rethinking decentralization in developing countries

Rethinking Decentralization in Developing Countries

The World BankWashington, D.C.

Jennie LitvackJunaid AhmadRichard Bird

S e c t o r S t u d i e s S e r i e s

© 1998 The International Bank for Reconstructionand Development / THE WORLD BANK

1818 H Street, N.W. Washington, D.C. 20433, U.S.A.

All rights reservedManufactured in the United States of AmericaFirst printing September 1998

The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s)and should not be attributed in any manner to the World Bank, to its affiliated organizations, or tomembers of its Board of Executive Directors or the countries they represent.

The material in this publication is copyrighted. Requests for permission to reproduce portions of itshould be sent to the Office of the Publisher at the address shown in the copyright notice above. TheWorld Bank encourages dissemination of its work and will normally give permission promptly and,when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy por-tions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222Rosewood Drive, Danvers, Massachusetts 01923, U.S.A.

Jennie Litvack is senior economist, Public Sector Group, Poverty Reduction and Economic Management,The World Bank. Junaid Ahmad is the Bank’s deputy resident representative, South Africa ResidentMission. Richard Bird is professor of economics, University of Toronto.

ISBN 0-8213-4350-5

Library of Congress Cataloging-in-Publication Data (CIP) has been requested.

iii

Foreword v

Abstract vii

1. Why Should We Worry about Decentralization? 1

2. Decentralization: A Complex Phenomenon 4Why Is Decentralization Happening? 4What Is Decentralization? 4What Are the Implications of Decentralization? 6Successful Decentralization Depends on Institution-Specific Design 7

3. Designing Decentralization: The Building Blocks of Fiscal Federalism 10Assigning Expenditure Responsibilities 10Raising Revenue 11Overseeing Subnational Borrowing 12Designing Intergovernmental Transfers 12Assembling the Building Blocks: A Systems Approach 13

4. Designing Decentralization: Incorporating Institutions 16Adjusting the Regulatory Framework and Decentralizing Borrowing 16Organizing Service Delivery 20Establishing Information Systems and Competitive Governments 21Adopting Asymmetric Decentralization 23Synchronizing Policy 24

5. Rethinking Decentralization in Developing Countries 26Incentives for Accountability 26The Role of Capacity 27Policy Dialogue and Project Design 28More Case Studies, Data, and Research 29

Annex. World Bank Economic and Sector Work on Decentralization, 1988–Present 32

Notes 35

References 37

Contents

v

Foreword

Masood AhmedVice President

Poverty Reduction and Economic Management Network

Central governments around the world are decentral-izing fiscal, political, and administrative responsibili-ties to lower-level governments and to the privatesector. Political pressure probably drives most de-centralization efforts. But whatever the origins, de-centralization can have significant repercussions forresource mobilization and allocation and ultimately formacroeconomic stability, service delivery, and equity.For these reasons, the World Bank is actively involvedin understanding the design and impact of decentral-ization policy in many developing countries.

Much of the literature on decentralization, norma-tive and empirical, is based on industrial countries.Developing nations, however, have very different insti-tutional frameworks. Given that the World Bank’s workis exclusively in developing countries, it must careful-ly consider what unintended consequences these insti-tutional differences might have on decentralizationpolicies and what the implications are for projectdesign and policy dialogue.

We do not know enough empirically to make defin-itive recommendations about which types of decen-tralization are best for which services in which institu-tional settings. But we do know that the best design willvary depending on circumstances and institutions andthat this complexity has sometimes been overlooked inthe haste to offer policy advice. If we can develop acommon framework, we can begin to document casesof decentralization, so that within a few years we may

be able to identify more precisely what works and whatdoesn’t in particular institutional settings. And by iden-tifying what institutions are important for successfuldecentralization in industrial countries and how theseinstitutions differ in developing countries, we can findways to compensate for weak institutions in the shortrun and to build the basic elements of key institutionsin the long run.

This paper, a product of the Public Sector Group andthe Decentralization Thematic Group in the PovertyReduction and Economic Management (PREM) Net-work, is intended to stimulate thinking and encouragea more nuanced approach to decentralization policy.This approach recognizes the importance of institu-tions and policy design in determining the impact ofdecentralization on efficiency, equity, and macroeco-nomic stability. It implies the need to address decen-tralization within a broader institutional assessment.The paper draws on a wide body of literature andrecent experiences documented by the World Bankand proposes a framework for improving the design ofdecentralization by incorporating information aboutcountry-specific institutions. The paper is the productof a partnership between an academic researcher, oper-ational staff, and network staff, an exchange that hasalready enriched the policy dialogue in several coun-tries. It is expected that PREM and other networks willcontinue to join in the policy dialogue through similarpartnerships in the future.

vii

Abstract

This paper draws on the literature and growing expe-rience with decentralization in developing countriesto explore how a wide range of variables can affectdecentralization efforts and how policies and incen-tives can be designed to improve outcomes. Thepaper highlights the fact that decentralization is nei-ther good nor bad for efficiency, equity, or macroeco-nomic stability; but rather that its effects depend oninstitution-specific design. It discusses the buildingblocks of fiscal federalism (expenditure and revenueassignment, intergovermental transfers, and subna-tional borrowing) and then discusses five meansthrough which decentralization policy and institu-tions interact. These are the regulatory framework forsubnational borrowing, the financing and delivery ofservices, information systems and competitive gov-

ernments, asymmetrical decentralization, and policysynchronization.

The paper’s starting point is the traditional fiscal fed-eralism approach. But the primary measures for localand central accountability assumed in most discus-sions of decentralization may not hold or are differentin many developing countries. Drawing on the evi-dence from the World Bank’s operational work, there-fore, the paper suggests the need for a stronger focuson institutions in designing decentralization policies.This broader agenda suggests an enhanced focus onaccountability, governance, and capacity in the contextof designing policies for decentralization. Thisapproach has strong implications for the Bank’s projectdesign and policy dialogue and calls for a reinvigorat-ed research effort focused on developing countries.

Central governments around the world are decentraliz-ing fiscal, political, and administrative responsibilitiesto lower-level governments and to the private sector.Decentralization is particularly widespread in develop-ing countries for a variety of reasons: the advent of mul-tiparty political systems in Africa; the deepening ofdemocratization in Latin America; the transition from acommand to a market economy in Eastern Europe andthe former Soviet Union; the need to improve deliveryof local services to large populations in the centralizedcountries of East Asia; the challenge of ethnic and geo-graphic diversity in South Asia, as well as ethnic ten-sions in other countries (Bosnia and Herzegovina,Ethiopia, Russia) and the attempt to keep centrifugalforces at bay by forging asymmetrical federations; andthe plain and simple reality that central governmentshave often failed to provide effective public services.

Political pressure probably drives most decentral-ization efforts. But whatever its origins, decentraliza-tion can have significant repercussions for resourcemobilization and allocation, and ultimately macroeco-nomic stability, service delivery, and equity. Since

decentralization can greatly affect economic develop-ment and poverty reduction, it is no surprise that theWorld Bank is actively involved with decentralizationpolicy in many developing countries.

The Bank is involved in decentralization issues in avariety of ways: • Since many countries are undergoing some form of

decentralization either by design or by default, agrowing number of Bank-funded projects are ineffect supporting sectoral decentralization strategies.Twelve percent of Bank projects completed between1993 and 1997 involved decentralizing responsibil-ities to lower levels of government. A sectoral andregional breakdown is provided in Table 1.

• The Bank is also supporting decentralizationthrough loans to subnational governments.Although these loans are (necessarily) guaranteedby the central government, they are otherwise nego-tiated and undertaken by independent local author-ities. Such loans are used for both specific projectsand state-level structural adjustment efforts (forexample, in Andhra Pradesh state in India). A new

1

1. Why Should We Worry aboutDecentralization?

Table 1: World Bank Projects with a Decentralization Component, by Sector and Region

Sector Percentage Region Percentage

Urban 43 Africa 19Health 27 South Asia 14Social funds 26 Latin America and the Caribbean 13Environment 16 Middle East and North Africa 11Water 13 Europe and Central Asia 6Agriculture 11Transport 10Education 9Energy 3Finance 1

Note: Given an average project cycle of five years, the database used for this analysis does not include projects that went to the Board after 1992. Thus thenumbers presented here likely underreport current decentralization-related projects.Source: Portfolio Review, PRMPS, 1998.

loan instrument was established in late 1997 to facil-itate subnational adjustment lending.

• Structural adjustment loans to central governmentshave supported the design of intergovernmental fis-cal relations and the process of decentralization (forexample, in Kyrgyz Republic, Russia, and Vietnam).

