the problem of allocating the subsidy among deserving and undeserving claimants) rather than to any change in the market price of land. If gover-nent land banking is limited to buying and selling of undeveloped land, as only one participant among many in the conversion process, the potential for reducing the general level of land prices is greatly reduced. If the holdings of private land specu- lators are not simultaneously reduced, the government's entry into the market as a buyer of undeveloped land would presumAbly raise rather than reduce the demand for land and thus raise rather than lower the price of undeveloped land. 1/ The government cannot sell land to reduce prices without first having bought it, and since a secular rise in land prices is the perceived problem, counter-cyclical buying and selling would presumably have little effect on the long run price trend. In any case, the chief concern should be for the supply and price of serviced sites available for building, because it is on expectations of future value in urban use that raw land values should depend, not the other way around. 180. One way in which land banking operations in the raw land market can affect the supply of building sites is by assembling land and then reselling it with restrictions on the type and timing of the uses to which the land may be put. If, for instance, land is sold with the requirement that construction of specified improvements be comwleted within a given period, the government has greater control over the market than conventional land use regulations normally permit. However, even government ownership of land may be insufficient to guarantee control over use, as evidenced by squatter settlements on government land and by the frequently noted inability of the government to prevent original recipients of sites in site and service projects from selling their sites to others. It should also be noted that there are alternative ways to influence the use of land, such as by vacant land taxes to stimulate early development of serviced sites, zoning to control use, building codes to specify construction requirements, land pooling to assemble sites of appropriate size, etc., although each of these has its own problems of administration and enforcement. 181. The obvious availability of alternatives to government banking of land to control development and reduce land value appreciation suggests two important points: (a) the reluctance to use control measures that do not involve government purchase of land implies a lack of political will or concensus which would also extend to proposals for extensive land banking, and (b) many observers feel that the government land use controls that are exercised may hinder more than help the low income population in their search for housing. 2/ By setting inappropriately high standards for 1/ Carr and Smith (1975) argue that a land bank will reduce land prices only if it causes a reduction in speculators' holdings which exceeds the size of the land bank itself. Otherwise, the price of raw land would be increased by a land bank, which they consider to be more likely. 2/ See especially John F. C. Turner (1976). D- 6t -- permitted castructior.z density, and services. governments in less developed nations ha74e, it is argued, redu ced the a:tersatives open to the large proport.tor Ef the. population who cannoi- alord hotusing of such high standard, and any policy sucth as land banking, bmhich strengthens the government's cor-rrol over lantt development may aoctally work a greater hardship on those who now secure their housing outs_ide formal. haanels. 182. Thusz 9hle inastitution of a land baak which secures greater govern- ment 5ccntrl over land use must be carefu lly Considered in light of the probable uses -:hich government off icials will try to make of the land that is banked. In particular, there is the great potential problem of using .Land wh c9 is in the bankk9 when azother non-owned location would actually be a better site. This has allegedly been a problem with some sites and services projects, whose inappropriate locations were determined by the fact that the government already owned one sit.e and therefore failed to consider others which might have beer, superior in terms of cost, accessibility and other ifactors. All these considerations suggest the importance of ensuring that all government lang use pclici ies are appropriate for the affected popula, io. if mosre effective enforce ent tools are being advocated. 1 83 One way a land bank may se rve the purpose of reducing the general market price of serviced sites is throngh appropriating some of the better- ment which acco-ipanies the provlsion of services. If all land to be ser- viced is first purchased by the government, and if compensation for raw land is made at a price not far above its agricultural value, most of the better- ment associated with the provision of services accrues to the government; this revenue can be used to iacrease the pace at which serviced sites are brought onto the market. If successful, thls process would both appropriate the rural-to-urban lanid value appreciation to the government, and permit use of the funds to increase the supply of serviced land, which should lower its price. 184. This concept of land banking implies the use of government power of eminent domain to acquire land at curreat use value. If the government is the only purchaser of raw land for conversion to serviced sites, and if compensation in compulsory transactions is set at the land's agricul- tural, timber or other current use value, one would not expect significant appreciation in raw land values prior to government acquisition; any appre- ciation would be limited to the premium above current use value which the government chooses to pav for land it acquires. The government offer price, backed up by the power of compulsory purchase, will set a ceiling price for private transactionas. This process also implies, however, that all or most of the land brought into urban use must pass through the urban land bank before being provided with public services and sold or leased as a building site. 185. The ability of urban governments in developing nations to carry out such a scheme of land banking is certainly questionable. Government monopoly of the conversion process ls most closely approached in Sweden and the Yetherlands (Neutze, 1973). However, the institutional framework - 66 - and administrative resources in these countries are completely different from those in most of the rapidly urbanizing developing nations, and it is very doubtful if the same government control over land conversion could be successful without these institutions and resources. Doebele's analysis shows that the favorable results in Sweden have resulted not only from the use of large-scale public ownership, but also from an "intricate comp.Lex of interlocking and mutually self-supporting institutions all focused on commnon objectives" (1974, p. i). In particular, the important role of secrecy in the operations of the public land acquisition agency places a premium on the incorruptibility of the civil servants involved. 186. The introduction of a land bank operation to buy raw land at its opportunity cost in non-urban uses would involve a reduction of all raw land values to their current use values. It would also greatly increase the rewards to illegal subdivision of unserviced land, already a severe problem in many rapidly growing low income areas. Thus, success of this sort of land bank would require strict enforcement against illegal sub- division. If the choice is between sale of land to the government at agricultural value and illegal subdivision at urban values, the owner will certainly have very little incentive to wait until the government is ready to purchase his land; he would be greatly tempted to subdivide illegally if at all possible, or else vicariously share in the profits by selling the land at high price to someone else who would subdivide it illegally. The price of such illegally subdivided land could almost approach the price of legally subdivided land if the government pursues a policy of subsequently "' upgrading" the illegal settlements by providing clear titles and public services. The prospect of this future upgrading would presumably lead buyers to pay more for illegally subdivided unserviced plots and thus further increase the profit from (and incentive to) illegal subdivision. The extent to which values in illegal subdivision approach those in legal subdivisions depends on how quickly titles to land are legalized and how and when the public services are provided; greater government leniency in providing titles and services to pirate settlers may have the unintended effect of increasing the value of land in the pirate settlements and thus further increasing the rewards to illegal subdivision. One example of the incentives provided by subsequent legalization of squatter settlements is suggested by a 1967 program in Seoul, Korea to give subsidies and legal permits to owners of squatter houses to which specified i=provements had been made. This short-lived program obviously increased the incentives for illegal building, and in one area of the city 1,000 new houses were noted within one month of the program's announcement (Doebele, 1976a, pD 81). 