I'1 The World Bank Policy, Planning and Research Staff Infrastructure and Urban Development Department Report INU 48 FINANCING URBAN SERVICES IN LATIN AMERICA: Spatial Distribution Issues by Glan Carlo Guarda June 1989 DISCUSSION PAPER This is a document published informally by the World Bank. The views and interpretations herein are those of the author and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting on their behalf. Copyright 1989 0 The World Bank 1818 H Street, N.W. ALl Rights Reserved First Printing June 1989 This is a document published informally by the World Bank. In order that the information contained in it can be presented with the Least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The World Bank does not accept responsibility for the views expressed herein, which are those of the author and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitations of its boundaries or national affiliation. The main author is Gian Carlo Guarda, Principal Urban Planner, Infrastruct-e & Urban Development Department. Many of the report conclusions are based on observations and contributions of Bank colleagues and sector experts, principally K. Mera, W. Dittinger, D. Mahar, J. Linn, R. Baht, R. Bird, J. Macon, F. Rezende, M. Casanegra, and many others who have covered much of the same ground previously. Several staff members (R. Buckley, W. Dillinger, J. Hicks, B. Renaud, and S. Whitehead), my division chief, P. Ljung, and my director, L. Poutiquen gave helpful suggestions on a rough, first version of the paper. The World Bank FINANCING URBAN SERVICES IN LATIN AMERICA: Spatial Distribution Issues DISCUSSION PAPER INTRODUCTION This paper discusses urban services finance in the Latin American context of rapid urbanization, severe fiscal constraints and democratization of administrative systems. Analyzing the situation in thirteen different countries, it draws several conclusions which task managers may find helpful in the design of new lending operations. The report's main theme is that the World Bank's approach to the urban sector needs to be re-directed. The long held dilemma between efficient pricing and poverty alleviation has often precipitated in the inconsistent application of a "user-fee" rationale to isolated project loans provided to particular cities. These "windfall" urban loans establish an unsound basis for Bank-Borrower dialogue by calling for special appropriations of central government funds. They favor discriminatory treatment of particular localities, distorting the regular pattern of resource allocation. The position taken is that a broader fiscal perspective is necessary to develop a systemic approach to urban lending. In this regard, the ongoing improvement of the existing systems of revenue sharing and institutional borrowing practices is seen as a promising area for dialogue with LAC governments. The structure of the report is as follows: first, it reviews the RisLng Demand for Urban Services (Section 1) noting the positive correlations of rapid urban growth with national development, but also the corresponding, incremental fiscal pressures. The Macro-Economic Constraints to Urban Expenditures are summarized in Section 2, which mentions the adverse circumstances limiting new capital formation and service provision in Latin American cities. Section 3 poses the question: Are LAC Governments Truly Decentralizing? Compared to those prevailing in industrial market economies the expenditure and revenue autonomies of subnational governments appears meager in the countries of Latin America, which are also plagued with regional disparities. These disparities foster political debates and widespread initiatives of administrative decentralization at a time when the critical state of the economies would rather call for centralized fiscal management. The strategies currently used in the Region to relieve the Fiscal Gap of Subnational Governments are reviewed in Section 4, including changes in the existing systems of revenue sharing, reassignment of public service functions, pricing adjustments, deregulation or privatization of certain services, technical assistance, and re-structuring of institutional credit. Transfers of some type appear inevitable in most countries, but there is also a need to offset the present trend of increasing reliance on top-down grants. The report's Main Findings and Recommendations are organized in Section 5 according to three main groups of concerns: those more frequently voiced by our LAC Borrowers, (Country Concerns); those stemming from economic and fiscal principles (Theoretical Concerns); and those emerging from the Dank's lending activities (Operational Concerns). For easy reference of the reader, the contents of each section have been condensed in the concluding Summing-Up paragraphs. The text is accompanied by three case studies, which illustrate achievements and shortcomings of experimental lending operations in Argentina, Brazil, and Mexico. Some discussions of a more general character (e.g., on decentralization, revenue sharing, etc.) reflect current thinking on fiscal problems and are condensed in form of inserts (boxes). To support some of the points made in the text, Betty Alvarado has carried out the necessary elaboration of the statistics available from the IMF data base, including comparisons between LAC countries and industrial market economies. - Wi - FINANCING URBAN SRVICES IN LATIN AMERICAt Spatial Distribution Issues TABLE OF CONTENTS Table References ABSTRACT SUMMARY OF THE REPORT Box 0.1 I. RISING DEMAND FOR URBAN SERVICES 1.1 The Region's Urbanization Process 1.2 Population Distribution Table 1.2 1.3 Population Movements 1.4 Urban Hierarchies Table 1.4 1.5 Urban Services Defined 1.6 Types of Urban Services 1.7 Assignment of Public Service Functions Among Levels of Goverriment Table 1.7 1.8 Diversity of Urban Service Demaid 1.9 Increasing Servici Standards Table 1.9 1.10 Variety of Institutional Responses Table 1.10 1.11 Reasons for Public Intervention 1.12 Political Overtones 1.13 Fiscal Impact of Urban Services Provision Table 1.13 1.14 Urban Service Finance as Policy Concern Table 1.14 1.15 Positive Features of Urbanization Table 1.15 1.16 Resource Mobilization Potential 1.17 The New 'Growth Management' Approach 1.18 Summing Up II. CURRENT CONSTRAINTS TO URBAN EXPENDITURES 2.1 The Debt Crisis Tables 2.1 (a, b, c & d) 2.2 Domestic Inflation Table 2.2 2.3 External and Domestic Borrowing Table 2.3 (note) 2.4 Expenditure Effects of the Recession 2.5 Progressive Budgetary Displacement Tables 1.13, 2.5 (a, & b) and 3.10b 2.6 Revenue Effects of the Recession 2.7 Centralized Economic Policy and Political Democratization 2.8 Growing Fiscal Dependence of Subnational Authorities 2.9 Planning by Negotiations 2.10 Political Use of Negotiated Transfers 2.11 Erosion of Political Responsibility 2.12 Summing Up III. ARE LAC COVERNMENTS TRULY DECENTRALIZING? 3.1 Structure and Role of the Public Sector Table 3.1 3.2 Politics of Intergovernmental Relations 3.3 Devolutive rather than Aggregative Governance 3.4 Parastatal Sector 3.5 Management of Public Enterprises - iv - 3.6 Government Structures 3.7 LAC Countries vs. Industrial Market Economies 3.8 Relative Urbanization and Industrialization Table 1.2 3.9 Public Sector Dominance? Table 3.9 (a) & (b) 3.10 Are LAC Countries Centralized? Tables 3.10 (a & b) and 3.10 (c & d) in Annex 3.11 Are LAC Governments Welfare Minded? Tables 3.11 (a, b & c) 3.12 Inter-Regional Disparities 3.13 Intra-Regional Disparities 3.14 Regional. Equalization Policies 3.15 Spatially Uniform Development? 3.16 Top-Down Transfers 3.17 The New Federalism 3.18 Rethorics of Decentralization 3.19 Ambiguities of Decentralization 3.20 Fiscal Reform Initiatives Boxes 1 & 2 3.21 Summing Up IV. THE FISCAL GAP OF LAC SUBNATIONAL GOVERNMENTS 4.1 Reasons of Fiscal Imbalance 4.2 Deficitary Trends Table 4.2 4.3 Revenue Side Factors Table 4.3 (a b) 4.4 Main Revenue Sources Tables 4.4 (a), and 4.4 (b, c, d, &e) in Annex 4.5 Taxes on Goods and Services 4.7 Taxes on Property Table 4.8 (a b) 4.9 Other Taxes 4.10 Non-Tax Revenues 4.11 Intergovernment Transfers and Revenue Sharing Tables 4.12 (a & b) 4.12 Dependence on Transfers Box 3 Ca & b) 4.13 Borrowing (Mexico, Brazil, Colombia Cases) Table 4.14 and Box 4 4.15 Expenditures Side Factors 4.16 Subnational Participation in Public Expenditures Table 4.16 (a) & (b) 4.17 The Composition of Subnational Expenditures Tables 4,17 (a), and 4.17 (b, c, d, & e) in Annex 4.18 Constraints to Reducing Expenditures Table 4.15 (a, b, c d) in Annex 4.19 Spatial Cost Differentials 4.20 The Fiscal Gap of LAC Subnational Governments Table 4.20 (a b) 4.21 Current Strategies to Relieve Fiscal Stress 4.22 Changing the Scope and Sc4le of Expenditures 4.23 Restructuring the Territorial Base of Jurisdictions 4.24 Adjusting the Level of Fees and Charges 4.25 Implementing Tax Reforms 4.26 Relying on Intergovernment Transfers Box 5 4.27 The Use 4i Transfers Boxes 6 & 7 4.28 Stabilization Objectives 4.29 Allocation Objectives 4.30 Distribution Objectives 4.31 The Design of a Transfer System Boxes 8 & 9 4.32 Summing Up -v - V. CONCLUSIONS 5.1 General Box 0.1 A. Issues 5.2 Country Concerns 5.4 Theoretical Concerns 5.6 Operational Concerns B. Recommendations Country Approach 5.8 The Territorial Distribution of Population and Economic Activities Table 1.2 5.9 Intermediate Cities Programs 5.10 Small Towns Programs Box 10 5.11 Large Cities Programs 5.12 Reviewing Fiscal Assignment Table 1.9 5.14 Containing Central Government Transfers 5.15 Decentralizing Fiscal Authority 5.16 Improving the Existing Revenue Sharing Systems 5.17 Improving Subnational Administrations 5.13 Substituting Borrowing for Grant-Transfers 5.19 Assessing Local Capacity and Willingness to Pay 5.20 Containing the Growth of Fiscal Deficits Theoretical Approach 5.21 Intersectoral Allocation of Resources 5.23 Spatial Allocation of Resources and Factor Mobility 5.25 Equity and Efficiency in Spatial Policy 5.26 Fiscal Equity 5.27 Approaches to Inter-Area Equalization 5.28 Delegating Some Distributive Activities 5.29 A Regional Perspective on Investment Distribution 5.30 Direct or Indirect Cost Recovery? 5.31 Should the Cost of Urban Services be Fully Recovered? 5.32 Setting Targets of Overall Fiscal Performance Operational Approach 5.32 Revising the Traditional Bank Approach to Urban Lending 5.33 The Conventional Cycle of Project Development 5.34 Project Specific or Poverty Alleviation Objectives 5.35 The Concern with Project Replicability 5.36 The Volume of Urban Development Lending 5.37 Leveraging the Impact of External Funding 5.38 The Concern with Counterpart Funding 5.39 Non Additionality of Bank Lending 5.40 A Fiscal Perspective on City Size 5.41 Selecting Target Municipalities 5.42 Should the Worst Come First? 5.43 A Gradualistic Strategy 5.44 Inducing Resource Mobilization with Matching Grants 5.45 Developing Local Planning Capabilities Boxes 10, 11, 12 5.46 The New Lending Approach: Country Dialogue, Transfer of Responsibilities, Financial Intermediation; Alternative Lending Tools; Research Focus - vi LIST OF BOXES Box Para. Box 0.1: Policy Concerns in Urban Sector Lending Box 1: Arguments on Decentralization 3.20 Box 2: Brazil: A Recent Case of Fiscal Decentralization 3.20 Box 3: Case Study: Subnational Finances in the Northeast Region of Brazil 4.12 Box 4: Institutional Credit in Brazil, Colombia, and Mexico 4.13 Box 5: Types of Transfers 4.26 Box 6: Arguments on Transfers 4.27 Box 7: Intergovernment Transfers by Objectives 4.27 Box 8: Revenue Sharing by Formula 4.31 Box 9: Revenue Sharing Criteria in Argentina, Brazil, 4.31 and Mexico Box 10: Brazil: Parana' Market Towns Improvement 5.10, 5.47 Project (Loan 2343-BR) Box 11: Argentina: Municipal Development Project 5.47 (Loan 2920-AR) Box 12: Mexico: Low Income Housing Project 5.47 (Loan 2612-MEX) STATISTICAL TABLES A complete list of Tables is given with the accompanying Statistical Annex. Tables have been numbered according to the paragraphs making a first reference to them. Only a few general Tables have been incorporated in the report together with the regional summaries. The latter have been derived as weighted averages of individual country data (on the basis of GDP or population). For country specific reference, one must consult the corresponding complete table in the Annex. A topical Index is suppLied at the end of the report. - vii - FINANCING URBAN SERVICES IN LATIN AMERICA: Spatial Distribution Issues SuMary of the Report i. Urban Services are the Man -upiness of Subatioul 2oyernments. Municipal or Regional governments would not exist if central authorities did not recognize that certain public services are better organized and provided at the local level. Most of these services are city-based, but they benefit also the rural owellers of territorial jurisdictions. Because all members of the communities concerned are potential consumers of urban services and because their availability also involves brtader concerns of national welfare, they are financed through a mixture of user charges, local taxes, revenue sharing, inter-governmental transfers and institutional borrowing. The budgets managed by subnational governmerts are therefore composite and can be seen as those of "social enterprises" producing and distributing services for the resident clientele of certain geographical areas. They are responsible to provide these services not "a-la-carte", but in whole "packages" that must satisfy diverse needs, preferences and income groups. These "enterprises" may be quite numerous within the same country, and will differ from one another not only in the range and type of functions they perform, but also in their capacities to deliver services and in the fiscal efforts they devote to recovering their costs. And, because customers move from one jurisdiction to another, subnational governments are engaged in some form of competition. This paper discusses how the regulatory frameworks of LAC governments and the lending practices of international donors do influence the patterns of this competition and the comparative performance of subnational authorities. ii. Large and Constant Aggregate Numbers. Government finance statistics of Latin American countries show that the activities of these regional and local "enterprises" account for about 20 percent of total government outlays, the equivalent of eight percent of GDP. Moreover, observing budgetary trends over time, one notices that the expenditures of subnational authorities did not contract measurably, in proportion to GDP, even during the 1980's crisis, when the overall volume of fiscal resources declined in almost all countries. The declining revenues of local and regional authorities were buttressed by the central governments through incremental transfers, subsidies and deficit absorptions. Public credit also played a role, rising to an all time high level of 15 perent in 1984. In order to sustain the volume of subnational expenditures during the recession, the central governments had to shift resources from the productive to the social sectors, from "tradeables" to "non-tradeables". Ad-hoc allocations were introduced to supplement the mandatory systems of revenue sharing, opening contentious issues of inter-jurisdictional equity. - viii - Nevertheless, most subnational governments of the Region are virtually bankrupt, contributing a major share (about a th:Lrd) of the public sector deficits. This gap can be as large, if not larger, than that of all national public enterprises combined, reaching in some countries up to five percent of GDP. iii. Fiscal Politics Explanations. Various factors can be quoted to explain the above. Eirst, strong budgetary pressures arise from the sustained pace of urbanization of the Latin American countries. In fact, the service needs and expectations of the cities' dwellers grow much faster than their revenues. In per capita terms, the revenues of the subnational governments continue to decline despite the large subsidies they receive. is is due tj some extent to the poor fiscal effort of local administrators by so to the low taxable base of their cities and to the progressive erosion ox iersonal incomes. Second, Latin America-a countries are plagued with territorial development disparities, which foster claims for central government measures of interregional equalization. Thir , the fisc-il politics of Latin American countries continue to be dominated by a tradition of governance which vests all powers and responsibilities in the central authorities and explains the prevalen,t "top-down" approach to the financing of urban services. Fourth, the fiscal crisis occurred in most count.-ies at the very incipient stages of a process of democratization, giving clout to the new voters' claims for financial assistance, especially in the larger metropolitan areas where greater muscle could be mustered. iv. A Relative Perspective on Developing Countries. The explanations offered above are rooted in the fiscal politics of the Latin American countries. However, a different perspective can be gained by comparing their structure with that of the industrial market economies. While their per capita income is much lower, many LAC countries are nearly as populated, urbanized, and industrialized as the developed countries. But in most of them, the share of the economy that is controlled by the public sector is generally much smaller (by more than ten percent of GDP). Furthermore, the proportion of expenditures and revenues corresponding to their subnational gr -:nments is far below that prevailing in Part One countries. Nor can LAC countries be said to be especially welfare minded, because e.e share of public resources they devote to social services is consistently lesser than that of the "capitalistic" industrialized economies, where the total amount of related subsidies and other transfers is also proportionally much larger. The governments of LAC countries are also more involved in concessionary credit to subnational governments than are those of the developed world, where the lending function is performed according to commercial criteria by more buoyant financial markets. -ix - Significantly, the subnational authorities of LAC countries devote to capital improvements a proportion of resources that is three times larger than that of the market economies. V. Position. This paper argues that the territorial distribution of intergovernmental transfers and the pattern of institutional credit should be more dir'ctly linked to th% contributions different communities can make to meeting che social costs of urban services and infrastructure. Presently, IAC governments are hard pressed to reduce long standing disparities in living conditions between regions, between urban and rural areas, between large and small cities. They try to do so with fiscal measures, rationalizing the existing s s of revenue sharing or reassigning the responsibility for specific services between different levels of government. They also try - with little success - to promote a different pattern of territorial development through costly programs of direct investment or "spatial" incentives. The paper suggests that both the horizontal distribution of transfers and the vertical reassignment of responsibilities can be improved without promoting an unwanted expansion of central government outlays. This could be achieved by delegating to regional authorities the assignment of transfers to municipal governments, facilitating their access to credit and devolving to them some independent source of revenue. It also suggests that granting to the subnational governments some borrowing autonomy and greater authority to set and collect charges could promote initiatives of resource mobilization and foster more responsible and self-reliant local administrations and that the consumers' willingness to pay for services received can only be developed through their direct participation in local expenditure decisions. vi. The Ouest of Horizontal Equalization. The extent to which territorial equalization of living standards can be pursued in LAC countries is presently a matter of resource availability. However, the level of improvement that can be effectively achieved in different areas depends also on the approach adopted in the spatial distribution of the fiscal resources available. The most current approach is one of "basic needs", which invariably leads to staggering estimates of capital requirements, no matter how modest a service standard is assumed. It also tends to scatter resources in areas of lesser development potential or to concentrate them in areas of poor fiscal performance. In turn, a "fiscal capacity" approach would seem more appropriate to a situation of resource constraints in that it rations the available capital according to the individual capabilities of jurisdictions to contribute to the payment of their own services. - x - Neither approach, however, can achieve equalization of living standards between "rich" and "poor" areas. It must be recognized that the idea of uniform territorial development is not an attainable nor a desirable goal, and should not be confused with the stated purpose of public policy, which is to maximize welfare for the national population as a whole. In fact, the pursuit of spatial equalization may increase significantly the average unit cost of the services produced (thereby lessening overall output) because of regional cost differentials and scale diseconomies. vii. Focusing on Fiscal Effort Rather Than Basic Needs. As with the basic-needs approach, a straight fiscal capacity approach to revenue 5h&ring would lead to favoring areas where presumably most people are poor, and could discourage fiscal effort in area of potentially greater buoyancy. For this reason, one should focus on the extent to which different communities effectively tap their own taxable base introducing some measure of reward for fiscal efforts. Much of the current controversies over the revenue sharing systems used in LAC countries stems from the fact that their present design blurs together conflicting issues of distributional equity, allocative efficiency and fiscal neutrality. Therefore, the first step to improve the situation would be to clarify the objectives of intergovernment transfers. If the main purpose of the government is to achieve a better distribution of personal income, it should be recognized that transfering more funds to "needy" jurisdiction is only a coarse approximation to this objective: not all residents are poor in the ill-serviced cities, nor are all rich in those which are comparatively better served. More direct measures would be needed to improve the earning power of the poor. If the main purpose is simply to even out the burden of self-financing, achieving greater equity in the fiscal treatment of different jurisdictions, the territorial distribution of resources should also take into account the comparative ability of each community to raise resources and their performance in this respect. This implies acquiescence with the fact that richer communities may endow themselves with better services as long as they choose to pay for them, and that the revenue sharing system exists only to ensure that some socially acceptable level can be provided everywhere. One must recognize that, ir the short run, area disparities within developing countries can be lessened only through the mobility of labor and that households, as well as firms have very good reasons to move to the larger cities. viii. Ouestionable Investment Policies. The governments of LAC countries try to correct regional disparities not only through generic transfers, but also through direct investments in local infrastructure and services. This might improve local living conditions (and possibly marginal income) for those who remain in the benefiting jurisdictions but is unlikely to curtail the inflow of migrants to the larger cities or to modify substantially the location behavior of firms and households. - xi - Similarly, agglomeration economies and the comparative advantage of more attractive areas defeat the use of spatially determined credit or fiscal incentives and the introduction of regulatory constraints to divert productive enterprises towards areas of the country where profitability is less apparent. These public policies are more likely to be wasteful than effective. Nevertheless, because of their concerns with the uneven territorial distribution of their population and economic activities, the governments of LAC countries frequently seek external assistance for special programs of capital improvement in cities, which are singled out according to size categories (large, medium or small) on the assumption that a different pattern of spatial development would result. More often than not, this approach leads to special treatment of individual urban centers, isolated from the context of fiscal, financial and economic flows that link them to neighboring cities or to their rural hinterland. Programs of this kind should be reassessed not merely in view of their urban development objectives but in the broader framework of national stabilization and economic recovery policies. ix. The Additionality of Windfall Project Loans. From an operational point of view, the report recommends revision of the traditional city-specific approach to urban projects in LAC countries. This approach tends to concentrate an occasional windfall of investment funds on particular jurisdictions distorting the regular patterns of domestic resource allocation and revenue sharing. Given tne scant resource generally available to the urban sector, these "spot-lending" operations warrant a concern with their equity, efficiency and replicability on a larger scale. They also raise the critical question whether or not the borrowed funds represent an increase in aggregate public sector debt and whether the central government would not be better off by directing credit merely to those local communities which can raise incremental resources to amortize at least the external component of an urban investment loan. In practice, external support should not be directed to those jurisdictions which present the most acute needs, but rather to those which show greater creditworthiness, implementation capacity and willingness to undertake reforms. The needs of the weaker communities would be better addressed through appropriate targeting of the mandatory revenue sharing. x. Can the Cost of Urban Services be Recovered Directly from their Users? By and large the tendency has been to repeat for Latin American cities formulae which are fashionable in Part One countries: privatization wherever feasible, establishment of single purpose, semi-autonomous enterprises, and better recovery of costs through tariffs, betterment charges and reassessment of property values. For these "get-the-prices-right" recipes to work out at the local level, one must assume that the distribution of personal income is fairly even, or that sophisticated systems of cross-subsidization can be introduced and that appropriate distributional trade-offs take place at the national level. One must assume also that both the traditional and the new consumers of urban services - xii - are capable and willing to pay more than they do at present and that some stability is already achieved for the economy and the financial markets. In the present context of Latin American countries the feasibility of full cost recovery through direct charges sounds like a rhetorical question. The conventional approaches to the pricing of transport, energy, water supply, sanitation and telecommunications are difficult to extend to the whole range of urban services. Many urban services involve externalities and complex issues of income distribution which justify some form of financial assistance by the central government to ensure a basic provisian even to communities or social groups which cannot pay back. If one were to apply literally the fiscal principle of exclusion, few of these services would fall in the category of "private goods" and be conducive to the application of marginal pricing. In fact most of them are neither "private" nor "public" in a strict sense, but fall between the two categories. Mentioning that user charges or property taxes are not sufficient to finance all urban services is not to suggest further growth of intergovernment transfers: this would quickly dissuade all local administrations from applying any form of direct imposition. The point made here is that merely focusing on isolated price adjustments does not address more systemic issues of local finance. What really matters in LAC cities is not whether financial discipline takes the form of direct or indirect levies but that the local community be induced to support a proposed "package" of services through a fiscal effort of some kind. The real issue is that of targeting financial assistance to jurisdictions in an equitable manner, according to equitable and defined criteria, and of ensuring that transferred resources are used effectively at the local level. In this respect it might be more expedient to set overall targets for the fiscal performance of local administrations as a condition for further transfers and to use credit or matching grants as mechanisms of reward for incremental efforts. This more comprehensive approach would leave a degree of freedom for the local administrations to decide which services to provide and how best to raise their revenues. It would also encourage the supply of those services which are most wanted by the local constituencies rather than promote expansion of those for which direct cost recovery is most feasible and earmarked funding available. One can argue also that the local authorities are more likely to raise payment for services they provide with the concurrence of consumers than for services decided upon at higher levels. xi. Regionalizing the Assignment of Fiscal Transfers. The report suggests that some gains could be achieved in both the cost effectiveness and the fiscal equity of transfers if the central governments were simply to allocate resources on a regional basis, leaving to intermediate authorities their further assignment to local jurisdictions. Breaking away from a tradition of centrally earmarked capital grants, such "regionalized" pattern of assistance has allocative and institutional implications. However, it would be consistent with the current claims of fiscal, rather than merely administrative decentralization. - xiii - Partisans of decentralization argue that only by reducing the present reliance on central government prescriptions one could achieve a more responsive pattern of service delivery. However, what is heralded as "decentralization", in many countries consists only in the proliferation of regional branches of sector ministries, or in the creation of new administrative jurisdictions, which produce only further budgetary fragmentation, higher current costs and little improvement in the effective delivery of urban services. In most Latin American countries, the regional governments function as mere pass-through agents for centrally directed transfers, but there quite a few cases in which they complement these flows with revenue sharing schemes of their own. Financial assistance seems to operate on a better informed basis, when coordinated by regional authorities, because these authorities are politically more responsive to their own jurisdictions and possess a comparative knowledge of their development potentials and pockets of poverty. The real question, then, is whether the regional authorities could become instrumental in gearing the allocation of grants to some combination of local needs and fiscal effort and in improving the local governments' access to institutional credit in a manner that helps containing the growth of budgetary deficits. xii. An Intermediate Step towards Decentralization: Operational Implications. The above considerations suggest a possible evolution of urban service finance in LAC countries, which would vest increasing planning and management responsibilities in the regional authorities. This approach is particularly suited for the larger countries, which already have a federal or quasi-federal organization, but could be applied also in smaller countries, which have some form of regional administration. The State or Provincial Banks of many LAC countries combine the function of channeling the flow of intergovernment revenue sharing with that of making occasional loans to municipalities. In this respect, they can link the repayment of credit with the secular flow of top down grants, acting as financial intermediaries for subloans. The local administrations would have to pledge a portion of their entitlements to mandatory transfers as collateral for urban infrastructure loans (a practice already in use in many countries). The crucial element in the suggested shift of responsibilities from central to subnational authorities is the agreement of the former on the criteria for the horizontal allocation of both the credit funds and the intergovernmental grants. This involves a commitment to revising inconsistent regulations and to substituting long term subnational borrowing for incremental revenue sharing. These are systemic issues which cannot be dealt with on a city-by-city basis and may require adjustment of national fiscal policies. This approach requires also the definition of generalized, minimal criteria for the preparation and integration of local investment plans, because the responsibilities of project identification, design and implementation, conventionally carried out at the center, would be delegated to subnational authorities, which will need to receive appropriate training and technical assistance. Tailored programs to improve local capabilities for fiscal - xiv - management and capital improvement budgeting thus become essential components of the urban development programs. The regional authorities should develop their own institutional capabilities for efficient financial intermediation and for the planning and supervision of programs involving numerous cities and projects at the same time. Since expenditure decisions would be brought closer to the beneficiaries, one would require also some tightening of budgetary and auditing controls at both the regional and the local levels. xiii. A Gradualistic Strategy for Urban Lending. The above considerations suggest adoption of a gradualistic strategy, based on the establishment of a line of credit for municipalities, in the form of a revolving fund managed by regional authorities. Subloans would be assigned first to those communities which show greater potential for fiscal improvement and progressively extended to other jurisdictions which, although initially reluctant to assume the corresponding obligations, may be induced to do so by the demonstration effects of the use of borrowing in more enterprising communities. Over time, it would become possible to restrict grant funding to the truly needy jurisdictions, while borrowing is emphasized in those with a more buoyant economic base. The proposed approach implies also a different perspective on city size. As long as population size reflects economic opportunities, the larger cities would receive more credit funds but less grants per capita not because their expenditure needs are less visible than those of medium size or smaller cities, but because their greater capacity should be better exploited to raise revenues for self-financing. One positive feature of this approach is that it may support the formulation of an integrated policy of urban development: at the national level, in implemerting a broad redistribution of fiscal resources among regions; at the regional level, in considering the comparative performance of different cities as parts of an interdependent regional system. xtv. Conclusion. Reviewing the World Bank experience of urban lending in LAC countries, one can see a progressive move from discrete project loans toward comprehensive operations, which combine a concern with the replicability of initiatives with increasing attention to the regular flows of intergovernmental transfers. This trend shows also a sharpening focus on the macro-economic issues involved in financing urban infrastructure and on the need to leverage the impact of Bank lending with some policy revisions on the use of domestic resources. Particularly in the fast urbanizing LAC countries, the economic and fiscal contribution of cities are an important component of the adjustment process. Urban development lending should therefore be based on a review of those stabilization, allocation and distribution policies of the central governments which directly affect the performance of cities. The Bank needs to clarify its approach to this sector of investment, developing with LAC Borrowers - xv- a common understanding of the relevant operational issues, of the impacts of the spatial distribution of intergovernmental transfers and of the effects of subsidies for infrastructure, housing and urban services. This paper outlines how external lending could be linked to the national systems of revenue sharing. Improvement of the latter is seen as a fruitful and productive area for policy dialogue with LAC Borrowers upon which a n,w rationale for external assistance could be developed. The main objective should be to ensure that at least a minimum of urban services can be provided in all communities, while sustaining local motivations toward fiscal effort. BOK 0-1: -MICY ODNWRNS IN URAN SECTOR LENDIN Cowtry Concarns Thearetical Concerns Operaticeal Comes (Related to Fiscal Politics) (Related to Economic/Fiscal Theory) (Related to Lending) Territorial Distribution of Population and Factor Mobility (Labor, Capital. Technology) Impact of Bank Lending on Overall Urban System Economic Activities (Voting with the Feet, Mobility of Enterprises) Size of Very Large Cities Economies/Diseconomles of Agglomeration Large Cities Absorb Most Credit bgiected Rural Areas and Towns internal Linkages of Settlement Systems Credituorthiness Independent of City Size Horizontal Equity (Neds) Fiscal Equity (Effort/Capacity) Selecting Borrowing Communities Deterioration of Levels of Service Cost-Effectiveness of Expenditures Replicability of Lending Operations Spatially "Unif.ormn Development? Area Comparative Advantage Strategic Use of Credit Sectoral Coordination Allocative Efficiency (Intersectoral) Managing Multi-Sectoral Loans Fiscal Assignment (Vertical Balance) Externalities/Econamies of Scale Shortage of Counterpart Funds Development & Poverty Alleviation Functions Growth of Subnational Fiscal Gap Budgetary Predictability of Linking Lending with Revenue Sharing Revenues/Expenditures/Sub sidles Short-term Liquidity Constraints Growth of Domestic Deficit Credit in Lieu of Incremental Grants Solvency of Institutions Impact on Nhtional Financial Markets Suitable Financial Intermediaries Willingness/Capacity to Pay Resource Mobilization/Efficient Pricing Cost Recovery (Direct/Indirect) Limited Local Capabilities of Responsiveness to Local Ngeds/Preferences Curbing Growth of Parastatals Management/Planning/Implamentation Local Autonomy (Participatory Planning) Accountability for Use of Resources Decentralizing Project Development Growth of the Informal Sector Promotion of Private Sector Initiative Revising the Regulatory Framewrk Fragmentation of the Urban Sector Urban Investment Policy/Tax Reform Nature of Conditionality Limited Factual Knowledge Theoretical/Operational Questions Sector Research Needs FINANCING URBAN SERVICES IN LATIN AMERICA: Spatial Distribution Issues I. THE RISING DEMAND FOR URBAN SERVICES 1.01 The Region's Urbanization Process. The proportion of total population living in cities has been rising at a markedly faster pace in Latin America than it has in other regions of the world. In 1960, less than half of the Region's population (49 percent) was urbanized. By 1980, the proportion had reached about two thirds (65 percent) and UN projections suggest an overall urbanization ratio of 77 percent by the end of this century.11 In year 2025 the population of LAC cities would attain an average ratio of 83 percent, matching the values already achieved today in Venezuela and in the three countries of the Southern Cone (Argentina, Chile and Uruguay). In most other LAC countries, the overall structure of the population is young and the fertility rates quite high in both the cities and the countryside, resulting in sustained rates of demographic increase. Especially prominent are the positions of Brazil in Tropical South America and Mexico in Central America: the combined urban populations of these two countries alone (128 million) represent more than half the regional total (54 percent) and they are still growing at average rates of over 3.2 percent per annum. 1.02 Population Distribution. In LAC countries, the territorial distribution of the population is characterized by two main features: the polarization of demographic settlement on a number of large metropolitan areas (see Table 1.2) and, at the same time, the dispersion of population in tens of thousands of smaller cities and towns, scattered over vast territories. At the root of these settlement patterns are the inappropriate terms of trade between the urban and the rural sectors (an enduring legacy of the colonial past) and the corresponding, sharp income differentials which drove millions of poor immigrants to the cities. To these factors, one must add the persistence of adverse conditions in the countryside which is plagued with poor living standards, generalized unemployment, land tenure problems and recurrent natural disasters.IF 1/ United Nations Center for Human Settlements (HABITAT): Global Report on Human Settlements, Oxford University Press, New York, 1987. 