• Given the importance of decentralizaton in manycountries and its potential impact on a range ofdevelopment objectives, a number of recent Bankreports have focused on decentralization and relat-ed issues in specific regions and countries (seeAnnex 1 for a list of economic and sector work ondecentralization).1

• Finally, a growing number of the Bank’s countryassistance strategies are giving greater prominence todecentralization (such as those for Hungary, Russia,South Africa, Thailand, and Uganda, among others). One reason decentralization has attracted so much

attention is that it is often a cross-cutting reform thatcan relate to such important Bank concerns as the rela-tion between fiscal and financial development; macro-economic stability; poverty alleviation and the socialsafety net; institutional capacity, corruption, and gover-nance; investment in infrastructure; and the provisionof social services. Other international and national insti-tutions are currently studying various aspects of theseissues from their own perspectives.2 Given the impor-tance of decentralization for economic management,the Bank also needs to invest in understanding it.

Much of the literature on decentralization, norma-tive and empirical, is based on industrial countries andassumes the existence of institutions that are usuallyvery weak in developing countries. For example, muchof the traditional public finance literature related to fis-cal decentralization relies on “voice” and “exit” toensure local accountability and achieve the allocativeefficiency gains expected from decentralization.According to this view, if people are dissatisfied withdecisions made by local leaders, they can vote them outof power. If they do not like the package of local taxesand public services offered, they can exit (or “vote withtheir feet”) by moving to a jurisdiction that bettermatches their preferences (Tiebout, 1956).

Yet governments in many developing countries areoften not responsive to their citizens, and decision-making is rarely transparent and predictable. Opport-

unities for voice and exit are limited because of weakinstitutions. Democratic systems are often frail, ren-dering the electoral system a highly problematicmethod of achieving accountability. Strong, broadlybased local participation can overcome weak formalelection systems, but powerful elites make this diffi-cult in many places. Mobility is often constrained bypoor information, infrastructure, and legal frame-works, which result in weak markets for land, labor,and capital. Although all countries have at least someurban areas, in smaller municipalities and rural areasit is often unrealistic to expect a family to sell theirland, learn of employment opportunities in otherjurisdictions, physically move to the new area, andborrow money in a new locality where they areunknown. The chances of overcoming these obstaclesare further reduced because low incomes and weaksocial safety nets often make households risk averse indeveloping countries.

A third channel for local accountability is throughhierarchical relationships within the public sector(World Bank, 1997), including the formal rules andoversight arrangements that exist within government.This is the primary channel for accountability in manydeveloping countries, but it is often ineffective becauseof poor information and monitoring systems. In short,the primary measures for local accountability assumedin most discussions of decentralization may not holdin many developing countries.

Given that the World Bank’s work is exclusively indeveloping countries, it must carefully consider whatunintended consequences these institutional weak-nesses might have on decentralization policies, andtheir implications for project design and policy dia-logue. This paper is intended to initiate this type ofanalysis and help those working on decentralizationin developing countries develop a more institutional-ly sensitive perspective when analyzing decentraliza-tion in a particular country.3 We draw on the literatureand growing experience with decentralization indeveloping countries to explore how a wide range ofvariables can affect decentralization efforts and howpolicies and incentives can be designed to improveoutcomes.4

We do not know enough empirically to make defin-itive recommendations about which types of decen-

2 Rethinking Decentralization in Developing Countries

Why Should We Worry about Decentralization? 3

tralization are best for which services in which institu-tional settings. Much of the discussion of decentraliza-tion reflects a curious combination of strong precon-ceived beliefs and limited empirical evidence. But wedo know that the best design will vary depending oncircumstances and institutions, and that this complex-ity has sometimes been overlooked in the haste to offerpolicy advice. If a common framework can be estab-lished, we can begin to assemble documented casestudies of decentralization, so that within a few yearswe may be able to identify with more precision patternsof success and failure. Also, by identifying institutionsthat are important for successful decentralization inindustrial countries and that may be weak in develop-ing countries, we can also begin to identify ways to

compensate for the weak institutions in the short run,and build key elements of important institutions in thelong run.

The next section describes the complexity of thedecentralization process. It concludes that the debateon whether decentralization is “good” or “bad” is unpro-ductive since decentralization is a political realityworldwide—one that varies greatly in form within andamong countries. The third and fourth sections focuson the lessons learned about the design of decentral-ization, with an emphasis on the building blocks of fis-cal federalism and on the importance of institutions indecentralization efforts. The final section draws impli-cations from this analysis and describes the researchagenda that remains to be explored.

4

2. Decentralization: A ComplexPhenomenon

Decentralization—the assignment of fiscal, political,and administrative responsibilities to lower levels ofgovernment—is occurring worldwide for different rea-sons, at different paces, and through different means.The “why” of decentralization is as varied as the “how”(Box 1). The complexity inherent in the decentraliza-tion process is further aggravated by its cross-cuttingimpact. For policymakers, who often have little controlover the political genesis or pace of decentralization, thechallenge is to implement it in a way that ensures thestability, efficiency, and equity of the economic system.

Why Is Decentralization Happening?

In most countries decentralization reflects a broaderprocess of political and economic reform (WorldBank, 1997). Political changes worldwide have givenvoice to local demands and the need to bring eco-nomic and political systems closer to local communi-ties. In some cases the very preservation of a nationalpolitical system has required the decentralization ofpower. In addition, technological changes and globalintegration of factor markets have changed the size ofgovernment needed to manage economic systems. Onone hand, economic mobility has led to the creationof supranational bodies to manage the growing eco-nomic integration among nations. On the other, anincreasing number of public services can be efficient-ly provided by decentralized (and often private) orga-nizations (World Bank, 1995a). For example, in met-ropolitan Buenos Aires the entire water and seweragesystem is operated, maintained, and invested in by aconsortium of private companies (Triche, Mehia, andIdelovitch, 1993). Finally, the collapse of central eco-nomic systems has encouraged regional and local gov-

ernments to participate in the political and economicprocess.

What Is Decentralization?

Decentralization is not easily defined. It takes manyforms and has several dimensions. Indeed, a wide vari-ety of institutional restructurings are encompassed bythis label, and several variants may be operating at thesame time within a country, and even within a sector.Thus care must be used in labeling, and labels—including those used in this paper—must be inter-preted with care.

One widely used distinction is among deconcentra-tion, delegation, and devolution (Rondinelli, 1981 and1989). Deconcentration occurs when the central govern-ment disperses responsibilities for certain services to itsregional branch offices. This does not involve any trans-fer of authority to lower levels of government and isunlikely to lead to the potential benefits or pitfalls ofdecentralization. The “decentralization” that has oc-curred in many unitary countries is actually deconcen-tration, since independent local governments (which arelegally accountable to local constituents) do not existand local field offices of the central government are sim-ply used to improve the efficiency and effectiveness ofservice delivery. This is the case in many East Asiancountries and, until recently, was the rule in EasternEuropean countries (Kornai, 1992). Deconcentrationcan also exist for some functions in federal countrieswhen the central government maintains a strong inter-est in ensuring delivery of a particular service.

In contrast, the central issue for both delegation anddevolution relates to the balancing of central and localinterests. Delegation refers to a situation in which the

Decentralization: A Complex Phenomenon 5

Many rationales for decentralization may be discerned inthe literature and in practice. Not all will be relevant inany one country, nor are they all necessarily consistent orequally important to all relevant parties. As noted, mostare based on normative or empirical work in industrialcountries.

The most common theoretical rationale for decentral-ization is to attain allocative efficiency in the face of dif-ferent local preferences for local public goods (Musgrave,1983; Oates, 1972; Tiebout, 1956). Problems may arisewith respect to coordination—which is itself costly(Breton and Scott, 1978)—where interjurisdictionalspillovers are important, including stabilization (Tanzi,1996; Wildasin, 1997) and distribution (Tresch, 1981).Most World Bank work on decentralization has focusedon this rationale (World Bank 1988, 1994, and 1997).

Related issues, on which there is neither theoretical norempirical agreement, concern the direction and impor-tance of the relation between decentralization and the sizeof the public sector (Mueller, 1996; Ehdaie, 1994) andbetween decentralization and the rate of economic growth(Martinez-Vazquez and McNab, 1997; Zhang and Zou,1998).

Equity and distributional concerns about decentraliza-tion cut both ways. Some analysts argue that in some cir-cumstances local governments achieve such goals moreeffectively than central governments (Pauly, 1973). Othersargue that central redistribution is needed both for effec-tiveness (Musgrave, 1983) and to overcome biases of localelites (Wilensky, 1974; Inman and Rubinfeld, 1997).Another view is that, regardless of what local governmentsmay attempt to do in the way of redistribution, theirattempts will be frustrated by resource mobility and theopenness of the local economy (Buchanan and Wagner,1971). All these views may be found in different Bankcountry studies in recent years.