187. Although a land bank that involves the purchase by local govern- ment of all land prior to provision of urban public services has the poten- tial to capture the rural to urban betterment and to provide a controlled supply of building sites, this potential may be very unlikely to be realized in most cities with a rapidly growing low income population. Govern- ment control over both squatting and illegal subdivision may be weak or to- be effective may involve very harsh measures, such as in Seoul, where - 67 - aerial photography is used to make frequent checks on illegal building, which is then quickly torn down after detection. Where such harsh control measures are not pursued, there would almost certainly be breakdowns in the government acquisition process caused by the great incentive to illegally subdivide rather than to sell to the government land bank at agricultural value. For this reason, a superior alternative to government purchase and later resale of all land undergoing rural to urban conversion may be the previously discussed technique of government supervised land readjustment, which involves the pooling of numerous individual parcels, subdivision of the assembled holdings into building sites with all public facilities and streets, return of some building sites to the original pooling owners, and government sale of enough sites to pay for all the infrastructure and planning costs. This method, as exemplified in Korea, promises most of the benefits of coordinated land use without requiring any form of land purchase by the government. 188. A final consideration regarding the feasibility of a land bank concerns the behavior of those entrusted with its operation. Managers of a land bank that is intended to influence the pattern of urban develop- ment would obviously have to make what are essentially political decisions about what is and is not desirable in future development. Dishonest prac- tices in both the acquisition and disposition of land would be difficult to control, especially where the justification for deviation from market prices in purchase or sale is claimed to be justified on the basis of achieving social objectives. If it is believed that dishonest practices in the land acquisition process are difficult to detect or control, there should be even more concern about a land banking project, the activities of which would consist mainly of land transactions. Private Development with Public Permission 189. Quite obviously, the present source of land for much (in some cities, most) new urban development is squatting and illegal subdivision. Because tenure is uncertain on such land, neither the occupants nor the landowners are certain of obtaining the rewards of investment in improving sites, and yet the vitality of self-help housing production by and for poor people on such land is readily apparent. One response to this realization has been the growing popularity of the site and service concept of govern- ment housing aid, which is intended to combine the potential of self-help construction with the government's ability to acquire raw land, lay out sites, and provide services. 190. The emphasis in site and service projects has so far been on government entrepreneurship in providing the sites, services, and perhaps core construction, with individual self-help effort devoted to actual housing construction. The use of private enterprise to lay out site and service projects on privately owned land has not been emphasized as a desirable policy, possibly because it has been assumed that private landowners who subdivide their land for sale or lease do not have the - 68 - proper incentives or capacity to provide sound planning and good services, such as a good physical layout of sites, water supply, sewage, electricty, etc., and that only public authorities can or will provide such services. Or perhaps it is the expectation that more private subdivision of land would serve only higher income groups, despite the fact that illegal private subdivision already is a major source of building sites for poor residents. Another explanation is that it is more convenient for international lending organizations to deal with government agencies in dispensing housing aid, even when these government agencies are not always well attuned to the demands of their intended clients. 191. Given the continued "uncontrolled" settlement, it is imperative to ask how the private land market might be better employed to provide the advantages sought in publicly sponsored site and service schemes, but with less use of the very limited public resources available. One way to do this is to facilitate servicing of land at private landowners' expense wich the resulting increase in land value (akin to betterment when the services are provided at government expense) as the necessary incentive to provide the services. This is, of course, a familiar land development practice irn the United States, where private subdividers are often responsible for provision of most on-site infrastructure, such as streets, utilities, and public recreation and school sites as well. An analysis of developers' incentives to provide services that residents want and can afford will suggest how in developing countries the private subdivision process might become an acceptable substitute for public land development where population growth has overwhelmed the government's capacities for servicing land. 192. As mentioned earlier in the discussion of land readjustment, an important aspect of the conversion process is the assembly of small parcels of raw land into one large holding, which is then serviced and subdivided into many small sites appropriate for urban use. The only point at which government participation in the land readjustment process is essential is in the land assembly, where public management is used to assure each participating landowner of his "fair share" of the resulting betterment, and thus induce him to participate in the process. This may not be necessary in all cases, however, for in many cities there already exist land holdings that are large enough to permit privately financed site and service projects. The size of the land holding is important, because if the scale of develop- ment is sufficiently large, many of what would otherwise be neighborhood external effects (providing betterment or worsenment to other landowners) are made internal to the total land development. Many factors, such as the amount of neighborhood open space to be preserved and maintained, the physical street layout, drainage, etc., which a small scale developer may sometimes neglect, become explicit decision variables for large scale developers. This is so because the value that residents attach to these public services and amenities should increase the demand for residential and commercial sites in the development, and thus increase the value of land for sale by the developer. Therefore, the profit maximizing land developer has - 69 - an incentive to provJde the amount and kind of services and other neighbor- hood amenities that potential residents desire and are able and willing to pay for through this market mechanism. The larger the scale of the new development, the more nearly do the public and private interests in pro- viding urban amenities coincide, because more of the value to residents of the amenities is capitalized into the value of the land owned by the de- veloper, who also pays the cost. 193. A site purchase is a complex transaction in that it involves not only the site itself, but also a level of accessibility and a package of amenities including everything from the road layout and level of public services to the pleasantness or otherwise of the neighbors. It is a bundle of purchases, and, recognizing this, the profit-maximizing entrepreneur will devote a certain proportion of land to uses which do not directly yield revenues. Up to a point, the provision of non-revenue-producing uses will raise the total return from revenue-producing uses because of the value site buyers place on such facilities. The revenue-producing use in the case of residential developments is mainly housing sites but may also include commarcial sites. The non-revenue use is mainly streets, any open space, and public facility sites. 194. It may be assumed that the developer will attempt to divide land between non-revenue uses and revenue uses to maximize the total value of the land in the development. Thus, he must simultaneously consider both the increase in the price of housing sites caused by increasing the amount of public land that does not directly produce revenue, and the decrease in the number of housing sites caused by dedicating land to public uses. For a maximum prof4t, the addition to total land value (or betterment) caused by the diversion of one more unit of land from revenue-producing to non- revenue-producing use must be equal to the opportunity cost of the addi- tional unit of land in revenue-producing use, which is simply the price of land for housing. 