2/ Severe droughts, floods, hurricanes, landslides, volcanic eruptions or earthquakes have hit Argentina and Brazil (1985), Chile (1939 and 1986), Colombia (1984 and 1986), the Dominican Republic (1979 and 1982), Ecuador (1987), El Salvador (1987), Haiti (1982), Guatemala (1976), Mexico and Nicaragua (1973) and Peru (1970 and 1983). - 2 - Table 1.2: DEGREE OF URBAN POLARIZATION IN LAC COUNTRIES (1985) Percentage of Urban Population Cities With More Than In Cities Over Total Larger Than In The 2 Million 500,00 Countries Population 500,000 People Largest City People People Larger Countries 73% 49% 26% 14 37 Colombia 67% 60% 23% 1 5 Argentina 85% 56% 42% 1 5 Brazil 73% 49% 16% 6 14 Mexico 70% 48% 31% 3 7 Peru 67% 48% 42% 1 2 Chile 84% 41% 41% 1 1 Venezuela 87% 37% 25% 1 3 Smaller Countries 45% 46% 39% 15 Dom. Republic 56% 71% 51% - 2 Jamaica 54% 66% 66% - 1 Paraguay 44% 66% 66% - 1 Costa Rica 50% 57% 75% - 1 Ecuador 52% 49% 20% - 2 Uruguay 85% 47% 47% - 1 Nicaragua 56% 42% 42% - 1 Panama 52% 41% 41% - 1 Honduras 40% 38% 38% - 1 Guatemala 40% 38% 38% - 1 Bolivia 48% 33% 33% - 1 Haiti 27% 31% 31% - 1 El Salvador 39% 24% 24% - 1 19 LAC Countries 69% 49% 27% 14 52 19 Market Economiesa/ 75% 49% 18% 32 163 Source: UNITED NATIONS: "The Prospects of World Urbanization", New York 1987 UNCHS - HABITAT: "Global Report on Human Settlements 1986", Oxford United Press a/ These include: Canada and United States; Austria, Belgium, Denmark, Finland, France, Germany (F), Ireland, Italy, Netherlands, Norway, Spain, Sweden, Switzerland, and United Kingdom; Japan; Australia and New Zealand. - 3 - 1.03 Population Movements. While still a significant component, the rural- urban exodus is no longer the major factor of the continued growth of Latin American cities. By now, the main force driving their expansion is the natural increase of large numbers of residents and their mobility from one urban area to another. Cities continue to grow even in temperate South America, where the pace of natural increase has long subsided, because people keep moving from the metropolitan core areas to the peripheral communities, from the smaller to the larger cities and from the rural hamlets to the nearby towns. A trend towards more localized movements is gradually replacing the historical patterns of massive inter-regional migrations, as can be observed also in Brazil, Mexico and Colombia. Therefore, the incremental demand for urban services is likely to remain high in all countries over the next decades particularly at the fringes of metropolitan areas and in the medium size cities. 1.04 Urban Hierarchies. Forty percent of LAC urban dwellers live now in twenty five metropolitan areas with more than one million inhabitants; a fourth of them are found in "megacities" with ten million people or more (Mexico City, Sao Paulo, Rio de Janeiro and Buenos Aires), of which the first two will be the world-largest urban agglomerations by year 2000.1i While the prominence of these megacities is assured by the combination of their present size with relatively high growth rates, several other cities will graduate into the million rank. However, the share of total urban population accounted for by these 'secondary' metropolises seems relatively stable over time, fluctuating between 15 and 18 percent of the total urban population. Settlements below this size (intermediate cities and small towns) are far too numerous to be counted. As an indication, the number of Latin American centers that are important enough to be seats of local government is now in the order of 13,000. (See Table 1.4). 1.05 Urban Services Defined. This expression has been used throughout the report in a broad sense, to identify those services which have a discrete spatial incidence within an urbanized area (the benefit region) where they are provided to the general public. Within this area, urban services (such as sidewalks or street lighting) are virtually available to all members of the resident community and their consumption is neither rival, nor exclusive. For this reason, consumers need not reveal their individual preferences through price competition and tend to set out as free riders. Thus, the correct amount and appropriate kind of urban services to be produced in a given jurisdiction does not depend on horizontal aggregates of effective demand, as is the case for private consumption goods, but on budgetary decisions taken for the end users by their ,2/ UNCHS Projections: Mexico City: 25.8 million; Sao Paulo: 24 million; Rio de Janeiro: 13.3 million; Buenos Aires: 13.2 million. Cities over one million inhabitants will pass from the present number of 25 to 46. - 4 - Table 1.4: NUMBER OF SUBNATIONAL GOVERNMENTS IN LAC COUNTRIES (1987) Countries Regional Local Larger Countries Argentina 22 Provinces, 1 National Territory 1617 Municipalities Bolivia 9 Departments Prefectures 9 Capital Municipalities & Several Others Brazil 24 States, 1 Federal District 4112 Municipalities Chile none 317 Municipalities Colombia 23 Departments, 1 Capital District 967 Municipalities Mexico 31 States, 1 Federal District 2376 Municipalitits Peru 25 Departments, 155 Provinces 1567 District Councils Venezuela 20 States, 2 Fedoral Territories 188 Municipalities, 1 Federal District Smeller Countries Oahamas none none Barbados none none Belize none 7 Town Boards & Belize City Council Cayman Islands none none Costa Rica 4 District Councils 81 Municipalities Dominica none 2 Town Councils, 47 Village Councils Dominican Republic none 31 Municipal Districts, 95 Municipalities Ecuador 19 Provincial Councils 136 Municipal Councils El Salvador none 262 Municipalities Grenada none none Guatemala none 328 Municipalities Guyana none 6 Town Councils Haiti none 137 Communes Honduras none 281 Municipalities Jamaica none 13 Parish Councils Netherlands Ant. 5 Islands Governments none Nicaragua none 135 Municipalities Panama none 66 Municipalities Paraguay none 192 Municipalities, I Capital District St. Lucia none 1 City Council St. Vincent none 1 Town Board Suriname none none Trinidad & Tobago 7 County Councils 4 Municipalities Uruguay 19 Departmental Governments 1 Superintendency (Montevideo) Source: IMF, Government Finance Statistics Yearbook, 1987 -5- political representatives, who may or may not be elected. This is why local and regional authorities are regarded as the natural agents of urban services supply and become directly involved in their production and distribution by constructing, purchasing and operating the necessary facilities, franchising these activities, releasing vouchers to consumers, or simply by regulating the performance of private enterprises. 1.06 Types of Urban Services. The types of services which come under the purview of subnational governmentsil are no different in Latin America than elsewhere. They may be in the nature of general public services, such as the regional coordination of the investments of national sector agencies or the supervision of local government activities, the maintenance of law and order, the administration of justice, the regulation of the use of land and of private construction activities, the registration and assessment of real estate property and a variety of other administrative, fiscal, and support functions. More specifically, they usually include the provision and operation of social services, such as health care, ed&c: .-, and welfare; the control and regulation of urban traffic; the lighting :f clty streets; the collection and disposal of solid wastes; the implementation ana maintenance of public works (construction and upkeep of certain roads, street pavements, sidewalks, storm-drainage, parks, playgrounds, and public buildings). They generally comprise also services that are not "public" in a strict sense, such as the local provision of and connection to utilities (water, sewerage, town gas, electric power, telephone); the establishment and operation of general facilities such as slaughterhouses, covered markets, garbage dumps, transportation terminals. Finally, subnational budgetary provisions may also involve the supply of merit goods, such as subsidized housing for the poor or meals and bottled milk for school children, which may be provided through government because specific official policies recognize this need. In general, the subnational governments perform a limited ~/ Throughout this paper, Subnational Governments indicates Regional (State, Provincial or Departmental) Governments and Local (Municipal) Governments as distinct from National (Federal or Central) Governments. -6- role also in the promotion and regulation of economic activities like agriculture, mining, manufacturing, commerce which account for a minor share of local expenditures.Ai 1.07 AssiUpment of Public Service Functions Among Levels of Government. Legal statutes generally assign the performance of specific seivice functions to different tiers of government. In general, the logic of these assignments is that functions with larger externalities or economies of scale should be performed by higher level authorities. However, the real situations are seldom clear cut because the definitions used are often too broad and because higher level authorities are frequently called upon to supply a given public good which the lower authorities are either financially or technically unable to provide. As a result, the Drovision of many public services is shared by different levels of government and the very proportion of this sharing may vary considerably for different services from place to place even within the same country. Since the analysis of statutory responsibilities is often of little help, a better grasp of the factual situation can be derived from examining the expenditures effectively made by each level of government for specific functions. Table 1.7 provides a fair illustration of such expenditure breakdown in the case of Brazil for fiscal year 1984. As can be seen, apart from expenditures on external affairs and defense, which are obvious concerns of the central government, general services remain an important expenditure item for the subnational governments, with the bulk of public outlays for the legislative and the judiciary being carried out at the regional (state) level, while both the regional and thxe local budgets are heavily burdened by the costs of public administration. Subnational governments, especially at the municipal level appears principally involved also in the provision of social services, mostly 5/ In its publication of government finance statistics, the IMF follows COFOG (the United Nations Classification of the Functions of Government) which considers four basic categories: (a) General Government Services, including all activities indispensible to the existence of an organized state, which cannot be allocated to particular groups of beneficiaries; (b) Community and Sonial Services, including all services supplied to the community or the household directly; (c) Economic Services: including those associated with the efficient operation of business, and (d) Other Functions: including interest payments or underwriting of public debt as well as transfers to other levels of government, which disappear in the consolidation of accounts of the levels and bodies of government. Thus, Housing, Community Amenities, Water Supply, Sewerage, Garbage Collection, Street Lighting, and Recreational, Cultural and Religious Facilities are classified as social services, while Street Construction and Maintenance, Urban Transport, Town Gas, Electricity and Telephone Connections are classified as economic services. -7- Table 1.7: ASSIGNMEMN OF PUBLIC SERVICE FUN0TIONS AMONG LEVELS OF GOVERNMENT AS EVIDENCED BY 1984 BUDGETS IN BRAZIL General Central Regional Local Capital Other Functions Government Government Governments Governments Municplts MunicpIts VERTICAL ASSIGN4ENT General Services 100.0 41.8 47.0 11.2 4.5 6.6 Legislature 100.0 20.8 26.8 52.3 6.4 45.8 Judiciary 100.0 22.6 75.3 2.1 1.5 6.2 External Affairs 100.0 100.0 - - - - Administration & Planning 100.0 36.6 51.4 12.1 5.2 5.8 Defense & Public Order 100.0 63.2 36.5 .3 .1 .2 Social Services 100.0 38.2 45.5 16.3 9.5 6.8 Education & Cultural Affairs 100.0 34.6 53.7 11.7 6.9 4.7 Health & Sanitation 100.0 24.6 58.8 16.6 ;0.7 5.8 Social Security & Welfare 100.0 55.8 36.2 7.9 4.8 3.1 Housing & Community Amenities 100.0 8.9 16.5 18.9 39.8 34.8 Economic Services 100.0 60.4 28.0 3.8 2.2 1.6 Regional Development 100.0 75.6 24.2 .2 .1 .1 Agriculture 100.0 67.3 31.5 1.2 .3 .9 Industry, Commerce & Services 100.0 47.1 44.4 8.4 5.1 3.2 Energy, Mineral Resources 100.0 83.8 16.0 .2 .1 .1 Transportation 100.0 53.8 37.7 8.5 5.0 3.5 Communications 100.0 71.9 26.1 2.0 .- 2.0 INIERSECTORAL ASSIGIENT TOTAL EXPENDIURES 100.0 100.0 100.0 100.0 100.0 100.0 General Services 34.8 30.2 39.8 36.7 28.7 45.3 Legislature 2.1 .9 1.4 10.5 2.5 19.1 Judiciary 1.7 .8 3.1 .3 .5 .2 External Affairs .5 .9 - - - - Adminstration & Planning 22.6 17.1 28.2 25.6 25.6 25.7 Defense & Putlic Order 7.9 10.4 7.0 .2 .1 .3 Social Services 33.1 26.3 36.6 50.8 57.2 43.9 Education & Cultural Affairs 14.3 10.3 18.7 15.8 18.2 13.3 Health & Sanitation 4.9 2.5 7.0 7.7 9.6 5.6 Social Security & Welfare 11.2 13.0 9.8 8.3 9.8 6.7 Housing & Community Amenities 2.7 .5 1.1 19.0 19.6 18.3 Economic Services 32.0 43.5 23.4 12.5 14.2 10.7 Regional Development 9.4 14.7 5.5 .2 .2 .1 Agriculture 3.2 4.4 2.4 .4 .2 .6 Industry, Commerce & Services 1.2 1.2 1.3 1.0 1.2 .8 Energy & Mineral Resources 4.4 7.7 1.7 .1 .1 .1 Transportation 13.6 15.2 12.4 10.8 12.5 ,.1 Communications .2 .3 .1 Source: Elaboration of Data from: Secretaria de Economia e Financas: Finances do Brazil, Exerciclo de 1984, Volume XXIX, Ministerlo da Fazenda, Brazilia DF, 1986. - 8 - for education, health, sanitation, housing and other urban-based activities. Note in this respect the relative importance of these activities in state-capital municipalities (Table 1.7). The brunt of economic services is carried instead by the central government, but are also partially delegated to regional authorities, remaining only a secondary concern in the local budgets, except for road construction and maintenance. As we shall see in Section 4, this generalized picture of Brazil, is to some extent repeated in Argentina, Chile, Colombia, and Panama, for which less complete information is available on expenditures by function (See Table 4.17 a, and Annex Tables 4.17 b & c) 1.08 Diversity of Urban Service Demand. Demand for urban services is not homogeneous. It varies from country to country, among regions of the same country and across the urban hierarchy. A variety of urban conditions exists, as cities differ greatly in size and extension, from the large megalopolises like Mexico City, Sao Paulo, Rio and Buenos Aires, to the countless "urban" settlements with a few thousand people.11 The very territorial extension of the great agglomerations creates specific problems of travel time, congestion of central areas, environmental pollution, shortages of drinking water, electric power and even food supply: the lower the overall density of population, the larger the cost of maintaining and operating utility networks and service systems. Topography and Climate originate different costs and different needs: recurrent flooding and drainage problems (Buenos Aires), landslides (La Paz), earthquakes (Mexico City), air pollution (Sao Paulo), short or extended street lighting hours, variable heating or air conditioning requirements. As Populatiox Characteristics change from city to city, so do the demand and cost of providing specialized services for particular social groups (the children, the elderly, the sick, the handicapped, the poor, the illiterate, the jobless, the unskilled, the homeless). 1.09 Increasing Service Standards. As national populations become increasingly urbanized, their service expectations rise. Public social expenditures tend to increase not only in absolute terms but also per capita. This is quite evident in the 1975-1980 trends of all LAG countries up to the sudden contraction of their economies in 1981-84 (see Table 1.9). Larger cities call for capital intensive and technically sophisticated investments, which may be provided only by the public sector. Furthermore, the country's more urbanized areas exert a demonstration effect on the less urbanized: the higher average i/ The lower threshold of 'urban' centers varies from country to country, according to the statistical definition adopted. Thus, in Argentina, Bolivia, Colombia, Mexico, Panama, and Venezuela this limit is set between 1500 and 2500 inhabitants. The most inclusive quantitative definition is that of Peru, where all settlements with 100 or more occupied dwellings are considered urban. In most other LAG countries the definition embraces all localities which are seats of local administrations and may include areas of suburban or rural character. -9- incomes of metropolitan areas command services of better quality and this, in turn, changes also the expectations of smaller urban centers, regarding the levels and types of services considered appropriate. Table 1.9: GENERAL GOVERNMENT EXPENDITURES IN THE SOCIAL SECTORS a/ -In US Dollars- 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE b/ Millions 41970 43177 43999 49765 52006 54586 62046 54410 43188 40303 Variat.Index 1975=100 100 103 105 119 124 130 148 130 103 96 Dollars Per capita 180 181 180 198 202 205 228 195 151 138 Variat.Index 1975=100 100 100 100 110 112 114 126 108 84 77 MARKET ECONOMIES c1 Millions 728097 787019 868617 1039564 1201856 1403882 1410800 1459511 1496096 1188182 Variat.Index 1975=100 100 108 119 143 165 193 194 200 205 163 Dollars Per capita 1645 1672 1722 1904 1964 2006 1811 1753 1731 1484 Variat.Index 175=100 100 102 105 116 119 122 110 107 105 90 I/ Social expenditures comprise education, health, social security and welfare, housing, recreation and other coomunity services. Road construction and maintenance are not included. a/ Argentina, Colombia and Panama include central and subnational governments expenditures while only central government expenditures are reported for the other countries. S/ Australia, Austria, Denmark, Iceland, Luxembourg and Switzerland include central and subnational governments expenditures, while only central government expenditures are reported for Canada, France, FR Germany. UK and USA. Source: IMF, Government Finance Statistics Yearbook, various years. 1.10 Variety of Institutional Responses. The performance of certain tasks (e.g., garbage collection or urban transport) may be shared differently by the public and the private sector, the formal or the informal sector of urban economies. In different places, the citizens expect fulfillment of different needs by the central or the local governments or may rely to a varying extent on independently organized action and individual initiative.Zi Most urban Z/ In Lima, about 60 percent of urban services (including trade) are managed by the informal sector (de Soto), while this percentage may be as low as 10 percent in Buenos Aires. - 10 - services are typically labor-intensive: their ratio of recurrent to investment costs is high, causing high payroll costs for public administrations, especially where job-protection policies are in effect.A' Responsibilities for service provision can be budgetarily fragmented among a variety of public institutions, departmental companies and controlled enterprises: the activities of many different agenci s are not easy to coordinate and this makes it difficult to control their investment and operation costs and to utilize their manpower efficiently.21 In metropolitan areas, this picture may be further complicated by Jurisdictional subdivision.LiY These institutional differences are often rooted in historical, cultural, social and political factors, which are not easily removed. Therefore, the rising costs of urban services are not only determined by population growth or by the trends of the local and national economy; in fact, the evidence is that per capita expenditures of local governments tend to rise faster than population and GNPU (see Table 1.10). 1.11 Reasons for Public Intervention. Residents of urban areas generally interact among themselves more thar people living in rural settings. Their business and household activities engender greater cost and benefit spillovers Table 1.10: PER CAPITA GNP AND SUBNATIONAL EXPENDITURES -Variation Index, 1975=100 a/ 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE Per Capita Expenditure 100 102 110 128 142 158 156 174 168 143 GNP Per capita 100 108 117 132 154 177 197 171 141 139 Population 100 103 105 108 110 114 117 119 122 125 I/ The expenditure and GNP variation index for the individual countries is based on constant national currencies time-series while the variation index for the sample average is based on constrant US dollars. Source: IMF, Government Finance Statistics Yearbook, various years. A/ 'The Municipality of Buenos Aires employs some 85,000 people, that of Montevideo nearly 14,000 people. 2/ This point is elaborated in Para 3.4 (Parastatal Enterprises). 10/ The conurbated area of Mexico City overlaps territories comprised in 5 states, surrounding the Federal District. Greater Buenos Aires includes 50 percent of the urbanized population of Buenos Aires Province, adjoining the Federal Capital. Similar inter-jurisdictional problems (more frequently inter-municipal) occur in most large metropolitan areas. 11/ Again one must note in Table 1.09 and 1.10 the sudden reversal of trends caused by the 1981 debt crisis. - 11 - throughout (and without) the urban community in which they live. As the share of national population living in cities increases, so does the need of social adjustments for these externalities. Moreover, in an urban area the increasing density of consumers gives rise to congestion costs, as the available service capacities soon become surtaxed. The technical solutions which are required may be beyond the scope and capabilities of private initiative. These issues need to be addressed through some form of government intervention. Furthermore, since the consumption of social goods is technically open to all residents, consumers tend to behave as free-riders rather than as competitive bidders. And, because individual demand schedules tend to differ according to income and tastes, consumer preferences are not reconciled through price competition but through a political process which ultimately results in budgetary allocations. Further complications arise from the fact that end-users are not called upon to express their preference on the public decision to increase the supply of one particular service, but rather on that of a 'package' of different services. This makes it rather difficult for the non-initiated citizen to give an informed judgement on the quality and quantity of services he will receive as a consequence of his endorsement of the proposed budget.1 Public intervention may also be dictated by considerations of general welfare, to ensure that certain services (like education or health) which are considered relevant to national development are provided everywhere with the desirable quantity and quality. 12/ Normative theories of intergovernment finance have labored on these concepts for some time. The idea of horizontal equity would require that all equals be treated equally. In 1950, Buchanan suggested the concept of fiscal surplus, as the difference between the fiscal contribution of the residents of one jurisdiction and the value of the services they actually receive. Thiebout suggested that the selection people make of the place to reside (voting with the feet) represents a form of competition or revelation of preference for the public sector, which reflects the relative cost and quality of local services (assuming that different jurisdictions have an option as to the kind of services they provide). The concept of public externalities was formulated by Pigou: it recognized that the services provided to taxpayers in one region may be enjoyed by residents of another. This led to the theory of optimal regional spaces (Breton, Reyes and others) which tried to define the geographical reacn of different public services, including their externalities. Buchanan's theory of the clubs showed that, even with decreasing costs, the emergency of congestion determines an optimal scale for different public services, suggesting their subdivision into different jurisdictions. Finally Tullock argued that centralization of public services is not justified, because scale economies are relatively scarce. He also pointed out that in the public sector, which is managed by the political apparatus, consumers can only choose among packages of goods with few options. - 12 - 1.12 Political Overtones. Government involvement in the provision of urban services may also be influenced by contingent political considerations, especially in metropolitan areas and provincial capitals which represent large and visible pools of voting power. Numerous countries of Latin America have recently introduced democratic institutions and their governments are increasingly dependent on the outcome of popular elections.1-1 But even governments which are apparently less conditioned by democratic consensus show a concern with the political restiveness of urban communities, with the organized claims of urban workers for minimum wage increase or larger fringe benefits, with the street riots engendered by proposed adjustments of bus fares or utility tariffs or from the mob protests stemming from the service deficiencies or uncertain tenure of large squatter settlementc.1 1.13 Fiscal Impact of Urban Services Provision. As the subnational provision of urban services acquired increasing weight in public budgets it also became the object of growing concern by fiscal economists. Examination of public finance statistics reveals that in some countries, like Argentina, Brazil and Colombia,1- about one third of all government spending is financed through subnational government budgets (see Table 1.13) and that this type of expenditures accounts for a share of aggregate government deficit which is as large, and often larger than that of all public enterprises combined (Table 3.1). In addition, a new perspective on the performance of urban economies has gained ground, recognizing that they play an essential role in national recovery and financial adjustment. This realization shed a new light on the status of intergovernmental fiscal relations. The main objectives of the reforms under consideration are to achieve (i) better control and predictability over the volume of annual transfers to subnational governments; (ii) greater accountability of local authorities for the use of the resources transferred; (iii) improved financial planning, budgeting, and programming activities of the regional and local governments; and (iv) promotion of fiscal effort and cost awareness among subnational authorities. 13/ Argentina, Bolivia, Brazil, Ecuador, Guatemala, Honduras, Peru, Uruguayhave adopted democratic rule during the last decade. 14J/ Hernando de Soto provides a vivid account of social struggles in Lima in his book: El Otro Sendero (Editorial El Barranc., Lma 1986). 15/ By contrast, the share of subnational expenditures is very small in all other countries examined, including Mexico and Venezuela which, despite their size remain highly centralized in this respect. - 13 - Table 1.13: EXPENDITURES BY LEVEL OF GOVERNMENT -Variation Index, 1975-100 a/ 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Average LAC SAMPLE AVERAGE b/ Subnational Government 24.4 23.8 22.8 23.0 23.0 23.0 21.0 20.1 19.8 20.4 22.1 Central Government 73.6 76.2 77.2 77.0 77.0 77.0 79.0 79.9 80.2 79.6 77.9 MARKET ECONOMIES Subnational Government 26.5 26.5 25.4 25.5 25.1 25.4 24.7 24.7 23.8 24.0 25.2 Central Government 73.5 73.5 74.6 74.5 74.9 74.6 75.3 75.3 76.2 76.0 74.8 pi Total expneditures comprise expenditures and lending minus repayment. Expenditures have been consolidated eliminating transfers (interest payments, subsidies, lending-repayment) among levels of government. hl Weighted averages. Comparability over time affected by the availability of data. Source: IMF, Government Finance Statistics Yearbook, various years. 1.14 Urban Service Financing as a Policy Concern. As a result of political, fiscal and economic considerations, the financing of urban services has become a central issue in the social policy of most LAC governments. The economic recession and the stabilization measures required by the external debt crisis have imposed new constraints to urban spending but have also brought into sharper focus the need to complement adjustment policies with programs which may alleviate their impact on the masses of urban poor. These come at a time in which, almost everywhere, subnational government finances have fallen into severe deficitary conditions, after years of steady erosion in the real value of local revenues and unrelenting pressures for overspending. (See Section 2). Table 1.14: SUBNATIONAL GOVERNMENTS DEFICITS (as Percentage of General Government Deficit) Years 74 75 76 77 78 79 80 81 82 83 84 World 28 15 13 1 11 12 14 11 12 na na Industrial Countries 22 12 15 3 12 14 15 11 14 5 1 Developing Countries 47 16 4 + 6 - 4 9 10 na na LAC Countries 48 35 28 20 26 29 14 16 8 35 na Source: Staff elaboration of data derived from IMF Government Finance Statistics. Note: See Table 4.2 for country data in percentage of GDP, Table 3.1 for relation vith public enterprise sector. - 14 - 1.15 Positive Features of Urbanization. On the positive side, the progressive concentration of population in urban areas, especially in the larger cities, has contributed to the development of national human resources. It has afforded economies of scale from large sunk invescments (at least in the early stages). It has expanded domestic markets, with vast pools of demand stimulating the growth of new agricultural as well as non-agricultural activities. It has promoted specialization of labor and induced technological innovation. In fact, strong correlations can be drawn between urbanization and per capita income (.70). As can be seen in Table 1.15 the more urbanized countries in LAC tend to have higher per capita GNP. Moreover, per capita value added tends to be higher in the larger cities. 1.16 Resource Mobilization Potential. On the revenue side, it is also increasingly recognized that urban governments could make a greater contribution to domestic resource mobilization. True enough, the increase in overall taxable capacity associated with rising urbanization is not easily tapped by the central government, especially in Latin America, as the city dwellers tend to resist payment of new taxes for which they cannot readily identify local benefits. It must be recognized also that comparisons with the much higher returns from urban property and service taxes which are being reaped in some of the developed countries, could never be attained in LAC countries. This is due not only to their comparatively lower fiscal effort (which could be improved) but also to an objectively lower taxable capacity. These points will be developed in Section 4 of the report. Table 1.15: PER CAPITA INCOME AND URBANIZATION TREND -GNP in Current US Dollars- 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE GNP Per Capita g/ 1307 1406 1596 1862 2136 2373 2058 1699 1673 X Urban/Total Population 58.9 59.5 60.0 60.5 61.0 61.6 62.2 62.7 61.6 MARKET ECONOMIES GNP Per Capita 1/ 7229 7940 8811 10046 11350 12638 13153 13812 14962 % Urban/Total Population 76.9 77.0 77.1 77.2 77.3 77.4 77.6 77.7 77.8 al Weighted average. Sources: United Nations National Accounts. The World Bank, Bank of Economic and Social Data. - 15 - 1.17 The New 'Growth Management' Approach. Traditionally, Latin American governments have opposed urban concentration. This attitude often found expression in explicit policies discouraging new settlers through 'slum eradication', police repression of 'spontaneous' service or trade activities or deliberate neglect in extending public provision to 'illegal' settlements. It also motivated several LAC governments to pursue rather ineffectual schemes to retain excess labor in rural areas. At different times, Argentina, Brazil, Mexico and Venezuela undertook massive investment programs with the stated objectives of re-directing migration away from their main cities towards 'alternative' growth poles. These policies have influenced even the outlook of smaller countries where some degree of urban concentration was instead economically desirable.16 The results were often mixed, while cities continued to expand. But, as the popular understanding of the root causes of urbanization got deeper, a new 'enabling' approach made its way in official policy statements of Latin American countries and in their professional literature. Countering or diverting immigration into the large cities is no longer advocated, while attention is focusing on more pragmatic approaches to the problems of their growth. Expressions like 'demand management', 'cost-effective investments' and 'participatory planning' have now become staple. Moreover, investments in the urban infrastructure of Latin American cities and in the development of their human resources are no longer viewed as purely distributive policies, but as critical intermediate inputs for the effective functioning of urban productive activities. 1.18 Summing Up. This first section of the report has analyzed the Rising Demand for Urban Services in LAC countries, focussing on the changing spatial distritution of their population and the higher service expectations associated with rapid urbanization. Urban growth management has become a major concern of national fiscal policies. This concern is reflected in the current attempts of LAC countries to restructure the existing systems of subnational government financing, to privatize the provision of some urban services and to support the initiative and contribution of a burgeoning informal sector. It is less apparent, however, in the persisting advocacy of central government policies trying to curtail or re-direct the inflow of migrants to the larger cities. The positive correlations of the urbanization process with national development have been noted, but also the fact that new fiscal pressures for incremental public expenditures arise from the spatial concentration of demand for infrastructure and services, especially in the metropolitan areas. In theory, the relative concentration of personal wealth and income in these larger cities should afford new opportunities for fiscal Imposition. However, recovery of the social costs of urban services is proving increasingly difficult in LAC countries because of the steady erosion of household incomes. The declining ability to pay of the J./ It should be noted that his theme was prominent also in the European planning literature and practice of the 1950's and 1960's and was widely disseminated by technical assistance agencies during that period. - 16 - end-users tends to reaffirm the rationale for subsidized or free provision of urban services in order to alleviate the impacts of economic adjustment. Nevertheless, there is Increasing awareness that this approach reduces even further the governments possibilities to bridge existing service gaps and that urban investments should be strategically directed to support the cities' productive contribution to the stabilization and recovery of national economies. - 17 - II. CURRENT CONSTRAINTS TO URBAN EXPENDITURES 2.01 The Debt Crisis. The late 1970's and the early 1980's were difficult years for Latin America. In virtually all countries, the level of national savings (shown in Tables 2.1a & b as the balance of their current accounts) declined precipitously, leaving little resources to finance public investments and intergovernmental transfers. Chronic budgetary deficits set in at both the national and the subnational levels, because current expenditures could be cut only gradually and some investments in capital formation had to be continued ::.n order to sustain growth of the national economies, to contain unemployment with4-n tolerable proportions and to support exports in view of a gaping balance of payments. After years of relative stagnation, the mere contraction of public expenditures proved no longer dependable as an instrument of adjustment and strong political pressures eventually built up to stimulate the economies. While financial systems were finding it difficult to attract private savings, several Central Banks lowered the cost of borrowing for both the subnational governments and the parastatal enterprises by subsidizing (if not waiving) interest payments. As public debt continued to swell, interest payments reached a proportion of up to five percent of GDP, crowding out public programs (Tables 2.1c & d). 2.02 Domestic Inflation, which acquired extraordinary momentum in many Latin American countries, (Table 2.2) worked only to a limited extent as a hidden form of debt repudiation, because the bulk of public debts were in short term maturities. This rendered debt obligations increasingly liquid and close money substitutes, making of them an attractive medium of investment for financial institutions (especially for the money market funds, the commercial banks, and the insurance companies). Investment capital had to bear all residual pressures on disrupted financial markets and could be gotten only short-term, at prohibitively high rates. Capital flight and 'refuge' investments became preferred avenues to protect household and corporate savings. Inflation also took a heavy toll on the bases of several important local revenues, such as the property tax, and the progressive erosion of real wages made it difficult to adjust public utility prices. The consequent, progressive decline in local revenues could only partially be offset by contingent increases in top-down transfers or, occasionally, by concessionary credit. 2.03 External and Domestic Borrowing. Under the circumstances described above, official financial assistance (from IBRD, IDB, and bilateral agencies) - 18 - TabLe 2.1a: CENTRAL GOVERNMENT SAVINGS AS PERCENTAGE OF GDP P/ 0 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 ARGENTINA -5.6 -0.5 4.8 0.0 1.4 0.7 -2.0 -2.2 -4.0 -0.4 BOLIVIA 1.8 1.3 1.1 0.1 -2.0 *4.7 -3.6 NA -2.5 -28.4 BRAZIL 3.7 4.3 5.3 4j 5.3 3.4 5.9 6.0 5.4 3.0 CHILE 4.5 3.6 1.4 2.4 5.9 6.7 3.7 -2.9 *2.2 -1.0 COLOMBIA 3.4 4.4 4.3 3.5 1.1 1.5 0.6 1.9 1.2 NA COSTA R 1.6 0.6 0.9 -0.4 *2.1 -3.5 -0.8 -0.5 0.9 2.0 DOMINICAN REP 10.6 6.9 6.1 4.6 1.8 2.9 2.3 -0.1 0.9 1.8 GUATEMALA 1.8 2.1 3.6 3.2 2.u 2.0 0.9 0.9 NA NA MEXICO 0.9 0.7 1.0 2.0 2.9 3.9 0.9 -7.1 -2.6 -2.4 PANAMA 0.3 -2.8 0.2 -0.9 -3.8 0.4 -1.0 -4.6 -0.4 *2.6 PARAGUAY 2.1 2.4 3.6 4.4 4.3 3.4 2.2 1.8 1.3 1.1 URUGUAY -3.7 -2.5 2.9 1.9 3.3 2.1 0.7 -6.2 -1.6 -2.8 VENEZUELA 19.2 14.4 11.6 8.7 8.3 8.7 13.0 8.5 6.2 9.4 A Savings equal current revenue minus current expenditures. Total expenditures include expenditures and lending minus repayment. Table 2.1b: SAVINGS INVESTMENT RATIO g/ 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 ARGENTINA -2.0 -0.1 1.3 0.0 0.5 0.3 -0.5 -0.8 -1.8 -0.3 BOLIVIA 1.1 0.7 0.6 0.1 -1.1 -5.6 -3.6 NA -3.5 -61.0 BRAZIL 1.5 1.6 3.6 2.5 4.4 2.2 3.4 3.1 2.9 2.8 CHILE C.7 0.8 0.4 0.6 1.7 2.5 1.4 -1.3 -1.0 -0.3 COLOMIA 0.8 1.2 1.0 1.0 0.4 0.4 0.1 0.5 0.3 NA COSTA RICA 0.4 0.1 0.2 *0.1 -0.4 -0.7 -0.2 -0.2 0.2 0.6 DOMINICAN REP 1.2 1.0 1.1 0.9 0.4 0.6 0.5 0.0 0.3 0.6 GUATEMALA 0.7 0.5 1.0 1.0 0.5 0.4 0.1 0.2 NA NA MEXICO 0.3 0.2 0.3 0.5 0.6 0.7 0.2 -1.1 -0.5 -0.5 PAUMA 0.0 -0.4 0.0 -0.2 -0.4 0.1 -0.2 -0.9 -0.1 -0.6 PARAGUAY 1.0 0.8 1.3 1.2 1.4 1.4 0.8 0.9 0.7 0.5 URUGUAY -3.1 -1.1 1.2 0.8 1.3 1.2 0.4 -2.9 -0.9 -1.4 VENEZUELA 2.8 1.5 1.0 0.8 1.5 1.8 1.5 1.0 1.1 2.8 t/ Current Revenues * Current Expenditures) / Capital Expenditure Source: IMF, Goverrment Finance Statistics Yearbook, various years. - 19 - Table 2.1c: CENTRAL GOVERNMENT INTEREST PAYMENT AS PERCENTAGE OF TOTAL EXPENDITURES f/ 1975 1976 1977 1978 1970 1980 1981 1982 1983 1984 LAC SAMPLE 9/ 5.5 7.1 8.1 8.2 8.3 7.4 9.0 12.2 16.6 22.0 j/ Total expenditures incLude expenditures and kending minus rep,ment. TabLe 2.1d: CENTRAL GOVERNMENT INTEREST PAYMENTS AS PERCENTAGE OF GDP S/ 1975 1976 1977 1978 1979 1950 1981 1982 1983 1984 LAC SAMPLE t/ 1.1 1.3 1.7 1.8 1.7 1.6 2.2 3.5 4.6 5.3 p/ Interest payments by subnationaL governments are generally negligible except for Brazil, where subnational government interest payments reached about 1% of GDP in 1983. / Weighted average. Source: IMF, Government Finance Statistics Yearbook, various years. Tabe 2.2: DOMESTIC INFLATION IN LAC COUNTRIES -Annual Percent Change of Consumer Price Index* 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 ARGENTINA 182.3 443.2 176.1 175.5- 159.5 100.8 104.5 166.8 343.8 626.7 672.1 BOLIVIA 8.0 4.5 8.1 10.4 19.7 47.2 28.6 133.3 269.0 -81.4 11749.6 ORA21L 29.0 42.0 43.7 38.7 52.7 82.8 105.6 97.8 142.1 if'.0 226.9 CHILE 374.7 211.8 91.9 40.1 33.4 35.1 19.7 9.9 27.3 19.9 30.7 COLOMIA 22.9 20.2 33.1 17.8 24.7 26.5 27.5 24.5 19.8 16.1 24.0 COSTA RICA 17.4 3.5 4.2 6.0 9.2 18.1 37.1 90.1 32.6 12.0 15.1 DOMINICAN REP 14.5 7.8 12.8 3.5 9.2 16.8 7.5 7.6 4.8 27.0 37.5 GUATEMALA 13.2 10.7 12.3 8.3 11.3 10.8 11.,4 0.3 4.5 3.4 18.7 MEXICO 15.2 15.8 29.0 17.5 18.2 26.4 27.9 58.9 101.8 65.5 57.7 PANAMA 5.5 4.0 4.5 4.2 7.9 13.8 7.3 4.3 2.1 1.6 1.0 PARAGUAY 6.8 4.6 9.3 10.6 28.3 22.4 14.0 6.8 13.4 20.3 25.2 URUGUAY 81.4 50.6 58.2 44.5 66.6 63.5 34.0 19.0 49.2 55.3 72.2 VENEZUELA 10.3 7.6 7.8 7.1 12.4 21.5 16.2 9.6 6.3 12.2 11.4 Source: IMP, Internationt Finance Statistics, 198?. - 20 - became an attractive source to finance needed urban investments.U However, the overwhelming propDrtion of foreign loans was requested for central, rather than for local government agencies. In Lat.i American countries, national laws virtually exclude subnational governments from direct access to external borrowing. In fact, since foreign exchange must be rationed, both the national planning authorities and the foreign lenders prefer to direct external credit to public enterprises because their functions generally involve a foreign component and because they view the supervisory activities associated with external loans as a means to introduce better pricing and financial discipline. Therefore, little external borrowing was made available to subnational entities and it was mostly limited to capital expenditures. Domestic borrowing was also viewed as inappropriate for city governments because it could induce improvident local administrators to undertake expenditure decisions which would restrict flexibility in their future budgets and leave them vulnerable to the uncertainties of the market and exchange rates. 