Decentralization’s potentially destabilizing effect onthe macroeconomy has caused much concern to some(Prud’homme, 1995; Tanzi, 1996; Ter-Minassian, 1997).Others suggest, however, that such effects are more like-ly to reflect inappropriate incentives than any probleminherent to decentralization (Spahn, 1997a and 1997b).In countries that have “decentralized” to offload fiscalimbalances from the center, it is of course not surprisingto see a strong association between decentralization andfiscal imbalance at lower levels (Wallich, 1994). At leastthree World Bank studies—on Argentina (1996a), Brazil(Dillinger, 1997), and Colombia (1996b)—have exam-ined this issue and suggested that destabilization effects

arose mainly from design problems, including a soft bud-get constraint between levels of government.

An interesting argument asserts that a primary econom-ic rationale for decentralization is to improve the “compet-itiveness” of governments—that is, decentralization willmake local governments try to satisfy the wishes of citizens(Breton, 1996; Salmon, 1987). Although suggestive, therehas been little empirical examination of this idea (Kenyon,1991; Mueller, 1997). Some Bank work has occasionallyrecognized “horizontal” competition, particularly withrespect to taxation (both at the national and local levels),although it has generally been viewed as a problem ratherthan a potential benefit. Little attention seems to have beenpaid to the more troublesome question of “vertical” inter-governmental competition and the nature and enforcementof the rules that may make intergovernmental competitionbeneficial (Breton, 1996). However, Weingast (1995) pro-poses that competitive local governments act as checks onthe central government to prevent it from confiscatingwealth; thus federalism preserves markets.

Another political rationale for decentralization is thatgood governments are those closer to the people (Inmanand Rubinfeld, 1997; World Bank, 1997). Since JohnStuart Mill, this argument has been closely related to ques-tions of political participation and democracy, but it is stillundeveloped in theoretical (Breton, Cassone, Fraschini,1998) or, especially, empirical terms (Martinez-Vazquezand McNab, 1997). Although there has been some indi-rect discussion of this proposition in relation to such gov-ernance concerns as corruption (Prud’homme, 1995),little has yet been done to link decentralization, participa-tion, and good governance apart from an interesting gen-eral empirical paper (Huther and Shah, 1998), one exam-ining electoral accountability and economic policy choicesby U.S. governors (Besley and Case, 1995), and a reveal-ing case study on Colombia (World Bank, 1995b).

A quite different political rationale for decentralizationis to accommodate pressure for regional autonomy and,hence, perhaps increase the legitimacy and sustainabilityof heterogeneous national states. In a sense, there maysometimes be a tradeoff between political and economicstability (Treisman, 1998). Although touched on in a Bankstudy of Russia (Litvack, 1994), this idea has—perhapsunderstandably, given the nature of the Bank’s clients—not been much developed in Bank literature. Still, theemphasis in many studies on the role of “equalization”transfers can perhaps be attributed to political stability aswell as a more magnanimous desire for interjurisdiction-al equity (Ahmad, 1996).

Box 1: Rationales for Decentralization

central government transfers responsibility for deci-sionmaking and administration of public functions tolocal governments or semiautonomous organizationsthat are not wholly controlled by the central govern-ment but are ultimately accountable to it. These orga-nizations usually have a great deal of discretion in deci-sionmaking. This form of decentralization can becharacterized as a principal-agent relationship, withthe central government as the principal and the localgovernment as the agent. From this perspective, themain design issue is to ensure that the self-interestedagent (the local government or semiautonomous orga-nization) faces incentives that induce it to act as close-ly as possible in accordance with the wishes of the prin-cipal (the central government).5

Finally, devolution, a more extensive form of decen-tralization, refers to a situation in which the centralgovernment transfers authority for decisionmaking,finance, and management to quasi-autonomous unitsof local government. Devolution usually transfersresponsibilities for services to municipalities that electtheir own mayors and councils, raise their own rev-enues, and have independent authority to makeinvestment decisions. In a devolved system, local gov-ernments have clear and legally recognized geograph-ic boundaries over which they exercise authority andwithin which they perform public functions (Ron-dinelli, 1998).

The shift in responsibility between tiers of govern-ment is underpinned by several fiscal, political, andadministrative instruments. These define the extent towhich intergovernmental relations are deconcentrated,delegated, or devolved. Fiscal decentralization—who setsand collects what taxes, who undertakes which expen-ditures, and how any “vertical imbalance” is rectified—has been especially prominent in recent discussions inmany countries, but as just indicated many of the morefundamental questions relate to political and adminis-trative decentralization. Political decentralization refers atone level to the extent to which political institutions mapthe multiplicity of citizen interests onto policy decisions(Inman and Rubinfeld, 1997). Administrative decentral-ization is concerned with how political institutions, oncedetermined, turn policy decisions into allocative (anddistributive) outcomes through fiscal and regulatoryactions. The political decision to devolve powers from

central government, for example, can only be translatedinto actual powers being shifted if subnational govern-ments have the fiscal, political, and administrativecapacity to manage this responsibility.

What Are the Implications of Decentralization?

For policymakers and their advisers, the multidimen-sional aspect of decentralization—the dispersion of fis-cal, political, and administrative powers—suggeststhree implications that heavily influence the context forthinking through decentralization. First, becausedecentralization can change the mobilization and allo-cation of public resources, it can affect a wide range ofissues from service delivery to poverty reduction tomacroeconomic stability. (Thus we refer to it as a“cross-cutting” issue.) Second, the management ofdecentralization requires intimate knowledge of localinstitutions and a nuanced understanding of theprocess of decentralization—that is, what is drivingdecentralization in a country (and sector) and whichstakeholders are involved. Third, limited empirical evi-dence exists about what works and what does not.Together these three factors pose a daunting challengefor those responsible for designing and managingdecentralization.

Broadly based reform and cross-cutting impacts

Decentralization is often implemented as a broadlybased reform that affects various sectors and levels ofgovernment. Outcomes reflect the interaction andcoordination of policies between different tiers of gov-ernment. In many countries it is a considerable chal-lenge to coordinate sectoral reforms undertaken by aministry of the central government with decentraliza-tion of fiscal, political, and administrative responsibil-ities to local governments.

The cross-cutting influence of decentralization hasa clear implication for the World Bank. Understandingthe evolution and status of a nation’s decentralizationprocess is essential to understand and analyze policyissues across sectors. For example, the successful deliv-ery and financing of education or welfare systems, asin South Africa, depends on the public finances of mul-


titiered governments rather than the edicts of a centralline ministry. In general, major policy reform of indi-vidual sectors—such as health, education, or infra-structure—cannot be undertaken without an assess-ment of a country’s intergovernmental system, or viceversa. This outlook needs to be embedded in the coun-try team framework at the Bank.

Country context

Many factors affect both the ideal and the actual formof decentralization adopted in any country for any ser-vice at any time: • The number of subnational units as well as their

absolute and relative sizes and wealth.• The distribution of functions (relative to the “span”

of public goods, externalities and jurisdictionalspillovers, and so on).

• The nature of the “common” institutions (with par-ticular attention to their effects on government com-petition).

• The role and status of the constitution (for example,with respect to the independence of the judiciaryand collective rights).

• The technical characteristics and policy objectives ofspecific public services.

• The current political situation.Much of the discussion of decentralization pre-

sumes that it is a matter of choice or deliberate design,but in many countries it may equally well be either apolitical necessity (for example, to hold a fractious het-erogeneous country together) or a default option (thatis, a way to try to execute more satisfactorily some nec-essary functions of the state in the face of central fiscalcrisis and political weakness). Indeed, all these ratio-nales may be at play at once, possibly to varyingdegrees, in a single country. For these and other rea-sons, decentralization often encompasses an extreme-ly nuanced set of activities that can only be understood,analyzed, and—to the extent possible—guided on thebasis of thorough knowledge of local institutions.

Strong beliefs but limited empirical evidence

Many people hold strong beliefs about various aspectsof decentralization: its intrinsic political and economic

merits, its potential problems, and its effects on staticand dynamic allocative efficiency, income distribution,macroeconomic stability, institutional demands on localcapacity, potential for corruption, governance, and soon. But the actual empirical evidence on these proposi-tions is either nonexistent or conflicting. In somerespects this is not surprising. Given the complexity andmultiple dimensions of the concepts involved and thecontext-specific nature of decentralization, it is to beexpected that studies will show considerable uncer-tainty about the strength and even the direction of therelation between decentralization and, say, growth, gov-ernance, distribution, and stability. Conflicting evi-dence and interpretations may be found with respect todecentralization and local tax efforts (Bird and Fiszbein,1998), the relationship (if any) between increased localdemocracy and allocative efficiency (Martinez-Vazquezand McNab, 1997), the relationship between subna-tional expenditures and growth (Zhang and Zou, 1998;Davoodi and Zou, 1998), and the ability to target thepoor under different forms of decentralization (Ra-vallion, 1998; Alderman, 1998; Inman and Rubinfeld,1997). A particular challenge in this respect is to under-stand how best to match fiscal, political, and adminis-trative arrangements to achieve the potential benefits ofdecentralization for any given service in any givencountry.