1/ 195. Where large parcels of land are privately subdivided into building sites the owners do have an incentive to provide a level of services consis- tent with what residents want and can afford because benefits of the services they provide are capitalized into the value of the sites they sell. It would be logically inconsistent to hold both the view that installation of on-site infrastructure on raw land provides betterment greatly in excess of the infrastructure's cost, and also the view that private developers would never themselves install such infrastructure when subdividing large sites for urban uses. In many countries, however, legal private subdivision for profit is discouraged by government regulations or subject to appropriation by squatters; illegal subdivision inevitably results, and the resulting insecurity of- title to the subdivided land is a strong disincentive for the investment of private capital in site planning and infrastructure. Bose (1975) describes how in Delhi the government's introduction in 1959 of the 1/ In this analysis it is assumed that the land subdivider is not a monopolist in the sense that the quantity of land supplied by him af:ects the overall price of land. Rather, the subdivider affects the price of his own land only by varying the amenities of the sites for sale. - 70 - Large-Scale Accjuisition, Develonment and Dis,osal of Laad Scheme, which was iceanded to "fraeze" the price of development land, effectively ended the legal private subdivision, which had provided a substartial ?art of the growing demand for housing. One result was a great increase ia the extent of unauthor'zed colonies with lower quality layout and infrastruccure and less security of tenure than were formerly provided in legal subdivisions. 196. One bit of evidence that sugests the potential of private sub- divisior is the considerable amount of pr-vately organized ard financed land readjustment that has taken place in Japan since early in this cen- cury. The process is sim"ilar to that described ina Korea, except that it is voluntarily organized by cte par:icipacing landowtners, who are themselves responsible for all the steps (including installat:on of infrastructure) performed by the govemeact in Korea (Doebele, 1976b). Private readjustment is, of course, unnecessary where the land is already i- large-parcel owner- shin and can be subdivided by one owner. 197. An alteruative process is for the provision of services by the government on an upgrad'ag basis only afeer the land Is already illegally subdiv-ded or squatted and occupied by residents who can vote or bring ocher oc" :ical pressure to bear on public authoritias. I' an expected sequence is for the provision of services at public expense afte: colonriza- :1on has :aken place, potential residen-s may even be wil tng 0o pay al-ost as high a pr-ce 'cr the initially unserviced land as they would for land chat already has services. The ironic rasult of this process is that some or all of the expected net benefi;s provided to residents by the evencual upgrading is thereby capħtalizad into land value and shifted to original landowners in c~e for= of betae-rment. -aat his does hannen In practice is suggested by the observation that the price c unserviced land in ilegalI subdivisaons Is often as high as that for serviced land in legal sub c.v_s :os. 193. ~:ea phencmena c0 squat:e; and Illegal subdivis'in -ave i=mLica- cions for land use reforms parallel to the reforms that are now popular in publIc housing policy. The realization ahat v ?h available resources few people could be publicly provided iwth housing in "standard" units and that those who did obtain whatever high qualcty publ c hous'-g was built were rarely poor, has comvinced ma-y of the necessity to "lower" the housIng standards to simple provision of serviced sites on which residants would build cheir own nouses. A parallel shift i. land Doolicy wculd be to locsen :he recuirad standards for legal private suadivisi_3, a: lens: _. ::ose s_:_a:iams *w;ere the inevitable alternat_ve Is squatting r 'llegal sub- diivs_on wi:thout whatever advantages frmal size planning and pre-supply of nublic services bring. Just as self-nelp aousing on servi:ed sites is now seen as a vay of utilizing individual initiat4ve in hous'ng construc- tion, legally santizoned privace servici-g and subdivision of land can, at private expense, Increase te su-ply of serviced urban land when :;he governaeat's own rasources are insuffIcie -t fcr the task. - 71 - 199. Since privately supplied illegal subdivision is now an important source of land for very low income migrants (especially for those too poor to obtain housing in higher standard public programs), "decriminalized" private subdivision activity could continue to serve a low income resident even when developers are subdividing for profit. The price for serviced land high enough to justify private investment in on-site infrastructure would almost certainly exclude the poorest residents, but increases in the supply of land affordable by higher income families can indirectly benefit even the very poor. In some government subsidized site and service pro- jects, the sites are pre-empted by families with income above the poverty level; thus, increasing the supply of privately serviced land could free public resources for aid to those in greater need. Further, legal sub- division should result in clearer identification of land ownership for taxation purposes, and the resulting property tax revenues could in part be used to extend publicly financed site and service programs. 200. In order to induce the private market investment in effective site planning and infrastructure, it is essential that the returns cover the cost, and therefore that the serviced land be protected from squatting. Since increase in land value is the only return for privately supplied infrastructure and site planning, this return must accrue to the owner to provide the necessary incentive to service the land. This increase in land value is in a very real sense "earned," as opposed to increases in value that result from the actions of others, and when the rate of urban popula- tion increase outstrips the government's ability to subdivide and service land, the possibility of improving the private sector's performance in the task should not be neglected. 1/ 201. Because of the advantages of unified control in the land develop- ment process,the use of private subdivision as a means of increasing the supply of serviced housing sites is most suitable where ownership of urban fringe land is already concentrated in large parcels. But because political and economic power is often correlated with land ownership in such situations, the challenge is to devise a policy to encourage the conversion of such to fringe land for use by the rapidly expanding low income urban population without unduly enriching landowners. The situation is comparable to that in the case of land readjustment, except that only one or a few initial owners rather than many are involved in the negotiations; this fewness of owners can greatly simplify and expedite the process of subdivision and installation of infrastructure, but it also raises more sharply the question of the distribution of the betterment associated with conversion of land to urban use. If all betterment accrues to the few initial landowners, there will be a strong incentive to supply serviced sites for urban use, but another result will be an increase in the inequality of wealth distribution. Recognition of this problem should not lead to attempts to suppress private subdivision (which may be futile anyway) but rather to modify the subdivision 1/ One important step would be the drafting and adoption of model land subdivision codes appropriate to the income, density, cultural traditions, climate, and other relevant factors for each area. - 72 - process, by either tax policy, regulation, or requirements for contribution of some of the privately subdivided sites to the governe-nt for allocation to the great number of low income families unable to purchase land in the private market. This sort of intervention must, however, strike a balance between maintaining sufficient incentives to increase the supply of serviced sites by private subdivision and also achieving other government objectives, including income distribution objectives. 1/ 202. If private entrepreneurs are to participate more effectively in increasing the supply of urban land, appropriate land taxation policies can be used not only to ensure an equitable distribution of the betterment, but also (as with special taxes on vacant land) to stimulate the conversion of land to urban uses when private owners have failed to respond to market demands or where the monopoly power of large landholders has led to supply restrictions. Any policy change that alters basic market conditions, as this proposed change would, calls for a reevaluation of related land policies. Legal private development of this sort, with regularized titles and officially recorded transactions, would enhance the prospects for more accurate assessment of land values, and this would change the environ- ment for tax policy. Thus, while more reliance of legal subdivision would call for taxation to redistribute some of the resulting land value in- creases, the process would at the same time generate some of the data that have heretofore been lacking for fair assessment of such taxes. 