2.04 Expenditure Effects of the Recession. Yet, the increasing demands of urbanization and the job protection policies in force in most Latin American countries allowed for no significant reduction in recurrent expenditures. The initial reluctance of many national governments to participate in financing the basic operating expenditures of local authorities was gradually overcome by the realization that this would disproportionately penalize the poorest areas of the countries. In Argentina or Brazil for instance, it soon became a necessity to avoid unwanted reductions of nationally valued activities like education and health services or social security payments. New capital formation, which was more easily deferrable, was restrained for an extended period, aggravating the already sizeable backlog of unmet demand for physical facilities. Moreover, the reduced volume of new investments was accompanied by a decline of maintenance expenditures of the existing capital stock, accelerating replacement needs. 12/ The progressive increase in World Bank Loans is shown below: Table 2.3 World Bank Lending in Social Sectors to LAC countries 1978-82* 198 1986 1985 1986 1987 1988 Urban Development 177.6 46.2 191.2 86.9 369.5 325.0 764.0 Water Supply and Sewerage 204.5 424.2 28.6 163.9 175.0 64.0 252.3 Education 62.6 59.8 68.0 195.8 10.0 84.9 88.3 Population, Health & Nutrition 7-6 3 .5 - 9 .-.Q 109.0 452.3 563.7 345.3 446.5 650.5 483.9 1.213.6 * Annual Average Source: The World Bank Annual Report, 1988. - 21 - 2.05 Progressive Budgetary Displacement. Incremental urban investments are now urgently needed in LAC countries. However, a sudden escalation in the volume of subnational investment would remain neutral to the objectives cf macro- economic adjustment only if fiaanced out of marginal local revenues, lest it become an additional source of inflation. Our analysis shows that LAC rapid urbanization had a substantial impact on the structure of expenditures by governments. This effect of progressive budgetary displacement is visible in national accounts (Tables 2.5a & b). It shows in the proportional increase of expenditures by subnational authorities (Table 1.13) in response to growing pressures by their constituencies, and in the corresponding increase of central government outlays, both in terms of transfers (Table 4.12a) and of direct investments (Table 3.10b). Ultimately, it shows in the progressive swelling of the subnational deficit (Table 4.2). 2.06 Revenue Effects of the Recession. In practice, the aspiration to preserve existing standard of service in LAC cities, confronted all levels of government with severe financial problems. This was mainly due to the slow growth of their economies in recent years, and to the steady erosion of the local revenue bases under the combined effects of inflation and declining incomes.La1 At the same time, the volume and predictability of resource transfers to subnational governments was challenged by the contraction of the central government revenues, which are even more directly related to the cyclical fluctuations of economic activity. Nevertheless, intergovernmental flows (generally the largest item in local revenues) did not decline in real terms as drastically as one would expect (Table 4.12b), even when the sources of institutional crtdit were gradually dried up by financial constraints or official restrictions (See Section 4.13 on Borrowing). 2.07 Centralized Economic Policy and Political Democratization. In a number of LAC countries, the contraction of public revenues at all levels of government came at a time of explosive urban growth and often in coincidence with the claims of greater subnational autonomy engendered by the re-establishment of elective governments. Yet, the national situation called for reductions in transfer payments, tightened credit controls and adjustments in public service rates. Crisis management also required central direction of all instruments of fiscal policy, including the authority to tax or not to tax certain activities, 1 / To this one could add the diminishing tax compliance by city dwellers in the expectation of tax amresties (blanqueos) or simply of the inflation- induced depreciation of arrear payments. For instance, in Argentina, delinquent tax obligations were totally or partially forgiven by new governments on repeated occasions. - 22 - Table 2.5 : PROGRESSIVE DISPLACEMENT IN SUDGETARY STRUCTURE OF LAC COUNTRIES *In Percentage- 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 AVERAGE ARGENTINA CENTRAL GOVERNMENT 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GOVERNMENT SERVICES 14.3 17.1 19.5 17.4 20.0 20.4 20.0 19.0 16.0 13.7 17.8 SOCIAL SERVICES 39.3 34.7 37.3 45.2 47.0 47.3 43.2 37.0 39.0 50.3 42.0 ECONOMIC SERVICES 21.4 29.9 26.9 19.0 20.0 18.5 17.6 17.7 20.4 20.3 21.2 OTHER SERVICES 3.6 18.1 16.3 19.0 12.0 11.1 19.2 23.0 24.5 13.6 16.0 SURNATIONAL GOVERNMENTS 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GENERAL PUBLIC SERVICES NA NA NA NA NA MA NA NA 20.3 24.8 4.S SOCIAL SERVICES 50.0 37.9 32.5 43.2 38.5 44.8 50.0 46.8 47.8 55.3 44.7 CHILE asWWWW CENTRAL GOVERNMENT 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GOVERNMENT SERVICES 25.7 26.2 24.6 24.1 23.1 23.5 23.7 22.3 23.5 22.7 24.0 SOCIAL SERVICES 49.2 50.9 53.3 53.8 57.6 59.8 63.3 67.2 66.3 65.2 58.6 EC.SIC SERVICES 14.9 14.6 12.6 16.8 15.5 13.8 11.6 9.0 6.3 7.5 12.3 OTHER SERVICES 10.2 8.3 9.5 5.3 3.8 2.9 1.4 1.6 3.9 4.6 5.1 SUSNATIONAL GOVERNMENTs 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GENERAL PUBLIC SERVICES MA NA NA NA NA NA 79.2 63.7 56.6 59.7 64.8 SOCIAL SERVICES NA NA NA NA NA NA 20.8 36.3 43.4 40.3 35.2 COLOMBIA m.uuum CENTRAL GOVERNMENT 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GOVERNMENT SERVICES NA NA NA NA NA NA NA 26.1 26.7 26.8 26.5 SOCIAL SERVICES NA NA NA NA NA NA NA 47.3 49.0 48.2 48.2 ECONOMIC SERVICES NA NA NA NA NA NA NA 28.8 27.1 19.3 25.1 OTHER SERVICES NA NA NA NA NA NA NA 5.7 5.3 5.7 5.6 SUSNATIONA: GOVERNMENTS 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GENERAL PU8LIC SERVICES 17.3 18.1 17.9 15.6 18.2 17.6 15.2 18.3 19.1 20.6 19.4 SOCIAL SERVICES 70.0 66.7 68.8 70.5 69.0 69.0 65.3 64.7 65.0 65.8 67.5 ECONOMIC SERVICES 12.8 15.2 13.3 13.8 12.8 13.3 19.5 16.9 15.9 13.6 14.7 OTHER SERVICES 0.4 1.1 2.2 3.5 2.6 1.4 2.0 2.6 2.3 3.0 2.1 PANAMA CENTRAL GOVERNMNT 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.3 GOVERNMENT SERicS S/ 18.8 20.8 19.9 22.4 19.7 18.3 20.7 29.0 22.1 23.1 21.5 SOCIAL SERVICES 43.8 43.7 46.7 43.9 40.6 40.4 39.7 37.1 45.6 46.0 42.7 ECONOMIC SERVICES 28.5 26.2 22.0 21.4 26.9 21.9 18.4 13.5 12.4 12.0 20.3 OTHER SERVICES 8.9 9.3 11.4 12.3 12.9 19.4 21.2 20.5 19.8 18.9 15.5 SUBMATIONAL GOVERNMENTS 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.3 (%MERAU PUALIC SERVICES 57.3 63.6 68.7 73.1 74.1 75.9 69.6 66.4 66.6 63.9 67.9 SOCIAL SERVICES 35.1 28.8 24.1 22.5 22.2 22.4 29.3 32.5 32.1 27.9 27.7 ECONOMIC SERVICES 3.2 5.1 6.2 3.8 2.7 1.3 0.8 0.7 1.0 5.6 3.0 OTHER SERVICES 4.3 2.5. 1.0 0.5 1.1 0.4 0.4 0.3 0.3 2.6 / Government Services at the centraL goverrmnt Level incLude expenditures in defense except in Panama. Sources IMF, Coverment Finance Statistics Yearbook, various years. - 23 - TMbt 2Lgb: TRENDS IN BUDGETARY STRUCTURE OF LAC COUNTRIES *Variation index, 1976a100* p1 t976 1977 198 1979 1980 1981 1982 1983 '984 ARGENTINA uuuWWaaWWS CENTRAL GOVERNMENT 100 88 96 88 95 107 99 121 89 GOVERNMENT SERVICES 100 99 96 10 111 123 108 101 70 SOCIAL SERVICES 100 94 125 fi9 129 133 105 136 129 ECONOMIC SERVICES 100 79 61 59 59 63 59 83 60 OTHER SERVICES 100 79 101 $8 58 114 125 164 67 SUBNATIONAL GOVERNMENTS 100 125 f13 114 126 119 100 139 141 GOVERNMENT SERVICES 100 109 107 72 108 106 80 323 384 SOCIAL SERVICES 100 107 129 1is 149 157 123 175 205 CHILE "Aax CENTRAL GOVERNMENT 100 124 149 159 160 167 171 158 164 GOVERNMENT SERVICES 100 116 137 141 14 151 145 142 142 SOCIAL SERVICES 100 130 157 181 189 207 226 206 210 ECONOMIC SERVICES 100 107 171 169 151 733 105 68 85 OTHER SERVICES 100 142 94 72 55 28 32 75 90 SUSNATIONAL GOVERNMENTS 100 78 104 108 130 254 364 284 280 COLOMBIA =assaum= SUSNATIONAL GOVERNMENTS 100 98 118 132 162 158 152 160 184 GENERAL PUBLIC SERVICES 100 97 102 133 15 133 154 169 210 SOCIAL SERVICES 100 101 125 137 168 155 148 156 182 ECONOMIC SERVICES 100 86 108 112 142 203 170 168 165 OTHER SERVICES 100 188 364 305 199 279 355 319 489 PANKMA CENTRAL GOVERNMENT 100 99 108 145 139 147 772 ?58 170 GOVERNMENT SERVICs b/ 100 95 117 137 122 146 240 168 189 SOCIAL SERVICES 100 106 108 134 128 133 146 765 179 ECONOMIC SERVICES 100 83 86 148 116 103 88 75 78 OTHER SERVICES 100 122 143 201 290 336 379 338 347 SUBNATIONAL GOVERNMENTS 100 94 84 79 88 92 99 100 99 GENERAL PUBLIC SERVICES 100 102 97 92 f04 101 104 104 100 SOCIAL SERVICES 100 79 66 61 68 94 112 111 96 ECONOMIC SERVICES 100 167 200 117 83 50 33 33 50 OTHER SERVICES 100 63 25 13 25 13 13 13 13 P/ The variation index is based on constant istional currencies time series. W/ Centret goverment expenditures do not include expenditures on defense. Source: IMF, Governfent Finance Statistics Yearbook, various years. - 24- irrespective of the impact such measures would have at the local level.191 Similarly, the principal decisions on public investment in cities and in key sectors such as transport, energy, housing, and public sanitation became the responsibility of the national agencies in charge of managing the corresponding sources of funds. 2.08 Growing Fiscal Dependence of Subnational Authorities. The progressive weakening of the financial position of regional and local governments resulted in a gradual loss of decisional autonomy. Subnational authorities were caught in a cycle of fiscal dependence with respect to the central government which went through several stages: (i) initially, the subnational governments provided their traditional services mostly with locally raised resources; (ii) in the course of time, the growth of their obligations outstripped that of own revenues; (iii) a widening financial gap led subnational authorities to depend on central government support to straighten out their accounts; and, lastly, (iv) the cycle of financial dependency closed and became chronic, as the local authorities began to rely on bilateral negotiations not only for the realization of investment programs or to even out their debts, but also to meet the monthly costs of direct administration. 2.09 Planning by Negotiations. The ultimate result of this trend was that the expenditure pattern of subnational authorities came to depend on a process of negotiations. They must struggle for resources, funds and programs defined by central authorities. They must keep informed on the funding sources available for specific investment programs and projects, go through the bureaucratic routine of application, produce suitable information on local development problems and potentials, and submit to the supervisory discipline of the monitoring and evaluation procedures set out in corresponding guidelines. At all phases of this process, the final outcome of the transactions remains uncertain.21 12/ A typical instance was the Brazilian decision to exempt many export-based activities from payment of the value added tax (ICM) which is the main revenue source for the State and Municipal governments. 20/ In the case of Brazil, "the present system revolves around negotiations of federal funds - negotiations in which the states, because of their precarious financial situation, find themselves in inferior bargaining positions. The amount of investment funds to be available therefore, has become difficult to predict, and specific sectors or projects to benefit from these funds are more apt to reflect national priorities than those of the states." (D.J. Mahar and W.R. Dillinger: Financing State and Local Governments in Brazil, World Bank Staff Working Paper No. 612, page 24). - 25 - 2.10 Political Use of Negotiated Transfers. Statutory intergovernment transfers are usually ruled by set formulae, but the fiscal treatment of particular areas has been modified in most LAC countries by a parallel, extensive system of discretionary transfers allocated without established explicit criteria.11i As a result, the actual distribution of negotiated transfers does not always bear a relation to the social and economic variables that dictate spending needs. The trends of subnational finances during election years often reveal the influence of political patronage. 2.11 Erosion of Political Responsibility. The public accountability of subnational authorities was progressively eroded as they found themselves dependent on upper level decision even to cover the salaries of public employees. The diminished importance of locally raised receipts weakened the critical social mechanism through which taxpayers can control the decisions of their administrators and protect their interests. The lack of transparency of public accounts made it virtually impossible to understand the real scope of expenditure options and locally motivated leaders became less and less attracted to public office jobs. Institutional credit became a virtual lifesaver thrown by the central government to local leaders for their political survival, reproducing at the level of subnational finance the uncontrollable growth of public debt, which characterized national accounts. Although in theory the law set limits to subnational liabilities in most LAC countries, in practice different types of bilaterally negotiated public debt escaped effective monitoring and controls, and no one knew exactly how many times the same securities had been put up for different borrowing agreements. 2.12 Summing U. Section 2 has shown how the adverse economic circumstances of most LAC countries impose macro-economic constraints on urban expenditures. The recent trends towards democratization of the systems of governance run counter the centralization of major policy decisions which would be required by external developments such as the declining world market prices of export commodities and the rising interest payments on foreign debt. To these factors one must add the slow pace of growth of the economies, the burden of domestic debt, the disruption of the financial systems, the severe inflationary pressures and the budgetary displacements caused by the operating costs of a burgeoning public sector, in presence of job protection policies. The combination of these problems allows little room for transfers supporting new capital formation and maintenance of the present levels of service in urban areas. Nevertheless, 2.0 In Argentina, the growth of discretionary transfers to the Provinces (Aportes del Tesoro Nacional) increased drastically as a proportion of total transfers: 1977 1978 J 1980 1981 1982 1983 1984 10.9% 11.4% 7.7% 7.6% 35.0% 25.4% 69.7% 65.0% Similar trends occurred in Mexico and Brazil (see Box 2). - 26 - economic adjustment through contraction of this kind of expenditures meets obvious political resistance, as evidenced by the growth of discretionary transfers and the central absorption of subnatIonal debt in several countries (Argentina, Brazil, Mexico). This gives rise to a pattern of increasing dependence on the outcome of ad-hoc negotiations which reduces even further the fiscal accountability of lower level authorities. - 27 - III. ARE LAC GOVERNMENTS TRULY UECENTRALIZING? 3.01 Role and Structure of the Public Sector. The range of activities encompassed by public budget measures in LAC countries is fairly wide, as total expenditures and revenues by government range between 30 and 40 percent of GDP. Besides the general administration, the justice system and the legislature, the public sector comprises an array of other functions which are performed at different levels by government-controlled, non-financial entities (decentralized sector agencies and non-market public corporations) or by financial institutions (development banks, trust funds insurance companies, social security funds). The scope and budgetary autonomy of this parastatal sector varies from country to country, making it difficult to draw accurate comparisons of its relative importance.Ul (See Table 3.1). Because the size and composition of the public sector have an impact on the efficiency with which revenues are raised and distributed, as well as on the purposes for which resources are transferred and used, it may significantly affect the achievement of social objectives and ultimately determine the patterns of national development and economic growth. And, because the very structure of the public sector does influence its policies, it may also influence the pattern of urbanization. 3.02 Politics of Intergovernmental Relations. The distribution of financial and administrative authority among different layers of government is the result of public choices made in the past and reflects the countries' concerns, their political systems and their basic values. Therefore, it is also constantly evolving over time. This historical and cultural dimension is essential to understand the nature of intergovernmental financial flows as they take place at present, the assignment of functional responsibilities to different tiers and sectors of government and the degree of decisional autonomy left at each level for determining which services to provide, to what extent to cover costs and which revenue bases should be tapped. These issues have become increasingly political in Latin America because the progressive reduction in the real volume f local revenues has strained the relations between the lower and the upper level governments, sharpening disputes over the extent to which current fiscal laws allow for local discretion on the scope and volume of urban revenues and expenditures. 3.03 Devolutive rather than Aggragative Governance. In general, LAC countries were created through the progressive disaggregation of centrally controlled colonial domains. Their legal and fiscal statutes still reflect today an original conception of the national government as the principal unit of social organization, so that any decentralization of authority below this level calls for a special decision. While the evident need to provide at least some services at decentralized level led to the creation of subnational governments, all their 2./ See Methodological Annex for details. Table j.: COMPARATIVE FINANCIAL STRUCTURE OF THE PUBLIC SECTOR _W -Public Sectorul0OX- CURRENT INCOME CURRENT EXPENITURE INVESTMENT SURPLUS/(DEFICIT) Administ. Enterprise Administ. Enterprise Adafist. Enterprise Administ. Enterprise ARGENTINA (1983) 58.9 41.1 72.7 27.3 56.5 43.5 (88.4) (11.6) Central Level 35.3 36.1 45.9 24.2 21.3 36.2 (45.3) (9.8) Regional Level 20.5 5.0 22.9 3.2 30.3 7.4 (26.1) (1.8) Local Level 3.1 NA 3.8 MA 4.4 MA (4.6) NA BRAZIL (1983) 38.5 61.5 61.9 38.1 28.4 71.6 (63.3) 163.3 Central Level 28.4 58.3 42.7 35.0 16.8 64.4 (28.6) 164.7 Regional Level 8.7 3.1 14.1 3.1 7.4 7.2 (17.0) (1.4) Local Level 1.4 KA 5.1 MA 4.2 MA (20.0) MA OD COLONBIA (1981) 49.2 50.8 63.5 36.5 42.5 57.5 (128.8) 28.8 Central Level 38.0 49.3 39.9 29.1 32.4 56.4 5.3 86.9 Regional Level 7.9 1.4 18.7 7.4 4.8 1.1 (98.6) (58.1) Local Level 3.3 NA 4.7 MA 5.3 NA (30.1) NA NEXICO (1980) 58.5 41.5 66.0 34.0 57.0 43.0 (105.0) 5.0 Central Level 6.0 41.5 55.4 34.0 43.9 43.0 (93.4) 5.0 Regional Level 11.1 NA 8.8 NA 12.0 MA 1.5 NA Local Level 1.4 NA 1.8 NA 1.1 NA (2.9) MA g/ Public sector includes government administration and decentralized financial and non-financial public enterprises at alt levels of government. Sources: For goverinment administration.- INF, Government Finance Statistics Tearbook, 1987. For decentralized public enterprises.- various public documents from national sources. - 29 - administrative and fiscal authority emanated from the top. This essentially devolutive (unitary) model differs fundamentally from the aggregative (pluralistic) model of countries that were formed through the federation of several autonomous governments and the voluntary transfer of some local prerogatives to the new composite entity.3 Therefore, the mindset and behavior of Latin American countries starts with the presumption that all public sector activities within their political boundaries will be performed by the central government apparatus and that specific activities will be turned over to lower administrative units only when required by some sort of technical efficiency rationale. 3.04 Parastatal Sector. The process described above may explain the proliferation of special purpose agencies in LAC countries. These agencies are essentially branches of larger central government counterparts which perform specific, delegated functions (such as road construction, water and sewerage provision, power distribution, transportation, telecommunications, rural development, public housing) which could be carried out autonomously by the lower levels of government, if not by the private sector. This gives rise to an expensive bureaucratic apparatus and an intricate pattern of financial flows. Since resources are mostly channeled through this apparatus, subnational administrations are not in the position to respond promptly to locally perceived needs, nor to make plans and to function effectively, because they have no real authority to direct the activities of national agencies. On the other hand, the recurrent attempts to contain the growth of parastatals, to reduce their inefficiency or to limit the drain they cause on the national budgets usually meet with considerable resistance. In fact, by funnelling funds through its regional branches, the central governments can retain more effective control over new investment decisions than it would by passing funds directly to the budgets of subnational administrations. 3.05 Management of Public Enterprises. National planning authorities apparently prefer to operate through specialized sector agencies or public enterprises, because these institutions could raise resources through user charges without encroaching upon the revenue bases of the central government. They also like to believe that the tariffs of public enterprises can be used as price signals to allocate and mobilize funds for investment or to influence short-term inflationary expectations. However, the overdevelopment and budgetary fragmentation of public enterprises has reduced their accountability and responsiveness to these objectives. In some LAC countries public enterprises initially conceived as instruments of sound financial management became mere 2A/ The only LAC country which approximates this pattern is Argentina, where a federal system was adopted by the provinces of the interior, which retained many of their own constitutional powers but pooled others, in ordet to counterbalance the influence of Buenos Aires. - 30 - vehicles to borrow abroad in support of the questionable exchange rate policies of the central government.2ii Eventually, as the debt crisis set in, the National Treasuries had to assume repayment of the external debts of 0^he enterprises, soon after having re-financed their domestic obligation it subsidized rates. 3.06 Government Structures. Four countries in Latin America have a Federal form of government: these are Argentina, Brazil, Mexico and Venezuela, which have explicit constitutional arrangements providing for intermediate level governments (States and Provinces). But the vast majority of countries has a Unitary organization, with local governments (Municipalities) as statutory entities directly linked to the central government, which can create or abolish them at its discretion. Yet other countries (Bolivia, Colombia, Costa Rica, Ecuador and Peru) present an intermediate structure, with regional bodies (Departments or Districts) which are granted some expenditure and supervisory responsibilities over the local units of particular areas, and mainly function as agents of the central governments. In relatively large countries, this device reduces the central governments' difficulties in dealing with large numbers of local units, which have different needs, fiscal bases and service capacities. Thus, the local governments become dependent upon these regional administrations for the provision of certain services and for the pass-through of national grants as they would be in a Federal country. Both in unitary and in federal countries, however, the largest cities (especially the capitals) tend to enjoy special status and are directly linked to the center.25 24/ Here one must note that the concern of international lending agencies with the technical, financial and administrative viability of the utility sector in particular has often resulted in the establishment of new enterprises. This was generally based on the assumption that their managerial and professional staff would be more competent and motivated than the cadres of local governments insofar as they would be: (a) better paid; (b) less politically influenceable; (c) more securely tenured; and (c) more technically motivated (Bird, Bahl and Linn, Macon). 2/ Bogota, Brazilia, Buenos Aires, Caracas, Mexico City, Santo Domingo are all examples of Capital Districts. In Colombia also the Cities of Cali and Medellin enjoy special districts status. - 31 - 3.07 LAC Countries vs Industrial Ma ket Economies. If one compares the group of LAC countries with the twenty industrial market economies of the WDR,-1 one notes first the cleavage that tells LAC countries apart in terms of GNP per capita (Venezuela, which is the richest is about four times poorer than the average of the second group and the poorest, Bolivia, is twenty five times below that level) (Table 3.7). Moreover, their income distribution is such that only about 20 percent of the combined income of households accrues to those in the bottom six deciles, against 37 percent in the more affluent countries. The largest part of these poor households is rural, and their low income can be explained by the comparatively lower remuneration of agricultural labor in most LAC countries and by the fact that cropland per head of farm population is only 1.7 ha. in Lati.n America compared to 2.2 in the developed countries.3U Yet, one cannot say that Latin America is comparatively overpopulated, since its average territorial density is of 18 people per kM2, i.e., 25 percent below the 24 people per km average of the developed countries, although the LAC value is strongly influenced by the vast uninhabited regions of Brazil and Argentina. 3.08 Relative Urbanization and Industrialization. In fact, LAC countries are relatively urbanized, with Argentina, Chile, Uruguay and Venezuela showing a higher proportion of urban to total population than the average developed country. Their urban population is also concentrated in cities of sufficient size to yield considerable economies of agglomeration. Ten countries (Argentina, Brazil, Colombia, Costa Rica, the Dominican Republic, Ecuador, Jamaica, Mexico, Paraguay and Peru) have nearly half their urban population living in cities with more than 500,000 inhabitants, if that is taken as the average measure of the 26/ Obviously, substantial country differences exist within the generalized scenarios traced here. These differences can best be gleaned from the attached Table 3.7, which shows the characteristics taken into conside:ation for a sample of LAC countries for which the central/local government breakdown is available from the IMF. The sample includes eight countries of South America and five in Central America and the Caribbean, giving a fairly wide representation of general conditions. It includes all major countries of Latin America (Argentina, Brazil, Colombia, Chile, Mexico and Venezuela) and a few smaller economies (Bolivia, Costa Rica, Dominican Republic, Guatemala, Panama, Paraguay and Uruguay). Together they account for 84.0 percent of Latin America population, 88 percent of its land area, 91 percent of its GDP, 88 percent of its urban population and 92 percent of the population in cities over 500,000 inhabitants. The Industrial Market Economies considered are listed on Table 1.2. j1/ Only Argentina, Uruguay and Chile, with 7.6, 3.1 and 2.8 ha per farmer respectively supercede the average ratio of the developed countries. The reported averages refer to 19 Industrial market economies and to 13 LAC countries. If all Latin America is considered the corresponding average would be of 1.41 ha/farmhead. - 32 - Table 3.7: INDICATORS OF INCOME, URBANIZATION AND INDUSTRIALIZATION IN LAC COUNTRIES Inoome Urbanization Industrialization GNP p.c. Share of Six 1980 Urban In Cities Share of 1980 Share of 1985 Countries 1985 US $ Lowest Deciles Population 500,000 Labor Force GNP a) b) c) d) e) f) Venezuela 2550 23 85 31 28 42 Argentina 2130 28 84 47 34 42 Panama 2100 18 50 21 18 18 'exico 2080 22 69 31 29 35 Uruguay 1650 NA 85 41 29 33 Brazil 1640 16 73 33 27 33 Chile 1430 NA 83 34 25 40 Colombia 1320 NA 67 24 24 30 Costa Rica 1300 25 45 28 23 29 Guatemala 1250 NA 41 15 17 NA Paraguay 860 NA 41 26 21 25 Dom. Republic 790 NA 56 35 15 31 Bolivia 470 NA 44 14 20 25 13 LAC Countries 1677 NA 68 32 24 32 9 IND NWMr. ECON. 11,810 37 75 35 35 36 Sources: Values In columns a), b), c) e) and b) derived from World Development Report 1987, World Bank no. ISBN 0-19-520563-4. Values in columns d calculated from UNCHS data Global Report 1986 Habitat. Values in column b) apply to different years. - 33 - urban polarization in developed countries (See Table 1. 2). Finally, LAC countries are also fairly industrialized: the contribution of their industrial sector to GNP formation exceeds the average 36 percent value recorded for the developed countries in Argentina, Chile, Peru and Venezuela and is already close to this value also in Brazil, Mexico and Uruguay (see Table 3.7). Thus, conventional indicators of development such as industrialization and urbanization do not by themselves explain the wide discrepancy of urban service levels between LAC countries and the developed market economies. 3.09 Public Sector Dominance? The importance of the public sector, both in terms of revenues and expenditures, is no greater in LAC countries than it is in the so called market economies, which are in fact "mixed" economies. In the average, the expenditures of a Latin American government account for 31 percent of GDP, against 41 percent in the more developed countries (see Table 3.9a). In this respects, one could argue that the role of government is less significant in LAC countries than it is in the industrial market economies. But, if one were to measure fiscal effort as the proportion to which GDP is taxed by the general government, LAC countries' proportion falls below 25 percent Table 3.9a: GENERAL GOVERNMENT EXPENDITURES AS PERCENTAGE OF GDP a/ Average 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 - LAC SAMPLE bi 26.2 25.5 26.3 27.3 25.9 27.1 30.0 33.7 32.4 31.0 28.5 MARKET ECONOMIES k/ 39.2 38.7 38.3 38.9 39.1 40.9 41.3 42.8 42.8 41.0 40.3 AI Total expenditures include expenditures and lending minus repayment. General Government comprises central, regional and local governments. This is different from 'Total Public Sector' which includes decentralize public enterprises. k/ Weighted averages. Comparability over time affected by availability of data. Table 3.9b: GENERAL GOVERNMENT REVENUES AS PERCENTAGE OF GDP &/ Average 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 = LAC SAMPLE h/ 22.3 21.9 24.1 24.7 24.4 24.4 24.5 26.3 25.8 24.6 24.3 MARKET ECONOMIES h/ 36.3 36.3 36.8 36.5 36.7 37.9 38.7 39.4 38.9 39.0 37.7 1/ Total Revenues include revenues and grants. k/ Weighted averages. Comparability over time affected by the availability of data. Source: IMF, Government Finance Statistics, various years. - 34- overall against the average 39 percent value recorded in industrial market economies (see Table 3.9b). This can be explained by the generally lower taxable base of LAC countries and the more developed fiscal systems of industrial market economies, but remains a determining factor of the lower standard of social welfare so far achieved in Latin America. 3.10 Are LAC Countries Centralized? Given the picture outlined above one would not conclude that LAC economies are more government-controlled than the more developed countries. However, another characteristic can be examined: the share of general government expenditures and revenues which is directly handled by their central governments. In LAC countries, central authorities collect in the average 80 percent of total tax revenues (against only 73 percent in industrial market economies) (See Table 3.10a) and implement 80 percent of their Table 3.10a: REVENUES BY LEVEL OF GOVERNMENT a/ -In Percentage, General Govt-100i- Average 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 m*- LAC SAMPLE AVERAGE b/ Subnational Government 20.1 20.4 20.6 20.0 19.8 21.5 19.8 19.8 18.6 20.5 20.1 Central Government 79.9 79.6 79.4 80.0 80.2 78.5 80.2 80.2 81.4 79.5 79.9 MARKET ECONOMIES Subnational Government 28.8 28.3 27.7 27.2 26.3 26.6 26.4 26.3 26.6 27.0 27.1 Central Government 71.2 71.7 72.3 72.8 73.7 73.4 73.6 73.7 73.4 73.0 72.9 g/ Total expenditures include expenditures and lending minus repayment. k/ Weighted averages. Comparability over time affected by the availability of data. Table 3.10b: CAPITAL EXPENDITURES AS PERCENTAGE OF TOTAL EXPENDITURES BY LEVEL OF GOVERNMENT Average 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 - LAC SAMPLE AVERAGE a/ Subnational Government 25.7 28.3 23.5 23.6 22.7 24.5 21.3 17.8 15.4 19.0 22.2 Central Government 19.0 23.0 18.5 19.0 17.0 17.9 19.2 16.9 14.2 11.7 17.6 MARKET ECONOMIES Subnational Government 25.0 23.7 21.9 22.9 22.2 18.7 17.1 16.3 15.2 14.8 19.8 Central Government 8.8 8.5 9.8 8.5 7.8 8.3 9.2 9.4 8.8 8.0 8.7 g/ Weighted averages. Comparability over time affected by the availability of data. Source: IMF, Government Finance Statistics, various years. - 35 - public expenditures (against 75 percent) (see Table 1.12). These proportions reveal the lesser degree of autonomy of LAC subnational governments. The average per capita revenue of subnational government (including transfers) is in fact twenty times lower than it is in industrial market economies (about US$100 against over US$2000, see Table 4.3a) while their average personal income is only soven times smaller (about US$1700 against US$12,000, see Table 3.7) One contradictory feature must also be noted: in the developed countries, the central government direct involvement in capital expenditures accounted for only 8.7 percent of their total expenditures during the 1975-1984 period while the corresponding share in LAC was two times as high (16.6 percent) (See Table 3.10b and Annex Tables 3.10c and 3.10d). 3.11 Are LAC Governments Welfare Minded? The budgets of central governments in developed countries show clearly that they are far more redistributive than those of LAC countries. In fact, as a proportion of GDP, transfers from the central to the other levels of government are three times higher in Part One countries (see Table 3.11a). During 1978-84, this type of outlays accounted Table 3.11a: INTERGOVERNMENTAL TRANSFERS AS PERCENTAGE OF GDP 1975 11976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE a/ 1.36 1.19 1.08 1.08 0.98 0.90 1.23 1.4.1 1.41 1.11 MARKET ECONOMIES a/ 3.51 3.59 3.55 3.51 3.48 3.45 3.41 3.36 3.28 3.12 &/ Weighted averages. Source: IMF, Government Finance Statistics Yearbook, various years. about 13 percent of total expenditure in market economies (declining to 11 percent over the period) but averaged only 8 percent in LAC countries, fluctuating sharply in the last few years (Table 3.11b). The percentage of total expenditures devoted to subsidies and other transfers by the central governments amounted during the same period to about 56 percent in the average market economy and only 33 percent in LAC countries. Table 3.11c shows that expenditures in the social se2tors (Social Security, Welfare, Education, Health, Housing and Community Amenities) accounted for about 56 percent in the central - 36 - Table 3.11b: TRANSFERS, SUBSIDIES AND LENDING INDIFFERENT COUNTRY GROUPINGS (as Percent of Central Government Expenditures and Lending Minus Repayment Years 1978 1979 1980 1981 1982 1983 1984 Tran sfer s to Other Govt. Level s Industrial Market Economies 14.48 14.17 13.35 12.60 11.63 11.44 11.42 Developing Countries 10.28 10.72 10.15 10.38 10.76 10.84 n.a. LAC Countries 8.16 8.66 7.53 8.26 8.57 9.11 7.95 Subsidies & Other Curr. Transfers Industrial Market Economies 56.26 56.41 55.94 55.71 55.83 56.26 55.60 Developing Countries 26.46 27.66 28.69 26.34 28.08 26.90 27.54 LAC Countries 33.02 33.47 36.16 32.04 35.98 33.51 33.40 Lending Minus Repayment Industrial Market Economies 6.40 6.20 5.90 5.20 5.20 4.70 4.50 Developing Countries n.a. n.a. 10.37 n.e. n.a. n.a. n.a. LAC Countries 12.60 13.10 12.60 15.40 13.60 16.60 14.90 TOTAL 1RAIFERS Industrial Market Economies 62.68 62.61 61.84 60.91 61.03 60.96 60.10 Developing Countries n.a. n.a. 39.06 n.a. n.a. n.a. n.e. LAC Countries 45.62 46.57 48.76 47.44 49.58 50.11 48.30 Source: IMF: (a) 18 Countries, (b) Scattered Sample, (c) Includes Argentina, Bolivia, Brazil, Chile, Colombia, Mexico, Uruguay, & Venezuela Table 3.11c: SOCIAL SECTORS EXPEIDITURES BY COUNTRY GROUPINGS (as Percent of Total Expenditures of Central Governments) 1978 1979 1980 1981 1982 1983 1984 Industrial tarket Economies (a) 56.35 56.11 56.06 56.00 55.76 55.65 54.61 Education 5.12 4.96 4.73 4.56 4.30 4.13 3.97 Health 11.18 11.46 -11.52 11.50 11.41 11.30 11.53 Social Security 37.66 37.35 37.41 37.65 37.77 38.03 36.76 Housing & Community 2.39 2.34 2.40 2.29 2.28 2.19 2.35 LAC Countries (c) 45.05 45.84 43.33 43.15 39.97 40.85 41.42 Education 12.17 11.99 11.29 10.87 9.70 9.51 9.88 Health 5.51 5.47 4.93 4.84 4.80 4.67 4.92 Social Security 25.85 27.37 25.52 25.59 24.05 24.83 24.66 Housing & Community 1.51 1.01 1.59 1.85 1.42 1.84 1.96 Developing Countries (b) 33.09 33.57 32.73 32.79 31.19 31.92 32.22 Education 10.76 10.76 10.55 10.03 9.48 9.60 9.80 Health 4.36 4.34 4.14 4.02 4.05 4.07 4.22 Social Security 16.06 16.65 15.91 16.31 15.77 16.21 16.07 Housing & Community 1.91 1.82 2.13 2.43 1.89 2.04 2.13 Source IMF: (a) 18 Countries, (b) Scatterd Sample, (c) Includes Argentina, Bolivi., 3razlf, Chlie, Colombia, Mexico, Uruguay, and Venezuela - 37 - government outlays in Part One Countries against 43 percent in those of Latin America (respectively 14 percent and 9 percent of GDP). On the other hand, the central governments of the industrialized countries spend far less in institutional credit (5.5 percent of their total expenditures), which indicates that this function is mostly performed by their financial markets, while it absorbs some 14 percent of public resources in LAC countries. (See Table 4.14). 3.12 Inter-Regional Disparities. At the root of the current debate over the pattern of intergovernment relations in LAC countries are the pronounced spatial differences in per capita incomes, employment opportunities and levels of services which exist between different regions, between the rural and the urban areas, between the large and the small cities. The larger the countries, the more these differences are brought to the fore, often aggravated by other factors of cultural discrepancy such as ethnic, linguistic and political allegiances. 