Successful Decentralization Depends onInstitution-Specific Design

Designing decentralization policy is difficult in anycountry because decentralization can affect manyaspects of public sector performance and generate awide range of outcomes. But it is particularly difficultin developing countries because institutions, informa-tion, and capacity are all very weak. The cross-cuttingnature of decentralization, the importance of localinstitutions in influencing the impact of decentraliza-tion, and the limited empirical evidence on what worksand what does not make the design and implementa-tion of decentralization a considerable challenge (forexample, matching expenditures and revenues at eachlevel of government, providing a regulatory frameworkthat imposes a hard budget constraint on subnational


governments, and incorporating local participationand accountability in decentralization). Evidence sug-gests that the problems associated with decentraliza-tion in developing countries reflect flaws in design andimplementation more than any inherent outcome ofdecentralization. The following sections illustrate thisimportant point by discussing how design can alter theimpact of decentralization on equity and macroeco-nomic stability.

Decentralization and equity

The impact of decentralization on interregional andinterpersonal equity can vary greatly depending on insti-tutional arrangements and policy design details. If thecentral government makes no effort to redistributeresources to poorer areas, fiscal decentralization willresult in growing disparities. Similarly, if provinces orstates do not redistribute within their jurisdictions, poorpeople may lack access to public services. But decentral-ization need not produce such outcomes. Depending onthe preferences of the national population, horizontalequity—that is, ensuring some level of comparability inability to provide public services throughout the coun-try—can be achieved through intergovernmental trans-fers that include equalization components. How inter-governmental transfers are designed (block or specificpurpose, matching or nonmatching, flat rate matching orprogressive matching) and monitored largely determinesthe extent to which decentralization results in greater bal-ance or imbalance in service provision between regions,states, towns and villages, and households.

Decentralization can affect interpersonal equitythrough public expenditure policy, tax policy, and thedesign of intergovernmental transfers. If decentraliza-tion of expenditure decisions leads to a greater shareof public resources being spent on services used by thenonpoor, equity will suffer. Similarly, if decentraliza-tion leads to a higher tax burden for the poor (throughhigher user fees and local taxes and no offsettingchanges in other taxes), then this too can weaken equi-ty. Nonetheless, if the central government is con-cerned about preserving equity and protecting thepoor, it can do so in part—despite decentralizedexpenditures and taxes—through the design of inter-governmental transfers.

How decentralization affects equity also depends inpart on the extent of local accountability and localpolitical participation by the poor. Accountability canbe enhanced when local leaders are elected and areconcerned about providing services to their con-stituents. When the poor participate in the politicalprocess, they can exert influence on leaders. The mereexistence of a democratic political system is insufficientunless there is meaningful political participation by allgroups. For example, India has a long democratic tra-dition but since local participation depends on socialcaste, the poor often have little influence. In contrast,in Cuba and Vietnam there is considerable participa-tion at the grassroots level (World Bank, 1996c). InOaxaca, Mexico, the targeting of poverty expendituresdepends heavily on how representative local govern-ments are (Fox and Aranda, 1996).

Decentralization can also enhance access by thepoor if it increases competition in the delivery of ser-vices (such as water and electricity) and this drivesdown prices. In addition, if policy reform separatesefficiency and equity objectives, decentralization canenable service deliverers to charge user fees and focuson efficiency and then use the revenues to expand cov-erage or improve quality, which can benefit the poor.For example, the introduction of earmarked user feesfor primary health care in Cameroon led to an improve-ment in quality and an increase in utilization—partic-ularly for the poor, who previously lacked access toalternative providers (Litvack and Bodart, 1993).

Decentralization and macroeconomic stability

The stringent conditions for successful decentraliza-tion have recently been emphasized with respect todeveloping countries (Prud’homme, 1995; Tanzi,1996). In particular, it has been argued that not onlycan decentralization fail to improve local service deliv-ery, it may even risk national destabilization. Argentinain the 1980s is a commonly cited example, but othersare not hard to find in the transition economies ofCentral and Eastern Europe (Bird, Ebel, and Wallich,1995). Similar fears appear to have played a role inChina’s fiscal reforms (Bahl, 1998).

Concerns about decentralization and macroeconom-ic instability are fueled by instances where governments


have decentralized without adequate planning to con-trol the deficit behavior of local governments. Forexample, international experience suggests that if coun-tries decentralize more expenditure responsibilitiesthan revenue resources, service levels will fall or localgovernments will press—sometimes successfully—formore transfers, more loans, or both. One of the clearestand most analyzed cases of this phenomenon is theRussian Federation (Wallich, 1994). On the other hand,if more revenues than expenditures are decentralized,local revenue mobilization may decline because of taxcompetition, and again macroeconomic imbalancesmay emerge—as in Argentina, Brazil, and Colombia.This risk is greatest when revenues are decentralizedwithout adequate steps to ensure that local revenuemobilization is maintained and that local authorities arecapable of carrying out the corresponding expenditureresponsibilities. Even if both sides of the budget aredecentralized in a balanced fashion, local governmentsmay not have adequate administrative or technicalcapacity to carry out their new functions. Such prob-lems may give rise to particular concern in developingcountries where local governments are charged withimportant social and economic infrastructure invest-ments (Bird, 1994).

Regardless of their empirical validity, concernsabout macroeconomic disaster ensuing from fiscaldecentralization rank high in many countries. Thuscare must be taken to avoid unwanted outcomes inthis respect. Some analysts fear that unchecked sub-national governments may increase current expendi-tures well above their capacity to finance them out of

current revenues and then attempt to close the gapthrough borrowing. Others argue that since macro-economic stabilization is properly a national govern-ment task, it is important that the national govern-ment have full control over all the policy instrumentsit needs to carry out this task, including borrowing—particularly borrowing abroad. The key to addressingthis problem is to ensure that decentralization isundertaken in a way that increases rather thandecreases accountability. This can be done by imple-menting stronger rules and regulations (such as thebalanced budget rules in the United States) and bychanging the incentives facing various actors.

If a national government wants to avoid macroeco-nomic problems arising from subnational debt, it cando so by not subsidizing such borrowing and by let-ting subnational governments that borrow too muchgo bankrupt. One way to approach this is to make therules of the game clearer and more transparent. Thisis exactly what was done in Morocco, where the gov-ernment changed the subsidy scheme for local gov-ernments from one of budget-balancing grants, inwhich capital and interest payments on loansincreased transfer receipts, to a formula-based equal-ization transfer that takes no account of borrowing(Vaillancourt, 1998). In addition, lenders were explic-itly told not to count on financial bailouts. As a gen-eral rule, however, the difficulty of envisioning, letalone carrying out, bankruptcy in the public sectorprovides good reason to require fairly stringent con-ditions on subnational borrowing to ensure local gov-ernment accountability.


10

3. Designing Decentralization: The BuildingBlocks of Fiscal Federalism

Much of the literature on decentralization, includingthat generated by the World Bank, reflects the “fiscalfederalism” orientation of the traditional publicfinance approach to this issue (Walsh, 1992; Shah,1994). The emphasis is on setting the appropriateexpenditure and tax assignment for each tier of gov-ernment and on designing intergovernmental trans-fers. The framework is driven by the Musgravianprinciples of efficiency, equity, and stability(Musgrave, 1983). Although this literature has con-sidered institutional issues, particularly in recentyears (Huther and Shah, 1998), the main emphasishas been on economic policies. Less attention hasbeen paid to how these policies might be implement-ed in disparate institutional settings and what impactthey might have on the design of institutionally sen-sitive policy. This section reviews the traditionalissues of public finance, while the next sectionexplores the institutional issues further.

Assigning Expenditure Responsibilities

Both theory and experience indicate that it is impor-tant to state expenditure responsibilities as clearly aspossible in order to enhance accountability and reduceunproductive overlap, duplication of authority, andlegal challenges. In theory, decisionmaking shouldoccur at the lowest level of government consistent withallocative efficiency (for example, the geographic areathat internalizes the benefits and costs of decision-making for a particular public service). Thus the opti-mal size of jurisdiction for each service and servicecomponent would likely differ, although in practiceeconomies of administration and transactions costswould presumably lead to the grouping of roughly

congruent services at the local (street lighting, refuseremoval), regional (rural-urban roads, refuse disposal),and national (intercity highways, environmental poli-cy) levels.