1/ A variant of large scale private subdivision with private supply of infrastructure to create site and service projects would be a mix between public and private subdivision, where the government supplies some or all infrastructure to privately owned land in exchange for a sufficient number of finished building sites to repay the cost. - 73 - VI RECOMMENDATIONS 203. It is difficult to recomnd policies that will be useful in every developing country, but there are some specific recommendations that may be appropriate in a wide variety of circumstances. Rather than recapitulate the preceding discussion, it seems more useful to highlight the land policy alternatives that look most promising in terms of achieving the objectives of efficient land use, equity in the distribution of the benefits of land, and providing government revenue. Tax Administration 204. In many countries property assessments are far below market value, and the ratio of assessed value to market value varies widely among proper- ties. Poor assessment practice limits the revenue capacity of the property tax, and the wide dispersion of the ratio of assessment to market value (and thus of effective tax rates on similar properties) is a serious viola- tion of the principle of horizontal equity in taxation. 205. Property tax collection performance is often poor; for instance, in Jakarta, property tax collections average only half of tax obligations despite the fact that the effective property tax rate is only about one- tenth of one percent of market value. A more uniform distribution of effective tax rates (on either rental or capital value) and more effective sanctions for nonpayment are important for improving both the equity and the revenue performance of the existing forms of property taxation. They are also indispensable for the success of any proposal to modify the property tax to include more complex tax bases, such as increases in land values caused by public expenditures. The inability or political unwillingness to manage a conceptually simple property tax system in some countries suggests the difficulties that would be encountered in extending the tax base to include various forms of betterment. Poor assessment and collection per- formance is not an inevitable problem, however, as evidenced by the favorable experience in some countries, such as Colombia, where the property tax administration is reasonably adequate and the revenue yield is substantial. 206. One prominent recommendation for improving assessment practice is self-assessment, in which the owner declares the value of property for tax purposes, with sanctions (such as possible compulsory government pur- chase at the self-declared value plus a fixed premium) to induce owners to declare true market value as the tax base. In countries where this policy has been tried (especially in Latin America), it has usually proved to be ineffective because the threat of purchase at the assessed value has not been made credible. In most countries it seems politically too difficult to enforce the sanctions against underassessment to make self-appraisal of Yalue an effective method of improving assessment performance. 74 - 207. The most pressing task in improving tax administration in many countries is to develop or improve a fiscal cadaster showing property boundaries, property values, and the names and addresses of all owners. Some countries, such as Ghana, tax only building value because of the ambiguous ownership status and ill defined property boundaries of much urtan land. Once a land registration system is in operation, however, it may even be easier to assess land values than building values because of the greater uniformity of land than building values in each area. Increased Taxation of Urban Land Value 208. At least in theory, there are advantages to be gained from heavier taxation of urban land value, and from a decrease in tax rates on building value. One argument is that taxation of buildings is a disincentive to construction including that of housing, and this is especially undesirable during rapid population growth. A second argument is that if land is taxed according to its capacity to produce revenue (even if the land is in fact undeveloped), owners will have an incentive to develop the land in order to pay the land taxes; this would put greater pressure on owners to build on land that already has public services, and relieve the pressure to build on unserviced land. A third argument in favor of heavier taxation of urban land is that this would help capture as tax revenue some of the betterment that is associated with the provision of public services. If there is an annual tax on the value of land, any increase in land value causes an increase in the annual tax payment. A fourth potential advantage of heavier taxes on urban land is that a land tax which reduces the market value of land may encourage savers to accumulate their wealth in the form of produced physical capital rather than in land value; in countries where land value is now an important means of holding wealth, this reduction in land value may be a way to stimulate capital investment and accelerate economic development. Although the theoretical advantages of site value taxation are quite clear, in practice it has been difficult to demonstrate that these advantages are very important. 209. Since site value taxation is already used in some developing countries, it would be useful to conduct comparative research on the land use effects of the two types of tax. One research method would be to examine the experience in countries, such as Jamaica, which have changed their tax base from total property value to land value alone. Another method would be to compare land use patterns in areas that are similar except for the tax system, as Clark (1974) was able to do in New Zealand. Taxation of Betterment 210. There is considerable evidence that the benefits of many site- specific government services are shifted to landowners by means of higher rents and capital values at benefitted sites. At the same time, governments are often prevented from more rapid extension of basic services to raw land by an inability to finance public expenditures. If so, a lack of serviced land for housing sites may be partly attributed to the failure to tax the betterment caused by public services. - 75 - 211. Although attempts to tax betterment have often proved disappoint- ing, there is evidence of success when the betterment tax is perceived as a price that owners must pay in order to receive an otherwise unpriced or underpriced public service (such as roads or sewers). This introduces some discipline into the political process of allocating public expenditures, and permits a more extensive program of servicing raw land for urban uses. If the "shortage" of serviced land in developing countries is due to an unwil- lingness or inability to charge a cost-recovery price for public services, a tax on increases in value caused by site-specific public expenditure may be the only feasible way to increase the rate of installation of public services. 212. There are, however, compelling theoretical and administrative grounds for not attempting to tax the full betterment associated with public services provision. The great difficulties in measuring betterment accur- ately, suggest that the goal should be less than 100% recovery of betterment in order to reduce the risk of taxing mDre than 100% when the estimate of betterment errs on the high side; the normal assessment errors in estimating land values can be greatly magnified when estimating the difference between land values. 213. Given the well known problems that many developing countries have in administering tax systems that are conceptually much simpler than a betterment levy, it seems prudent to begin modestly in attempting to tax betterment associated with provision of public services. An obvious start would be with public projects where land value gains are large in relation to project costs to be recouped, or where the full project costs do not have to be recouped by the betterment levy. User Charges as an Alternative to Taxes 214. Property taxes and user charges are alternative finance mechanisms for some public sector services that are site-specific (such as water, electricity, and sewerage). Just as the value of public services may cause an increase in land values, associated charges for the use of those services may reduce land values. That is, it should be the net benefit (net of user charges) of the services that would raise land rents and be capitalized into higher land values. Charges for services may then be shifted to an impor- tant extent from occupants, who are the service users, to landowners (if they are different persons). Thus, to the extent that property taxes are also shifted to landowners, the ultimate distributional consequences of the two finance methods may be quite similar. If progressive user charge rate structures, like those that permit an initial unit