2i Certain areas differ not only in their absolute level of economic activity but also in their rate of economic growth and in their relative capacity to internalize the benefits of national development. The characteristics of the workforce (age, education, skills) are not necessarily constant across regions and this has an impact on the occupational status and incomes of their inhabitants.291 3.13 Intra-Rezional Disparities. References to average income per capita in different regions often mask strong disparities in interpersonal income distribution, since not all people are poor in the 'poor' regions nor are they all rich in the 'rich' regions. Moreover, even within regions, certain areas may be industrialized, while others remain eminently rural and both per capita incomes and access to public services may be closely correlated to the relative extent of urbanization. There are also significant correlations between the fiscal capacities of different territorial units and their relative degree of urbanization, with marked differences between rural and urban areas, between non-industrialized and industrialized communities. The ever present concern in LAC countries is that large cities are somehow receiving too large a share of total central government transfers. But also the fiscal contribution of large cities is far greater because of their higher incomes, more productive labor force and the concentration of taxable enterprises attracted by the scale economies of sizeable sunk investments. Furthermore, the larger their population size, the greater is also the latitude these cities have in raising taxes and user charges, in expanding their tax base and, occasionally, even in borrowing money. 28/ Regional concentrations of ethnic minorities are found in Bolivia, Brazil, Colombia, Ecuador, Guatemala, Mexico, Panama, and Peru. 29/ Research on Colombia shows that: a) educational attainment and rural-urban condition are more sirificant explanatory variables of personal income level than regional factors and b) that greater income inequality (over 80 percent) exists within rather than between regions. (See G.T. Fields and T.P. Schultz "Regional Inequality and Other Sources of Income Variation in Colombi t" in "Economic Development and Cultural Change"). - 38 - 3.14 Regional Equalization Policies. At different times, several Latin American governments have adopted explicit policies aiming to reduce regional disparities. A number of approaches have been experimented with: (i) deliberately locating new public enterprises in less developed regions, often based on the assumption that a self-sustaining process of growth could be triggered by the establishment of capital good industries (steel, petrochemicals); (ii) intensifying central government expenditures in poor regions, either for basic infrastructure, (transportation, energy), for social services (e.g., housing, education, health) or for both; (iii) launching ambitious agricultural support programs in typically rural regions; (iv) charging different prices or levying different taxes in different regions; (v) adopting macro-economic or sector measures (tax incentives, subsidies, credit policies) so as to favor the less privileged areas. Usually a combination of sectoral investments and fiscal policies was used. Although such measures have occasionally shown some visible result, they were mostly taken in isolation and lacked sufficient continuity over time to offset the deeper systemic causes of the backwardness of the target regions. Private sector participation was slow to come and costly to induce, and it remains doubtful whether the market distortions and the marginal public costs of such programs were actually compensated by the benefits obtained. 3.15 Spatially Uniform Development? The fundamental ambivalence of Latin American spatial equalization policies lies in the overlap between economic development objectives (which call for optimizing the allocation of resources among regions) with redistributive objectives (which call for assuring at least a minimum level of welfare to all citizens of a nation, wherever they live). In large and diversified countries as those of Latin America, interregional differences are bound to persist, because different areas have different degrees and qualities of resources, both natural and man-made. One must confront the fact that certain areas may achieve the desirable level of welfare only over the long term, on condition that investment and institutional development be supported there, but also that demographic pressure be relieved, by encouraging excess labor to migrate to regions where it can find productive employment. The application of scarce resources with the short-term objective of achieving better standards of living in overpopulated and resource-poor regions is likely to result in greater marginal costs per unit of output with little visible impact on general welfare standards. Consequently, this objective is not easily reconciled with that of maximizing national growth and development. Pursuing spatially uniform policies in all areas of a country would be tantamount to reducing the rate of development for the country as a whole, and there are few countries in the world that have adopted this expiicit objective and even less that have achieved it even among developed countries. 3.16 Top-down Transfers. Statutory transfers from the central government to lower levels of government play an important role in the fiscal systems of LAC countries. Taking five larger countries into consideration (Argentina, Brazil, Colombia, Mexico and Venezuela, see Table 3.16) one can infer some generalizations which are to some extent applicable to most of the others. The five countries considered are large, structured in three layers of govt rnment and have a diversified and complex system of decentralized parastatal agencies. In all of them, the central government has a dominant role, both as a collector of tax revenues and as a supplier of funds to subnational authorities or - 39 - decentralized sector agencies in fotm of transfers, subsidies and conceisionary credit. Focusing on transfers, these accounted for 22 percent of central government expenditures in Colombia, over 13 percent in Venezuela, possibly the same in Mexico, over 8 percent in Brazil and 5 percent in Argentina. Subsidies and other current transfers averaged respectively 46 percent in Brazil, 39 percent in Colombia, 36 percent in Argentina, 24 percent in Mexico and 20 percent in Venezuela. Credit by governments (net of repayments) was 24 percent in Brazil, 12 percent in Argentina, 10 percent in Venezuela. Adding up the three items one realizes the full scale of this effort: 70 percent of central government expenditures for Brazil, 48 percent for Argentina, 42 percent for Colombia, 31 percent for Venezuela and 29 percent for Mexico. 3.17 The New Federalism. The existence of regional and intra-regional disparities, the pressures of urbanization and the widening fiscal gap of the subnational governments give much steam to the current debate over the reassignment of fiscal functions among different levels of government. This debate must be seen in the broader context of a political evolution, which in several LAC countries has been called 'new federalism' and revives traditional claims for greater municipal and state autonomies. This new philosophy of locally-based solidarity, supported by the growing voting power of urban communities, reflects a major change in the philosophy of fiscal relations between central and local authorities, suggesting major swaps of responsibilities, consolidation of categorical grants and revision of revenue sharing systems. One line of arguments, obviously resisted by central planning authorities, leans to unconditional transfers based on considerations of regional balance. The opposing line, presuming superior wisdom of the central planners, emphasizes that only tied grants can ensure that local authorities provide certain public services in the right amounts. 3.18 Rhetorics of Decentralization. The main character of this debate is political, because fiscal reassignment would result in a larger or lesser public budget and therefore in greater or lesser powers of political influence. Therefore it is often difficult to tell when decentralization is intended in a political or merely in an administrative sense. In the case of administrative decentralization, decisions would be made on the basis of delegated authority, but would remain revocable; while, in the case of political decentralization, the initiatives of the local governments would not be revocable by the central authorities. In the case of fiscal decentralization, some additional tax base would be actually surrendered to lower levels of government or else greater discretion would be left as to the way transferred funds are spent. Furthermore, some countries are mostly concerned with achieving demographic or economic decentralization (Mexico and Brazil for instance) which they try to promote by channeling government investment through regional branches of central government Ministries or Enterprises. But few countries in LAC are seriously considering a radical redistribution of taxing authority between their different layers of government. More often than not in Latin America, the central government's commitment to decentralization amounts to political rhetoric, since the proposed reforms do not necessarily envisage a more prominent role of the subnational governments in either the financing or the delivering of services. - 40 - Table 3.16: CENTRAL GOVERNMENT TRANSFERS, SUBSIDIES AND DOMESTIC CREDIT IN ARGENTINA, BRAZIL, COLOMBIA, MEXICO, AND VENEZUELA (as Percent of Total Expenditures and Lending Minus Repayment) Countries/Groupings 1978 1979 1980 1981 1982 1983 1984 1978-84 ARMENINA 40.58 45.12 48.04 43.06 41.16 62.25 56.73 48.13 (Transfers) (4.59) (3.72) (4.22) (5.58) (5.10) (8.45) (5.61) (5.33) Subsidies & Transfers 32.90 36.57 39.47 32.93 31.42 39.22 41.11 36.23 Not Lending 7.68 8.55 8.57 10.13 9.74 23.03 15.62 11.90 BRAZIL 68.94 70.72 72.12 72.66 69.22 72.22 66.56 70.35 (Transfers) 8.94 9,64 (8.42) (8.60) (8,44) (7.55) (7.93) (8.50) Subsidies & Transfers 48.94 48.86 53.02 47.41 44.34 40.59 37.92 45.87 Net Lending 20.00 21.86 19.10 25.25 24.88 31.63 28.64 24.48 COLOMBIA N.A. 42.39 39.19 41.39 40.98 44,26 N.A. 41.64 (Transfers) N.A. (22.87) (19.06) (21.16) (22.23) (25.17) N.A. (22.10) Subsidies N.A. 39.43 36.35 36.97 41.07 44.40 N.A. 39.64 Net Lending -2.08 2.96 2.84 4.42 -.09 -.14 N.A. 2.00 MEXICO 28.64 27.60 30.67 25.50 40.77 25.25 23.59 28.86 (Transfers (est.) (11.06) (11.76) (14.34) (14.05) (8.12) (12.79) (13.43) (12.22) Subsidies & Transfers 23.87 20.98 23.41 17.35 35.91 22.71 21.82 23.72 Net Lending 4.77 6.62 7.26 8.15 4.86 2.54 1.77 5.14 VENEZUELA 19.18 22,49 34.52 36.87 31.84 29.73 36.45 31.15 (Transfers) (12.19) (16.38) (12.44) (12.08) (13.00) (15.16) (12.16) (13.34) Subsidies 12.24 17.53 18.22 19.92 19.23 23,29 28.93 19.91 Net Lending 6.94 4.96 16.30 16.95 12.61 6.44 7.52 10.24 LAC COUNTRIES 45.62 46.57 48.76 47.44 49.58 50.11 48.30 48.05 (Transfers) (8.16) (8.66) (7.53) (8.26) (8.57) (9.11) (7.95) (8.32) Subsidies 33.02 33.47 36.16 32.04 35.98 33.51 33.40 33.94 Net Lending 12.60 13.10 12.60 15.40 13.60 16.60 14.90 14.11 IMD,. W. ECONOMIES 62.68 62.61 61.84 60.91 61.03 60.96 60.10 60.87 (Transfers) (14.48) (14.17) (13.35) (12.60) (11.63) (11.44) (11.42) (12.73) Subsidies & Transfers 56.28 56.41 55.94 55.71 55.83 52.26 55.60 55.43 Net Lending 6.40 6.20 5.90 5.20 5.20 4.70 4.50 5.44 Source: IM Government Finance Statistics Yearbook, 1986 and own elaboration. - 41 - 3.19 Ambiguities of Decentralization. In reality, central governments of LAC countries are trying to retain as much power at the center as is feasible. This is not only because decentralization tends to dilute the political influence of the national authority, but also because it may become difficult to cope with a hard pressed economy when expenditure decisions and priorities are independently set by subnational authorities. Therefore it is not always clear whether the broad declarations by many a LAC government reflect a real determination to let people reflect their preferences in local government budgets (which suggests more decentralization) or are merely promoting a more articulated structure for their civil service (which would remain centrally directed). Nor is it clear, in other cases, whether the main objective is to balance off through transfers the uneven fiscal capacity of different areas or rather to reserve to the central government the flexibility required to engage in resource mobilization and stabilization efforts (both of which suggests a central pattern of decisions). Real decentralization would consists in the autonomous right of different geographical jurisdictions to provide different types of services and in different amounts, as may be warranted by cultural and historical reasons, but also by other no less important factors, related to differences in climate, city size, geographical extension, agricultural practices, land tenure, topography etc. 3.20 Fiscal Reform Initiatives. As illustrated by the previous discussion, there seem to be as many arguments against fiscal decentralizations as there are in favor of it (See Box 1). In nearly all LAC countriesill the current state of intergovernmental fiscal relations is being questioned by financial authorities, political parties, provincial and local governments and business communities, all of which have different agendas, priorities and interests. In fact the subject itself is multi-dimensional and tends to cut across geographical groupings as well as capital and labor divisions. Changing attitudes, evolving social structures and political alignments result in insistent claims for a rising share of transfers and for new, distribution oriented programs. Nearly all countries have formed a commission on intergovernmental finance relations which is at work. In general, it remains difficult to pass judgement on these efforts in terms of principles, without considering the particular constraints that reform initiatives intend to remove (i.e. their particular objectives) or the detailed features of the proposals (i.e. their specific design) (See Box 2). 30/ Argentina, Brazil, Colombia, Mexico and Peru being the more glamorous examples. - 42 - Box 1: - ARGUMENTS ON DECENTRALIZATION in Favor Ag! nst STABILIZATION: The devolution of some fiscal STABILIZATION: The central government must retain authority to the local level may provide Incentives sufficient flexibility and control on fiscal policy: for independent fiscal Initiatives: local governments greater local fiscal autonomy could stifle national can select their oe buoyant tax bases better than Initiatives of tax relief. Also, the use of public the central government and people are are willing to Investment resources for Incremental local pay back locally for receiving tangible services. In expenditures may have undesirable Inflationary particular, local governments Could more effectively effects. increasing the real volume of transfers or charge for the marginal costs of urbanization in concessionary credit would aggravate thi already rapidly growing cities. Expanded access to credit critical position of Treasuries. Borrowing by could compement grants to support local Investments. subnational authorities, as well as the rollover of their debts, must comply with central restrictions. ALLOCATION: Local budgetary outcomes refSlect directly ALLOCATION: Promotion of new Investment must reeain voters preferences. Bringing closer to the people the under central control because capital Is short and power of deciding over expenditures and soe authority ought to be used to maximize economIc growth and to determine the level of taxation nould permit support production of tradeables. More local autonomy visible Improvements In local services and restore Is likely to result In greater claims for local confidence In government performance. Greater benefit expenditures. Local goods and avices ould confidence and satisfaction with local government may absorb public resources vtach should be steered to the also have some Indirect effects on the flows of labor production of goods with short-term national benefits. and capital. SPATIAL oISRIBUTIONt Subnatlona governments have SPATIAL olSTBUTION: Although sore knowledgeable on more detailed and accurate Information on the nature local technical and managerial matters, subnational and magnitude of demand for local services; they can authorities tend to be partisan and narrow-focussed, better adjust their own budgets In accordance to local hereas a wider persective is needed. They cannot needs and idpiement a more efficient and resonsive make sound decisions on the efficient distribution of vsatibl distribution of services. Disparities are resources among regions bause of the existence of groeat across region but also within regions and externalities and the mobility of factors of within cities. A satial distribution policy ought to production. Only the central government can sake be detailed at local level to better approximate regional policy decisions oifch maximize national existing discrepancies. Presently. most of the shared uelfare because of Its better knowledge of the country revenues and Investments accrue to large cities and as a Wtole and greater capacity to plan over the long more Industrallsed areas, end little resources trickle term. Increasing local tax poers Mould mostly down to the small towns and rural communities whare benefit larger cities or industrlized areas, leaving social srvices are most neded, virtually untouched the poorer rural corunites. ACCONT4ABILITY: Feed-backs with the local communities AOCOUNTABILI TY: To emphasize the redlistributlve would be mre effectively maintained, if local element of revenue-saring schems may be undesirable offitals a" made responsibie for raipsing funds to In times of economic recession: It may dnntivate sPare the costs of new Investments. Locai Initiatives fiscal effort both In the areas that contribute more mnd retonsible participation could i promoted, taxes and In those which receive larger transfers. particularly of iocai Interests and bus ness, result- Public expenditures ould tend to Increase In the Ing better decisions. Strengthening local democratic latter simply because more mosy I made avaIlable: Institutions mould help achievement of national goals unit output costs are ikely to Increase. It may be by educating voters to ont what might be done for batter to resume eMhasis on epatial redistribution them anyhoss people tend to identify e easily with an the economy Is more buoyant. The political decisions in tich t participate than with those stem Is less selective at to local than at the taken by higher levei authorities. Although local national level and less likely to produce the kind of voting outsmes may be less than perfect, periodic re- leadership that,is required. elections provide for corrections. AMNIIRATIVE EFFICIEICY: Decentralized Ae4eNsSIRATIVE EFFICIECY: In the short term, administration would encourage pubic aweness of administration of taxes and delivery of services are costs and benef Its of urban expenditures and concern better done by the centrai government which disposes over the design of new expenditures. One must start of siAlied professionals for them tess. Experienced this process by delegating responsibility._ With time, planners, analystn fIscal managers, assesso and tax technical assistance end adequate remneration, Inspectors tare too few to be di uted among dIfforent officials anywhere In the coun" can become competent ters of government. especially in the less developed to make rational decisions. The activities of countries. Some degree of ad einistrative centrally controlled buresucsts are not necessarlyv decenralizaton can be achieved by opening regional Mo efficient or responsibpe. but simply less subject branches of central governent ministries. to scrutiny. - 43 - @94_A; BRAZIL: A RECENT CASE OF FISCAL DECENTRALIZATION Background. The new Constitution promulgated on September 20, 1988, bears significant fiscal consequences for Brazil. The whoke tax system is being restructured and the revenue position of both the State and Municipal governments is expected to inprove substantially due to increasing transfers and the devolution of considerable fiscal authority from the federal to the subnationat levels. This is especially interesting because Brazil, by Latin American standards, did not present an extreme case of centralization. Main Features. The federal level is now restricted to seven taxes: Income (IR), Industrial Production (IPI), Rural Property (ITR), Imports (II), Exports (IE), Financial Operations (IOF), and Large Patrimonies (IGF). The yield of the first two sources is earmarked for revenue sharing with the subnational governments. The State level gains from the inclusion of fuel, energy, transport and cammunications in the base of Value Added Tax (ICHTC) and acquires the faculty of piggy-backing the federal Income Tax (IR) with a 5 percent surcharge. The Minipat Level increases its share of the ICMTC from 20 to 25 percent and has the authority to charge real estate transactions (ITIVBI), to introduce a fuel tax (IVVC) and to levy betterment charges in connection with capital inprovements. The new fiscal system is expected to produce an overall increase in tax revenues on the order of 3.2 percent (about half point of GDP) by FY 1993, but the federal government's disposable income (after revenue sharing and social security cortributions) wilt contract by 15.6 percent, the equivalent of 1.05 percent of GDP. In turn, the States revenues will increase by 12.9 percent (.73 percent of GDP) and those of the Municipatit as by 29.5 percent (.81 percent of GDP). The displacement of 1 1/2 points of GDP in the general government budget represents a rather drastic reversal of the historical trend from 1957 onwards which saw the federal government increase its participation in general tax revenues from 50 to 62 percent while the States and Municipal shares had declined from 44 to 34 percent and from 7 to 4 percent respectively. The new proportions wilt be 36.5, 40.7 and 22.8 percent for the Federal, the State and the Municipal levels. Spatial Ifmact. Simulations of the 1993 distribution of transfer resources, coupled with the expected increases in local revenues indicate measurable gains for the Northern and Northeastern States which wilt receive consistently higher increases than the 12.6 percent average State of Brazil, although also the State of Rio (+23.7 percent) and the Federal District (+17.8 percent) wilt fair exceptionatLy well. The average Municipality of Brazil wilt record a +32.2 percent increase in revenues, based on additional federal and state transfers (+33.1) and on new independent tax sources (+23.9). Per capita tax income in sample Municipalities would rise between 25 and 50 percent (especially in small and mediun size cities). There is, however, a danger that much of the new resources be used to foster growth of current expenditures. Impact on National Accounts. The heavy requirements of revenue sharing and the loss of some tax bases wilt limit the budgetary ftexibility of the central government. Additional budgetary pressures wilt arise from the new Constitution: On the expenditure side, incremental public outlays will be required by the following: (a) universal social security coverage; (b) more fringe benefits granted to workers and civil servants as wel as to non active personnel; (c) higher tax shares earmarked for education, health and famity/children assistance; (d) adminitrative costs related to new, state-Level functions, and to the establishment of four new States; (e) expenditure growth associated with the 1988 and 1990 elections. On the revenue side, shortfalls may be produced by: (a) tax exemptions, rebates and deferred payment altowances; (b) reduced control by the central fiscal authorities; (c) credit subsidies to the productive sectors (a 3 percent share of the main federal taxes (IR and IPI); (d) failure to incorporate many services in the State collected ICHTS; and (e) doubtful effectiveness of the federal tax on large patrimonies (:GF). Possible Fiscal Adjustments. The loss of resources of the Federal goverranent could be offset by the foLtowing: (a) expanding the IR and IPI coverage so as to compensate for the loss of other subnational contributions; (b) reducing current exemptions from these taxes (especially on agricultural income and capital gains); Cc) conditioning transfers to States and Municipalities to the payment of the corresponding credit obligations; (d) eliminating "negotiated" or "discretionary" transfers; (e) tying social security contributions to the gross profiz of enterprises; (f) staging the social benefits expansion over several years. Conctusions. Many questions remain, which need to be settled through the elaboration of a new regulatory framework. White no fixed calendar has been established for this second stage (except for the revenue sharing system), the following still await definition: (a) norms for the imposition of the income tax (IR) (b) rationing criteria for the territorial distribution of the participation funds; (c) establishment of a new Tax Code; (d) rules and procedures for the decentralization of public service functions; (e) redefinition of the budgetary categolies in public accounts; (f) establishment of a Public Finance Code, especially for the control of debt of States and Municipalities. j/ Abstract of IPEA paper No. 121 by F. Rezende and J.R.R. Afonso, November 1987. IPEA produced the initial design upon which the wide ranging constitutional debate was subsequently based. - 44 - 3.21 Summing Up. This Section has raised the question whether LAC governments are truly decentralizing. Several Latin American countries, especially the larger ones, are nearly as populated, urbanized and industrialized as the so called industrial market countries. However, not only is the absolute size of their economies considerably smaller in terms of GNP (which explains the low per capita incomes) but also the share of the economy that is controlled by their public sector is lower. LAC governments also spend a relative smaller portion of GDP in the social sectors. The most striking difference, however, is the comparatively low proportion of revenue and expenditure authority delegated to their subnational governments. During the 1980's the national economies have been contracting, while LAC cities have continued to grow at a sustained pace. The resources available to provide urban services at regional and local levels have declined dramatically in per capita terms and have not significantly increased even as a proportion of GDP. Therefore the proliferation of jurisdictional units and of special purpose urban service agencies should not be misinterpreted as a sign of growing local autonomy or more effective delivery of urban services. In most cases, it has only produced further budgetary fragmentation and a confusing pattern of fiscal flows between government levels and parastatal institutions. The corresponding growth of the public sector's current expenditures can also be traced to the politics of inter-governmental relations prevailing in LAC countries which are rooted in a devolutive (,:op- down) rather than pluralistic (bottom-up) tradition of public administration. 3.22 In addition, most LAG countries present sharper inter-regional disparities, due to the marked contrasts in factor endowment within their territories. These disparities bring about a public concern with inter-area equalization which frequently blurs the objective of inter-personal income distribution. The above contradiction feeds rhetorical debates on the reassignment of powers and functions among different levels of government in the name of a new 'Federalism' or 'Municipalism' into which disparate political, administrative, economic and fiscal issues only apparently converge. They also engender official redistribution schemes aiming to satisfy basic needs in the poorest areas of the countries or to compensate particular jurisdictions for their lower fiscal capacity. A brief review has been made of the arguments in favor and against decentralization and revenue sharing policies. 3.23 Looking from some distance at the current initiatives of fiscal reform of several LAC governments, one is Inclined to question them on efficiency grounds. One reason for this is the critical state of the national economies, which would rather call for centralized fiscal management. Another reason is the yet immature state of the democratic institutions through which local officials are expected to make budgetary decisions more responsive to the expressed needs of their constituencies. Sound fiscal policy would require that incremental public services expenditures be made only to meet effective demand (on the assumption that the taxpayers' personal income actually grows and with it also the willingness and ability to pay for services). It also implies that local governments dispose of sufficient Installed capacities to expand services so as to meet total demand by end-users and that they possess a degree of efficiency in capturing a share of the increased income of the community. An obvious corollary of this position is that fiscal autonomy cannot start with small towns and rural jurisdictions, where one is unlikely to meet the conditions - 45 - described above, but with the larger cities and Iae areas of greater fiscal capacity. Nevertheless, one cannot ignore that ur,an growth, industrialization and political evolution have altered the broad correspondence between the revenue authority and the functional responsibilities of the lower tiers of the government which had been envisaged in the original constitutional design of LAC countries. Some adjustments of the present fiscal structure is unquestionably needed. These adjustments are bound to be gradual, because of the contentious nature of fiscal politics, which makes radical reforms the least likely to be accepted and the most difficult to implement. - 46 - IV. THE FISCAL GAP OF LAC SUBNATIONAL GOVERNMENTS 4.01 Reasons for Fiscal Imbalance. Ideally, the current revenues of each level of government should be sufficient to cover its current expenditures as well as the debt service on outstanding loans, leaving adequate margin for capital investment or incremental borrowing. In fact, this is not Che situation for most regional and local governments in Latin America. The reasons for their current financial gap are many, as it will be pointed out below, and are often related to circumstances beyond the control of their budgetary authorities. Furthermore, the individual subnational units of each country present striking contrasts of fiscal capacity and fiscal effort and their expenditure decisions may reflect different degrees of wisdom and different priorities. Yet, some generalizations may be drawn across countries which are to a varying extent applicable to all their subnational communities. 4.02 Deficitary Trends. Observing the trends of revenues and expenditures over time, one can see that the subnational governments in LAC have run chronic deficits since 1974 and that these have been only aggravated by the 81-84 recession, reaching the level of several points of GDP in some countries (see Argentina, Mexico, Chile in Table 4.2). The explanation of this ten-year trend is to be found in both the revenue and the expenditure structure of the lower tiers of government and in the obvious mismatch between the resources available to them and the service responsibilities they have to confront. A. Revenue Side Factors (See Tables 4.3a and b) 4.03 General. First of all, a good portion of the tax bases assigned to regional and local governments are basically of a direct nature and are both politically and administratively costly to administer and collect. Second, the subnational revenue structure is not sufficiently buoyant, that is, not very well suited to afford a prompt response to the cyclical fluctuations of the economy. Third, those subnational taxes which are directly related to production, sales and income remain also quite vulnerable to the impact of centrally decided stabilization measures. Fourth, any stagnation of the economy affects different subnational communities to the extent to which they depend on revenue sources of this type: the sales and turnover taxes in particular tend to favor the jurisdictions where goods and services are purchased rather than those where they are produced. Fifth, subnational communities are allowed little latitude to adjust the rates of existing taxes, which are usually subject to central control, as are the corresponding procedures of imposition and collection. Sixth, the central governments maintain the same rate schedules and assessment procedures over extended periods, with the consequence that some taxes (e.g., the property taxes in particular) are capable of very low yields at current rates. Seventh, locally elected authorities show little commitment to adjusting taxes or user charges to the pace of inflation, which are obviously unpopular activities during - 47 - Table 4.2: OVERALL SURPLUS/DEFICIT BY LEVEL OF GOVERWMENT AS % OF GDP (Government Levels: G - General; C - Central; S a Subnational) 74 75 76 77 78 79 80 81 82 83 84 World G -2.05 -4.96 -4.62 -3.28 -3.75 -3.29 -3.85 -4.45 -5.25 N.A. N.A. C -1.47 -4.21 -4.00 -3.26 -3.33 -2.89 -3.33 -3.97 -4.63 -5.51 -4.81 S - .58 - .75 - .62 - .02 - .42 - .40 - .52 - .48 - .62 N.. N.A. Industrial G -1.88 -5.24 -4.85 -3.40 -3.84 -3.51 -4.06 -4.41 -5.04 -5.96 -5.06 Countries C -1.47 -4.51 -4.13 -3.30 -3.36 -3.03 -3.46 -3.91 -4.36 -5.67 -4.99 S -. 41 - .73 - .72 - .10 - .48 - .48 - .60 -z 50 -. 68 -. 29 - .07 Developing G -2.72 -3.87 -3.71 -2.83 -3.42 -2.43 -3.06 -4.62 -6.06 N.A. N.A. Countries C -1.45 -3.26 -3.58 -3.15 -3.22 -2.43 -2.92 -4.19 -5.48 -4.99 -4.24 S -1.27 - .61 - .13 + .32 - 20 .00 - .14 - .43 - .58 N.A. N.A. LAC G -3.39 -5.20 -4.11 -2.61 -3.03 -2.03 -2.67 -5.25 -7.41 -9.13 N.A. Countries C -1.77 -3.39 -2.94 -2.08 2.24 -1.45 -2.31 -4.43 -6.79 -5.97 -4.51 S -1.62 -1.81 -1.17 - .53 - .79 - .58 - .36 - .82 - .62 -3.16 N.A. Argentina G -10.25 -15.84 -9.88 -3.68 -4.51 -3.44 -4.86 -10.91 -8.37 -16.29 N.A. C - 6.1 -10.33 -7.11 -2.77 -3.21 -2.60 -3.57 - 8.15 -7.10 -11.84 N.A. S -4.10 -5.51 -2.7/ - .91 -1.30 - .84 -1.29 -2.76 -1.27 -4.45 N.A. Brazil G .71 -1.68 -1.54 -1.41 -2.68 -1.17 -2.83 -3.30 -3.9. -4.65 -5.01 C +1.27 - .50 - .18 - .84 -1.67 - .55 -2.32 -2.34 -2.'4 -3.44 -4.12 S - .56 -1.10 -1.36 - .55 -1.01 - .62 - .51 - .96 -1- 9 -1.21 - .89 Colombia G -1.57 - .21 + .98 + .65 + .75 - .74 -1.91 -2.70 -1.32 - .23 N.A. C -1.26 - .23 +1.00 + .63 + .68 -1.04 -1.85 -2.96 -1.64 - .79 N.A. S - .31 + .02 - .02 + .02 + .07 + .30 - .06 + .26 + .32 + .56 N.A. Mexico G -5.87 -5.73 -5.11 -3.68 -3.20 -4.17 -3.52 -7.71 -17.09 N.A. N.A. C -3.81 -4.87 -4.67 -3.30 -2.68 -3.32 -3.12 -6.68 -15.44 -7.95 -7.28 S -2.06 - .85 - .44 - .38 - .52 - .85 - .00 - .03 -1.65 N.A. N.A. Chile G -5.33 + .17 +1.41 -1.07 - .15 +4.87 +5.53 +2.38 -2.27 -2.61 -2.94 C -5.33 + .14 +1.37 -1.11 - .11 +4.82 +5.41 +2.59 - .98 -2.63 -2.97 S - .00 .03 .04 .05 - .04 + .05 + .12 - .21 -1.29 + .02 + .03 Source: IMF Government Finance Statistics Yearbook, Vol. X, 1986 Note: Subnational Government values obtained by difference. - 48 - Table 4.3as PER CAPITA REVENUES OF SUBRATIONAL GOVERNMENTS ai -In constant 1980 US dollars- 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE P/ 94 95 99 101 104 109 103 89 69 74 MARKET ECONOMIES h/ 2064 2190 2265 2383 2417 2327 2097 L009 2030 2046 a/ Revenues include own revenues and grants. hi Weighted Averages. Comparability over time affected by the availability of data. Table 4.3b: PER CAPITA REVENUES OF SUBNATIONAL GOVERNMENTS -Variation Index, 1975*100- ai 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE 100 100 105 107 110 115 109 94 73 78 MARKET ECONOMIES 100 106 110 116 117 113 102 97 98 99 a/ The variation index for the individual countries is based on constant national currencies time series while the variation index for the sample average is based on constant US dollars. Source: IMF, Government Finance Statistics Yearbook, various years. a period of steady erosion of personal incomes and wage control. Eighth, a large informal sector 1-s growing in the metropolitan areas of IAC countries, which remains outside the reach of fiscal authorities. Finally, the less urbanized areas are negatively affected by the loss of active population and enterprises to other regions. 4.04 Main Revenue Sources. For simplicity, the revenue sources of IAC regional and local governnments will be reviewed in four conventional categories: (a) Tax Revenues; (b) Non-Tax Revenues; (c) Intergovernment Transfers; and (d) Borrowing. The first two categories, which group the only truly independent sources of subnational revenues, are significant in those regional and local jurisdictions which are more urbanized, but have little weight in local rural comaunitics. Transfers become the most important category of revenues in predominantly rural areas. Borrowing has never been a significant source of revenues for local authorities as we shall see below. Table 4.4a shows the main structure of revenues of regional and local governments in Argentina, Brazil, Colombia and Mexico. The variation over time of the yield from different sources in constant national currencies is shown in Tables 4.4 b, c, d, & e in Annex). It is interesting to note the instability of the grant elements over the years - 49- (striking in the case of Argentina) and the sustained efforts by the governments of Brazil, Colombia and Mexico to maintain a substantial volume of transfers even during the recession. Table 4.41: REVENUE STRUCTURE OF SUBRATIONAL GOVERNMENTS -In Percentage- ARGENTINA BRAZIL COLOMBIA MEXICO 1975-84 1975-84 1975-83 1975-84 TOTAL REVENUE & GRANTS 10C.0 100.0 100.0 100.0 I. TOTAL REVENUE 74.6 70.6 49.3 40.1 1. CURRENT REVENUE 74.1 70.6 49 40.1 1.1 TAX REVENUE 62.4 60.2 39.8 21.4 TX ON INC.,PROFITS & K GAINS 21.5 2.1 2.2 3.4 TAXES ON PROPERTY 12.5 3.9 5.3 6.2 DOMESTIC TAXES ON GDS+SERVS 24.4 52.1 23.9 9 1.2 NONTAX REVENUE 13.4 10.9 9.2 18.7 ADM.FEES + CHGS NONIND SALES NA 1.9 0.4 18.7 OPERATING SURPLUS OF ENTERPRS NA 2.9 NA NA 2. CAPITAL REVENUE 0.6 0.2 0.4 N1. II. TRANSFERS 25.4 29.4 50.7 59.9 1. TX COLLECTED BY CENTRAL GOVT NA NA NA 52.3 2. GRANTS NA 29.4 50.7 7.6 GRANTS FROM OTHER NATNL GOVT 25.4 29.4 47.9 5.8 Source. IMF, Government Finance Statistics Yearbook, various years. 4.05 Taxes on Goods and Services. These may be levied by regional and local governments within their jurisdictions and include general sales, turnover or value added taxes; excises on goods, vehicles and passengers entering or leaving the jurisdictions; profits on fiscal monopolies (such as tobacco or alcoholic beverages); taxes imposed on the use or the permission to use specific goods (such as motor vehicles) or to perform business and professional activities; also in this broad category fall the taxes on entertainment, gambling or catering activities. In general, this group of taxes represents the main source of independent income for subnational governments in Latin America. In Brazil, for instance, the value added tax (ICM) accounts for 68 percent of the States revenues after sharing one fifth of its yield with the municipalities. In Colombia, taxes on goods and services (especially coffee, alcoholic beverages, tobacco and gasoline) constitute 55 percent of the Departments' income. The - 50 - Argentine Provinces derive some 26 percbnt of their total receipts from taxes of similar nature. In Mexico, where they are collected by the Federal Government, they represent the mainstake (63 percent) of State budgets after revenue sharing. 4.06 The main advantage of this category of taxes is that they are indirect and, remaining implicit in the purchase price of the goods and services, they are the least conducive to grassroots resistance. However, because their basic incidence is on general consumption, they tend to be regressive, which leads many LAC governments to exempt staple food items from this form of taxation. Where the tax has been converted into a value added tax (as in Brazil), most of the revenue tends to accrue to the jurisdictions which produce the goods and services and are the most industrialized and urbanized. Otherwise, they tend to shift the fiscal burden to the place of purchase rather than to that of consumption. A buoyant sub-category of these taxes are those on vehicle ownership, circulation and motor fueLs, which are of special significance to urbanized areas, where automobile usage is more intensive, especially among social groups with higher incomes. However, problems of registration arise for these as for all taxes on goods and service, which are difficult to apply in small municipalities, where an adequate census of local activities is lacking, and in metropolitan areas, because of the existence of a large informal sector, which does not appear in the fiscal records. 4.07 Taxes on Property. A second group of taxes is considered suitable for subnational use in practically all LAC countries, because they raise no issue of horizontal double-imposition. 