Decentralized decisionmaking enlarges possibilitiesfor local participation in development. However,accountability is often best promoted by establishing aclear and close link between the costs and benefits ofpublic services, so that the overall amount of expendi-ture responsibility assigned to a level of governmentideally will correspond to the amount of revenues thatlevel has at its potential command.

Some national allocative objectives can be carriedout either directly by the central government or indi-rectly by local governments responding to incentivescreated by national grants and regulations (as well asinterlocal or interregional agreements). Moreover,national governments have obvious roles in formulat-ing stabilization and distribution policies, and atten-tion must be paid to possible local conflicts with thesepolicies. In many instances it seems appropriate forsome functions to be shared in the sense that higherlevels of government may exercise a regulatory or pol-icy role, while lower levels of government are respon-sible for delivering services.

Sectors need to be unbundled because some aspectsof service delivery are appropriately undertaken by high-er levels of government while other components are bestdone locally. For example, there are four broad functionsin water supply and sewerage (usually considered a localservice), and each has several components:• Policymaking (central and local, financial, econom-

ic, environmental, health-related, subsidies). • Regulation (economic, environmental, water quali-

ty, service quality).• Investments (planning, financing, execution).

Designing Decentralization: The Building Blocks of Fiscal Federalism 11

• Operations (production and treatment, distribu-tion, commercial operations and consumer rela-tions, maintenance).

The characteristics of each function differ and thushave different implications for the role of the govern-ment and private sector, and the level of governmentinvolved (Triche, 1993).

A key to designing good policy is a clear under-standing of what outcomes are important for the cen-tral government, and what outcomes can be deter-mined by local governments. The central governmentshould retain control (either by directly providing ser-vices or by creating incentives for local governments toact according to central preferences) over functions forwhich it desires certain outcomes (such as family plan-ning), and should relinquish control on issues wherelocal divergence in priorities is not a big issue. Localgovernments must be held accountable to the centerfor nationally mandated services, while for otherdevolved functions they must be held accountable totheir local constituents.

Finally, experience has shown that effective decen-tralization requires complementary adaptations in avariety of institutional arrangements for intergovern-mental coordination, planning, budgeting, financialreporting, and implementation. Such arrangementsmay encompass both specific rules (for example, in thedesign of fiscal transfers) and provisions for regularintergovernmental meetings and periodic reviews ofintergovernmental arrangements. Detailed central con-trol over local use of funds is seldom appropriate.Instead, what is needed is better monitoring and over-sight of local fiscal performance.

Raising Revenue

Two broad principles of tax assignment are the need foreconomic and administrative efficiency and the desir-ability (for accountability reasons) of having each gov-ernment finance its own expenditures out of its ownrevenues. Unfortunately, these two principles general-ly conflict. • Economic and administrative efficiency suggests

that taxes on mobile factors (such as corporate andpersonal income taxes), value-added taxes, and

taxes on international trade should be assigned tothe central government. Similarly, local govern-ments should presumably tax only immobile factors(such as land and real estate) in addition to apply-ing user charges wherever feasible.

• Because the central government has a large role toplay in achieving distribution and stabilizationgoals, it seems logical for it to be responsible forprogressive (redistributive) taxes (such as those onwealth and personal income), as well as those taxesmost sensitive to economic fluctuations (such ascorporate income taxes and taxes on naturalresources) and any specific taxes (for example, car-bon taxes) related to national objectives.6 On theother hand, to ensure accountability to local resi-dents, local governments should to the extent fea-sible be responsible for financing local services.When local governments are responsible for suchmajor expenditures as education, accountabilitythus suggests more taxing power for local govern-ments than efficiency—or central policy—permitsin most countries.An important national policy consideration in coun-

tries where subnational governments have indepen-dent taxing power is to ensure an adequate degree ofinternal tax harmonization and coordination to pre-serve the internal common market. This concern alsoargues against giving lower-tier governments thepower to impose taxes on corporate income, value-added, or natural resources. For corporate taxes, suchgovernments may engage in inefficient tax competitionor in equally undesirable (for accountability reasons)interjurisdictional tax exporting. Tax exporting (fromproducer to consumer regions) and administrativecomplexity provide strong arguments against subna-tional value-added taxes. Tax exporting and instabilityboth argue against assigning resource revenues to localor state governments. The same arguments can bemade with respect to tax-sharing arrangements underwhich local governments receive a fixed percentage ofcertain national taxes collected in their jurisdiction.Such arrangements create undesirable incentives fortax exporting and bias national tax policy (in the direc-tion of raising taxes that do not have to be shared).Moreover, tax rates in such systems are invariably setby the central government, and the sharing rate is often


applied uniformly throughout the country. Thus theaccountability link is broken and subnational govern-ments have no incentive to ensure that the amount andpattern of their spending is efficient.

As a rule, the allocation of revenues in most coun-tries provides most (and often all) lower-tier govern-ments with insufficient own-source revenues to financethe expenditures for which they are responsible. Apartfrom borrowing, there are two ways to bridge thisfinancing gap, either or both of which may be applic-able to varying extents in different countries.

The first is to supplement local revenues with inter-governmental fiscal transfers—without undesirablyreducing local efforts to collect taxes (see below). Thesecond is to permit subnational governments to levytheir own broadly based taxes—so long as they burdenlocal residents only. In principle two such taxes arepossible: a retail sales tax and a personal income tax.Administrative difficulties rule out retail sales taxes inpractice. Thus the only efficient broadly based subna-tional tax that seems both feasible and desirable is like-ly to be a flat-rate surtax (or surcharge) on a nationalpersonal income tax—provided that such a tax existsand works moderately well.7

To ensure local accountability with respect to sucha surtax, a local property tax, or local taxes in general,local governments must be restricted (to the extentpossible) from exporting taxes and allowed to set theirown tax rates. For efficiency it may be desirable toassess the base of a tax centrally and even to have it col-lected by the central government. But for accountabil-ity it is critical that the local authorities be responsible(perhaps within limits) for setting the tax rate.

Overseeing Subnational Borrowing

Subnational governments can be permitted to borrow,but a great deal of caution is required. Unchecked sub-national governments, particularly those highly depen-dent on national transfers, may increase current expen-ditures well above their capacity to finance them out ofcurrent revenues and then close the gap through bor-rowing (especially if, as in Argentina, local govern-ments control provincial banks), often pledging futurerevenue transfers to service the loans. More generally,

it is sometimes argued that subnational governmentsshould be restrained from accentuating cyclical pres-sures by borrowing—especially borrowing abroad andthus adding to the national obligation to service foreigndebt (Ter-Minassian, 1997). On the other hand, alloca-tive efficiency and intergenerational equity oftenrequire that long-lived investment projects, especiallythose that will increase productive capacity, befinanced by borrowing rather than relying solely oncurrent public savings or transfers from above. Thequestion, then, is how to distinguish good borrowingfrom bad borrowing.

In the past subnational borrowing was an issue offiscal federalism because local borrowing was guaran-teed by central governments or secured through cen-tral intermediaries. This has changed significantly inrecent years with the rise of local and global capitalmarkets, increasing political decentralization, and theimportance of the regulatory environments in whichsubnational borrowing occurs. For these reasons, sub-national borrowing is addressed in greater detail insection 4.

Designing Intergovernmental Transfers

Since transfers are the main source of revenue for sub-national governments in most countries, the design oftransfers is of critical importance to the success ofdecentralization. Transfers can be broadly divided intononmatching (lump sum) and matching transfers.Nonmatching transfers can in turn be divided intoselective (conditional) and general (unconditional).Conditionality ensures that the recipient governmentspends at least the amount received on the designatedfunctions, though the fungibility of funds means thatsome money that would have been spent on suchfunctions in any case may be diverted elsewhere.Empirical studies of industrial countries suggest thatas a rule even unconditional grants are fully spent byrecipient governments rather than used to lower localtaxes (the so-called “flypaper effect,” meaning thatmoney sticks where it hits; Hines and Thaler, 1995).Similar results have been noted in developing coun-tries, such as Colombia (World Bank, 1996b).Matching transfers require that the funds be spent on


specific purposes and that the recipient to some extentmatches the grant out of its own funds. Such transfersmay distort local priorities and be consideredinequitable in that richer jurisdictions can raisematching funds more easily. But the latter problem canbe offset to some extent by varying matching rateswith jurisdictional wealth, and the former may be thedesired outcome when the transfer is intended to, forexample, internalize spillovers or achieve overridingnational policy objectives.