of service at a low fee with a higher charge for additional use (increasing bloc rates), are more progressive than a property tax of uniform rate, the well-known efficiency advantages of user charges as a rationing device would suggest greater reliance on them to cover at least marginal cost. In addition, to the extent that user charges are incident on land values, they should not be neglected as a betterment recapture device. - 76 - 2!5. Particularly in an inflationary environment, the use of historical (,osts as the basis for setting user charges in already completed projects can Iead to financial cost recovery, but also make it difficult to charge higher prices for similar projects completed later, which are more expensive because of inflation or real cost increases. To achieve some degree of horizontal equity of user charges among projects completed at different times, user charges should be at least indexed to the general price level so that cost- recovering user charges in future projects will not be greatly above the charges for similar services in existing projects. Land Readlustment 216e The discussion of land readjustment suggests that this is a land -1-ioy that works very well in the handful of countries where it is prac- -.id rTt is particularly well suited to planning and.financing of urban fringe growth, and appears worthy of being tried in other countries where rapid urban population growth requires the conversion of land from rural and urban uses. 17.1 gX Though land readjustment is almost by definition self-financing :. tDerms of current costs and prices, this is not so in terms of present ailese This is because the government must pay for all planning and _a*f%astructure costs (frequently long before) any revenues are derived from land sales. For instance, the large Yeong Dong project in Seoul, Korea, reqvurad a total capital investment of $49,000,000, and the typical Korean land readjustment project takes four or five years to complete (Doebele, X 976a, Ch. 2, p. 41). On the other hand, there is a tendency for land to be removed from productive use for only a short time during readjustment, so taat some of the capital investment is recouped even before the entire project is completed. 213. In considering a commitment of funds for readjustment, an impor- tant issue would be the distribution of the net benefits. As mentioned, in Morea. when the process is initiated by a private association and/or farmers aze holding part-time urban jobs and perceive well the value of urban land, the net benefits accrue largely to the initial landowners in the project area. This distributional impact could be altered in a more favorable direction by (1) goverrment appropriation of a larger proportion of the increase in land values; (2) increasing the scale of the problem, which onuld br-ng down both site price and the gains by owners from readjustment; (3) changes in the design of the finished sites to ensure that some would be appropriate for and affordable without subsidy to lower income residents; or g4) provisiou of readjusted land below market prices to those with limited capacity to pay. The initial design of the building sites in terms of size and amenities largely determines the income groups who can afford and will choose to pay to live in the project area; thus, some influence over the aatual design of the projects as well as over the distribution of the increases in land value is necessary to ensure that low income residents benefit from the readjustment process. - 77 - Subdivision Regulations and Service Level Standards 219. Inappropriate land subdivision regulations and public infrastruc- ture standards are in some cases an impediment to the supply of serviced residential land. Public service standards borrowed from higher income countries may, by aiming too high, reduce the supply of land provided with basic services that many residents could afford. In recent research on the optitum layout of site and service projects, Caminos and Goethert (1976) noted that their recommended forms of land subdivision, at the minimum service standard, would be illegal in most cities of the developing world. 1/ This suggests that it is very important to design land planning and sub- division regulations more closely attuned to the demands of low income residents. Bogota's recently implemented "Normas Minimas" program to align land planning and public infrastructure regulations more closely to what low income residents can afford to pay is an example of a move in this direc- tion, although it is too soon to evaluate the results. Land Transactions in Site and Service Projects 220. In site and services and squatter upgrading projects, initial owners who are given sites at a price below market value may wish to sell their sites to higher income groups rather than to retain them. In such instances, the subsequent higher income residents would appear to be the major beneficiaries of the project, while in reality much of the benefit of the project may have been shifted back to the original lower income owners by means of a sale price higher than what they paid for the site. Thus the project would have a higher target efficiency than would be apparent from observation of the subsequent occupants. Prohibitions on private sale of sites, though often ineffective, are an attempt to ensure that the original occupants receive the benefits of the project in terms of housing rather than in terms of a capital gain on the land transaction. In fact, prohi- bitions on sale of sites may simply act to lower the sales prices that the original owners receive, since they may often be unable to convey clear legal title to property sold in illicit transactions. In some cases, the prohibition on sale of property may thus actually lower rather than raise the target efficiency of the project. 221. One interesting aspect of the subsequent transactions experience of sites in site and services projects is the evidence it provides of how well adapted the project is for its intended occupants. Success in retain- ing the tenure of original occupants without compulsion is an indication that the design of the project is appropriate for the circumstances of the intended occupants; a large turnover by sale of sites to other occupants is evidence that the original occupants themselves did not find the sites appropriate. Prohibition on sale tends to suppress this evidence, or at least legal recognition of it, without necessarily improving the target I/ Further work on the implications of legal restrictions for low income households is being pursued by Alain Bertand for PADCO (Planning and Development Collaborative, Inc., Washington, D.C.) in a study of design criteria, costs and affordability trade-offs for urbanization projects. - 78 - efficiency of the project. Data on the rate of site transfers would be particularly useful in making ex post evaluations of site and services projects, and would help in tailoring new projects for specific clients. 222. Restraints on sale of sites are usually justified on the grounda of external benefits or donor preferences. However, the opportunity costs tc the residents should be fully understood before such restraints are adopted, and it should also be understood that, insofar as the net benefits of the project are capitalized into site values, the departing residents who sell their sites can take their benefits with them, in cash. The argument against permitting original occupants to sell their land derives from a suspicion that they will "waste" the proceeds on consumption (a funeral or wedding) and will soon be back on the streets, creating problems for the government, and demanding another site. An important area of research would be to trace the use of funds that low income residents receive from "cashing in" benefits initially received in kind. If the funds are typically used for productive investment that creates employment and provides the means to obtain other unsubsidized housing, the case for restricting the sale of sites is further diminished. - 79 - BIBLIOGRAPHY Aaron, Heary. Who Pays the Property Tax? Washington, D.C.: The Brookings Institution, 1975. Adams, F. Gerard, et al., "Underdeveloped Land Prices During Urbanization," Review of Economics and Statistics, May 1968, pp. 248-258. Anderson, Robert J., and Thomas D. Crocker. "Air Pollution and Residential Property Values," Urban Studies, 1971, Vol. 8, pp. 171-180. Archer, R.W. "Site Value Taxation and Redevelopment in the Brisbane Central Business Ara," The Valuer (Sydney), Vol. 22, No. 2, January 1972. Archer, R. W. "Land Pooling for Planned Urban Development in Perth, Western Australia", Canberra, Metropolitan Research Trust, 1976. Bahl, Roy. "Comparing and Evaluating the Use of Urban Property Taxes in Less Developed Countries." IBRD, April 1975. Bahl, Roy; Coelen, Stephen; and Jeremy J. Warford. "Estimation of the Economic Benefits of Water Supply and Sewerage Projects," Syracuse, N.Y.: Metropolitan and Regional Research Center, Syracuse University, October 1973. Baker, Harold L. "The Land Situation in the State of Hawaii," Land Study Bureau Circular No. 13, University of Hawaii, Honolulu, Hawaii, November 1961. Bauer, Raymond. "The Study of Policy Formation: An Introduction," in Raymond A. Bauer and Kenneth J. Gergen, The Study of Policy Formation, New York: The Free Press, 1968. Beier, George; Churchill, Anthony; Cohen, Michael; and Bertrand Renaud. "The Task Ahead for the Cities of the Developing Countries," IBRD Staff Working Paper No. 209, July 1975. Bird, Richard M. "The Incidence of the Property Tax: Old Wine in New Bottles?" paper presented to the Symposium on Property Taxation sponsored by the Ministry of State for Urban Affairs, Toronto, February 17, 1976. Bose, Asish. "Some Aspects of Rising Land Prices and Land Speculation in Urban Delhi," paper prepared for Habitat, United Nations A/CONF. 70/RPC/BP/17, May 15, 1975. Brown, J. Bruce. "The Incidence of Property Taxes Under Three Alternative Systems in Urban Areas in New Zealand," National Tax Journal, September 1968, pp. 237-252. - 80 - Carr, Jack and Lawrence B. Smith, "Public Land Banking and the Price of Land," Land Economics, November 1975, pp. 316-330. Cheo, Ih-Kwei. "A Preliminary Study of Taxation of Urban Vacant Land in Taiwan." In A. M. Woodruff and J. R. Brown, Land Use for thae Cities of Asia, Hartford, Conn.: The John C. Lincolu Institute, 1971. Cheung, Steven N.S. 