11 These are the taxes on the property of land or real estate, on individual or corporate wealth and on capital transactions, inheritance or bequest. In theory, they would permit partial recovery of the public costs of urbanization, and could represent a buoyant source of revenue for rapidly growing cities without discouraging private development. In practice, a variety of structural reasons impede effective use of these instruments: first, LAC countries have little tradition of direct imposition and the very local and direct nature of the property taxes makes public officials the immediate targets of complaints from a large number of contributors; second, the fact that they are mostly levied on residential property makes them appear as levies on housing consumption, suggesting a regressive distribution of the fiscal burden; third, use of property taxation implies the hurdles of establishing a cadastre of the properties, (which means sorting out the peculiarities of tenure), keeping records up to date, establishing an assessment system and collecting the corresponding levies, all of which entails considerable II/ Double imposition, i.e., the possibility that different fiscal authorities may tap the same tax base (either vertically or horizontally) is a traditional subject of debate among Latin American experts who have constantly argued for a rigid separation of sources. Jorge Macon offers a more balanced view in his essay: Federalismo y Decentralizacion Fiscal en Lotinoamerica, CIEDIA, Buenos Aires, 1987. - 51 - administrative efforts and costs; fourth (a peculiar Latin American problem), coping with yearly assessments becomes an intricate operation during periods of galloping inflation and rapid urban expansion and engenders interminable appeals by taxpayers and reluctance, if not active opposition, by local politicians. 4.08 As a result, the effective yield of property taxes in IAC countries remains minor, compared to that of alternative sources (see Table 4.8a). In Mexico overall receipts from property taxes are 3 percent at the local level and less than 4 percent at the State level. In Brazil these taxes account for 1.3 percent of State and for 7.3 percent of municipal receipts. The only significant Table 4.8a: PROPERTY TAX YIELD AS PERCENTA'.E OF TOTAL REVENUES IN LAC COUNTRIES a/ -Subnational Governments- Average 1977 1978 1979 1980 1981 1982 1983 1984 - ARGENTINA Regional Government NA 25.0 12.5 12.7 13.0 13.7 16.3 16.1 15.6 BOLIVIA Local Government NA NA NA 13.3 11.1 11.5 6.2 3.9 11.8 BRAZIL Local Government 11.3 10.8 13.4 13.3 11.7 11.4 8.4 7.3 11.0 Regional Government 1.6 1.4 1.5 1.4 1.4 1.7 1.7 1.3 1.5 COLOMBIA Local Government 24.7 22.3 18.7 19.5 20.7 22.4 20.9 NA 21.3 COSTA RICA Local Government 31.5 33.4 32.7 23.9 27.1 25.4 18.5 19.9 26.5 MEXICO Local Government 3.2 3.5 3.2 3.0 1.7 1.4 1.5 6.0 2.9 Regional Government 10.0 9.2 8.3 5.4 3.1 3.2 2.6 NS 6.0 pa Total revenues include own revenues and grants. Source IMF, Government Finance Statistics Yearbook X987. - 52 - Table 4.8b: PROPERTY TAX AS PERCENTAGE OF TOTAL REVENUE IN MARKET ECONOMIES / -Subnational Goverments- 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 AVERAGE zWaamas HIGH TAXATION COUNTRIES AUSTRALIA Local Goverment 53.17 52.87 55.95 52.83 52.69 52.47 50.00 50.44 44.86 42.92 50.82 Regional Government 8.06 8.01 8.22 7.54 7.35 7.60 7.38 6.83 5.78 6.29 7.31 NEW ZEALAND Local Government NA NA NA 36.52 37.55 38.47 38.25 NA NA NA 37.70 CANADA Local Goverrmnsnt NA NA NA 33.62 34.79 35.99 35.92 35.20 34.95 35.94 35.20 Regional Government NA NA NA 1.22 0.24 0.21 0.21 0.21 1.31 1.30 0.67 UNITED STATES Local Goverrnant NA NA NA NA NA 28.75 28.40 29.09 29.67 29.91 29.16 Regional Governmnent NA NA NA NA NA 2.35 2.16 2.07 2.05 2.03 2.13 UNITED KINGDOM Local Goverment 26.54 25.49 28.03 28.04 28.65 29.62 31.03 32.04 30.78 30.59 29.08 LOW TAXATION COUNTRIES 2awsumm maunasan IRELAND Local Government 17.54 18.43 15.17 9.68 8.79 7.89 6.61 5.24 5.10 5.54 10.00 SWITZERLAND Local Goverrent 7.75 7.37 7.72 7.61 7.43 7.32 7.24 7.10 6.75 6.50 7.28 Regional Government 7.99 7.84 8.14 8.44 8.52 8.51 8.54 8.65 8.71 8.59 8.39 AUSTRIA Local Goverranent 5.90 5.85 5.57 5.69 5.65 5.60 5.38 5.27 5.50 5.60 5.60 Regional Goverment 0.44 0.43 0.44 0.42 0.40 0.40 0.41 0.42 0.81 0.79 0.50 GERMANY, FR Local Governent 5.03 5.46 5.31 5.19 5.12 4.78 4.83 4.89 4.90 4.83 5.03 Regional Goverrmnent 3.98 4.51 4.66 4.16 3.96 3.94 3.95 4.04 4.47 4.55 4.22 DENMARK Local Goverment 5.22 4.68 4.24 4.45 4.37 3.98 3.81 NA NA 2.68 4.18 LUXEMBOURG Local Government 3.60 4.67 4.33 3.94 3.48 3.51 3.22 3.01 3.16 3.15 3.61 SPAIN Local Government NA NA 7.48 8.98 7.08 1.63 1.76 1.40 0.92 0.85 3.23 NORWAY Local Goverinment 2.32 2.08 2.06 NA NA 2.45 2.45 2.51 2.78 3.05 2.65 ...................................................................................................... p/ Total Revenue comprises own revenues and grants. Source: IMF Goverment Finance Stntistics Yearbook, various years. - 53 - exceptions are those of Argentina, where property taxes account for 16.1 percent of provincial revenues and Colombia where Municipalities derive 21 percent of their receipts from this source.3 4.09 Other Taxes. Besides the taxes on goods and services and those on property, a variety of other taxes is applied by LAC subnational governments. These may include employers' taxes on payroll or manpower, taxes on construction activities or public utility enterprises, customs and export duties, profits on money exchange, licenses, poll taxes, stamp taxes. Many levies in this category are actually difficult and costly to collect relative to their yield. However, . when taken together, these sources of revenues may be more important and more successfully collected in LAC than the property taxes. In Mexico, for instance, they represent 37 percent of local and 6.3 percent of State revenues. In . Colombia they reach 22 percent of Municipal receipts and in Argentina 7 percent of provincial receipts. In a few LAC countries the subnational governments are also allowed to apply levies on personal and corporate income, profits and capital gains. While this latter types raise issues of double imposition with the central governments, they may be of some significance, as in Argentina (19 percent of provincial receipts) and Mexico (about 2 percent of both State and Municipal receipts). Finally in some countries (e.g., Brazil, Colombia) subnational governments may administer contributions to decentralized social security funds and workers' pension funds. 4.10 Non Tax Revenues. This broad category of receipts ranges from the operating surpluses of departmental enterprises, income from public properties or establishments, to other fees and charges on departmental sales, fines, forfeits and a variety of other levies. For reasons of simplicity one can include in this category also the capital revenues (generally small) which may derive from sales of public land or transfers of fixed assets and stocks. This group of sundry revenues accounts for considerable income in the Municipalities (38 percent) and States (15.5 percent) of Mexico, but are far less important in Colombia and Brazil, for instance. In nearly all countries, inflation has severely eroded the contributions of departmental enterprises, whose operating .2/ In spite of the copious literature advocating intensified use of property taxAs in rapidly urbanizing developing countries, it must be noted that the above statistics are not out of line with the actual yields of this type of imposition in market economies. (See Table 4.8b) Apart the cases of Australia, Canada, New Zealand, the U.K., and the U.S., where property taxes represent the backbone of local receipts (43,36,38,31 and 30 percent respectively), the contribution of property taxes to the municipal budgets is quite limited in countries like Austria (5.6 percent), Denmark (4.0 percent), Ireland (5.5 percent), Germany (4.8 percent), Luxemburg (3.2 percent), Norway (2.6 percent), Switzerland (6.5 percent). These figures are no better than those encountered in Latin American countries. (See Table 4.8b). - 54 - surplus has vanished with the rapid devaluation of user charges on public utilities such as water, sewerage, electricity, and telephone services. The IMF records show that, except for few exceptional cases (thay still allow for some 3 percent of total revenues in Brazilian states and municipalities), this item has virtually disappeared from Latin America subnational finance statistics. A more buoyant form of non-tax revenues and one directly related to urban development are betterment charges, or lump-sum payments through which beneficiary landlords must compensate the public authorities for the social costs of infrastructure improvements. The relative importance of these receipts and the extent to which they meet investment costs is difficult to determine since betterment charges basically accrue to the departmental enterprises involved and do not appear explicitly in government statistics. Although most Latin American cities are entitled to make use of betterment charges, which are reportedly applied with some success at the local level in Colombia, Argentina, Brazil and elsewhere, their administration is not always simple and experience with this instrument remains rather mixed. 4.11 Intergovernment Transfers and Revenue Sharing. Besides their own independent sources of tax and non tax revenues, briefly reviewed above, transfers from the higher levels of government constitute the most relevant item in the budgets of regional and local authorities in Latin America. These may take place in form of outright grants or in form of revenue-sharing. In practice, the central (or regional) government reserves the right to collect specific taxes on the local base and to remit a set proportion of their yield to the originating jurisdictions.21 As most of these resources are derived from the more affluent urban areas or from industrialized regions, redistributive criteria may be incorporated in such schemes. The amounts granted to particular jurisdictions may be: (i) directly proportional to their contributions (origin based schemes); (ii) independently decided by superior authorities on the basis of perceived needs (discretionary schemes); (iii) attributed on the basis of pre-determined criteria (formula-driven schemes); or (iv) simply refunding local expenditures on programs previously approved by superior authorities (reimbursement schemes). In all cases, intergovernmental transfers imply a 13J In Venezuela, as in Colombia, revenue sharing takes the form of a tax allowance (Situado Constitucional and Situado Fiscal respectively reverting to the States or the Departments). These Tax allowances represent a fixed percentage of all ordinary tax revenues of the central government, with certain exceptions. - 55 - degree of fiscal dependency of the local communities from an upper layer of government which may be central or intermediate when a regional authority is charged with the internal re-distribution of the funds.2-1 4.12 Dependence on Transfers. Table 4.12a shows how central government transfers have maintained their importance as a source of income for subnational budgets, particularly at local level, in spite of the cyclical fluctuations of the economies. Expressed in constant domestic currencies, the national trends show a variable pattern with a progressive contraction in Chile, Panama, Paraguay and Uruguay, but significant upward movements in Argentina, Brazil, Colombia, Costa Rica and the Dominican Republic (Table 4.12b). However, the influence of rapid demographic growth largely offsets gains in per capita terms also in the latter countries (Table 4.12c). A detailed case study for Brazil (see Box 3) shows the differential impact of transfers on state budgets and on municipalities of various categories. Table 4.12a: INTERGOVERNMENTAL TRANSFERS AS PERCENTAGE OF TOTAL REVENUES 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE g1 39.5 38.0 38.2 33.5 38.5 40.0 41.6 42.8 37.8 37.6 MARKET ECONUMIES c/ 35.1 35.4 34.8 35.05 35.1 34.8 34.2 33.2 32.5 31.3 aI Total Revenues - Own Revenues + Grants. kl Comprises Grants and Revenue Sharing. Sl Weighted averages. Comparability over time affected by the availability of data. Source: IMF, Government Finance Statistics, various years. L4/ Some alternative ways to classify intergovernmental transfers can be the degree to which the transfers are requited (i.e., involve total or partial repayment as in the case of loans or the contribution of a concurrent, "matching share" by the recipient in form of a quantifiable benefit, product or service to the payer in return for the payment). A second criterion of classification can be the extent to which the transfer funds must be put to a specified use, and monitored by the giving authorities ("earmarked" or "categorical" transfers vs. "unmarked" or "block" transfers). A third criterion can be the extent to which the pool of resources available for transfer in a given year is "open-ended" or "limited", i.e., whether or not there is a set maximum amount that must be rationed among all eligible parties according to an established "formula" or whether "reimbursement" funds will be made available to cover all technically eligible expenditures. - 56 - Box 3: CASE STUDY: SUBNATIONAL FINANCES IN THE NORTHEAST REGION OF BRAZIL The following indicators were used by Rezende to analyze the financial situation of 9 States and 90 Municipalities of the Northeastern Region of Brazil over the period 1979-1984: RC: Resource Commitment: Current Expenditures + Interest Payments + Amortization/Total Receipts - Borrowing ROR: Reliance on Own Revenues: Own Revenues/Total Receipts - Borrowing IER: Investment/Expenditure Ratio: Investments/Total Expenditures PCI: Per Capita investment: Investment/Total Resident Population SIR: Savings/Investment Ratio: Total Receipts - Borrowing - Current Expenditures - Interest - Amortization/Investment DTT: Dependence on Total Transfers: Total Transfers/Total Receipts - Borrowing DET:* Dependence on Earmarked Transfers: Earmarked Transfers/Total Receipts - Borrowing OUT: Dependence on Untied Transfers: Untied Transfers/Total Receipts - Borrowing ONT: Dependence on Negotiated Transfers: Negotiated Transfers/Total Receipts - Borrowing DST: Dependence on State Transfers: State Transfers/Total Receipts - Borrowing (Municipalities Only) In 1983 the situation of the 9 States was as shown below: States RC ROR IER PCI SIR OTT DET DUT DNT DST Plaul 0.56 0.29 0.22 24,769 1.82 0.71 0.05 0.27 0.38 .28 Mararhao 0.46 0.30 0.22 13,171 2.67 0.64 0.08 0.39 0.17 .32 Sergipe 0.38 0.47 0.24 43,510 2.86 0.53 0.03 0.25 0.26 .32 R.G. Norte 0.57 0.49 0.13 15,157 3.12 0.51 0.06 0.28 0.17 .41 Paraiba 0.69 0.52 0.11 13,394 2.23 0.48 0.03 0.29 0.15 .50 Ceara 0.76 0.62 0.01 11,201 15.65 0.38 0.02 0.28 0.08 .57 Pernambuco 0.48 0.68 0.13 16,083 4.23 0.32 0.04 0.15 0.13 .64 Alagoas 0.56 0.70 0.03 5,113 12,08 0.30 0.03 0.22 0.05 .66 Bahia 0.48 0.73 0.08 14,897 5.11 0.27 0.05 0.14 0.08 .63 Observations: (a) only Bahia, Alagoas, and Pernambuco can meet their resource commitments with internally generated revenues, as shown by the difference RC-ROR. (b) some States are very dependent upon transfers, with negotiated (discretionary) transfers representing a very high proportion of the total, followed by mandatory transfers. Earmarked transfers represent only a minor contribution everywhere. (c) States receiving more transfers invest the most in per capita terms, Irrespective of their poor coverage of recurrent expenditures through Internally generated revenues. (d) tax receipts are highest in the more developed States (Bahia, Pernambuco, Alagoas) because of the greater yields from the value added tax (1CM). Non-tax receipts supply lIttle income, except in Sergipe and Rio Grande do Norte. - 57 - Box 3: CASE STUDY: SUBNATIONAL FINANCES IN THE NORTHEAST REGION OF BRAZIL (CONT.) Also In 1983, the situation in the 90 Municipalities, classified by type,. was as follows: Municipalities RC ROR IER PCI SIR DTT DET DPT DNT DST GPCI Metropolitan 0.70 .27 0.21 23,675 0.90 0.73 0.01 0.22 0.06 0.38 (30.0) Industrialized 0.63 .25 0.24 47,850 0.35 0.75 0.01 0.13 0.02 0.46 (20.1) State Capitals 0.75 .41 0.16 16,498 1.11 0.59 0.02 0.15 0.07 0.36 (40.6) Residential 0.73 .17 0.24 6,678 1.24 0.83 0.02 0.39 0.10 0.33 (51.6) Agglomerations 0.70 .25 0.26 10,003 1.12 0.75 0.01 0.35 0.05 0.34 (36.6) Industrialized 0.72 .26 0.28 13,405 0.98 0.84 0.01 0.32 0.03 0.48 (24.8) State Capitals 0.86 .33 0.12 8,251 1.04 0.67 0.02 0.27 0.11 0.27 (61.3) Residential 0.52 .27 0.39 8,354 1.34 0.73 0.01 0.45 0.60 0.28 1.8 isolated Centers 0.76 .18 0.19 6,688 .65 0.82 0.01 0.40 0.12 0.28 (18.0) >100,000 inhabit. 0.75 .23 0.22 8,914 1.31 0.77 0.01 0.36 0.16 0.25 (9.6) <100,000 Inhabit. 0.78 .14 0.16 4,462 1.91 0.86 0.01 0.44 0.08 0.32 (30.9) Observations: (a) Independent revenues generally cover less than half of the Municipal resource commitments. (b) per capita Investment is quite high in metropolitan areas and secondary irban agglomerations, where their proportion is nearly four times as much as in the isolated centers. (c) per capita investment is highest in industrialized communities and administrative capitals. The less urbanized the municipalities, the lowest their average investment. (d) local revenue generation is more significant in the administrative centers of both metropolitan areas and urban agglomeration and lowest in non Industralized communities. (e) categorical and negotiated transfers do not represent an important share total transfers except in the administrative centers. (f) Federal transfers (tied) are relatively more important in small centers and residential communities. (g) State transfers are specially significant In the industrialized communities which receive important.shares of the value added tax. a/ The sample analyzed included 90 Municipalities with more than 30,000 inhabitants in 1980 out of a universe of 1,390 Municipalities. The sample was organized as follows: (a) 22 Municipalities belong to 3 Metropolitan Regions; (b) 35 Municipalities belong to 9 Urban Agglomerations; 33 Municipalities represe.t isolated centers of which 6 had populations greater than 100,000 inhabitants, 10 had populations between 50,000 and 100,000 inhabitants and 17 had 30,000 to 50,000 inhabitants six Mu6!cipalities were considered relatively industralized, six predominantly residential and nine were state administrative capitals. (Source: Brazil: Urban Development in the Northeastern Region, Special Sector Report, IBRD NO. 6585a-BR, May 6, 1986). - 58 - Table 4,12b: TRENDS OF INTERGOVERNMENTAL TRANSFERS - Variation Index, 1975 - 100 - 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE 100 99 108 118 128 152 134 154 128 133 MARKET ECONOMIES 100 105 110 116 118 118 109 106 106 106 Sourcei IMF, Government Finance Statistics Yearbook, various years. 4.13 Borrowing. Direct external borrowing is usually not allowed to subnational governments, which makes good sense in exchange short countries, but "onlending" arrangements are often adopted between the centrf government and subnational entities, especially regarding officially apprvved projects or programs bearing international financing. Since the exchange risk is usually carried by the central authorities and the same repayment of the subloans not infrequently forgiven, such concessionary lending can be considered as just another form of intergovernment transfer. Presently, the access of LAC subnational authorities to domestic capital markets is merely nominal. This is not only due to the fact that the money available is little, short-term and very costly, but also to the existence of regulatory constraints which were designed to prevent local authorities from incurring debt and possibly making unwise investment decisions. This restraints usually involve lenghty procedures of loan review and approval, unwieldy ratios of debt-service to revenues, matching funds contributions, short maturities and cumbersome disbursement procedures. The current financial and regulatory environment also discourages the so called "autonomous" intermediaries from tapping private savings on behalf of municipalities by issuing bonds or debentures, which are rarely used to finance local investment. In pracqLce, the central banks or the financial authorities set annual ceilings to domestic credit for the whole public sector with very low priority to subnational governments amongcometn official borrowers. 4.14 In most countries, different forms of institutional credit are being used to substitute for the scant development of commercial borrowing. In this fashion, substantial public resources are channeled to departmental enterprises and to regional and local governments (Table 4.14), mainly on a concessionary basis. Subnational authorities are often allowed to cover budget deficits with overdrafts or to undertake short term loans and other stop-gap arrangements with local banks (a feature which in a disrupted situation as that of Brazil led many State Banks to catastrophic results). Official credit 'windows' and/or Municipal Development Funds (MDF's) have been established nearly everywhere, capitalized through loans or deposits by the central banks or by publicly controlled pension and insurance schemes. External funding has also been used occasionally to - 59 - establish lines of credit or MDF's specially when subloans were to be used for capital improvements. However, the largest share of resources were usually provided through appropriations from the government budgets or direct allocations of a fixed percentage from the revenue of particular taxes. Because the initial capital is usually unrequited and the mix, origin and actual cost of subsequent contributions frequently unpredictable, "spreads" remain somewhat arbitrarily determined and are usually insufficient to assure significant internal generation of resources over time. Past experience with MDF's in Latin American countries (Bolivia, Brazil, Colombia, Costa Rica, Dominican Republic, Guatemala, Honduras, Mexico, and Venezuela) show that they generally failed to become self-supporting or to achieve a significant scale of operation. The three case studies reviewed Table 4.14: CENTRAL GOVERNMENT LENDING MINUS REPAYMENT AS PERCENTAGE OF TOTAL EXPENDITURE p/ Average 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 - LAC SAMPLE b/ 12.8 11.6 14.1 12.6 13.1 12.6 15.4 13.6 16.6 14.9 13.7 MARKET ECONOMIES = 5.6 4.4 6.8 6.4 6.2 5.9 5.2 5.2 4.7 4.5 5.;, g/ Total expenditures include expenditures and lending minus repayment. ki Weighted averages. Comparability over time affected by the availability of data. Source: IMP, Government Finance Statistics, various years. in Box 4 are rather typical of the use of institutional credit in LAC countries. Most subnational governments (especially the smaller municipalities) are yet unaccustomed to servicing debt and are reluctant to adopt borrowing as an instrument of budgetary policy. The overwhelming dominance of traditional patterns of revenue sharing systems has in fact discouraged them from doing so and has fostered their chronic dependence on decisions taken in the upper levels of government. Nevertheless, the acute shortage of fiscal resources to increase grants to the desired level over the 1980's has revived interest in the MDF's as an alternative means to finance urban investments. The international lending agencies (especially IBRD and IDB) have partly encouraged this trend because it affords a wholesale solution to the problem of supporting a spatially diffused pattern of small investments, which could not be funded through conventional project lending. Consequently, external financial assistance tends to associate - 60 - Box 41 INSTITUTIONAL CREDIT IN BRAZIL, COLOMBIA, AND MEXICO In Brazil, subnational borrowing is linked to tax revenues, which allows the wealthier jurisdictions to achieve considerably higher levels of borrowing. The domestic indebtedness of States and Municipalities is ruled by Federal Senate Resolution 62 of October, 1975, which states that: (1) Total debt outstanding cannot exceed 70% of the revehues of the previous fiscal year (adjusted for 'tflation). (1l) Incremental debt contracted in any year cannot exceed 25% of total revenues. (ill) Debt service (interest and amortization) cannot exceed 15% of the revenue in the previous fiscal year. However, a subsequent Resolution No. 93 freed many credit operations (Urban Development, Social Development and the National Housing Bank Loans) from the preset limits upon approval by SEPLAN, the National Monetary Council and the Federal Senate. The growth of indebtedness allowed by these exceptions (extra-limit credit reaching about 61% of the total debt of subnational institutions in 1983) was not spread evenly either in terms of type of debt or of category of jurisdictions. In general the smaller municipalities were less favored by the criteria used for extra-limit credit approval and were less accustomed to borrowing. The use of the statutory indicators of debt capacity would show that the legal limits have not been reached in the majority of cases. Nevertheless, a more rigorous analysis could show a different picture as small municipalities tend to stretch their spending beyond their available resources by turning to direct suppliers credit, by postponing due payments, by drawing advances on next year's receivables, by not collecting contributions from public employes. Although such short-term devices are also used by the larger municipalities and by the states themselves (which makes the actual records of their financial position less than transparent) It would seem po sible in many instances to finance urban improvement programs through credit, especially if the planned investments are likely to increase local revenues, so as to meet future demands of debt amortization.L/ According to the Central Bank statistics, in 1983 the external debt of States and Municipality of Brazil accounted for 33% of their total obligations and the internal debt for 70%. Between 1980 and 1984 subnational debt grew only 3% per year (.3% in per capita terms) at a slower pace than subnational revenues. In Calombia, because of the generally low level of the municipal budgets, only the largest centers are able to allocate a substantial proportion of their resources to capital investment. According to Bank staff estimates, many of the Intermediate-size municipalities have virtually exhausted their existing borrowing capacity. Widespread concern over the status of local finances and the underdevelopment of local ecorcmles prompted the National Congress to pass new laws (no. 14 of 1983 and No. 12 of 1986) increasing substantially the fiscal autonomy and revenue generating capabilities of the municipal governments. The debt servicing potential of smaller communities has been increased by earmarking an additional share of the value added tax allowance for decentralized investment but this requires an expanded role of public agencies both to assist these local units in the mobilization and management of resources and In actual lending to them b a/ See IBRD report No. 6585a-BR: Brazil: Urban Development in the Northeast Region, May 1986. b/ See lt;D report No. 6631-CO: Colombia: A Review of Recent Fiscal Decentralization Measures; January 1987. - 61 - Box 4: INSTITUTIONAL CREDIT (CONT) Municipalities can borrow for priority capital projects from the Territorial Credit Institute (ICT) and the National Urban Development Fund (FFDU). The first is a large public institution concentrating on the financing of low income housing and urban infrastructure. ICT derives its resources through mandatory credit from the financial system (commercial banks, insurance companiei and savings and loans associations), some annual support from the national budget and periodic transfers from the Treasury to replenish its capital. Continuing decapitalizaticn of ICT is caused by its use of highly concessionary interest rates (presently lower than inflation), high operating costs and occasional resort to more costly sources of financing in order to fulfill government mandated targets of invest- ment. Portfolio revenues are plagued by arrears and insufficient even to remunerate the financial system for its forced investments in ICT programs. FFDU is Instead a trust fund of the Central Mortgage Bank (supported by an IDB loan), which is providing credit to small municipalities for priority capital Improvements along with some technical assistance in financial and institutional analysis, project preparation, appraisal and implemen- tation. Municipalities can also expand their borrowing capacity through the purchase of government guaranteed bonds, which allow them to triplicate available investment resources. Bank staff projections indicate that FFDU will need substantial re-capitalization to assume a larger role of Intermediation In providing local investment credit. Legislation has been drafted to restructure FFDU as an independent financial Institution for urban and regional development. In Mexico, credit is supplied to States and Municipalities by the National Bank of Public Works and Services (BANOBRAS), which operates a number of trust funds through a network of regional offices (one in each State), using resources drawn from foreign loans (37), savings and demand deposits (47), allocations from the national budget and repayments from the existing portfolio (16%) _2 The main trust funds are FIFAPA (Water Supply and Sewerage Investment Fund) which lends primarily to medium size cities and is supported by a US$265 million World Bank loan: PACDU (Concerted Action Program for Urban Development) lending for Infrastructure works to four Gulf States (World 9ank loan of US$164 million); FONHAPO (National Low-income Housing Fund) lending to States, Municipalities and Building Associations (World Bank loan of US$150 million), FOMUN (Municipal Promotion Fund) primarily supporting community facilities and infrastructure provision in small municipalities (DB loan of US$80 million), and FORTAMUN (Municipal Strengthening Program: World Bank loan of US$40 million); focused on municipal technical assistance and basic infrastructure investments. In general credit terms and conditions are highly concessionary and the Central Government must offset the rapid decapitalization of BANOBRAS through yearly budgetary support or by forgiving due repayment. BANOBRAS uses different maturities and Interest rates in its loans depending on the type of investment, the relative need and size of the Municipalities and their classification according to nationally defined development priorities. Despite the high rates of subsidy involved, BANOBRAS credits are not being rapidly absorbed by States or Municipalities. This is due primarily to their lack of experience in borrowing, their difficulties in projecting debt service capacity in times of recession and macro-inflation and the alternative expectation of discretionary grants. Subnational authorities usually resist the prospects of indexation formulae or loan covenants Imposing tariff-upgrading. They also bulk at the procurement conditions and the bureaucratic procedures and delays involved in the approval and pass-through of the funds. c/ See 'BRD report No. 7157-ME: Mexico: Financing States and Municipalities: Trends, issues and Recommendations, World Bank, Ndrch 1988. - 62 - reforms of the local revenue sharing systems with the restructuring and establishment of Municipal Development Funds and aims 1o mobilize additional resources, to improve their utilization through technical appraisal of local investment proposals and to adjust the pricing of capital as well as that of urban services.5 B. Expenditure Side Factors 4.15 General. Different considerations apply to the growth of expenditures, which increase much faster than revenues. First of all, most subnational expenditures are made in labor-intensive sectors and therefore result in high recurrent costs, while offering low levels of recovery. This is the case of education, health and welfare services, but also of police and fire protection, general administration and a variety of other urban functions which are not directly chargeable to end-users. In rapidly urbanizing areas, the demand for these services grows faster than the taxable population. Local leaders are subject to constant social claims to increase payroll expenditures or social services and their short terms of office make them not always accountable for the consequences. Moreover, subnational authorities receiving capital improvement grants or loans from the central government, are subsequently saddled with the problem of staffing, operating and maintaining the new facilities under job-protection and minimum wage policies which are externally determined. Capital expenditures present a different set of problems. Certain urban investments (water supply, sanitation, power, drainage, transport facilities) are lumpy and require large sunk investments to achieve economies of scale. A conventional response to resource constraints is to neglect maintenance of the existing capital stock and to defer to better times their needed upgrading. Over time this results in a backlog of obsolescent housing and infrastructure, deteriorated environmental conditions, water rationing and intermittent power failures. Finally, categorical grants or project loans made to encourage new capital formation tends to concentrate counterpart funds on those subsectors for which external loans are more readily available. 4.16 Subnational Participation in Public Expenditures. Unfortunately, the allocation pattern of subnational expenditure by economic type or function is difficult to trace for the LAC countries as a group, because this information is only available for few individual countries. IMF data on expenditures by function are generally limited to central government operations and the distinction between recurrent and investment expenditures is not always available at the subnational levels. There are also problems in separating the expenditures made directly by the subnational governments and those made by decentralized sector agencies of the central government within their territorial jurisdictions. With these qualifications, one can still infer from IMF statistics, some broad comparisons of the subnational expenditure patterns prevailing in different groups of countries. A first observation is that LAC j5/ See the recent review by K. Davey: Municipal Development Funds and Intermediaries, PPR Working Papers, World Bank, 1988. - 63 - subnational government participate in general public expenditures in about the same proportion of GDP as that applicable to all developing countries (about 5 percent in Table 4.16a) but this percentage is considerably lower than the 10 percent value recorded for the industrial market economies, even when expenditures supported by borrowing are included (Table 4.16b). Expressed in percentage of general government expenditures, the split of responsibilities between central and subnational authorities is definitely unfavorable to the latter, which control directly only some twenty percent of total public outlays. (Table 4.16c) This picture gets even worse if one considers separately subnational expenditures in social services (Table 4.16a). During the year examined (1982), expenditures by lower tier governments for education, health, social security, housing and community facilities were much lower in Latin American couptries (3.5 percent of GDP) than they were in the industrial market economies (6. percent of GDP). 4.1.7 The Composition of Subnational Expenditures. Some indications in this respect can be derived from the time series available for Argentina, Chile, Colombia, and Panama. In Table 4.17a (Annex) one can note that social services generally account for the largest share of the subnational budgets, fluctuating between 28 percent (Panama) and 68 percent (Colombia). Also significant is the fact that this kind of expenditure has kept growing over the 1980's in constant terms, at an even if not faster pace than the resources of the subnational government (See the variation Tables 4.lFa to 4.18d in Annex). Expenditures for education command in general the first place and can be as high as 42 percent of total in the case of Colombia where they grew rapidly, as they did in Argentina, and hovered about 20 percent in other countries (except for Panama, where the subnational role in education appears minor). Health expenditures are of the order of 16 percent in Colombia but seem to be far less of a subnational charge in other countries. Social Security and Welfare are significant in Argentina and Brazil (13 and 9 percent respectively), due to the relative decentralization of their systems, but account for 5 percent or less of total subnational expenditures elsewhere. Housing and Community fa ilities grew quite rapidly in the subnational budgets of Colombia and Panama, and averaged some 5 to 10 percent of subnational budgets. General Services (Administration and public order) command a considerable share of subnational outlays, ranging from 19 percent in Colombia up to 65 percent in Chile. In real terms this typr- of expenditure has been growing quite steeply in Colombia, where their weight is still moderate, but remained constant in Panama where they are very high (68 percent). Outlays for Economic Services, generally in the order of 15 percent, experienced lower than average growth, with transportation investments accounting for the main share. Obviously, this generalized picture tends to blur structural differences between regional and local level budgets and between the expenditure patterns of metropolitan areas and those of small rural communities. Detailed expenditure studies for Brazil, for instance, show that significant correlations exist between the level of urbanization of local jurisdictions and the composition of their expenditures. - 64 - Table 4.16a: 1982 EXPEDI1TURES BY FUNCTION AND LEVEL OF GOVERIMENT AS PERCENT OF GDP Industrialized Developing LAC (a) Contries Countries Countries Argentina Brazil Chile Colombia Panama GENERAL GOVERiMENT EXPENDITURES 44.6 32.6 32.1 31.1 39.1 34.7 20.8 40.1 General Services (a) N.A. N.A. N.A. 2.4 6.7 5.9 3.9 11.4 Economic Services N.A. N.A. N.A. N.A. 6.3 3.1 5.3 5.1 Social Serv!ces 24.5 11.3 13.2 11.5 14.3 24.2 10.7 14.2 - Education 5.1 3.6 3.4 2.9 2.9 6.2 5.4 4.2 - Health 5.3 1.8 1.3 1.0 2.4 2.4 1.2 5.0 -Soc. Sec. & Welfarel I I I I I I -Housing & Community] 14.2 5.8 1 8.4 3 7.4 1 9.0 ] 15.6 1 4.0 1 5.1 Amenities 3I ] 3 3 1 Social Expenditures as % of Total Expend. (55.0) (34.6) (4s.0) (37.0) (36.5) (69.7) (51.4) (35.4) CENTRAL GOVERNMENT EXPENDITURES 34.1 27.2 27.2 22.7 27.7 31.2 15.1 39.4 General Services (a) .7 .8 .6 1.6 2.5 3.7 2.8 10.9 Economic Services 3.1 5.8 5.2 3.6 4.6 5.1 4.3 5.1 Social Services 18.4 7.7 9.7 7.6 10.0 22.8 7.1 14.0 - Education 1.4 2.3 2.3 1.3 1.0 5.0 3.1 4.1 - Health 3.8 1.0 1.2 .2 1.6 2.3 .6 5.0 -Soc. Sec. & Weltdre 12.4 3.9 5.8 5.8 7.4 14.3 3.0 3.1 -Housing & Community .8 .5 .4 .2 .1 1.1 .4 1.7 Amenities Social Expenditures as % of Total Expend. (54.0) (28.1) (35.5) (33.4) (36.1) (72.9) (47.4) (35.4) SUBNATIONAL GOVERNMENT EXPENDITURES 10.5 5.4 4,8 8.4 11.4 3.5 5.7 .7 General Services (a) N.A. N.A. N.A. .8 4.2 2.2 1.1 *4 Economic Services N.A. N.A. N.A. N.A. 1.8 N.A. .9 Social Services 6.1 3.6 3.! 3.9 4.3 1.2 3.5 .2 -Education 3.6 1.3 1.1 1.6 1.9 1.2 2.4 -Health 1.5 .8 .2 .9 .8 .1 .6 -Soc. Sec. & Welfarel 1 1 1 1.4 1.0 .- .2 -Housing & Community) 1 1.0 1 1.5 1 2.2 .- .6 .- .3 .2 Amenities ] 1 I Social Expenditures as % of Total Expend. (58.3) (67.2) (71.8) (47.0) (37.5) (36.0) (62.0) Source: Staff elaborat!on of data from IMF, Government Finance Statistics Yearbook, Volume X, 1986 Expenditure totals include Lending Minus Repayment and Transfer to Other Levels of Government. Expenditures data for specific functions exclude net Lending flows. (a) LAC regional averages keighted against GDP on a wider sample of countries. (b) Includes General Administration and Public Order but excludes Defense expenditures. - 65 - Table 4.16b: SUBNATIONAL GOVERNMENT EXPENDITURES AS PERCENTAGE OF GDP I 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE b/ 7.6 7.1 7.0 7.3 6.9 7.1 7.1 7.9 7.6 7.3 MARKET ECONOMIES h/ 10.4 10.3 9.7 9.9 9.8 10.4 10.2 10.6 10.2 9.8 Al Total expanditures include expenditures and lending minus repayment. hl Weighted averages. Comparability over time affected by the availability of data. Table 4.16c: SUBNATIONAL GOVERNMENT EXPENDITURES AS PERCENTAGE OF GENERAL GOVERNMENT EXPENDITURES a/ 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC SAMPLE b/ 24.4 23.8 22.8 23.0 23.0 23.0 21.0 20.1 19.8 20.4 MARKET ECONOMIES b/ 26.5 26.5 25.4 25.5 25.1 25.4 24.7 24.7 23.8 24.0 1/ Total expenditures include expenditures and lending minus repayment. Expenditures have been consolidated. h/ Weighted averages. Comparability over time affected by the availability of data. Source: IMF, Government Finance Statistics yearb-ok, various years. 