When local governments are expected to play amajor role in delivering social services, the design ofintergovernmental transfers is particularly importantbecause the central government usually retains a stronginterest in at least some of the outcomes. If the centralgovernment is, in effect, using local governments asagents in executing national policies—for example, toprovide primary education at a specified level through-out the country—then it is important to make thetransfer conditional on the funds actually being spenton education or on the attainment of some level of edu-cational performance (Bird, 1993).8

If key services are to be provided through decentral-ized governments, careful attention has to be paid to get-ting the prices facing service providers right (for exam-ple, through a well-designed system of matching grants),setting up an information and inspection system capa-ble of ensuring that the desired services are delivered tothe target groups, and devising some system (such as anational “fail-safe” provision) for dealing with the non-compliant without punishing the innocent.

As Bahl and Linn (1992) show, the most appropri-ate form of a transfer largely depends on its objective.Regardless of the particular design, however, goodintergovernmental transfer programs share certaincharacteristics:• Transfers are determined as objectively and openly

as possible, ideally by some well-established formu-la. They are not subject to hidden political negotia-tion. The transfer system may be decided by the cen-tral government alone, by a quasi-independentexpert body (such as a grants commission), or bysome formal system of central-local committees.

• Transfers are relatively stable from year to year topermit rational subnational budgeting. At the sametime they are sufficiently flexible to ensure that

national stabilization objectives are not thwarted bysubnational finances (Box 2). One approach thatappears to achieve this dual objective is to set thetotal level of transfers as a fixed proportion of totalcentral revenues, subject to renegotiation periodi-cally (say, every three to five years).

• The transfer formula (or formulas) is transparent,based on credible factors, and as simple as possi-ble. Unduly complex formulas are unlikely toprove feasible or credible in countries where, forexample, there is serious dispute about regionalpopulation sizes. If several of the objectives discussed earlier are

applicable—for example, some degree of equalizationis desired while at the same time there are clear nation-al policy objectives—it will generally assist both clari-ty and effectiveness if separate transfers are targeted ateach objective (see World Bank 1996b).

Assembling the Building Blocks: A SystemsApproach

The assignment of expenditures, revenues, transfers,and subnational borrowing together compose the sys-tem of intergovernmental finance. Though there ismuch to know about each of these components as wellas their sequencing, it is perhaps most important toknow that these issues must be considered together, aspart of a complete system. This system, combined withthe institutional environment (see the next section),determines impacts on efficiency, equity, and macro-economic stability. For example, an intergovernmentalfiscal system might assign local governments theresponsibility for financing primary school, collectinglocal school fees, and administering central funds totarget subsidies to poor schoolchildren. This set ofexpenditure, revenue, and transfer policies canincrease efficiency and equity. But if the three policiesare not designed with a common objective, the quali-ty of services could decline dramatically and the poorcould lose access to schooling.

Although the ultimate impact of decentralizing anyone fiscal instrument will depend on the design of allof them, all components should not be designed simul-taneously. Expenditure functions should be assigned


Any good system of intergovernmental grants must be sta-ble. Yet it must also be flexible. How can both character-istics be achieved simultaneously? From the perspectiveof a central government, the best system might be one inwhich the total amount to be transferred (sometimescalled the “distributable pool” or the “primary distribu-tion”) is determined each year in accordance with budgetpriorities. But unless the amounts to be received are tosome extent predetermined and predictable, recipientgovernments will not be able to budget properly and willnot face an appropriately hard budget constraint. One wayto provide some degree of stability to local governmentsand some degree of flexibility to the central governmentwithout biasing central tax policy is to establish a fixedpercentage of all central taxes (or current revenues) to betransferred—preferably in accordance with an objectivedistributive formula (the “secondary distribution”) thattakes into account such factors as the needs and capacityof recipient governments. If not too complex, such a sys-tem will also be transparent. Moreover, interjurisdiction-al equalization can be built into the formula.

Many transfer systems in developing countries possessfew or none of these characteristics. In Russia and other for-mer socialist countries transfers are negotiated every year inan obscure political process. In Bangladesh nominally pre-determined transfers are not made available in a reliablefashion. And in Argentina the factors in the formula arecomplex, contradictory, and not optimally equalizing.

Fortunately, better examples can be found around theworld. The Philippines, for example, distributes 20 percentof national internal revenues to local governments. Similarly,Colombia distributes 25 percent of (non-earmarked)national current revenues to departmental (intermediate-level) governments, in part in equal portions and in part onthe basis of population. Colombia also distributes 30 per-cent of value-added tax revenues to municipal governments,largely on the basis of population. Although these systemshave their flaws, they are objective, stable, and transparent.

Even these systems, however, make no formal provisionfor periodic reexamination of the total amount to be dis-tributed or the distribution formula (although Colombia’s1991 constitutional reform provided for a one-time reeval-uation of the municipal grant system after five years). Incontrast, formally federal countries such as India andPakistan have established fiscal commissions to carry outsuch evaluations and to recommend changes every fiveyears. India, for example, has had 10 such reports over theyears and has made numerous changes in the distributionamount and formula as a result. South Africa’s Financialand Fiscal Commission may play a similar role.

A number of developing countries distribute transfersusing a formula intended both to equalize public expendi-tures in localities with different needs and capacities and tostimulate local fiscal efforts, although severe data problemsoften constrain the parameters used in such formulas. Braziland India, for example, allocate some transfers in accor-dance with per capita incomes in the different states, but fewother developing countries do so owing to data difficulties.In the absence of such data, simpler approaches—like thoseused in Colombia and Morocco—based on, for example,population and a simple categorization of localities (by size,type, and perhaps region) are more likely to prove useful inmeasuring general expenditure needs. Few developingcountries include explicit capacity measures in their for-mulas. India and Nigeria include a measure of tax effort inthe basic distributional formula to states, and Colombia hassuch an element in one of its transfer programs. Chile goesfurther and actually “taxes” richer localities to some extentby reducing their transfers and raising those granted topoorer localities. Of course, this approach makes sense onlyif local governments have the ability to vary local tax rates.The absence of much local autonomy with respect to localtaxes, combined with data difficulties, probably explains thefew examples of transfer programs incorporating explicitcapacity measures in developing countries.

In principle, a system requiring local governments tocover some portion of service costs out of their own fundsis desirable for accountability and efficiency reasons. Inaddition, given that different localities have quite differ-ent capacities to finance services, it may be appropriate torequire different degrees of local finance (matchingratios). An example of such a system is in Zambia, wherelocal governments receive a transfer that equals the dif-ference between the estimated cost of providing a speci-fied level of local services and the expected revenues to beraised locally by applying a standard set of local tax rates.The basic problem with this approach is that it is quitedemanding in terms of information.

The design of intergovernmental transfers is particular-ly important for the social sectors, where central govern-ments maintain a strong interest in certain expendituresand outcomes (as in public health and primary education).Rules and incentives set by the center through intergov-ernmental transfers are the key to ensuring that centralmandates are realized. In Vietnam, despite central effortsto target more funds for primary education and basichealth care to poorer provinces, the funds do not neces-sarily reach the poorer districts and communes or even theintended subsectors, since provinces are not provided withspecific instructions on how to allocate the funds.

Box 2: Experience with Intergovernmental Transfers

first. Assignment of expenditure functions can broad-ly follow the subsidiarity principal, which addressesinterjurisdictional spillovers; however, fine-tuning offunctional assignments within countries will lead todifferent outcomes. Once expenditure functions aredetermined, revenues should be assigned to differentlevels of government to ensure that services can befinanced and there are no unfunded mandates. Localrevenues will come from a combination of intergov-

ernmental transfers and local taxes. The transfersshould address both vertical imbalances and horizon-tal inequities, but assigning at least minimal tax instru-ments to local levels is an important part of account-ability. The combination of transfers and taxes shouldcover local recurrent expenditures. Subnational bor-rowing (under the conditions described in the follow-ing section) should serve as the last source of financefor the capital budget.


The impact of the intergovernmental fiscal system oneconomic or service delivery outcomes depends onhow that system is designed and on the institutionalsetting in which it is implemented. Institutions can bedefined as the rules of the game in society (or the incen-tives and constraints that influence human behavior)and the organizations and other means to enforcethem. Both the rules and the enforcement mechanismsinfluence the design of decentralization.

Institutions have had an increasingly importantinfluence on decentralization in recent years because oftechnological change, political demands for greater par-ticipation, and the emergence of global and local capi-tal markets, among other factors (Box 3). This sectionidentifies five institutional factors that shape the designof decentralization and influence its economic impact:the regulatory framework, the organization of servicedelivery, information systems and competition, thepotential for asymmetric decentralization, and the needfor policy synchronization. Taken together with the tra-ditional public finance perspective, these factors reflectan approach to decentralization that is based on thedesign of institutional incentives and rules of gover-nance as instruments for better economic management.