'Roofs or Stars: The Stated Intents and Actual Effects of a Rents Ordinance," Western Economic Journal, March 1975, pp. 1-21. Clark, W. A. V. The I,act of Property Taxation on Urban Spatial Development, Report No. 187, Los Angeles: Institute of Government and Public Affairs, University of California, Los Angeles, 1974. Crecine, John P.; Davis, Otto A.; and John E. Jackson. "Urban Property Markets: Some Empirical Results and Their Implications for Municipal Zoning," Journal of Law and Economics, October 1967, pp. 79-99. Davis, Otto, and Andrew Whinston. "The Economics of Urban Renewal,' Law and Contemporary Problems, Winter 1961. Doebele, William A. "Land Policy in Seoul and Gwangju, Korea, with Special Reference to Land Readjustment," Urban and Regional Report No. 77-9 (Vol. I), International Bank for Reconstruction and Development, January 1976a. Doebele, William A. "Land Readjustment As An Alternative to Taxation for the Recovery of Betterment: The Case of South Korea." To be published in the proceedings of the 15th Annual TRED Conference. Doebele, William A. "Valorization Charges as a Method for Financing Public Works in Bogota, Colombia," Urban and Regional Report No. 77-10, International Bank for Reconstruction and Development, June 19.75. Doebele, William A., with Orville F. Grimes. Valorization Charges as a Method of Financing Urban Public Works: The Example of Bogota, Colombia. World Bank Staff Working Paper No. 254, March 1977. Doebele, William A., Orville F. Grimes and Johannes F. Linn. "Participation of Beneficiaries in Financing Urban Services: Valorization Charges in Bogota, Colombia." Forthcoming in Land Economics. Feldstein, Martin. "The Surprising Incidence of a Tax on Pure Rent: A New Answer to an Old Question," Journal of Political Economy, Vol, 85 No. 2, April 1977, pp. 349-360. Friedmann, John. "A Generalized Theory of Polarized Development." In Niles M. Hansen (ed.), Growth Centers in Regional Economic Development, New York: The Free Press, 1972. - 81 - Gaffney, Mason. "Adequacy of Land as a Tax Base," in Daniel M. Holland (ed.), The Assessment of Land Value, Madison: The University of Wisconsin Press, 1970. Gibson, John G. "The Interpretation of Property Price Chang.- for a Measure of the Welfare Benefits of Pollution Control: Two Simple Models," Urban Studies, 1976, Vol. 13, pp. 45-50. Gilbert, Alan. "The Arguments for Very Large Cities Reconsidered,," Urban Studies, 1976, pp. 27-34. Goldsmith, Raymond. The National Wealth of the Unites States in the Postwar Period, Princeton: Princeton University Press, 1962. Grierson, Ronald E. "The Economics of Property Taxes and Land Values: The Elasticity of Supply of Structures," Journal of Urban Economics, 1974, pp. 367-381. Grimes, Orville F. Housing for Low-Income Urban Families, Baltimore: Johns Hopkins University Press, 1976. Grimes, Orville F. Urban Land and Public Policy: Social Appropriation of Betterment, World Bank Staff Working Paper No. 179, May 1974. Harriss, Lowell C. "Transition to Land Value Taxation Some Major Problems," in Daniel M. Holland (ed.), The Assessment of Land Value, Madison: University of Wisconsin Press, 1970. Hicks, Ursula K. "The Taxation of Land and Buildings," in Milton Taylor (ed.), Taxation for African Economic Development, New York: African Publishing Corporation, 1969. Hicks, Ursula K. "Can Land be Assessed for Purposes of Site Value Taxation?" in Daniel M. Holland (ed.), The Assessment of Land Value, Madison: University of Wisconsin Press, 1970. Holland, Daniel M. and Oliver Oldman. "Jakarta - Real Estate Tax Study - Final Report," IBRD, October 1972. Kimoni, S.M. "The Structure of Landownership in Nairobi," Journal of East African Research and Development, Vol. 2, No. 2, 1972, pp. 101-24. Kamm, Sylvan, Land Banking: Public Policy Alternatives and Dilemmas, Washington, D.C.: The Urban Institute, 1970. Koopmans, Tjalling and Martin Beckman. "Assignment Problems and the Location of Economic Activities," Econometrica, January 1957, pp. 53-76. Lee, Goh Ban, The Pattern of Land Ownership in Central Georgetown, Monograph No. 2, Centre for Policy Research, Universiti Sains Malaysia, 1975. 82 - Lent, George E. "The Urban Property Tax in Developing Countries," Finanzarchio, Band 33, Heft 1, 1974. Lent, George E. "Taiwan's Land Tax Policy," Interuational Monetary Fund, Fiscal Affairs Department, Working Paper FAD1/76/2, May 7, 1976. Lind, Robert C. "Spatial Equilibrium, the Theory of Rents, and the Measurement of Benefits from Public Progrms," guarterlv Journal of Economics, May 1973, pp. 188-207. Lind, Robert C. "Spatial Equilibrium, the Theory of Rents, and the Measurement of Benefits from Public Programs: Reply," Quarterly Journal of Economics, August 1975, pp. 474-476. Linn, Johannes F. "Property Taxation in Bogota, Colombia: An Empirical Analysis of Poor Revenue Performance." Forthcoming in National Tax Journal. Liun, Johannes F. "The Incidence of Urban Property Taxation in Developing Countries: The Case of Colombia." To be published in the proceedings of the 15th Annual TRED Conference. Linn, Johannes F. "The Incidence of Urban Property Taxation in Developing Countries: A Theoretical and Empirical Analysis Applied to Colombia. World Bank Staff Working Paper No. 264, August 1977. Linn, Johannes and Nelson Valverde. "Public Service Costs and City Size: The Case of Urban Water Supply and Sewerage Cost Functions in Colombia," Forthcoming in Journal of Urban Economics. McClure, Charles E. "The Relevance of the New View of the Incidence of the Property Tax in Developing Countries." To be published in the pro- ceedings of the 15th Annual TRED Conference. Macon, Jorge and Jose Merino Manon. "Betterment Levies in Latin America: Nature,'\,Experience and Recommendations for Their Adoption in the Financing'of Public Works Projects," Economic and Social Development Department, General Studies Division," Economic and Social Development Department, General Studies Division, Inter-American Development Bank, Washingtin, D.C. March 1975. Manvel, Allan. "Trends in the Value of Real Estate and Land, 1956 to 1966," in U.S. National Commission on Urban Problems, Three Land Research Studies, Washington, D.C. 1968a. Manvel,\ Allan. "Land Use in 106 Large Cities," in U.S. National Commision on Urban Problems, Three Land Research Studies, Washington, D.C., 1968b. Maxwell, James A. and J. Richard Aronson. Financing State and Local Governments, 3rd edition, Washington, D.C.; The Brookings Institution, 1977. - 83 - Mills, Edwin S. "Do Market Economies Distort City Sizes?" in Harry Swain and Ross D. MacKinnon (eds.), Issues in the Management of Urban Svstems Papers and Proceedings from an International Institute for Applied Systems Analysis Conference on National Settlement Systems and Strategies, Shloss E-axenburg, Austria, December 1974. Mohring, Herbert and Mitchell Harwitz. Highway Benefits: An Analytical Framework, Evanston, Illinois: Northwestern University Press, 1962. Neutze, G. Max. "The Price of Land for Urban Development," Economic Record, Vol. 46, No. 115, pp. 313-328, 1970. Neutze, G. Max. The Price of Land and Land Use Planning: Policy Instruments in the Urban Land Market. Paris: OECD, 1973. Peterson, George E. "Federal Tax Policy and Land Conversion at the Urban Fringe," presented at the Conference of the Committee on the Taxation of Resources for Economic Development, Madison, Winsconsin, October 1974. Pines, David and Yoram Weiss. "Land Improvement Projects and Land Values," Journal of Urban Economics, pp. 1-13. Pollock, Richard L. and Donald C. Shoup. "The Effect of Shifting the Property Tax Base from Improvement Value to Land Value: An Empirical Estimate," Land Economics, Vol. 53, No. 1, February 1977. Richardson, Harry W. The Economics of Urban Size. Lexington, Mass.: Lexington Books, 1973. Richman, Raymond L. "The Theory and Practice of Site-Value Taxation in Pittsburgh," 1964 Proceedings of the Fifty-Seventh Annual Conference on Taxation, Harrisburgh, Pa.: National Tax Association, 