4.18 Constraints to Reducing Local Expenditures. It is quite plausible that the availability of transfer resources feeds the growth of expenditures, but it is unlikely that stricter rationing of this type of funds would achieve the desirable effect of balancing off local budgets, especially in rapidly urbanizing areas, because of the concern with maintaining a basic level of welfare for their growing populations.6' The common trends of Argentina, Chile, Colombia and Panama, (Tables 4.18 a, b, c & d in Annex) show that social expenditures continued to rise faster than those in economic sectors. The devolution of responsibilities for services like primary education or health to the lower tiers of government took place not so long ago and was based on the assumption that resource transfers supporting these expenditures would continue to grow over time. Additional problems arise from the repayment of capital loans received in the past by local governments and from the growing needs to replace a by now obsolescent capital stock. Certain capital improvements (e.g., water supply, sanitation, drainage, power) may be no longer postponed and are unlikely to take place through private sector initiatives. Deregulations of certain services may help eliminate gross local inefficiencies or restrictive practices and legitimize informal sector provision, but its applicability is limited and 26/ The legal/constitutional statutes of most LAC countries include the public obligation to provide specific local services. Several Latin American constitutions even state explicitly as a government responsibility that of providing all citizens with "decent" housing. - 66 - does not always reduce public spending or relieve the local budgets. Limiting the level of intervention of the public sector is a valuable objective, but it is difficult to deteLmine in practice what its appropriate scope should be.271 It must be noted also that all solutions aimed at strenghtening the local resource base may reduce the need for transfers to the relatively affluent jurisdictions but do not eliminate the need of transfers to the poorest ones. 4.19 Spatial Cost Differentials. From a national perspective, the priority should be placed on reducing expenditures per unit of production of given goods or services and on achieving a significant proportion of local financing for those social goods which bear specifically local benefits. But even disregarding the existence of economies of scale and of local differences in demand levels, one still wonders at the great disproportion in per capita expenditures recorded in different areas of a country for specific budgetary entries. These differences can only be explained by relatively higher expenditure/product ratios, especially in peripheral regions, where apparently too much is being spent to obtain a given unit of public good, suggesting that specific public expenditures might be expanding simply because earmarked transfers become available. It may also suggest that some jurisdictions make expenditures or implement investments which are not of the highest priority and incur greater costs than others for equal services.381 4.20 The Fiscal Cap of LAC Subnational Governments. The trends in own revenues and expenditures of the regional and local governments over the 1975- 1984 period (see Table 4.20a & b) shows that a serious gap in their balance sheet was already widening before the onset of the debt crisis. As can be seen, 11/ Two main considerations can be applied in determining which level of government is more suitable to provide which services: one is the extent to which the service involves externalities, the second whether scale economies are involved in its provision. 1 / This observation was made for the Argentine Provinces by J. Macon, who put in evidence the wide unit cost differentials encountered in the provision of similar services (see his essay Federalismo Fiscal v Sistemas Tributarios Subnacionales (1982). In fact, too little is known of these cost differentials within LAC countries in general. Discrepancies may be sectoral (of the rural vs. the urban sector), intra-regional (of the peripheral communities vs. the core city in a metropolitan area) or more often than not inter-regional (one region or city compared to another). Yet, these differential costs are seldom reflected in budgetary policies for urban service, which tend to be set uniformly throughout a country and even less in the formulation of special investment programs (such as those favoring medium size cities). Prices are not necessarily adjusted in these programs for cities where significant sunk capacities may already exist in given services and for cities where severe capacity constraints involve relatively higher marginal costs. - 67 - the expenditures of subnational governments traditionally supersede their own source revenues in practically all countries examined. In the larger countries (Argentina, Brazil, Colombia, Venezuela) the gap is of the order of 50 percent. (Table 4.20a Annex) In other words, the independent revenues of local authorities in these countries are barely sufficient to cover their basic administrative costs. Public services provision is being financed through other sources. Table 4.20a: FISCAL GAP OF SUBNATIONAL GOVERNMENTS AS PERCENTAGE OF EXPENDITURES */ 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC AVERAGE SAMPLE -37.34 -33.56 -29.49 -30.06 -28.07 -21.76 -24.67 -21.22 -22.67 -15.61 Table 4.20b: FISCAL GAP OF SUBNATIONAL GOVERNMENTS AS PERCENTAGE OF GROSS DOMESTIC PRODUCT A1 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 LAC AVERAGE SAMPLE -1.13 -0.85 -0.57 -0.72 -0.65 -0.61 -0.82 -0.76 -0.91 -0.72 al Fiscal gap - own revenues - expenditures and lending minus repayment. Source: IMF, Government Finance Statistics Yearbook, various years. C. Current Strategies to Relieve Fiscal Stress 4.21 General. As we have seen above, the fiscal imbalance of many subnational governments is due to the fact that the expenditure demands arising from urbanization have not been matched by sufficient revenues. This may be due to the failure of local authorities to engage in proportionate efforts of revenue collection but also to the limited buoyancy of their revenue sources to resto nd to the increasing needs and the fluctuations of the economy. As a result, almtst everywhere in Latin America the central governments are called upon to fill the - 68 - gap in the current accounts of subnational governments. A variety of strategies is being used to this end, which can be reviewed under the following headings: (i) changing the scope and scale of local expenditures, (ii) changing the territorial base of jurisdictions, (iii) raising the level of fees and charges; (iv) implementing tax reforms; (v) relying on transfers, (vi) relying on concessionary credit. Most often, a combination of the above strategies is being used. 4.22 Changing the Scope and Scale of Subnational Expenditures. A variety of approaches can be used by governments: (i) taking over the burden of some local expenditures at the regional or national level; which is accomplished by transferring specific functions"U or by direct financing of capital investments while local authorities remain responsible for the operation and maintenance of the facilities; (ii) privatizing or franchising the provision of certain services, which may be helpful if guided by commercial logic and if the regulatory framework (tariff control, operating rules) is appropriately modified; (iii) exerting direct administrative control on local expenditures to ensure consistency with national objectives, as long as the changing needs of local communities are adequately recognized by the measures introduced (ceilings to the growth of given expenditures or to payroll outlays, measures of the cost effectiveness or efficiency of expenditures); (iv) resorting to technical assistance and persuasion to improve local practices of expenditure planning, programming, budgeting and of revenue assessment, imposition and collection; diffusion of public information on comparative fiscal performance could also help in this respect; (v) applying monetary and interest rate policies so as to expand (or restrict) the supply of credit to finance public expenditures; and, lastly, (vi) limiting top-down transfers, which may be an effective approach insofar as sufficient local autonomy is granted to undertake fiscal initiative and to make independent decisions on the cost efficiency of expenditures. 4.23 Restructuring the Territorial Base of Jurisdictions. The jurisdictional problems arising in large conurbations can be addressed in several ways: (i) annexing surrounding areas and jurisdictions to a core city and uniforming the tax base so as to balance the metropolitan burden of provision of certain services and to internalize spillover benefits as well as costs; (ii) reducing the number of administrative units by consolidation to ward off a9/ This case is rare- the general tendency is towards decentralization of service responsibilities, especially in the education and health sectors. However, Bird points out how the Colombian government dealt with the general problems of departmental finance by taking away the most important departmental functions, i.e., the provision of education and health services. - 69 - the weaknesses of unviable communities;al (iii) creating a two-tier metropolitan government with revenue raising and spending authority; (iv) upgrading the level of government of the larger cities, conferring to them Special District status and prerogatives; and (v) establishing regional authorities or development corporations which assume certain expenditure responsibilities of the local government. The relative success of these solutions depends on the extent to which they consolidate activities which involve significant economies of scale and public externalities, leaving to the municipalities those services which require a direct and prompt response to local needs. 4.24 Adjusting the Level of Fees and Charges. To ensure that sufficient resources are raised to expand basic services in response to the rapid growth of cities, one could adjust their pricing so as to approach real costs and adopt standards of a level proportionate to the end user's ability and willingness to pay. Obviously, this approach applies only to services that are directly chargeable (e.g., utilities, public transport, shelter, garbage collection, etc.) while most others (e.g., education, health, welfare, recreation, fire and police protection, general administration) still remain to be financed through local taxes, transfers or other sources of revenues. In many LAC countries, where national or regional authorities set upper limits to user charges, price adjustments require legislative approval and political debate on whether the proposed measures do or do not preserve the social objectives of service provision. The usual approaches are (i) scheduling progressive price adjustments as conditions for grants or loan disbursement; (ii) delegating the authority to set tariffs to semi-autonomous enterprises providing specific services (e.g., utility companies, urban transport, garbage collection) on the expectation that they become self-supporting and (iii) privatizing the provision of particular services (through concessions, franchising, sub-contracting) - As already discussed (Paras. 3.4 & 3.5) few departmental enterprises or franchised companies in Latin America achieve a degree of financial viability and even fewer yield significant income to subnational governments. However, it ought to be recognized that direct cost recovery is not necessarily an overruling objective for all types of activities. A more realistic approach is to set appropriate targets of financial performance to ensure that local governments or enterprises remain responsible for raising at least part of the revenues necessary to cover service provision and develop a proper concern with the efficiency and the design of expenditures. (See Para. 5.31). 4.25 Implementing Tax Reforms. These may take place in a variety of ways. One is to improve the rates and collection procedures of existing taxes. The main focus may be brought on the taxes levied on goods and services, (upgrading AQ/ In reality, political pluralism and the peculiarities of some revenue sharing systems encourage a trend towards progressive disaggregation. This is particularly noticeable in Brazil, where the number of municipalities has grown from 3991 in 1985 to 4118 in 1986. - 70 - rates, extending registration, streamlining collection procedures, converting general sales taxes into value added taxes, adjusting fuel or motorvehicle taxes, increasing the fiscal burden on the consumption particular goods, etc.) These reforms are also frequently aimed at revamping property taxation (improving cadastral procedures, extending the tax base, updating assessments, improving collection). The most dramatic short term results are generally obtained by extending coverage of the property tax in rapidly expanding urban areas. However, once this is done, property values ought to be regularly re-assessed in order to guarantee a steady source of income for the future.A.l' A second approach is to allow the right to introduce new taxes to the subnational authorities: this may permit them to place emphasis on the fiscal instruments they consider more appropriate, to elect new ones or to modify the applicable rates within centrally mandated ceilings. A third approach aims to achieve better horizontal and vertical balance by reassigning or sharing some tax source between the national and the subnational authorities. 4.26 Relying on Intergovernment Transfers. All the above solutions (implementing tax reforms, restructuring the territorial base of imposition, changing th- scope and scale of local expenditures, improving the level of fees and charges) may reduce, the need for transfers to the more affluent jurisdictions of a country. But even if the majority of jurisdictions were to become self- reliant, national policy objectives or distributional considerations would still suggest the necessity of some transfers to the less fortunate areas. Besides that of achieving horizontal equity among jurisdictions, the main argument for intergovernmental transfers is one of vertical equity, because the national government controls the revenue sources wbich are more elastic to GNP trends and is presumably better equipped and more efficient in raising resources. The most current types of intergovernment transfers practiced in Latin American countries are reviewed in Box 5. 4.27 The Use of Transfers. The systems of intergovernmental transfers presently in place in Latin American countries, whether in form of revenue sharing, grants, or concessionary loans, came about over time through a series of discrete and contingent measures. It is therefore difficult to tell today why the transfers were originally established and what criteria were applied in their design. As they exist at present, they exhibit large margins of discretionality and a rather unpredictable volume over time. Much of the current arguments over revenue sharing derives from the lack of a explicit statement of objectives for the transfers and from the fact that they involve conflicting jl/ The CIATA Program experience in Brazil, whereby the Federal Government provided grants and technical assistance to local governments f,-r the establishment of cadasters, showed dramatic revenue improvements in the short term, but proved difficult to sustain once the external help was withdrawn, in face of macro-inflation. (See Mahar and Dillinger op.cit.) - 71 - Box 5: TYPES OF TRANSFERS The most current types of Intergovernment transfers practiced in Latin American countries are reviewd below: (a) Origin-Based Tax Shares: consist in the straight (e) Emergency Relief Grants: jurisdictions struck by devolution to the lowr levels of government of a national or man-made calamities (floods, pre-estabi1shed share of the amount collected by a earthquakes, droughts, gluts of agricultural superior authority within their jurisdiction. commodItles, bankruptcies of departmental Being strictly based on origin, these schemes enterprises, etc) usually received central leave out distributional considerations. A government subventions. These take place though typical example is the Value Added Tax (104) quick disbursements from earmarked accountsor collected in Brazil by the States, which retain special bond Issues and extra-budgetary transfers 80% and return 20% to the originating from the treasury. The disaster proneness of many jurisdictions. LAC oountriesmake thIstype of assistance a recurrent event. (b) Formula-Driven Tax Shares: redistribute the yield of certain taxes collected by a higher authority f Capital Grants: thew transfers represent among contributing jurisdictions according to some unrequited contributions to specific local formulae hich may combine considerations of government Initiatives dictated by the desire to fiscal capacity (income), disparities in service encourage capital formation rather than recurrent levels (need) and encouragement of tax effort (per expenditures. They are generally awarded upon a capita collection). It is however difficult to list of approved and localized projects, but may identify formulae which are equally suited to all also be used In form of bloc-grants to fund broad cities and situations: the more sophisticated the capital Improvements programs. Primary schools, formulae, the more likely they are to be local health center and a variety of other politically questioned (See Box 5). Argentina, community facilities are almost Invariably Brazil, and Mexico offer typical examples of financed this way In LAC countries. relatively elaborate systems of formula-driven revenue sharing. (See Box 6). (g) Categorical Cost-reimbursement Grants: widely used In almost all countries, they consist In (c) Unreported Transfers: one questionable type of partial or full reimbursement of the costs transfers must be mentioned because of its Incurred by area governments In carrying out widespread use of Latin American governments: specific activities. Usually some revenue sources this includes the non-payment of bills, are earmarked to finance particular activities contributions of taxes; the forgiving of arrear (e.g., fuel taxes being used to finance highway payments, the absorption of the operating deficits Improvements). of subnational authorities by the treasury; the quasi-fiscal support measure adopted by monetary h Project Loans since the termsof credit used In authorities and the many implicit subsidies that official lending are generally softer than characterize intergovernment transactions. commercial terms and the repayment of the corresponding debt Is often forgiven, project (d) Discretionary Grants: another rather loose form loans can also be considered as a form of of transfers, consists of ad-hoc allocations transfers. Even when external funding Is (generally from a reserve fund budgeted annually Involved, the foreign exchange riss Is normally for discretionary used by the upper levels of carried by the guarantor, Ich Isa central government). This form of central government government authority. Few of the Bank urban assistance has acquired extraordinary relevance in projects to Latin American countries have LAC countries, virtually overhelming the covenante8 full coverage of debt-service by redistribative impact of the parallel mandatory sub-borrowers. schemes. - 72 - issues of neutrality, efficient allocation or equity in distribution (See Box 6: Arguments on Transfers). This problem of internal consistency becomes explicit as one tries to correlate the different types of transfers in use with the possibly negative or positive impact they may have on the achievement of specific goals, as illustrated in Box 7. Types of transfers are grouped according to the extent to which they are conditional upon a specific use of funds (tied transfers) or unconditional (untied transfers). Objectives are grouped according to the three main functions of government policy: stabilization, allocation and distribution. 4.28 Stabilization Objectives. A first group of concerns has to do with the stabilization of the economy: these may reflect the intent to exert some budgetary control over the annual volume of resources transferred from the central to the local budgets; to avoid unwanted growth of central government parastatal agencies; to curb the budgetary displacement and inflationary effects of swelling social sector expenditures; to reduce subnational dependence upon top-down transfers; to maintain some predictability on regional and municipal budgets; to induce some tax effort and resource mobilization at subnational level by requiring matching contributions. Finally, transfers may be dictated by a desire to maintain stability to the political system, or by a feeling of national solidarity, which have obvious implications for economic growth and development. 4.29 Allocation Objectives. Another group of objectives may reflect concerns with resource allocation. These may be to ensure the provision of socially valuable, but non chargeable services, which subnational governments or the private sector may be reluctant to supply in adequate amounts and quality; another may be to support sectoral investment programs designed to achieve macro- economic or social targets (employment generation, export promotion, resource development, etc.); still others may be to promote the development of resource regions or to promote balanced development among regions; to ensure a margin of budgetary flexibility allowing subnational authorities to develop their own planning and programming capabilities; to foster awareness of the real costs of public service provision by making local governments more accountable for expenditure decisions; finally to encourage capital formation by earmarking transfers for specific purposes. 4.30 Distribution Objectives. A third group of objectives relates to the distribution of resources. One chief purpose of transfer systems is to reduce the vertical imbalance between the revenue authority and the service responsibilities of different levels of government so as to preserve some measure of subnational autonomy. Another is to ensure a horizontal distribution of resources which guarantees to all regional and municipal governments sufficient means to address local needs so as to achieve basic standards of welfare everywhere in the country. 4.31 The Design of Transfer Systems. The objectives that different forms of transfers intend to achieve are multiple and often conflict with each other. - 73 - Box 6: ARGUMENTS ON TRANSFERS in Favor Against Stabilisation: Revenue sharing allows to the central Stabilizction: Effective fiscal management is government sufficient budgetary flexibility to top the impaired, because the revenue structure becomes less most productive and buoyant revenue sources. flexible and public sector allocatiois correspondingly distorted. Mandatory revenue sharing provides subnational Irregularities and delays in the administrative authorities with a steady and predictable source of transfer of funds hampers subnational cash management, income, allowing for some budgetary latitude, especially in presence of high Inflation. If properly designed (matching grant schemes) a Transfers foster dependence on this source of revenues, transfer system may Induce the general public to pay reducing self-reliance and fiscal effort. They tend to for now taxes, promoting incremental resource remain in force long after the circumstances for which mobilization, they were established have Ceased. Reducing the fiscal gaps of subnational governments, There are difficulties in designing an efficient transftrs contribute to national unity and stability revenue-sharing system suited to all cities and with obvious benefits for country-wide growth and situations. Formula-driven schemes are politically davelopment. questionable and tend to foster divisiveness rather than unity. Earmarking may lImit possible Inflationary aspects of Tied or untied, any increase in subnational transfers. expenditures tends to be inflationary, unless matched by incremental revenue growth. Direct transfers to subnational levels avoids unwanted Additional revenues at subnational level supports growth of inefficient central government enterprises, alternative growth of departmental enterprises, which are not necessarily more efficient. Allocation: Transfers can be used to Induce Allocation: Greater availability of revenues does not subnational governments to provide specific services always result in subnational production of the desired according to nationally eflned priorities. goods or of appropriate services. Transfers are needed to offset externalities and Both externalities and congestion costs are not easily congestion costs. translated Into operational criteria. Transfers can be used to promote development of There may be better alternative uses of fiscal resource regions In line with national priorities. resources. The opportunity costs are likely to be high. Categorical transfers may promote capital formation at Earmarking may distort local budgets and possibly lead subnational level. to misallczation of resources and subsequent rises in recurrent costs. If properly structured, transfers may promote awareness Transfers provide little Incentive to improve of real costs of service provision, expenditure efficiency and deter recipients from facing up to the real costs. Distribution: Transfers are necessary to broaden the Distribution: Better achieved through explicit 6rant resource base of subnational governments to enable them schemes, supporting the earning ability of the poor or to address the soclo-economic needs of the areas where Improving conditions in the less developed areas of the they provide direc 1;nefits (vertical equity). country, than through broad-design revenue-sharing schemes, whereby different Jurisdictions end up receiving too much or Too little. Revenue-sharing permits to compensate for differences Fiscal capacity can be measured only crudely and in in local fiscal capacity (horizontal equiyy). relative terms (horizontal equity remains open to question). Transfers enable all Ilocal governments to satisfy at Transfers should be related to effective local demand least the *Asic needs of their constituencies. and fiscal effor4. People will move from one jurisdiction to another. Revenue-sharing results in a better spatial Large cities and Industrialized areas, which contribute distribution of fiscal resources between urban and more to the tax-pool, resist especiai treatment of rural areas. rural jurisdictions. Transfers increase subnational ability to provide Local authorities must be free to provide different servicles and to respond to local needs in the context types and amounts of services and to apply those of upper supervision (local autonomy). charges which they consider more appropriate (local autonomy). - 74 - Box 7: INTERGOVERNMENT TRANSFERS -, OBJECTIVES TRANSFER UNTIED (BLOC) UNTIED OR TIED TIED (CATEGORICAL) TYPES Discret- Cost Origin Formula Unreported Emergency lonary Capital ReImburs- Project OBJECTIVES Based Driven Transfers Relief Transfers Grants ment Loans ................ ..W..................=.. a... 86-...........a........... .MUSSONS... STABILIZATION Control of Total + + + Transfer Volume Limit Public Sector Growth Inflat. Impact of + + + + + Social Expenditure Subnational Fiscal + + + Dependence Predictable + + + + Subnational Revenues Subnational Resource + + - + + Mobilizatlon ALOCATION Non-Chargeable + - + + + + + Service Provision Development of + - + + + + Resource Regions Balance Interregional + - + + Development Subnational Planning/ Budgeting Flexibility Subnational Real Cost + + + Awareness Promotion of + + + + + Capital Formation DISTRIBUTION Revenue Expenditure + + + + + + Vertical Balance Local Autonomy Vertical Balance + Local Needs + + + + ResponsIveness Basic Needs + + + + + Horizontal Equity Fiscal Capacity + + Horizontal Equity Fiscal Effort 4orlzontal Equity - + - - + + scstly positive Ifpact + depends on transfers-destgn * mostly negative Impact - 75 - The relative effectiveness of any particular system in achieving different goals depends in a fundamental way on the specific design adopted for its implementation (see Box 8.). One special difficulty of revenue sharing is to determine the extent to which the ability to pay for social services of different jurisdictions should be equalized through transfers, without de-motivating the recipients from making some contribution to their cost. This involves some measurement of relative fiscal capacity and fiscal effort.L1 Since the main idea is not only to compensate for vertical imbalance but also to achieve some degree of economic welfare in a horizontal sense, almost all revenue-sharing formulae in use in LAC countries incorporate some re-distributive features (Argentina, Brazil, Colombia, Mexico, Venezuela) (see Box 9.) To a varying extent, they also try to make transfers partially dependent on local fiscal effort. However, the relative weight of these effort-based or redistributive corrections is-offset in practice by the overwhelming weight of origin-based components or by the overdevelopment of parallel, discretionary forms of transfers. Nevertheless, a new trend is apparent, which may give preference to those types of transfers which are more conducive to local resource mobilization. 4.32 Summing Up: This section of Lhe report has considered the importance of the subnational fiscal gap, which in several countries accounts for a significant portion of the aggregate deficit of the public sector. The discussion moved from the revenue side, analyzing the sources of independent income assigned to regional and local authorities, their relative buoyancy, administrative costs and effective yield. Among tax revenues, on:y those imposed on goods and services and on real property appear to be on some general A2/ Accurate measurement of the fiscal capacity of a given jurisdiotion would call for analysis of the "potential" revenue collection, but this involves in each community of the effective ability of the residents to pay taxes and other chargef:, the degree of monetization of the local economy, the availability of tax-handles, the ability and willingness of local adminis- trators to irpose and collect levies which are all factors that vary considerably from one part of the country to another and would require individual assessment and auditing of each jurisdiction. A gross, but widely used measure is the ratio of the per capita income of a particular community to the average national value. Other short cut indicators often used are the per capita value added produced in the jurisdictions, the consumption of fuel and electricity, the registration of motor vehicles, economic activities or property transactions, the er-ension of the area to be served or simply the size of the urban population as a proxy for all the above. Most revenue-sharing formulae in use in LAC countries rely on these or similar indicators. Relative fiscal effort is measured as a ratio between the actual a '.unt of per capita revenue collection (tax and non- tax) of a jurisdiction and the potential collection obtainable from the local tax base by application of the average national value. By nature, it represents a relative, rather than absolute measurement. - 76 - Box 8: REVENUE SHARING BY FORMULA A redistribution of fiscal participations by formula, (as opposed to a straight proportional devolution of the contribution made by each jurisdiction to the tax-pool) reflects the Intent of directing more resources to the areas of greatest need or lesser fiscal capacity. Formulae are meant to avoid the issues of vertical and horizontal equity raised by more discretionary systems of redistribution, which are subject to political manipulation. However the basic ingredients of the formula itself may have differential impacts. FitIA INGREDIIENTS: POSITIVE iPACTS NEGATIVE IWACTS Equal Shares: divides the tax pool rural areas, Irrespective of lower urban areas, Irrespective of greater in equal parts (only apparently tax contributions. tax contributions. equitable). Area: assumes that larger extensive jurisdictions may be smaller jurisdictions may be highly jurisdictions have greater underpopulated, underdeveloped (e.g., urbanized (e.g., core cities, capital infrastructure and services costs. peripheral areas). districts). Total Fbpulation: overhead service -e populous jurlsdlctions (urban less populous jurisdictions (rural or financing. .. s) sparsely settled). Urban Polmlation: focus on cities more urbanized arefs rural areas service needr Last Pfpulation Census: legally stagnant, declning areas dynamic growth communities reliable but often obsolete. Demogaphic Growth Rate: fastest growing areas/cities outmigration areas, slow growth controversial, subject to estimates. cities Need Yardsticks: some indicators may communities with greater gaps more endowed areas (generally the rtf lect an urban or rural bias. (housing, utility connections, more urbanized ones). education, health). Fiscal Performance: measures past more enterprising, wea.thier underdeveloped areas with poor expenditure against need. jurisdictions. expenditure record. tmem Far Caita: accounts for poorest jurisdictions measurements wealthier jurisdictions. demographic size, not for skewed often unreliable. Income distribution, low monetization of economies. Total Revenue Collected: neglects more Industrialized urbanized areas, less developed, rural areas demographic size, availability of tax able to impose/collect taxes and collecting little revenues. handles. charges. Per CaMita Revenu : accounts for tax-efficient urban areas, but large least tax efficient areas, especially demographic size not for income informal sector, urban poverty pose rural. distribution. limits. Fiscal Caaci1y: pro-rated to jurisdicilons with relatively low arass with above-average capacity potential revenue yield at average capacity (problems with chosen (I.oblems with indicators). collection rates. Indicators) Fiscal Etfwt: local collection vs. preserves fiscal motivation of stimulates incremental efforts in capacity (metching grants). performing communities. non-performing communities. - 77 - Box 0: REVENUE SHARING CRITERIA IN ARGENTINA, RAZIL, AND MEXICO (1985) Shared To Regional Governments To Local Governments Taxes $ Share Criteria % Share Criteria ANTMINA Sales Tax 65% Provinces' Population 40-100% Population Rtral Property Tax Varies 0-31 Population Growth Urban Property Tax 48.5 25% "Development Gap* a to 0-10% Land Surface Vehicle Tax (housing quality & no of cars) Province 5-30% Equal Shares Stamp Tax 10% Inverse of Territorial Density 0-25% Payroll Expenditures Other 3.0 General Provincial Fund 0-75% Fiscal Effort in addition to national transfer, Provinces may share own sour,.e revenues with Municip3lities RAZIL To Regional Governments To Local Governments Income Tax and 12.5 95% Inverse Per Capita Income 10.5 Educational Sector Mnfrng Sale Tax 5% Land Area of State Priorities 40% Fuel Consumption Fuels Tax 30.7 40% Population 8.0 Nat'l Highway Plan 20% Land Area Prioritie;l Vehicles Tax 45.0 100% Car Registrations 0 to State Discretion 50% Origin Highway Transp. Tax 20.0 30% Network Length a to State Planning 20% Population Priorities 60% Population Power Se.tor Investment Electricity Tax 50.0 20% Land Area 10.0 Priorities 15% Consumption 5% Production and affected areas Mineral Tax 70.0 100% Origin 20.0 State Priorities Education Salary 66.7 100% Origin a to -ste Priorities Rural Land Tax 0.0 Directly to municipalities 80.0 Origir. 75% Origin Value Added Tax 80 State Tax 20.0 25% State Discretion Property Transfer Tax 50 State Tax 50.0 J 'o Origin In addition to national tra"sfe-s, States share value added and property tax with municipa!lties MExICD Foreign Trade Tax 90% General Fund 20.0 min GF 505 Origin Income Tax but varies CF 25% Oopulation Value Added tax 17.4 t to stdte MDF 25% State Discretion Services Tax 4 to 8% Complementary Fund 70% Inverse of GF Transfers Mineral Extr. Fees origin 30% Equa Shares & effort 90% Inverse of GF Transfers formula 2% Municip. OvIpt Fund 10% Equal Shares AutomobIle Tax 80.0 State Tax 20.0 General Fund Rules Ioport (3%) and Export (2%) Surcharges 0.0 D!rectly to Collecting Munic. 95.0 C ic Origin H1drocerbons Fee 0S%) 0.0 Directly to Municipalities 63.4 I to Origin - 78 - significance, while betterment and user charges, income from departmental enterprises and a variety of other non-tax revenues seem to have become under- exploited forms of subnational imposition. Borrowing also appears to be a form of financing that is improperly used by LAC subnational governments because of little experience, restrictive practices or inadequate institutional procedures. Inter-government transfers, often the largest single source of revenue, are discussed at some length, because of their allocative potential. On the expenditure side, the analysis points to the relative weight of subnational outlays (about twenty percent of general government expenditures and eight percent of GDP) noting that the lower tier authorities carry out a significant proportion of national social expenditures (over forty percent). In addition, the traditional involvement of LAC subnational governments in capital investments is proportionally higher (about twice) than it is in industrial market economies. 4.33 A few general observations can be extracted from the review of typical strategies that LAC governments adopt to relieve subnational fiscal stress. Tax reforms (improving rates, Introducing new taxes, modifying revenue sharing) may help rationalize the existing fiscal systems and strenghten the local revenue base. Certain departmental enterprises can be made more self-reliant by linking loans and equity contributions to price adjustments on the corresponding urban services. The largest urban agglomerations of the continent (Greater Mexico City, for instance) could achieve greater efficiency and rationality through a territorial re-organization of the tax base (through annexation, reduction in the number of administrative units, establishment of regional services or creation of two-tier metropolitan governments). In general, much could be achieved by: (I) changing the scope and scale of expenditures, (ii) relieving the local budgets of some expenditure responsibilities, (iii) franchising or privatizing the provision of certain services, (iv) de-regulating particular activities, (v) legalizing the services provided by the informal sector, (vi) extending cechnical assistance to local governments, (vii) exerting direct administrative control, (viii) applying more transparent monetary and interest rate policies, and (ix) limiting the volume of top-down transfers in exchange for some devolution of fiscal authority. 4.34 Transfers of some type appears inevitable and may be justified by objective differences in the fiscal capacity of individual jurisdictions. Here the main concern should be to take a fresh look at the present systems, selecting those reforms which are more conducive to local resource mobilization, eliminating cross-transfers and discretionary grants and gradually substituting credit schemes for grants wherever feasible. Large cities, which have the greatest poteitial to become self-reliant, should be encouraged to do so, by allowing them greater autonomy to determine tax rates and to select which fiscal instruments they want to emphasize for resource mobilization. It appears most important to break the present cycle of general dependence upon top-down transZers and to ensure that subnational authorities become responsible for raising a more significant portion of the revenues necessary to pay for their expenditures. This appears to be the way to develop in both taxpayers and delision makers a concern with the efficiency and the effectiveness of local expenditures. Nevertheless, financial assistance from the central government will continue to be needed for those poorer jurisdictions, which because of their limited fiscal base, would need to engage in a disproportionate fiscal effort in crder to achieve even the minimum standards of basic services. - 79 - V. CONCLUSIONS 5.01 Geperal. The conclusions of this paper can be organized according to three main groups of concerns: those more often voiced in Latin American countries; others of a more theoretical nature, which stem from economic and fiscal principles; and those emerging from the lending operations of the Bank itself. This Section has been divided in two parts: first it summarizes the main issues .aised in the report (Part A), showing that obvious correlations exist between the three groups of concerns. These were visualized in Box 0.1. Part B addresses specific issues in form of recommendations. Some of the questions raised, which are more speculative or political in character, may not be answered here other than partially or in pragmatic terms. In fact, the intent of this paper is not to unravel theoretical knots with which scores of specialists have toiled for decades, but 3imply to take stock of the problems a Bank project officer faces when designing his operation and to examine some of the implications of empirical approaches. In the attached Boxes 10, 11 and 12, a few recent Bank projects are briefly summarized, to illustrate the type of situations encountered in urban sector activities in LAC countries, and to show how some of the concerns discussed in the paper may bear on the conception of projects. While the case studies presented reflect, no doubt, still partial and possibly inadequate approaches to the issues confronted, they indicate new trends in urban sector lending which are also evident in the current operations of other regions. Part A: Issues 5.02 Country Concerns. A first group of concerns, frequently expressed by our borrowers, are mostly related to national fiscal politics and developmental aspirations. Among these, one can list the constant preoccupation with the uneven territorial distribution of population and economic activities, with the size and primacy of vary large cities, with the relations between these cities and the smaller towns and the rural areas. The issues raised relate mainly to horizontal equity in the fiscal treatment of different communities or to the broader aim of interregional equalization. 