Adjusting the Regulatory Framework andDecentralizing Borrowing

There are many sound justifications for allowing sub-national governments to finance their own investments(for example, the intertemporal nature of large capitalinvestments). But experience has shown that main-taining a hard budget constraint for subnational gov-ernments is both essential and extremely difficult.Several approaches are possible:

• One common strategy is to restrain provincial andlocal borrowing by, for instance, limiting such bor-rowing to financing for capital expenditures, limit-ing debt service to a maximum percentage of cur-rent revenues, or requiring prior approval of centralgovernment for borrowing.

• A more fundamental approach is to remove the insti-tutional problems that give rise to unsustainable sub-national borrowing. This focus might include reas-signing revenues and expenditures to provide localgovernments with some own sources of revenue withwhich to finance local expenditures, revising thetransfer system, introducing transparent, timely, andreliable reporting systems, and establishing a stable,accepted periodic review process. These measuresare related to the fiscal system that binds differenttiers of government. Equally important would be toensure transparent reporting by and regulation of thefinancial sector—particularly as it relates to borrow-ing by subnational governments.International experience suggests an important objec-

tive in the design of an intergovernmental fiscal system:keeping the fiscal and financial systems separate (Box 4).Indeed, the lack of separation between fiscal and finan-cial systems—rather than the decentralization processitself—may have led to macroeconomic instability insome countries, as the web of implicit obligations sud-denly appeared as explicit budget commitments. Inaddition, government involvement in the financial sec-tor may dampen the economic role of capital markets inallocating credit and signaling creditworthiness. Thus itcould limit the financial sector’s ability to promote eco-nomic efficiency and local accountability.

In Argentina, for example, provincial governmentdeficits have been financed in part by provincial banks,many of which have gone bankrupt. As Wildasin

16

4. Designing Decentralization:Incorporating Institutions

Designing Decentralization: Incorporating Institutions 17

South Africa has embarked on a comprehensive decen-tralization of political, fiscal, and financial responsibili-ties to provinces and local governments. At the politicallevel, the apartheid system was formally ended with theelimination of racial jurisdictions, the holding of elec-tions on a nonracial basis for all tiers of governments, andthe writing of a new constitution. Legislation is beingenacted to implement a new fiscal system for subnation-al governments, to provide a framework for subnationalborrowing, and to provide institutional incentives for effi-cient service delivery. In addition, the central governmenthas initiated a multipronged strategy for capacity build-ing to enable successful decentralization. The overallframework does not, however, take a uniform view ofdecentralization. Guided by the demographic and eco-nomic concentration of the urban sector (which accountsfor 60 percent of the population and 80 percent ofGDP)—and perhaps by its political clout—decentraliza-tion is being implemented rapidly for local governmentsin urban areas, particularly in metropolitan centers, butwith greater caution in the provinces.

A major objective of the fiscal system in the post-apartheid era is to improve the distribution of incomeimposed by apartheid. The policies implied by this objec-tive involve the efficient assignment of expenditures, cen-tral financing of redistributive measures, and the devel-opment of a stable and predictable intergovernmentalgrant system. Given their redistributive and spilloverimpact, education and health have been designated asprovincial responsibilities but are financed from the cen-ter. The central government uses its “divisible pool” ofnational revenues to support transfers to poorer jurisdic-tions through two channels: one for provinces and onefor local governments. Both systems are formula driven.For local governments the formula is based on householdper capita income and for provinces on household percapita income augmented by a measure of the rural econ-omy in a region. As a result the transfers have shiftedresources into poorer jurisdictions. The system has alsoeliminated the ad hoc and inequitable fiscal transfers ofthe apartheid era, which shifted resources from central tosubnational tiers for the purpose of balancing end-of-the-year local deficits.

To complement the transfers, all local authorities havebeen given access to user charges, property taxes, andsome business taxes, and can set rates for user charges andproperty taxes. But provinces do not have their own taxinstruments, creating potential problems of accountabili-

ty and governance. In addition, the central government isreforming budget systems. A uniform accounting systemis being established for municipalities in hopes of settingup clear budget rules to link all tiers of government to anintegrated budget system.

On the financial side, a regulatory framework is beingdeveloped to enable local governments to access localcapital markets directly. The framework will have clearrules for public sector bankruptcy. This policy measurecomplements the political and fiscal decentralization bydecentralizing borrowing powers. Borrowing for long-term infrastructure development will be more appropri-ate in large metropolitan areas, while smaller urban gov-ernments and local rural governments will still requireintergovernmental transfers to finance capital invest-ment. A capital grants program from the center to thelocal level has been established to accommodate thisneed. But in the case of the provinces, access to capitalmarkets has been prohibited. This policy is consistentwith the policy of not granting provinces any significantown-taxes because it avoids creating implicit liabilitiesfor the central government from provincial access to cap-ital markets. To facilitate capital investment at theprovincial level, a program of capital grants is expectedto be implemented.

In addition, legislation is enabling local governmentsto restructure the institutional organization of municipaldelivery systems. In particular, municipalities are legallyable to experiment with partnerships with the private sec-tor (such as concessions) and nongovernmental organiza-tions (such as community delivery systems). Thus thedecentralization framework not only assigns responsibili-ties to local authorities but also extends it to the privatesector and community groups. The institutional focus alsocovers metropolitan areas, where a policy debate is underway about the appropriate structure of metropolitan gov-ernments. Similarly, a debate is under way about theappropriate structure of welfare and health systems, cur-rently the responsibility of provinces, and its implicationfor provincial governance.

The central government has focused on capacity-building measures to support the overall framework ofdecentralization. For example, rather than impose top-down, supply-driven systems of capacity measures, thecentral government provides grants to municipalities tobuy capacity for projects involving the private sector inthe delivery and financing of municipal services.Similarly, the central government (in partnership with

Box 3: Decentralization in South Africa: Redesigning Intergovernmental Fiscal Relationsand Creating Institutional Incentives


(1997) points out, the central bank’s policy of manag-ing these banks and absorbing their losses providedprovincial governments with a circuitous mechanismof inflationary finance. It also weakened incentives forfiscal discipline at the provincial level. In Brazil too,

evidence suggests that the de facto structure of inter-governmental fiscal relations includes the use of statebanks, and their relationship to the central bankthrough the financial regulatory system shifted implic-it liabilities for state deficits to the central bank. In

donors and the private sector) has launched manage-ment training programs aimed at municipal authorities.A similar program for provincial treasuries is being for-mulated.

Finally, the central government is investigating institu-tional settings for coordinating economic policies in adecentralized system. Currently, national policies (such asthose for standards and personnel) are set by central lineministries, but financing and implementation are in thehands of subnational governments. This divergence isvery acute in the social sectors (such as education andhealth) and, therefore, in the relationship between thecenter and the provinces. How to coordinate nationalpolicies with an emerging multitiered system of govern-

ment is an important design issue in the formulation andimplementation of intergovernmental systems.

Decentralization is far from complete, and it is far tooearly to judge the outcome of the measures taken. Butwhat is striking is the comprehensive vision of SouthAfrican policymakers in the design of the decentralizationframework. Political, fiscal, financial, and institutionalchanges are being managed simultaneously and in differ-ent ways for different jurisdictions. Few countries havetaken such a fundamental approach to reforming theirintergovernmental systems. Given its potential to offerother countries a host of lessons in the reform of inter-governmental systems, South Africa will be watched veryclosely by researchers and policymakers for years to come.

Separating fiscal and financial systems helps the growth ofimplicit liabilities for the central government and soreduces the possible macroeconomic consequences ofdecentralizing borrowing powers. It also raises severalissues on the nexus between the sequencing of policies forseparating fiscal and financial systems and the institu-tional development of capital markets.

In countries without a domestic capital market, it maybe preferable to have central fiscal transfers (from a cen-tral government borrowing from international sources)provide additional resources to subnational govern-ments, while focusing central government policy effortson developing a private commercial banking system. Thisapproach avoids the creation of development financeinstitutions, keeps fiscal and financial systems separate,and potentially offers a better starting point for a munic-ipal finance system. The reform of the financial sector inEastern Europe is a case in point. As the private bankingsystem emerges and establishes itself, long-term financecan be provided through discount facilities funded bycentral governments. Discount facilities take only thematurity risk by stretching the term of commercial banklending to municipalities. Commercial banks, however,retain the credit risk for their retail lending, thus pre-serving the separation of risks between the public andprivate sectors.