1965. Risden, 0. St. Clare. "A History of Jamaica's Experience with Land Taxation Based on the Site Value System," presented at the Conference of the Committee on Taxation, Resources and Economic Development,. Cambridge, Massachusetts, October 1976. Scott, Randall W. (ed.). Management and Control of Growth. Washington, D.C.: Urban Land Institute, 1975. Shoup, Donald C. and Ruth P. Mack, Advance Land Acquisition by Local Governments. Washington, D.C.: USGPO, 1968. Shoup, Donald C. "Advance. Land Acquistion by Local Governments: A Cost-Benef-it Analysis," Yale Economic Essavs, Fall 1969, pp. 147-207. Squire, Lyn and Herman van der Tak. Economic Analysis of Projects, Baltimore: Johns Hopkins University Press, 1975. -84- Strotz, Robert. "The Use of Land Rent Charges to Measure the Welfare Benefits of Land Improvement," in Joseph E. Haring (ed.), The New Economics of Regulated Industries: Rate Making, in a Dynamic Economy, Los Aageles: Economic Research Center, Occidental College, 1968, pp. 174-186. Turner, John F. C. "Rousing as a Verb," in John F. C. Turner and Robert Fichter (eds.), Freedom to Build, New York: The Macmillan Company, 1972. Turner, John F. C. Housing bX People, London: Marion Boyars, 1976. Vargha, Louis A. "An Economic View of Leasehold and Fee Simple Tenure of Residential Land in Hawaii," Land Study Bureau Bulletin No. 4, University of Hawaii, Honolulu, Hawaii, January 1964. Vickery, William S. "Defining Land Value for Taxatin Purposes," in Daniel M. Holland (ad.), The Assessment of Land Value, Madison: University of Wisconsin Press, 1970. Watson, Peter L. and Edward P. Holland. "Congestion Pricing, the Example of Singapore," Finance and Development, March 1976, pp. 20-23. Weiabrod, Burton. "Collective Action and the Distribution of Income: A Couceptual Approach," in Robert Haveman and Julius Magolis (eds.), Public Expenditure and Policy Analysis, Chicago: Markham Publishing Company, 1970. Wheaton, William C. "Residential Decentralization, Land Rents, and the Benefits of Urban Transportation Investment," American Economic Review, March 1977, pp. 138-143. Woodruff, A.M. and L.L. Ecker-Racz. "Property Taxes and Land Use Patterns in Australia and New Zealand," in Arthur Becker (ed.), Land and Building Taxes, Their Effect on Economic Development, Madison: The University of Wisconsin Press, 1969. Woodruff, A.M. and J.R. Brown. Land for the Cities of Asia. Hartford, Conn.: The John C. Lincoln Institute, 1971. - 85 - SO3E PERSPECTIVES ON URBAN LAND USE REGULATION AnD CONTROL 1Salcolm D. Rivkin, Consultant Rivkin Associates, Inc. Urban Proje&cts Department - 86 - TABLE OF CONTENTS Page No. T. INTRODUCTION 87 II. RATIONALE FOR LAND USE REGULATION AND CONTROL 88 A Schma of Objectives 89 Fiscal Implications of Regulatory Approaches 90 II. CONSTRAINTS TO LAND USE REGLATION AND CONTROL 91 Lack of Priority and Policy 91 Lack of Information 92 Lack of Manpower and Managerial Skills 92 Economic Significance of Investment in Land and Buildings 93 Rapidity of Growth 93 Rigidity of Techniques 93 IV. SOME EXAMPLES OF INTERPLAY AND IMPACT 93 Macro-scale Impacts 94 Micro-scale Impacts: The Shaping of Land Use in a Given Community 98 Conclusions 103 V. THE ARRAY OF TECENIQUES 103 Playing the Game by Direct Intervention 104 Setting the Rules by Planning and Controls 113 Some Promising Additional Approaches 119 VI. TOWARDS A MORE EFFECTIVE POLICY OF LAND USE CONTROL: SOME CONCLUSIONS AND RECOMMMNDATIONS 123 Conclusions 123 Recommended Research 124 - 87 - I. INTRODUCTION 1. In any urbanizing area governmental policy, programs, and actions can affect the pattern of urban development. By pattern we mean the character, location, scale, and densities of land uses and the adequacy of their supporting utilities and services. 2. Many factors, however, play roles in shaping the pattern: market demand for the range of urban activities; physical characteris- tics of the land such as topography and soils; tenure arrangements; 1/ and, at any point in time, the pattern of man-made physical facilities already in place to which new growth must accommodate. Of particular importance is the price mechanism which serves to allocate available land among competing uses and densities. Governmental intervention is only part of the process. Indeed the nature and degree of govern- mental intervention varies from country to country, from city to city, dependent on the political and economic traditions of the urbanizing society. 3. Although the role of government differs widely, the subject of public policy to regulate and control the pattern of urban develop- ment is of considerable significance to the developing world. For many serious land use problems in Third World cities are capable of mitigation only through some form of governmental intervention. Continued immigration produces intense demands for land that cannot be satisfied by available sites. Extremely high densities of uses and people become concentrated on inadequately serviced land. Low- income workers are relegated to housing far from places of potential employment. Extreme traffic congestion and inadequate sanitation produce dangers to public health. The price mechanism is not noted for its altruism. And in rapidly growing Third World cities, it is a poor vehicle for achieving efficiency and equity in the pattern of physical development. 4. This paper indicates that government can work to produce greater efficiency and equity in the allocation of urban land use and services. However, the principal tools available may not be those generally considered in the category of formal land use regulation and control; e.g. master planning and various types of zoning, build- ing code, and subdivision instruments. Instead, the primary control devices may be instruments available to most urban governments but generally adopted for purposes not directly related to land use, namely devices associated with direct public capital investment and finance. 1/ See "Tenure in Relation to Market and Land Use Controls," Selected Issues in Urban Land Tenure by William A. Doebele. 88 - If governments can be persuaded to broaden their understanding of the development-shaping power in these devices, then the limited potential inherent in formal regulatory techniques can be applied selectively as an important supplement. 1/ II. RATIONALE FOR LANDUSE REGULATION AND CONTROL 5. The concept of "regulation and control" of land use implies a rulesetting function on the part of government. It is therefore appropriate to ask at the outset whv government should undertake to establish rules about how land is used (e.g. the location, scale, and densities of various land-consiting activities and the physical stand- ards for land development and construction). Why should not this be a matter left to the market place, with the role of government limited to providing essential public facilities and services to the pattern of activities established through free market competition? Simplistic though the question may be, it is an appropriate starting point to set a framework for analysis. 67 One answer is that the market place cannot be trusted to produce a rational, efficient land use system; a system which can both accommodate all legitimate needs for space at a given period of time and allow for long-term acco modation of future growth. A city evolves over time, hovever, and its needs for land constantly change. In this context, urban land is a scarce resource. As a resource it requires stewardship by public bodies who are able to balance short-term against long-term needs and balance the claims of one interest group against another. Thus even the most capitalistic, free enterprise-oriented societies have increasingly imposed some measure of public control over the use of urban land. 7. A 1973 United Nations study dealing with the subject of land use regulation and control throughout the world underscores the resource conservation aspect in stating: The demand for urban land is growing, yet the supply is both genuinely and artificially limited. The situation radically increases land costs and in turn, consumes scarce investment capital better used elsewhere. It also irrationally distorts patterns of urban growth and development. ....as the urban infra-structure becomes 1/ See Urban Land Use Regulation by John Courtney for a review and evaluation of regulatory techniques. - 89 - more costly and inefficient and institutions and facilities fail to provide adequate services to their populations, urban social and economic imbalances and injustices are intensified...the quality of the total urban environment erodes. 