5.03 There are also concerns with the deteriorating quality of the urban environment and the apparent reduction in the levels if service that the public authority can provide. The latter problem focuses attention on the burden of subnational deficits and calls into question the vertical assignment of powers and functions to different levels of government, pointing to the current mismatch between revenue authority and expenditure responsibility. Lack of liquidity at all levels of government may be imputed also to the limited capacity or willingness to pay of the taxpayers, and this in turn is linked to the poor quality and insufficient quantity of the services they receive. The above considerations often lead to the suggestion that some of the inefficiencies and - 80 - insolvencies of central government agencies could be offset by the local governments, were they granted greater participation and/or autonomy in expenditure decisions. This latter position is generally countered by national authorities, who point to the local governments' limited capacities to plan, implement and administer urban services, as "demonstrated" by the unregulated growth of self-help, informal sector activities. Moreover, the fragmentation of urban investment into a variety of authorities, budgets and programs makes it difficult to manage and even to conceive as a single, integrated sector. This may also explain why urban policies, expenditures and administration are often based on limited factual knowledge of relative needs, priorities and schedules of demand. 5.04 Theoretical Concerns. A second group of concerns relate to general economic and fiscal theory. Among these one should mention first the Bank's traditional concern with ensuring mobility of the main factors of production to create a suitable climate for free enterprise and investment. Country attitudes and policies concerning internal migrations, industrial location incentives and inter-regional terms of trade often interfere with the free flow of labor and capital. Other concerns relate to official disregard of agglomeration economies or indiscriminate emphasis on congestion costs. These stem from a segmented vision of the national urban system, which is seen fractionally, as a rank-size hierarchy of isolated settlements, rather than as an interdependent spatial network of productive, marketing and service flows. A comparative regional perspective is necessary also to confront the issues of horizontal equalization, recognizing that individual jurisdictions may differ substantially in fiscal capacity, but also in fiscal effort. Similarly, the problem cf efficient allocation of resources brings into focus the relative returns of specific sector investments but also the notion of comparative advantage among different areas. 5.05 The existence of economies of scale, decreasing costs or externalities conditions the assignment of functions among government levels. One can debate to what extent re-distributive functions can be delegated to the subnational authorities as agents of the central government, but there is little doubt this question has particular significence for the implementation of poverty alleviation programs. Related concerns are the transparency and targeting of subsidies and the budgetary predictability of both revenues and expenditures. The subnational governments contribute substantially to the growth of domestic debt through deficit spending and inflationary borrowing. Policies of debt absorption, terms of on-lending and targeting of credit become critical issues because they may induce distortions on national financial markets. Fiscal stabilization would require rigorous criteria to assess the creditworthiness of sub-borrowers, and more efficient pricing of public goods to achieve acceptable levels of cost-recovery from end users. But local cost recovery and accountability for expenditures cannot be achieved without enabling local authorities to respond directly to the needs of their constituencies, which implies a certain degree of fiscal autonomy. Similarly, the promotion of local entrepreneurship and private investment for the provision of public services is - 81 - tied to the removal of unnecessary regulatory constraints. Finally, the financial requirements of an urban investment policy for the country as a whole should be assessed with regard to its allocative and fiscal implications. This is an operational issue which may shed a different light on many theoretical questions. 5.06 ODerational Concerns. The third group of concerns refers to the relative impact of the Bank's lending operations on a national urban system as a whole. Questions can be raised on the overall effectiveness of our lending, its contingent character, its long term replicability. To make strategic use of the resources available implies selectivity in lending, which is not easily implcmented, given the multi-sectoral complexity of urban operations, which occasionally makes them difficult to instrument and to manage even within the Bank itself. Moreover, counterpart funding is not always clearly identifiable in budgetary terms and often leads to exceptional allocations of earmarked grants by the central government for particular communities. 5.07 Broadening the scope of Bank lending for urban development evidently requires the linkage to a more systemic pattern of top-down transfers and the gradual substitution of requited credit for unrequited grants. This raises questions on the suitable financial intermediaries for on-lending, on the lending terms themselves and on the pattern of cost recovery of both capital and recurrent costs from end-users. Since the implementation and management of projects is of necessity decentralized, a more flexible approach, relying on development of the independent administrative capacities of sub-borrowers, is viewed as a desirable alternative to an unwanted proliferation of parastatal executing agencies, but requires a substantial injection of technical assistance. One last set of operational concerns relates to the eventual need to implement changes in the regulatory framework of urban services provision, which may result in specific covenants for Bank loans. Finally, there is an evident need to pursue research on urban development, to clarify its intersectoral and macro economic linkages, the related fiscal and institutional problems, and to address the issues arising from the introduction of new operational approaches. Part B: Recommendations - Country Approach 5.08 The territorial Distribution of Population and Economic Activities. As discussed in Chapter 1, most LAC countries present a high degree of spatial concentration of urban population and activities. In Argentina, Brazil, Colombia, Chile, Mexico, Peru and Venezuela, a very large share of the economy is concentrated in but a few cities with more than half a million inhabitants. The same is true also for the smaller countries, often to an even larger extent (see Table 1.2). Because the larger cities offer greater opportunities of employment, income and access to public services, and because of their sustained, self-cumulative growth, LAC goverments are increasingly concerned with the apparent inequity between the social conditions of metropolitan dwellers and the rest of the population. For this reason, they feel they have to reduce these differentials not only through social investments but also by redirecting - 82 - productive enterprises to lagging regions in order to forestall outmigration. This perspective does not always account for the important contribution tne large urban agglomerations are making to the modernization and development of their own nation and for the fact that firms and households have very good economic and social reasons to gravitate towards the more urbanized areas. Experience shows that compulsory measures of deconcentration of business enterprises have only marginal impacts on overall demographic trends. Furthermore, countering the spontaneous trends towards agglo%aration often results in a distorted use of public resources. While it is a responsibility of the central government to promote growth and to sustain productivity wherever unexploited potentials exist in the country, population distribution objectives do not offer a sound basis for decisions on the decentralization of productive investments which should be guided by adequate feasibility studies and a specific statement of economic objectives. 5.09 Intermediate Cities Programs. The above considerations apply also to the indiscriminate priority often accorded to public investments in the so called "intermediate-size" cities or "alternative" growth-poles (Mexico and Brazil are zlassical examples), which are generally costly and not necessarily conducive to the desired effects. In fact, some medium size cities may deserve urgent attention because their rate of spontaneous growth is for greater than that of other urban areas or because of other factors, which are generally not the result of conscious government decisions. In those cases, or when an identifiable sectoral constraint is stifling the growth of a particular city, discrete government support may be justified. But this is different from singling out all medium size cities as targets for public investment (or external loans) merely on the basis of demographic size, trying to confer to them some comparative advantage which might induce sustained growth. The experience with broadly defined investment programs for intermediate cities has proven generally ineffective and sometimes wasteful because the sheer provision of infrastructure and services cannot assure to them a viable economic base, while living conditions are only marginally improved. 5.10 Small Towns Programs. A second instance of misguided policy is the tendency of LAC governments to scatter social investments in small towns and semi-rural hamlets, in the hope of improving the lot of farming populations in areas of rural exodus and stagnant agriculture. This type of investments may make good sense within the impact zones of broader programs, where systemic change is being introduced in the current patterns of land tenure, farming techniques, water management and crop choices. However, they provide only marginal and temporary relief in areas where structural conditions persist for endemic unemployment, declining prices and depressed incomes. Investments in infrastructure and services in market towns may become an important factor in the transformation of the rural environment where a perspective of growth and change exists (see Box 10) but one must also recognize its obvious limitations in recessionary situations. In most cases, small towns programs might improve the welfare of the population which will remain but will accomplish little in retaining excess labor. - 83 - 5.11 Large Cities Programs. Yet, in spite of the prevailing rhetorics on decentralization, the investment policies of LAC countries continue to favor the larger cities. This is because these policies are basically aimed at filling perceived service gaps and because the most visible, growing concentrations of urban poverty and political instability are found in metropolitan areas. When the rise of current costs and the decline of statutory transfers deprived the metropolitan governments of investment resources, these were the arguments used to justify exceptional measures such as debt cancellations or rollovers, external loans, ad hoc transfers, and direct investments by central government agencies. But the favor accorded to large cities is also justified by a giowing realization that large urban areas make a critical contribution to the national economy, and inadequate infrastructure may bottleneck effective resumption of economic activities, discouraging private investment and diverting savings towards unproductive uses. 5.12 Reviewing Fiscal Assignment. The arguments on equity and efficiency in resource allocation converge on the issue of fiscal assignment. As discussed in previous chapters, urban agglomeration calls for a variety of functions and services to be carried out by the public sector for one reason or another. Urban services provision involves economies of scale and benefit or cost spillovers from one jurisdiction to another. Jurisdictions with a low tax base would have to impose higher taxes and charges in order to raise sufficient revenues to achieve levels of service comparable to those of other communities. The need for a particular service may be greater or lesser in different jurisdictions. The cost of providing the same service may vary from one community to another. Some services can be directly charged to end-users, others cannot and must be financed through general revenues. 5.13 The above considerations imply that it is not always simple to determine what service should be provided and paid for by each level of government. In practice one has to consider, service by service and for each tier of government: (i) the current statutory arrangements; (ii) the spatial incidence of each service and the externalities involved; (iii) the corresponding economies of scale; and (iv) the availability of sufficient revenue sources to finance its provision. Table 1.7 offers an illustration of the functional assignment of responsibilities between central, regional, and local governments in Brazil. Quite often these responribilities are shared by different tiers of government, but it is the very proportion of this sharing that remains ambiguous and does not clearly tell when and for what support is provided by a higher authority. To grasp the realities in the assignment of public service functions among government levels, an inventory of the present inter-institutional and budgetary arrangements (in law and in practice) becomes necessary in each case. 5.14 Containing Central Government Transfers. In Para. 4.19, the observation was made that only when local governments are made responsible for raising at least part of the revenues necessary to finance their expenditures, they will develop a concern with the efficiency and effectiveness of public spending. The Latin American experience shows that excessive reliance on - 84 - transfers reduces the political responsibility of local leaders and possibly results in wasteful use of resources. Reducing the dependence on transfers calls for some decentralization of fiscal authority, not necessarily to match service responsibilities, but to an extent sufficient to make both decisionmakers and consumers aware of the political responsibility to impose and pay the corresponding charges. This paper is not advocating an increase in the volume of unrequited intergovernment transfers. Its central theme is rather to reduce the relative importance of the grant and revenue sharing systems to avoid too great a dependence on national sources of financing and its undesirable effects on the subnational public sector. The best approach to achieve efficiency in local expenditures is not to emphasize grants, but to assign some additional sources of revenue to the administration of subnational authorities or to guarantee to them a share of particular taxes, according to an explicit set of performance rules. 5.15 Decentralizing Fiscal Authority. To a considerable extent somae administrative decentralization of the responsibilities for providing specific urban services has already occurred in LAC countries and keeps occurring. However, as suggested above, along with the authority over certain expenditures many governments would need to decentralize to some extent also the tax system, at least in sufficient measure to discourage subnational governments from carrying out investments which are less essential or from spending inordinate amounts to provide virtually the same services. The traditional solution of establishing regional service branches of central government agencies has already expanded the parastatal bureaucracy to an extent that it is no more efficient or prompt in meeting public needs than the local governments. Besides, the incremental current costs of this overbloated parastatal sector impose constraints on other forms of budgetary assistance and claim resources which could be used for re-distribution. Rather than encouraging further expansion of LAC parastatal sector the fiscal system should be so modified as to offer sufficient motivation for local decisionmakers to take into consideration the costs as well as the benefits of urban services provision. 5.16 Improving the Existing Revenue Sharing Systems. Many LAC countries are presently trying to improve their revenue sharing systems. The recession of national economies has generally affected the yield of the tax sources on which these schemes were based, and spurred abnormal growth of compensatory flows of discretionary grants which add significantly to the fiscal burden of more conventional types of transfers as in the case of Argentina (footnote 2) or Brazil (Box 3), where they are generally perceived as arbitrary and inequitable and have occasionally become a dominant source of local revenues. What is needed instead is an analysis of the extent to which top-down transfers are instrumental to achieve: (i) country-wide objectives (growth resumption, fixed capital formation, employment, inflation control, human resource development); (ii) some predictability and control of the annual volume of transfers (to minimize unanticipated budgetary deficits); (iii) better accountability for their use; (iv) more efficient allocation of resources at the local level; (v) greater local - 85 - autonomy and self-reliance; (vi) greater interjurisdictional equity; (vii) improved local planning, programming and budgeting; (viii) greater mobilization of local resources; and (ix) a more focussed impact on the alleviation of urban and rural poverty. Reform of the present revenue sharing systems should aim at the elimination of discretionary transfers and move from a comprehensive reassessment of all Intergovernmental flows to the local governments, not only in the context of urban development policy, but in the context of national stabilization and economic recovery objectives. 5.17 Improving Subnational Administrations. One argument often voiced in LAC countries to explain shortfalls in cost recovery and general mismanagement of local services, is that the technical capacities of lower level officials are limited and this is seen as a serious obstacle to the devolution of either expenditure or revenue authority. True enough, administrative skills are frequently scarce, especially in the less developed areas of a country. But, the purported greater competence and integrity of higher level officials is only based on the fact that their actions are less closely scrutinized, and that more able officials are drawn to central government posts by the better pay, more secure tenure and higher status. It is difficult to believe that the officials of lower level governments in Latin America could not learn to make sound decisions, given adequate time, suitable training and technical assistance, appropriate salary policies and more secure tenure, or simply by being placed in a situation that is conducive to responsible management.- Given their greater and more direct knowledge of their jurisdictions, subnational adminis- trators may perform better than central administrators when it comes to responsiveness to local demands. 5.18 Substituting Borrowing for Grant-Transfers. Many urban services involve creation of long-lived capital assets (roads, water and sewerage mains, power grids, public building, housing), for which the long-term repayment of borrowed funds makes good sense. However, there are few LAC countries in which long term borrowing is available to local communities in sufficient amounts at affordable conditions. Thus it might be desirable to assist LAC governments in financing such works through lines of credit and through the establishment or adequate restructuring of Municipal Development Funds. Past experience with MDF's in several LAC countries (Bolivia, Brazil, Colombia, Costa Rica, Guatemala, Honduras, Mexico, Venezuela) has been disappointing as they generally failed to become self-sustaining or to achieve a significant scale of operation.iV However, the present efforts to strenghten existing MDF's or to create new ones on sounder footings should be encouraged under current circumstances. Institutional credit should be associated with active programs of technical assistance to the borrowing municipalities, to help them in project identi- fication and design, in planning their investments and in putting order in their financial accounts so as to reach minimal conditions of eligibility. Capital improvement credit should be made accessible to all jurisdictions meeting set criteria of eligibility, giving preference to revenue-generating investments and ! / Bird, Bahl and Linn, Macon, op. cit. A/ Kenneth Davey, Municipal Development Funds and Intermediaries World Bank, PPR Working Papers, 1988. - 86 - setting some conditions for direct and indirect cost-recovery. Municipal loans should be coordinated with the parallel flows of grant and revenue-sharing funds, which also must be made conducive to local fiscal effort. 5.19 Assessing Local Capacity and Willingness to Pay. It has been pointed out that "even very poor people are willing to pay something for service they value and moreover that people value more those things for which they have to pay something. "451 Shelter, water, education and medicine are cases in point. However the general public is likely to resist the attempts of a public authority to raise the level of current taxes or charges as long as its confidence in a corresponding improvement of services has been undermined by decades of poor perfo&mance in delivery. This attitude of end users is often generalized in the stance of many LAG communities when faced with central government promises of "conditional" improvement programs: it is difficult to expect of them an incremental effort before some service improvement is actually implemented. Not only is it difficult to start exacting payment for functions which the central government was traditionally expected to provide free of charge, but it is particularly difficult in those communities which in the past have seen their claims inadequately, if at all, attend3d by superior authorities. This tends to create a self-fulfilling cycle whereby no service improvement is feasible without mobilizing additional resources and no improvement in the charging system proves acceptable without visible progress in the delivery system. Nevertheless preliminary experience in a few Bank projects based on competitive access to credit (Boxes 10, 11, 12) would indicate that the extent of repressed demand for urban services is such, that attitudes towards public cost recovery can change quite rapidly even in those communities which were at first most reluctant, as soon as the demonstration effects of an going program begin to materialize in the neighboring communities. 5.20 Containing the Growth of Fiscal Deficits. In Section 4 it has been shown that the subnational governments of LAC countries account for a substantial share of general government expenditures and that the cumulative deficit of regional and local budgets represent a serious, destabilizing factor in their economies. It has been shown also, in Section 2, that the growth of demand for urban services makes incremental expenditures unavoidable. Gradual substitution of credit for the existing, alerming pattern of discretionary transfers would A5/ Richard Bird in A Review of Recent Fiscal Decentralization Measures, Sector Report p. 71 . See also Gabriel Roth's review of a variety of cost recovery experiences in his "The Private Provision of Public Services in Developing Countries", Oxford University Press, New York, 1987. - 87 - help making local authorities more concerned with the design and cost-effectiveness of their expenditures. As most of the works financed involve a low component of foreign exchange, external support to this type of operations could also have some beneficial impacts on the country's balance of payments, as long as its scale is significant and the cost-sharing ratio of the foreign lender is reletive ly high. The terms of on-lending should be realistic enough to avoid inflat,;anary or deficitary effects. The expectation is that the improved performance of local governments in delivering urban services would be rewarded by end-users through enhanced cost recovery. While some emphasis on directing credit to revenue generating investments may be helpful in this respect, it need not become an overruling criterion: the borrowing commur.ities would be induced to make rational choices if debt amortization payment were to be retained from the iuture flows of revenue sharing transfers as a form of collateral guarantee. Part C: Theoretical Aproach 5.21 Intersectoral Allocation of Resources. The urban investments carried out by subnational authorities mainly consist of social overhead capital, rather than directly productive investments. Although part of these investments, especially those in infrastructure and land development, represent intermediate inputs and support mobilization of productive capital, they are mostly categorized in government budgets as "social service" expenditures. Al Decisions on social sector expenditures and on corresponding transfers are set in view of long range policies of national development rather than in the context of short-term management of the economy and therefore remain vaguely (and more politically) defined. This means that the volume of social sectors, allocations can only tentatively be derived from macro-economic planning targets. This also makes the measurement of their relative efficiency a highly controversial matter. 5.22 In practice, some emphasis on allocative efficiency and some criteria to evaluate performance are specially required as long as the national economies are under stress and may give way to greater emphasis on re-distributive and poverty alleviation criteria when economies will be buoyant once again. But it is obvious that the financing of broad-design programs to spread social services across regions or categories of cities is a less effective way to pursue distributional adjustments than might be more focused programs to generate employment and improve earning power of social groups at the lower end of the income ladder. Nor can efficiency be ensured in the local use of resources by simply earmarking funds to sectoral programs which promise relatively higher returns or by imposing some perfunctory criteria of project evaluation. Resource transfers to subnational governments are typically subdivided in many different localities and small projects, covering a variety of sectors, many of which do not generate cash revenues or quantifiable benefits. In fact, this diffuse character of urban investments is at the same time an asset (in that it &/ See footnote 5 at page 6. - 88 - subdivides the risks of large, single-purpose projects, allowing for greater implementation flexibility) and a source of concern (because it makes it difficult to detail 'ex-ante' the eventual use of funds and therefore require intensive monitoring during the investment stage). A deterministic approach would call for rationing capital improvement grants or subloans among local jurisdictions after a comparative assessment of their present or potential contributions to national development, their fiscal performance, or the extent to which their past expenditures met actual needs: these are evaluations which are as difficult to make in technical as in political terms for any central government. 5.23 Spatial Allocation of Resources and Factor Mobility. Normative theory of fiscal assignment does not recognize a significant role for the subnational governments in stimulating area development, because they have no control over the flows of capital and labor from one part of the country to another and a preeminent concern with the provision of local public goods. Some time ago, Thiebout has argued that, assuming free inter-regional flows of factors of production, people would "vote with the feet" and move to those jurisdictions where their preferences for social services are met in more satisfactory and cheaper ways (lower costs or lower taxes) and that this form of competition in the provision of public goods could lead to the gradual elimination of interregional disparities. While this image suits best the situation of the industrialized market economies, where capital flows between one part of the country and another are definitely more pronounced, and where income levels and employment opportunities are more evenly distributed, there is no question that "voting with the feet" occurs also in LAC countries, even if the reduction of spatial disparities is progressing at an intolerably slow pace. 5.24 The misguided assumption underlying the medium and small size cities programs mentioned in paras. 5.9 and 5.10, is that a different size distribution of cities may result from carrying out iwprovements in urban services, while contradictory national policies continue to exist, which constrain or discourage the flow of entrepreneurial capital or try to force investment into specified areas. In reality, large cities remain more attractive for private investment, because of the higher returns afforded by their agglomeration economies and their large sunk investments. So far, the metropolitan areas of LAC have continued to absorb incremental labor productively and have performed a vital role in redistributing personal income and developing human resources. In a climate of public resource constraints, deployment of public resources in order to divert productive investments towards alternative urban centers (where profitability is less evident and potentials often vaguely identified) may fritter resources without achieving significant impacts on the inflow of migrants to the large cities. 5.25 Equity and Efficiency in Spatial Policy. Although it could be argued, from a broader perspective, that the efficiency of investments is not necessarily in conflict with interpersonal income distribution, one must acknowledge that the pursuit of efficiency does not by itself lead to interregional equalization. It may be argued also that territorially uniform development is not an attainable - 89 - goal, because areas differ in comparative advantage, both natural and man-made, and therefore a policy of spatial equalization would result in diminishing returns for the country as a whole (para 3.15). Finally it may be argued that supply-side policies of equalization would deprive municipalities of cost-recovery incentives and discourage self-reliant development initiatives. On the other end, severe regional disparities engender political disunity and unrest, which can threaten both economic growth and stabilization. Since the welfare of a nation must be measAred also for its discrete territorial units, considerations of spatial equity may actually clash with national efficiency and some trade-off between the two objectives becomes an inevitable concern of fiscal policy. The basic question is then to determine what resources can be devoted to the pursuit of territorial equalization and to define clearly the instruments that can be used to reduce disparities to an acceptable level in the less dynamic areas of the country, without unduly slowing down the overall growth of the economy. 5.26 Fiscal Equity. This paper is concerned with the efficient and equitable provision of urban services within a national territory. Since urban services are mostly local in character, one could argue that they should be provided and paid for by the residents of the jurisdiction 4i which their benefits accrue, so as to link directly the revenue and the expenditure side of public choices. However, this concept of benefit region is not easy to apply in practice, because the income and willingness to pay of the end-users may be unevenly distributed within the jurisdiction and many of them will tend to act as free-riders. Difficulties arise also from the existence of economies of scale (decreasing marginal costs) or congestion (increasing marginal costs) which grow with city size and density, and from the fact that certain essential social services ought to be provided everywhere in the country, irrespective of the local ability to pay. Therefore, national policy seeks to ensure that jurisdictions (and social groups) of uneven revenue potential are not burdened with equal expenditures and that different levels of government have the ability (and authority) to raise sufficient revenues to confront their expenditure responsibilities. The very initiative of revenue sharing, which stems from the dual objective of approaching vertical and horizontal balance, should preserve the cpcept of fiscal equity, distributing the fiscal burden in line with the economic capacity of different communities. 5.27 Approaches to Inter-Area Equalization. The current revenue sharing systems in LAC countries seek to achieve horizontal equity among various jurisdictions in different ways, depending on whether they distribute the pool of available rasources according to some measure of relative need for urban services, (such as the size of existing shortfalls from pre-established minimum standards of national welfare) or according to the fiscal capacity of jurisdictions, (so as to enable the poorer communities to Rchieve some acceptable, average level of service without a greater than average fiscal eftort). The first approach (basic needs) eventually slants the distribution of national government expenditures (through direct investments or closed-ended, earmarked grants) towards regions and local jurisdictions, which present the larger gaps. The second directs resourc,-i to those subnational governments - 90 - which do not have sufficient resources to provide their residents with a minimum standard of service which is considered appropriate in kind and correct ir amount for all areas. As illustrated by the examples presented in Box 9, many current formulas try to help the less urbanized areas by incorporating some factor of inverse proportionality to income per capita or population size. These corrections confer a character of ambivalence to the resulting revenue sharing pattern which combines considerations of basic needs and fiscal capacity by assigning different weights to diverse formula ingredients which often offset each other. The resulting distribution of resources is unclear and tends to overlook the fact that, in spite of their higher average income, the metropolitan areas may contain large concentrations of poverty and that economies of scale could be reaped by encouraging them to provide more basic services. The ambivalence of revenue sharing formulae is not cured by simply reverting to a straight origin-based devolution, nor by eliminati..- or playing down income related coefficients (reverting to a pure basic needs roach): the real issue is to provide adequate incentive for the larger ur . communities to improve their fiscal performance. 5.28 Delegating Some Re-distributive Responsibilities. National policies of nrban service financing are bound to be spatially differentiated. Large ci..es have a different fiscal base than small towns, different service needs, and different concentrations of poverty. These disparities can be generalized at country level, but are better un"erstood in the context of specific regions. Regional authorities are better aware of specific local situations tban central authorities can possily be and local authorities have a detailed knowle-ge of the spatial distribution of income groups within their own jurisdictions. While general criteria for the allocation and distribution of resources must be set by central planning authorities, regional and local authorities can implement these policies in a more responsive and equitable manner within their own constituencies. Even though the lower level authorities of most IAC countries are in practice agente of the central government, they have closer political ties to the people they serve. Therefore, they would be better pcsitioned to read their preferences and priorities and to determine how much should be spent for each service in different localities, which type of charges are more suitable to recover its costs from end-users, and how proposed changes in local imposition would affect the wellbeing of the poor. Regional involvement in redistributive policy is desirable because it can supply the more direct and detailed knowledge of the location of poverty than that afforded by the generalized, comparative statistical information available at national level. 5.29 A Regional Perspective on Urban Investments. An(-ther reason for relying on subnational authorities is that the territorial distribution of infrastructure invest-nts (e.g. transportation, drainage, water, sewerage and power) must consider tae dynamics of growth of individual centers within clusters of cities and then relations to the rural areas which surround them. This more focused regional ,'rspective is useful because regional economies are indeed subsets of tha national economy, but have internal trends and characteristics - 91 - of their own, which ought to be specifically considered when making decisions on the distribution of investment amng and Nithin urban subsystems. Moreover, in those countries where an intermediate level of government exists, the regional autherities not only have a responsibility in passing through national revenue sharing flows, but often complement these flows with parallel fiscal transfers o.k their own, which may be of equivalent importance. Strategies for the development of particular regions should not be merely conceived as sectoral interventions of the central government but should pay attention to the existing budgetary mechanisms within the regions themselves and their component Jurisdiccions. 5.30 Direct or Indirect Cost Recovery? From a public finance perspective, it could be argued that many important urban services should be fully charged to the direct beneficiaries because they are technically "private" rather than "public" goods according to the exclusion principle. This is the case of individaal connections to public utilities (water, sewerage, electricity, gas, telephone) which benefit directly households or business enterprises and can be rationed through user fees. By extension one could restrict to paying customers even the access to roadspace, recreational grounds, schools, health clinics and a host of other facilities, even if this may become technically difficult or politically undesirable. The point made here, however, is not that urban services should be treated as "public" goods (most of them are at best "mixed" or "merit" goods) but that their provision has a spatially defined range, and mainly benefit a community of residents in which all members are potential users. There is little aigument that in Latin America urban services still afford scope for a more rigorous application of user charges and for innovative Bank operations which may link directly expenditures with revenues, while this is not possible for most national functions. But it should be recognized also that this approach raises thorny distributional issues which in final analysis limit the effective level of cost recovery or the delivery of needed services to all members of the community. In practice It does not matter much whether or not particular services are charged directly to individual users or indirectly to the urban community as a whole, as long as sufficient consensus can be achieved, within the community itself, to bear the corresponding costs through some type of collective fiscal effort. 5.31 Should the Cost of Urban Services be Fully Recovered? In addition, the provision of urban services presents numerous exceptions to the rule of short-term marginal pricing. Second best pricing often falls short of full cost recovery. Therefore, the question can be raised, whether the full recovery of the costs of specific projects or programs should be regarded as a primary objective, especially in local jurisdictions which, because of their limited fiscal capacity, will continue to receive transfers on a predictable basis. It may be noted also that the pattern of foreseable cost recovery greatly depends on the nature of the activity that is being financed. When the national interest is significantly involved, full cost recovery becomes just one of many plausible objectives. The critical issue becomes not the level of cost recovery that can - 92 - be reasonably expected, but the more general decision by the central government whether or not to provide some subsidy, by weighing alternative uses of the resources involved. Since intergovernmental transfers are designed to help communities provide those public services which they need most rather than those for which cost recovery is more readily achievable, the Bank preference for the recovery through direct user charges of some specific costs of urban projects may lead to distortions in local budgetary allocations. 5.32 Setting Targets of Overall Fiscal Performance: One cannot ignore the fact that the success of a given program in achieving cost recovery may be considerably influenced by the nature of parallel intergovernmental fiscal flows. Direct and full recovery of the cost of specific urban service improvements may not be critical for sound financial management, as long as there is reasonable expectation that budgetary resource will become available to replicate needed investments. In this respect, i- appears more important to establish a sound financial framework for the overall management of the local budgets (including an element of explicit subsidy where necessary) and to make sure that sufficient resources are likely to become available and put to appropriate use, in response to an adequate assessment of priority projects and effective demand. Transparency of general accounts and public assessments of comparative fiscal performance may have more significant impacts on the accountability of local authorities than isolated records on specific tariff hikes or sectoral improvements, which are difficult to monitor in a large number of jurisdictions. To introduce some financial discipline in local accounts, higher level authorities should set realistic targets of overall performance and make transfers, credits or matching funds conditional upon the achievement of such topics. Part D: Operational Approach 471 5.33 Revising the "raditional Bank Approach to Urban Lending. The earlier urban loans of the World Bank to LAC countries were made to support capital improvements in individual cities or categories of cities, generally singled out by population size or servx-.e gaps. Since loan repayment was usually expected from the central government, even the credit share of project financing acquired the character of an equity contribution to the local governments. Concessionary credit would also be accompanied by ad-hoc national grants, complementing the local counterpart efforts which were frequently in kind (land, force account works and project administration). This meant that selected cities would receive a sudden windfall of incremental revenues to assist (and induce) them in carrying AZ/ Several observations reported in this last section were made by Koichi Mera (OPS Regional Planning Advisor) who, before leaving the Bank, became the most explicit advocate of chLage in our urban lending policy (see his draft paper: Intergovernmental Financial Relationships. Financing of Local Public Services and Regional Development, dated 3/24/86. - 93 - out particular projects. Thus, the availability of Bank financing would have a rather disruptive impact on the regular process of national resource allocation. Although more systemic approaches were developed over time, it cannot be said that, at the start, the urban lending approach of the Bank reflected particular concerns with the fiscal aspects of urban services financing besides those related to re-distributive policy. Not surprisingly, a case-by- case approach was welcomed by our LAC Borrowers, as it fell quite consistently within their conventional pattern of discretionary grant allocations and often proved expedient in meeting special political pressures. 5.34 The CInventional Cycle of Project Development. Usually this went through the foll.wing stages: first, the Bank Project Officer would carry a survey of the urban sector (or subsector). He would then proceed to identify target cities and investment priorities, following the Bank's recognition of sector issues and its perception of what was feasible among the opportunities pointed out by national authorities. Thirdly, he would undertake preparation of the project with a team of colleagues, some consultants and eager local authorities. Lastly, he would work out the financing arrangements, often requesting a special allocation of central and regional public resources in form of a tied grant, sufficient to cover the capital costs of the project and often also those of a single-purpose special unit established to administer and implement the project through a separate account. Project arrangements were consolidated in a legal agreement and extraordinary budget appropriations were often covenanted as conditions for effectiveness. 5.35 Project Specific or Poverty Alleviation Focus. The main emphasis was on targeting subsidies to low-income groups and reducing the design standards to affordable cost levels. More often than not, the financing arrangements were worked out only after the identification of target cities and physical improvements. User or betterment charges would be set at levels considered appropriate to the economic capacity of project beneficiaries, and public counterpart funds requested to make up for the differen e. Loan repayment responsibilities and complementary financing would be carried by the central authorities, who considered it unrealistic to expect independent financing of the "Bank" Projects by the local government concerned, irrespective of their ability to raise revenues. Occasionally, the distributional or developmental objectives of a project were perceived as overriding considerations even if some recipient communities were incapable or unwilling to contribute anything at all to project funding, it was felt that the project-specific developmental or distributional objectives constituted overriding considerations. A_/ There have been few cases in which the protests and competitive claims of other jurisdictions found their avenue in the political process of loan approval and temporarily stalled congressional or state ratification of the agreement with the Bank. - 94 - 5.36 The Concern with Project Replicability. The observations made above do not imply that in past urban operations Bank staff neglected to analyze the financial situation of the target municipalities. A certain care was always used to identify the local sources which could be used for partial financing of the proposed investments or to ensure coverage of operation and maintenance costs. But the resource requirement of a particular project are only part of a much greater fiscal problem of national urban development. Even when economic analysis may support the priority accorded to specific operations, the distortion it induces on the overall fiscal system can be undesirable. This more general concern with the fiscal impact of individual projects, with their replicability and with the magnitude of Latin American urban problems has led Bank staff in recent years to pay increasing attention to the country-wide implications of urban improvement financing. Given the number of cities which would "deserve" projects of the conventional type, the ad-hoc allocation of grants or concessionary credit to a few municipalities does not represent a significant solution unless the problem addressed is a unique and one-time event. 5.37 The Volume of Urban Development Lending. Since the inception of urban operations in 1972, the World Bank has made forty-nine loans to LAC countries totalling the equivalcnt of US$3.2 billion. To the Urban Development Projects one must add another US$2.5 billion for forty-five Water and Sewerage Projects over the same period. If one were to sum up also US$1.8 billion in Education, Health and Nutrition Projects, all of which have a direct bearing on urban welfare, one achieves the sizeable cumulative contribution of US$7.5 billion for social sector improvements in Latin American cities. However, one must note that this money was divided among two dozen countries and all sizes of urban centers. During the same years, the total capital investment of LAC subnational governments in the same sectors can be estimated to have been in the order of US$350 billion, more than fortyfive times the volume of Bank lending. Taken together, the four metropolitan areas of Buenos Aires, Mexico City, Sao Paulo and Rio make every year capital investments about two times bigger than the annual volume of Bank lending for all LAC cities. 5.38 Leveraging the Impact of External Funding. At best, the World Bank, the InterAmerican Development Bank and the bilateral donor agencies combined may contribute less than one billion dollars per year to LAC urban development. To expect a visible, direct impact of these external contributions on the physical environment of 280 million urban dwellers through a direct "project approach" may be asking too much. Our urban loans must therefore become less project focused and more systemic in character. This means that the Bank's urban staff as well as that of other donors should become increasingly involved in sector and country policy dialogue so as to leverage the impact of external lending with some policy adjustment on the use of domestic resources. As the nature of lending operations might evolve, so would the character of the attached conditionalities, calling for fiscal and financial adjustments and for administrative and regulatory change. - 95 - 5.39 The Concern with Counterpart Funding. As the above implies also an escalation in the volume of lending, it becomes increasingly apparent that the Bank loans should be linked to a sizeable and predictable flow of resource transfers in the borrowing countries. The annual flow of statutory revenue sharing between the national and the subnational governments of LAC countries represents a regular and important transfer of resources. (See Table 4.12c in Annex) In fact, its contraction or expansion might aggravate existing disparities among regions and levels of government and threaten the stability of the overall fiscal system. If the existing revenue sharing pattern is questionable, the Bank should be prepared to assist the government in its attempts to improve its overall design, rather than inducing the central authorities to make independent allocations to support specific projects. Intergovernmental finance could become a productive area for policy dialogue with LAG countries but this requires not only a change in focus but also adequate training and development of the Bank's urban staff. 5.40 Non Additionality of Bank Lending. Ideally, one should make sure that at least the cost of the external loan is repaid to the central government by the borrowing municipalities, which should in turn recover their costs from the consumers in form of general taxes or in form of user or betterment charges. Because the externally borrowed funds usually cover only part of the total investment costs, it is assumed that other counterpart funds supplied by central or regional authorities to complement local contributions correspond to some regular flow of revenue sharing or capital grants schemes. It is assumed also that these complementary flows follow rational criteria of allocation which are not inconsistent with the objectives of the Bank-sponsored program. This is because only art of the total system of top-down transfers (the portion used in counterpart financing) would become affected by specific loan conditionalities. In other words, the Bank's objectives of rationalization of the existing fiscal system can only operate at the margin and, because money is fungible, could be frustrated if an overwhelming proportion of the total trans- fers received by municipalities follows rules that are incompatible with those of the Bank-sponsored program. Some government commitment to the revision of inconsistent revenue sharing rules ought to be built-in in the loan agreement. 5.41 A Fiscal Perspective on City Size. The conventional distinction among cities according to population size has been dismissed as a useful perspective for planning decisions because it may lead to sepatate consideration of individual urban centers and tends to neglect the economic and functional linkages which operate within a given cluster of cities and their territory. Nevertheless the categorization of urban jurisdictions according to demographic size may be warranted in one respect: average per capita income and population size seem to go hand-in-hand. This consideration would lead to a design of the revenue sharing system along the lines recommended by Richard Bird in the case - 96 - of ColombiaA1 The more urbanized jurisdictions generally have a more buoyant tax base and therefore a greater potential for independent revenues. They should be treated as largely self-financing entities and be granted not relatively larger transfers in per capita terms, but simply greater authority to set tax rates and to collect revenues and possibly more autonomy to provide chargeable services on their own. Relative self-reliance could be promoted also in some intermediate size cities as an option rewarded with special grants matching incremental revenue raising efforts. Finally, financial assistance would continue to be given to the small municipalities enable them to achieve predetermined level of services, as long as they engage in some fiscal effort of their own, they revise and rationalize the local revenue system and improve their administration with the help of upper level authorities. The main objective of improvements in the current revenue sharing systems should be to stimulate tax efforts in all size cities, while ensuring that at least a minimum level of urban services can be provided in those of lesser fiscal capacity. 5.42 Selecting Target Municipalities. As we have seen, the tendency in traditional urban loans has been to allocate resources according to perceived urban service gaps. The main focus was on the expenditure side, how to deliver urban services best, in a cost-effective manner. Now the focus is veering on the revenue side: how to pay for urban services. While some years ago it was taken for granted that LAC economies would continue to expand faster than their population, and that the growth of personal income in their cities would support greater capacity and willingness to pay for more urban services, this is no longer the case: demand for public goods keeps expanding, but economies are stagnant, personal income is eroded and, with it, also the capacity and willingness to pay of urban dwellers. In this situation, the "basic needs" approach is only conducive to staggering estimates of capital requirements which central governments cannot meet, no matter how modest a service standard is assumed. Since repayment of the Bank loans can be reasonably expected only from creditworthy municipalities, the selection of target cities for our projects must be guided by some different criteria than a basic need approach. 5.43 Should the Worst Come First? As a Bank consultant pointed out in the case of Brazil,21 the "worst-first" approach may not work and may even result in the allocation of resources to areas that are actually losing population. Experience shows that the diffusion of sound management practices in poorer jurisdictions become easier only after change has been successfully implemented in a few pilot municipalities. Therefore, credit funds should not be assigned first to areas where poverty itself is highest, but to those where the potential for alleviating poverty is greater, because of their higher likelihood to achieve employment and income growth and because of the ability and willingness of the local administrations to implement reforms. L9/ Richard Bird: Intergovernmental Finance in Colombia, Harvard University Press, Cambridge, Mass. 1984. ,a/ David Vetter in: The Policy Framework for Cost Recovery for Infrastructure and Energy in Brazil, November 27, 1987 (Draft Report). - 97 - 5.44 A Gradualistic Strategy. A programmatic approach that can be taken is to accept that fiscal equity concern continue to rule the general distribution of grant money, while at the same time a complementary system is established, promoting criteria of efficiency in the distribution of borrowed funds. This can be summarized as a three-tier scheme. In any given region, there will be many municipalities, which are not at present or in the foreseeable future in the position to service incremental debt of any size and that, because of their weak fiscal base or stagnant economy, will need to continue receiving external help in form of grants. But there will be also some municipalities which can effectively absorb loan funds in limited amounts and are willing to implement the required reforms. Between these two extremes one is likely to find a number of municipalities which could technically borrow some resources for investment, but are as yet reluctant to submit to the discipline of the Bank-sponsored program or simply prefer to continue managing on the basis of the grants they receive. The strategy should be to ration capital grants (or revenue sharing funds) among municipalities according to their relative fiscal capacity, but to set aside a sufficient amount each year, to be used as a complementary watching fund for the credit scheme, in order to induce more municipalities to move from the third group into the second. This gradual strategy is likely to move forward to a more efficiency minded system without drastically altering the current pattern of revenue sharing. It also permits to narrow progressively the grant money to needy municipalities and, by inducing initially reluctant municipalities to borrow, it allows a gradual expansion of credit without jeopardizing either the principle of equity in grant distribution or that of financial efficiency. 5.45 Inducing Resource Mobilizatior with Matching Grants. The requirement of a local counterpart contribution (in kind if not in money) as a condition for receiving project grants or credits is seen as a useful device to stimulate local resource mobilization and to ensure committment to the realization of local projects. However, the level of matching grants should not become so high as to make them into a principal, rather than supplemental source of project financing. The required matching ratio (of the counterpart funds) may be varied with both the nature of the projects financed (for instance with the size of the local benefit component) and at least to some extent with the fiscal capacity of the jurisdictions concerned. This latter feature would represent a natural evolution of the present practice of grant distribution, which is often made in inverse proportion to the locall, generated value added or to the revenue collected during the previous year or simply to the urban population of different communities. Thus, matching grants could be used to reward local fiscal effort. 5.46 Developing Local Planning Capabilities. A coherent plan for regional (or national) urban investment should be based on a comprehensive (and comparative) assessment of the expenditure needs in local services and infra- structure in different areas. This is necessary to analyze financing gaps and to devise possible credit schemes. Some planning function for medium term capital improvement must therefore be developed at the local level, to identify and select the projects to be financed. In turn, this requires the establishment of a fiscal framework in which the future availability of resources can be reasonably forecast over, say, the next five years. Although some tradition of - 98 - budgetary programming exists in most municipalities (albeit imperfect and short-term) technical assistance from upper level authorities may be required to achieve some uniformity of criteria for project selection and some control over unit costs. Experience in a few Bank projects indicates that this can be done relatively rapidly and effectively, even when little traditional practice exists in this respect, through the establishment of regional pools of professionals, serving clusters of local communities. The work performed by these teams is both technical and political, as they must assist the mayors in their consultations of the city councils, to clarify the purposes of the program under consideration and to orient local options towards feasible and financeable initiatives. At the very least, municipalities should be required, as a condition for receiving capital transfers, to provide minimal investment plans as to the uses to which they propose to put the new resources, and some estimate of their own capacities to contribute to the investments and to carry the maintenance and operation costs of the facilities to be built. 5.47 The New Lending Approach. The recent reorganization of the World Bank was designed to make our approach more systemic and less project focused by regrouping professional resources on a country-focus and by supporting intersectoral coordination. In the urban sector, this new lending approach calls for: (a) pursuing country wide dialogue on those stabilization, allocation and distribution policies which affect directly the economic and fiscal performance of cities (i.e. issues of intergovernmental relations, local resource mobilization, tax reforms, pricing of urban services, interest rates, subsidies, investment criteria, assessment and monitoring of urban management performance, public/private sector roles, etc.); (b) transferring project development responsibilities to our Borrowers which implies more sector work (to appraise institutional capacities, regulatory framework and criteria of operation) and more technical assistance (to develop national or regional capabilities when needed). (c) selecting and strengthening financial intermediaries for the efficient pass-through of our loans, which means more attention to the on-lending criteria, the financial viability of the institutions and the impact of their operations on the national budget and the financial market; (d) adopting alternative lending tools, more comprehensive and supple than specific project loans, such as (i) sector wide investment or maintenance loans (umbrella operations extended to all cities meeting set performance criteria); (ii) financial intermediation loans (conceived and managed as lines of credit); (iii) restructuring loans for departmental enterprises (covering one or more subsectors); and (iv) sector adjustment loans (supporting the short term foreign exchange needs of large programs). - 99 - (e) refocusing urban research efforts on such themes as: (i) the national impact of intergovernment transfers, housing finance subsidies, urban infrastructure investments and land policies; (ii) the economic contribution of cities and, in particular, of the very large ones to national development; (iii) the functional, economic and fiscal flows which, irrespective of their size, link different cities one to another and to their rural hinterland within integrated regional systems; (iv) the public/private sector roles in urban service provision; and (v) the potential impact of de-regulation of certain services and legalization of informal sector provision. - 100 - Box 10: BRAZIL: PARANA MARKET TOWNS IMPROVEMENT PROJECT (LOAN 2343-BR) Parana' as Pilot State: Favored by a natural Implementation: After four years of environment well suited for export crops and by the Implementation, achient of the project objectives entrepreneurial 3kills of recent waves of Imigrants, Can be Considered satisfactory. Nearly all eligible Parana' accounts for nearly one-l'ifth of Brazilts municipalities have become sub-borrowers (280 against foreign exchange earnings with only 2.3 percent of its appraisal expectations of only 50) and some 2000 land area and 6.4 percent of its population. subprojects have been completed (another 500 are still Nevertheless, Pa.,anal is not one of the richest states In progress). Despite the 1983 political elections of Brazil: Its per capita income may be less skewed in and the financial and monetary crisis which caused a distribution, b)t is only slightly above the national belated project start and a climate of general average. Nor 4 the State Imune of social tensions: uncertainty, disbursement of loan funds matched the during the 1970's an estimated 15 percent of its average Bank profile, making this project a rare population left Parana' to other regions, mostly from exception In Brazilts urban portfolio of the Bank. Its rural areas, where convulsive changes occurred in land use, tenure and degree of mechanization. Sharp Project Impact: The Improvements made In contrasts exist in the endowement of resources and small market centers y!elded evident welfare gains for access to public services between the rural and the their residents, relieving social tensions and urban sector and between areas of the State. Parts of possibly some contributing factors of out-migration. Parana' are still in the process of dlonization, However, the most Important contribution of the PRAM others are pockets of poor, subsistence farming and Project was that of enhancing the capacities of rural others are overtaken by industrialized agriculture, municipalities, the lowest tier of Brazil's yielding surpluses which are reinvested in the main administrative hierarchy, to recover the costs of cities in comerce, real estate and Industry. These social Investments thus lessening their own dependence contrasts are reflected in deep political divisions, on State and Federal grants. The project did not causing the alternance of different factions in State Involve a Federal contribution, Its entire cost being and municipal governance. financed by the State of Parana@. Municipal loan repayments are used to capitalize a Municipal The Project and its Objectives: In August Development Fund. 1983, the World Bank approved a USS50 million loan to support and rationalize a program of financial The project concept had far reaching assistance to municipalities (the PRAM or Municipal Impact. Not only was It adopted In a subsequent Bank Action Program, which the State of Parana@ had loan to the nearby State of Santa Catarina (Loan No. undertaken since 1981. The main objectives of the 2623-M) but several other Brazilian States have project were to systematize the flow of earmarked expressed Interest In Bank support for similar fisca. transfers by introducing a 35 percent credit Initiatives. it has been adopted also by the component and cost recovery criteria, so as to pave the Government of Argentina for a five Province operation way to a progressive emphasis on borrowing rather than (see Box 11) and Is under active consideration in grant financing and to a gradual re-direction of Mexico and Colombia. unrequited transfers on the most needy municipalities. Project Data: Total cost US$150.2 million of Project Design: The project design which 35 percent financed by the World Bank, 65 Interlinked municipal investments in infrastructure and pqrcent by the State of Parana$ and Municipalities. services with fiscal improvements and selection of Infrastructure Investments are expected to amount to productive components, to expand the local revenue base 65 percent of total costs, community facilities ind cope with rising current costs. It promoted another 23 percent and vehicles and equipment 7 widespread awareness of economic and fiscal percent. Project design, administration, and constraints, fostering creative competitive approaches technical assistance account for 5 percent. Foreign by local authorities to develop comparative advantage costs were estimated at 35 percent of the total. The and attract private sector Initiatives to their project disbursement profile had been assumed to span jurisdictions. four years (to December 31, 1987), but a one year extension was required due to the delay in loan ____________effectiveness. Onlending terms to Municipal ities were See 19R0-4403 a-BR: BRAZIL: Parana' Market Towns at It percent interest over indexed principal, with improvement Project, Staff Appraisal Report, May, 1983. maturities up to ten years and 1-3 years grace periods. implmenatio, ahievmen ofthe rojct ojecive - 101 - Box 11: ARGENTINA MUNICIPAL DEVELOPMENT PROJECT (Loan 2920-AR) Objectives and Description: A US6120 million to the municipalities on "harder" terms (max 10 years loan was approved for this project on 03/22/88, with repayment Including up to one year of grace) with the main purpose of contributing to more effective spreads of at least 1.5 perCentage points over Interest public sector management at provincla! and municipal paid by the Provinces to the Central Government In level. Its specific objectives are: a) mobilizing order to capitalize Municipal Development Fund for external and internal resources to finance justified project replication in each Province. Municipal local investments in a non-deficitary and non- repayment of subloans Is guaranteed through the Inflationary way; b) strengthening municipal capacities Province's retention of the municipalityls debt to plan, finance and execute cost-effective programs of servicing obligations from provincial revenue sharing. capital improvements; and c) promoting structured, periodic consultations between municipal and provincial Benefits: Four main categories of benefits authorities for the formulation and evaluation of are expected from the project: (a) financial, where investment plans. The project provides subloans to the project's design criteria assure that Incremental five of Argentina's 22 Provinces (Buenos Aires, resources are mobilized by limiting oniending to Cordoba, La Pampa, Neuquen and Santa Fe) and to municipalities which are creditworthy and willing to eligible municipalities of those provinces, financing recover costs substantially In excess of the loan up to 505 of eligible costs for: a) physical amount and that partirpating municipalities are investments, including construction and rehabilitation financially strenght*ned to provide other services not of infrastructure and community facilities and financed under the project; b) distributive, where purchases of vehicles and equipment; and b) technical long-term financing makes cost recovery affordable to assistance and training. In order to implement the poor, besides creating approximately 58 thousands effectively the project objectives, only creditworthy person/years of construction work; 0 Institutional, municipalities (having a balanced current account and through providing financial predictabiiity necessary debt obligations within acceptable ifmits) are eligible for subnational governments to plan and program to receive subloans for physical investments, and their investments; and d) economic, with an average rate of annual capital improvement plans must allocate at least return of 38 percent on Infrastructure Investments. In 65% of total proposed Investment to components for addition, it Is anticipated that the lessons learned which investments costs will be fully recovered from from this operation will support replication of the direct payments to the municipalities by direct project to other Argentine Provinces. beneficiaries (betterment charges, connection charges). Risks and Safeguards: Possible risks Include Relanding Terms: The Central Government Ineffective provincial management and lack of onlends the proceeds of the Bank Loan to the Provinces continuity In project support. The main safeguards for under the same terms and conditions (15 years, this are the spreading of risk among the five including three years of grace at the standard variable Provinces. This has been sought primarily through the rate of interest) except that the Central Government condition that, If implementation delays occur In some would assume the cross-currency exchange risk and pass provinces, the Initial loan allocation by province on the US dollar exchange risk to the Provinces with would be adjusted In favor of those provinces with the cond;tion that any required adjustments in the demonstrated greater Implementation capacity. periodic, dollar-indexed debt service exceeding a plus or minus 10% deviation from the official inflation rate Project Data: Total costs USS240 million of will be credited to/or capitalized in outstanding which 50 percent financed by the Bank, 13.3 percent by provincial debt, in addition, Provinces pay a service the Provinces and 36.7 percent by the municipalities. charge for financial Intermediation of up to I percent Infrastructure Investments are expected to amount to 69 of loan account disbursed. Except for an amount up to percent of total costs, community facilities to iother 5-10 percent of the loan financing technical assistance 9 percent and vehicles and equipment 12 percent. and training, all proceeds of the Bank loan are onlent Project design and administration and technical from the provinces assistance and training each account for 5 percent. Foreign costs are estimated at 35 percent of total, An ___________eight years disbursement profile has been assumed as See IBRD-7093-AR: ARGENTINA: Municipal Development for the project. Project, Staff Appraisal Report, February 26, 1988, - 102 - Box 12: MEXICO: LOW INCOME HOUSING PROJECT (LOAN 2612-ME) Background: Mexico's housing problems are rooted in the Project Impact: A Housing Finance Study carried out rapid growth (2.5% p.a.) of its large demographic base under the project focussed Government attention on the (76 million), the pace of its urbanization (over 5% fiscal Importance of housing subsidies during a period of p.a.), and the uneven financial capacity of consumers, severe financial and economic adjustment. The Housing 70% of wich earn Insufficient income to afford the price Policy Statement of February 1987 spelled out strategic of conventional housing on adequately serviced land. goveriment commitments to sector reform, namely greater Therefore, about half of housing supply takes place on administrative coordination, legislative changes, cost illegally occupied and poorly serviced sites through the recovery targets and resource mobilization for rental and initiative of clandestine developers io infringe rural housing, paving the My for subsequent Bank building codes and planning ordinances. This "Inlormal" operations In the sector. FONHAP0 Improved It3 property cannot be used as an acceptable collateral for operational structure and financial capacity, home improvement loans and buyers are forced to resort to contributing to the continuation of Its activity through the disadvantageous conditions offered by of the the revenues of an already sizeable portfolio. The unregulated credit market. quantitative targets already achieved are Impressive, Including settlement of some 150,00 low Income famlIles The critical issues of the regulated housing supply are In over one hundred subprojects In 27 srates, despite the (1) the generally low capital recovery of public Inflation of construction costs and steady erosion of the investments (traditionally of the order of 30% in real beneficiaries ability to pay. terms); (11) the fragmentarlon of supply among disparate agencies, each practicing different mortgage terms, Witch On-lending Terms; The proceeds of the Bank loan are renders it difficult to control and predict the total onlent by FONHAP0 to Sub-borrovrs on terms and public subsidy involved; and (11) the inequitable conditions Wich vary with the type of Sub-Borrowar and distribution of such subsidy among social groups. The the unit cost of developments. As a first step In sector rationale for the Bank loan approved on 08/06/85 kes to Improvement, a minimal level of cost recovery of 50% In strengthen capital recovery during the Initial phase of real terms ws covenanted In view of FONHAPO's rigorous operation of FOMAPO, a national trust fund established targeting of Its production to beneficiaries of very low to finance land development and core housing for the Income. The Loan Agreement called for reviews of the urban poor. annual programs for commitment of loan funds and periodic revision of mortgage terms If needed. FONHAPO eventually Project Objectives and Description: Following a housing agreed to a solution that: Ca) maIrtaIns the level of sector review (IBRD No. 4993a-ME), Loan 2612-ME as mortgage payments In fixed proportions to the real Income conceived as a country-wide operation, supporting of beneficiaries (expressed In multiples of the FOMAPO's credits to States, Municipalities and Building prevailing minimum wage rate); (b) charges an Interest Associations. It tes Intended to assist the Government rate equal to the level of domestic Inflation; of Mexico in its objective to redirect an increasing (c) capitalIzes Interest charges In excess of monthly percentage of public subsidies to the poorest and largest payments, and (d) makes the fIscal subsidy explicit and segment of the housing market. The project also endorsed budgetable, In form of an up-trant grant. the Government's commitment to control and reduce the subsidy element of preferential credit, by making it Project Data: Total costs US$300 million financed explicit and moving towards its gradual reduction. through a US$150 million loan of the Bank (500, a US$40 million equity contribution by the Government (13.311), Implementation: Subloans include land preparation and USS35 million counterpart financing by Sub-Borrovrs basic infrastructure works, core dwellings and technical (1i.7%) and US$75 million (25%) by FONHAPO through own assistance. The sub-borrowrs prepare and implement revenues. About 78% of total project costs are In civi projects of a certain size (minimum 500 units) within korks and 12% In land purchases with administrative, established unit cost cellings, participate in their professional. and technical assistance costs accounting financing, provide collateral guarantees, and allocate for another 10%. The foreign component ins estimated at the facilities built under mortgage arrangements to about 25%. Project disbursement Is likely to be non-wage workers earning less than 2.5 times the minIt,aum completed In a considerably shorter period (four years) salary.. One important feature of the FOfiAPO programs is than the originally anticipated seven years profile. that beneficiaries' are pre-solec+ed and subloans are underwritten only when projects are completed and at least 80% occupied. Monitoring and evaluation studies are conducted annually, verifying compliance with income See lBRD No. 5473c - MEXICO: Low Income Housing Project targeting, unit costs and cost recovery. Staff Appraisal Report, July 18, 1985. ANNEX STATISTICAL TABLES Tables are numbered according to the text paragraph in which reference to them is first made. Tables marked with an Asterisk (*) are found in the body of the report. Several Table summaries (weighted averages) have also been incorporated in the main text of the report. - 104 - LIST OF TABLES Para. 1.2* Degree of Urban Polarization in LAC Countries (1985) 1.4* Number of Subnational Governments in LAC Countries (1987) 1.7* Allocation of Public Service Functions Among Levels of Government 1.9 General Government Expenditures in the Social Sectors (in 1980 US Dollars) 1.10 Per Capita CNP and Subnational Government Expenditures (1975-84 Trends) 1.13 Expenditures by Level of Government (in Percentage) 1.14* Subnational Governments Deficits (by Country Groupings: 1974-1984 (as Percentage of General Government Deficit) 1.15 Per Capita Income and Urbanization Trends (1976-1984) 2.1(a)* Central Government Savings as Percentage of GDP (1975-1986) 2.1(b)* Savings/Investment Ratio (1976-1984) 2.1(c) Central Government Interest Payments as Percentage of Expenditures (1975-1986) 2.1(d) Central Government Interest Payments as Percentage of GDP (1975-1986) 2.2* Domestic Inflation in LAC Countries (1975-1985) 2.3(note)* World Bank Lending in the Social Sectors to LAC Countries (1978-1987) 2.5(a)* Progressive Displacement in Budgetary Structure of LAC Countries (in Percentages) 2.5(b)* Trends in Budgetary Structure of LAC Countries (1975 -1984 variation) 3.1* Comparative Financial Structure of the Public Sector (Argentina, Brazil, Colombia and Mexico) 3.7* Indicators of Income, Urbanization and Industrialization in LAC Countries 3.9(a) General Government Expenditures (as Percentage of GDP) 3.9(b) General Government Revenues (as Percentage of GDP) 3.10(a) Revenues by Level of Government (in Percentage) 3.10(b) Capital Expenditures by Level of Government (as a Percentage of Total Expenditures) 3.10(b) Capital Expenditures by Level of Government (1975-86 variation) 3.10(c) Capital Expenditures by Level of Government: LAC Countries (in constant 1980 dollars per capita) 3.10(d) Capital Expenditures by Level of Government: LAC Countries (1975-1984 variation) 3.11(a) Intergovernment Transfers as a Percentage of GDP 3.11(b)* Transfers, Subsidies and Lending in Different Country Groupings 3.11(c)* Social Sector Expenditures of Central Governments by Colintry Groupings 3.16 Central Government Transfers, Subsidies and Lending in Argentina, Brazil, Colombia, Mex.co, and Venezuela 4.2* Overall Surplus/Deficit by Level of Government: World, Industrial Countries, Latin America, Sample LAC Countries (as Percentage of GDP) - 105 - 4.3(a) Per Capita Revenues of Subnational Governments (in constant 1980 US Dollars) 4.3(b) Per Capita Revenues of Subnational Governments (1975-1984 variation) 4.4 (a)* Revenue Structure of Subnational Governments (1975-1984 variation) 4.4 (b) Structure of Revenues of Subnational Governments in Argentina 4.4 (c) Structure of Revenues of Subnational Governments in Brazil 4.4 (d) Structure of Revenues of Subnational Governments in Colombia 4.4 (e) Structure of Revenues of Subnational Governments in Mexico 4.8(a)* Property Tax as Percentage of Total Revenues: LAC Countries 4.8(b)* Property Tax as Percentage of Total Revenues, Market Economies 4.12(a) Subnational Government Dependence on Transfers (in Percentage) 4.12(b) Trends of Intergovernmental Transfers (1975-186 variation) 4.12(c) Transfers to Subnational Governments: LAC Countries (in constant 1980 dollars per capita) 4.14 Domestic Lending Minus Repayment (as Percent of Total Central Governments Expenditures) 4.16(a)* 1982 Expenditures by Function and Level of Government: (Industrial Countries, Developing Countries, LAC Ceuntries) (as Percentage of GDP) 4.16(b) Subnational Government Expenditures: LAC Countries (as a Percentage of GDP) 4.16(c) Subnational Government Expenditures: LAC Countries (as Percentage of General Government Expenditures) 4.17(a) Subnational Government Expenditures: LAC Countries (as Percentage of General Government Expenditures) 4.17(b) Expenditure by Function an