1/ Government must set rules regarding the use of land to avoid economic, social and physical chaos. A Schema of Objectives 8. Our organizing assumption is, therefore, that government should undertake some form or forms of allocative and distributive effort regarding the use of urban land. What should, or rather, could be the specific objectives for government to pursue in a land regula- tion and control system? The following come to mind. 9. Efficiency. Land controls could be instituted to ensure that the city "works," e.g. that there is a sufficient supply of land avail- able and at reasonable costs for all necessary urban activities (housing, industry, services, recreation, etc.) and for the public facilities to support these (roads, sewers, hospitals, etc.); that these activities and facilities are appropriately accessible to each other; and that capacity -- in the serviced sites available -- is sufficient for the needs at any point in time. 10. Health, Safety, and Welfare. This is an oft-employed triad referring to the protective functions of government. In the area of land use it would apply to setting of rules or standards that might limit overcrowding on land, prescribe the adequacy of utilities (as to design and capacity), establish construction regulations geared to the safety of buildings, the nature of noxious air and water emissions allowable and a host of other possible measures that intervene on be- half of the quality of urban life. ll. Equity. Regulations to intervene on behalf of equity can be identified at two levels: a. To ensure that certain classes of activities or people are not denied land within the urban area or denied adequate access to other activities (e.g. employment) or services (e.g. transportation, water). b. To protect rights of all owners of land against arbitrary seizures, intrusions, or classifications by government or private parties which di=inish the utility or value of the land. 1/ United Nations, Urban Land Policies and Land Use Control Measures, Global Sum-ary (Vol. VII), 1973, p. iv. 12. Adaptability. Regulations and controls could be established to ensure that the urban pattern can adapt and change, readily, as new pressures of population growth and economic activity arise; e.g. to enable expansion of activities and services as the need arises, and to enable conservation of areas or facilities deemed of value to society as periods of change occur. 13. Conflict Resolution. A control system could be established to mediate and resolve conflicts between competing objectives at one level and competing uses at another. 14. The several objectives listed above are not discrete, in the sense that "adaptability" can be considered an aspect of "efficiency," that "health, safety and welfare" contains components of "equity," etc. They do, however, cover the range of potential purposes which a govern- ment could pursue if it wished to foster a land use pattern which was able to meet legitimate needs of urban society. Fiscal Implications of Regulatory Approaches 15. Let us presume, for the moment, that urban governments do accept these objectives and do proceed to establish and enforce rules about the use of land. The particular devices employed might be com- binations of land use plans, zoning and building codes, subdivision regulations or other less conventional techniques. Although not neces- sarily a direct objective of controls, their fiscal impact can become of extreme significance and could well be a consideration in the for- mulation of a control strategy. The fiscal consequences of such regula- tion may be felt in three areas: 1. The value of land. 2. The costs of site development and building construction. 3. The revenues derived by government from real property taxation. 16. The Value of Land. The allocation of use type and density to specific sites by government can -- if taken seriously by the market place - effectively determine much of its site values. By way of example, take two pieces of vacant land of equal size, equal accessi- bility to transportation and utilities, of equal physical suitability for a wide range of construction. A planning agency or a zoning authority determines that Parcel A should be designated, officially, as a site for a multi-story office building. Parcel B is designated as available only for individual, one-story houses on large lots. There is, in consequence, a rush of value to parcel A, because of its greater potential yield. Replicated many times in a planning and zoning process, this allocation of uses and densities can indeed become - 91 - a major force in establishing the pattern of land values within an urban area. 17. Clearly many other factors do influence land value -- the supply of sites of different types and the demand for such land, the physical c&aracteristics of the land, its accessibility, etc. -- but designations of use and density type can be a powerful instrument in the hands of government, providing the market place regards the regulatory actions as serious. 18. The Costs of Site Development and Building Construction. Regulations prescribing standards to be met can affect the costs of development and construction and therefore the value of improvements on the land. By prescribing certain materials to be used (e.g. con- crete rather than wood) or certain minimal internal space requirements per occupant, building regulations can establish cost levels for con- struction that may or may not be higher than private enterprise, left to its own devices, would incur. Any land development regulations which prescribed on-site installation of utilities, minimum open space on a lot per dwelling unit, requirements for parking space, etc. would have a similar effect on costs. Much of the cost (and therefore the value) of improvements depends on facility types, and labor and mate- rials costs, but regulations can -- if enforced -- add an increment as well. 19. Revenues to the Public Derived from Property Taxation. In most countries where property taxation serves as a form of muni- cipal or national revenue, some computation of value is made of the land and improvements as a basis for tax levies. To the degree that governmental land use regulation has affected the value of the land and construction costs of facilities, these impacts may be indirectly reflected in the valuation of the property for tax purposes and thus in the revenues derived from property taxes. III. CONSTRAINTS TO LAND USE REGULATION AND CONTROL 20. Although evidence is limited and research almost nonexistent (see below), formal governmental rulesetting over land use in develop- ing countries has been ineffective in achieving the objectives out- lined above. Numerous reasons may be advanced as to why. Lack of Priority and Policy 21. Few urban governments (or national administrations, for that matter) have accepted land use control as a major priority. Without a priority, policies towards land use have not been derived, and explicit objectives for control techniques have not been established. - 92 - The schema of objectives suggested above is a reasonable one considering the issues to be faced - especially in the great metropolitan areas. But few instances exist where these or other comparable objectives have been adopted as a guide to governmental action. Numerous planning activities have been commissioned which attempt to establish a physical form or de- sign for an urban area, without regard to questions of efficiency, equity, and adaptability. Numerous codes, ordinances and other instruments have been adopted as derivatives from developed countries with totally dif- ferent land and construction circumstances. These are unrelated to the local development setting and market. The planning and the codes thus become unenforceable, partly because of low priority and partly be- cause of irrelevance. Lack of Information 22. It is extremely difficult to establish a regulatory and control "system" without information on the matters to be regulated. Even the most unsophisticated and rudimentary control effort to be formulated and enforced requires information. At the very minlnmum are availability of records on existing land use, development density, and ownership; along with accurate physical information on soil charac- teristics, locations and capacities of existing utilities and services, etc. For planning purposes, information on economic and social characteristics of population and activities is essential. The infor- mation base available in many developing cities is improving, par- ticularly with the aid of aerial photography that eliminates much labor- intensive data gathering. By and large, however, the cities lack sufficient accurate and up-to-date data on many of the above items, a deficiency which inhibits application if not formulation of comprehen- sive regulatory techniques. Lack of Manpower and Managerial Skills 23. Improvisation in the face of inadequate data is possible, but requires skilled people to establish and interpret substitutes. However, the most sensitive long-range planning and development con- trol system in the world has no value if unsupported by skilled and honest property appraisers and inspectors in the enforcement process. 24. Here, developing countries are hampered both ways. Few trained professionals are available to prepare plans and formulate development regulations. Even in Comonvealth countries where appraisers represent a traditional skill for which training is sometimes available, municipal evaluators and inspectors are in very short supply. Often they are (as in some Latin American countries) administratively removed from the planning and ordinance formulation process. Such permit systems that do exist are, moreover, frequently subject to graft and corruption. - 93 Eomicicane of I 25. In many developi