As capital markets emerge with financial instrumentsoffering long-term finance, the need for discount facilitieswill disappear. And as long as the fiscal system is welldesigned, subnational governments will be able to accesslonger maturities. For fiscally weaker municipalities (suchas rural local governments), however, transfers wouldremain an important vehicle for access to funds.

Finally, institutional changes in the organization ofdelivery systems provide mechanisms for accessing capitalmarkets to expand the delivery of services (such as waterand electricity) while offering mechanisms to separate fis-cal and financial systems. For example, even where capitalmarkets are weak, the creation of privately managed deliv-ery systems (such as concessions for water and electricity)may provide access to equity and debt from private inter-national markets. Similarly, smaller municipalities maypool their service delivery systems by contracting out to aregional utility and take advantage of economies of scaleon the production side. This option may alleviate the needfor the public sector to catalyze or manage a financial pool-ing instrument to lower the cost of funds for smaller localgovernment units (for example, the Michigan Bond Bankused such a financial pooling mechanism). What remainsimportant, however, is to develop a regulatory frameworkthat ensures these delivery systems are not inherited by thepublic sector in the case of bankruptcy.

Box 4: Separating Fiscal and Financial Systems and the Evolution of Capital Markets


commenting on the Brazilian story, Dillinger (1997)points out that the hard budget constraint imposed onthe fiscal side was circumvented because states have—wittingly or not—used the financial system to obtainfederal resources. Similarly, in China the weak revenuebase of the center has created political pressures on thePeople’s Bank of China to offer credit to lower-levelgovernments. As Lall and Hofman (1995) suggest,such policy lending can have a negative impact onmacroeconomic price stability and, presumably, fiscaldiscipline at the subnational level.

These examples suggest various mechanisms forseparating fiscal and financial systems. An importantissue is, where possible, getting the public sector outof the financial system. This includes privatizing thebanking system. But by itself privatization is not suffi-cient to enforce the separation. As the U.S. savings andloan crisis and elements of the recent Asian financialcrisis suggest, private banks must be supervised with aclear prudential, regulatory framework that monitorsthe level and nature of their aggregate liabilities.

The central government also must clarify the fiscalrules of the game and follow them. For example, itshould first clearly assign expenditure responsibilitiesto each level of government. After these responsibilitieshave been assigned, tax instruments should be assignedto subnational governments to increase their access toown-revenues, and arrangements should be fixed forthree to five years to improve predictability. Similarly,ensuring the stability and predictability of the fiscaltransfer system and implementing the legal frameworkfor pledging revenues as debt payments are also essen-tial. Transfers and own-revenues provide the “collater-al” necessary for access to capital markets. In fact, thelack of access to financial markets by subnational gov-ernments often results in central government interven-tion in financial markets when the real issue is design-ing and implementing an appropriate fiscal framework.

On the other hand, if financial intermediation isundertaken by the public sector, as in the case of devel-opment finance institutions or municipal developmentfunds, the separation of fiscal and financial mechanismsmay need to be enforced in other ways. For example, itwould be important to ensure that the capital borrow-ing of public entities responsible for onlending to sub-national governments is an explicit line item on the cen-

tral government’s budget and a formal part of the cen-tral government’s liabilities. This measure will eliminatea possible source of implicit liabilities and may providea check through the budget system on the spendingbehavior of development finance institutions andmunicipal development funds.9 The credit allocationmechanism must be made transparent and binding toreduce ad hoc and politically influenced allocation offinance. In addition, the municipal development fundmust not provide subsidized funding since this will hin-der the development of local capital markets—whichultimately will be the sustainable, sound source of localinvestment finance (Peterson, 1997). Finally, a credibleexit strategy for the institution would be needed, espe-cially as local capital markets develop.

Macroeconomic crises linked to subnational financehave been caused more by a lack of institutional sepa-ration between fiscal and financial systems than bydecentralization per se. In fact, decentralization mayeven provide incentives for a central government toinstitutionalize the separation of these systems. Whensubnational governments have direct access to capitalmarkets, the risk of inheriting liabilities may provide animpetus for the central government to monitor bor-rowing, making the interaction between subnationalgovernments and capital markets—whether throughpublic or private intermediation—more transparent.

In addition, it may be useful to develop a mecha-nism by which any emergency central support requiredto work out such debt problems will carry with it theobligation to introduce and make effective all thesereforms under the supervision of a central review board(World Bank, 1996a). To avoid any moral hazardbehavior from this implicit insurance scheme, howev-er, it would be necessary to specify through legislationthe bankruptcy procedures.

Still, at least two basic guidelines for subnationalborrowing seem to be needed. The first is a limit onborrowing solely for investment (which may, of course,be difficult to enforce in the absence of strictly segre-gated capital budgets). The second is a requirement forexplicit national approval for foreign borrowing. At adeeper level, given the heavy dependence on centraltransfers of subnational governments in most develop-ing and transition economies, stronger restrictions onborrowing may be warranted—particularly if some


subnational units are large enough to affect nationalmacroeconomic variables (Wildasin, 1997). Unlesssubnational governments are able to save themselvesfrom fiscal crises by drawing on their taxing powers,their only options are bankruptcy or bailouts. In somecases the only way to reduce the moral hazard implic-it in this situation may be by imposing stricter limitson subnational borrowing than would be needed ifcapital markets functioned well and the central gov-ernment’s nonguarantee of local debt was credible.

Finally, it should be noted that there is almost no casefor central government lending for local investment ininfrastructure—other than the possible argument thatsuch lending may help develop a private capital marketto finance such investment. Central finance may beappropriate if the central government has an interest inlocal infrastructure, but it should be provided through,for example, competitive grants from a social investmentfund (Glaessner and others, 1994) rather than throughsubsidized credit. As experience with municipal devel-opment funds has shown, subsidized financing agenciesalmost inevitably become contaminated by political con-cerns and end up blocking rather than aiding the devel-opment of private capital markets. Thus they weakenrather than strengthen the potential regulatory role ofsuch markets with respect to subnational borrowing.

Organizing Service Delivery

The broader institutional approach to decentralizationhas several important implications. For example, to theextent that decentralization is intended to improve thedelivery of services, it is essential to consider in detailnot only the nature of each service but also the struc-ture of delivery. Not only does infrastructure invest-ment raise quite different questions than do social ser-vices, there are also important differences within eachcategory. As the World Bank’s World Development Report1994 shows, for example, roads, power, telecommuni-cations, and water and sewerage each have specificrequirements with respect to the most appropriateinstitutional structure for efficient and equitable ser-vice delivery (see also Kessides, 1993). Similarly, edu-cation, health, and social assistance—or, more broad-ly, the “social safety net”—may be very different with

respect to such critical matters as the justification for,and appropriate level of, public subsidy and the role ofdifferent tiers of the public sector in financing, deliv-ering, or regulating a service. Economic support ser-vices such as those involved in rural development mayraise still other problems.

In this connection, it is useful to distinguishbetween public delivery of services (production) andpublic financing of services (provision). It is alsoimportant to distinguish between production efficien-cy (that is, the cost of delivering a given quantity andquality of service) and allocative efficiency (that is, theextent to which public expenditures reflect localdemand). In principle, there is a wide—almostunbounded—variety of forms in which particular pub-lic services may be produced and provided, and thereis no necessary link between the two (Box 5).

Schools, for example, may be run by a local educa-tion authority, a local government, a nonprofit body, ora private company. This is the production side, to whichissues of production efficiency (such as economies ofscale and scope) relate. In some instances—such asschools—a case can be made that public production isa more efficient way of obtaining a desired output (say,uniform primary education) than is regulation andmonitoring of contracted production. In other casespublic production may be warranted to achieveeconomies of scale—as in the technical design of super-highways in countries that do not afford sufficient scopefor the development of competitive private producers(and that do not want foreigners to enter the market).It is important not to confuse the question of whoshould deliver a service with the question of how muchshould be provided and who should pay for it. Themore fundamental question from an allocative point ofview is who pays, not who delivers. And, as noted, thetwo can be quite different. It is also important to deter-mine who decides who delivers services. It is still anopen question whether local governments should beallowed to choose to deliver services or contract out.Where procedures are transparent and basic rules ofprocurement are in place, this may be a good idea. Butwithout these preconditions, it may not.

In one sense privatization is the ultimate form ofdecentralization. In recent years a wide variety of coun-tries have seen the private sector play an increasingly


large role in the delivery and to some extent the financ-ing of many public services, depending on the extentof externalities involved. New forms of service deliveryincorporating private, public-private, and horizontaland vertical combinations of public and private actorsalready exist, and more are being created all the time.Private delivery of services is important in developingcountries because it loosens the traditional monopolyof public service providers, which often provide ineffi-cient, poor-quality services.

Market solutions to service delivery offer manyadvantages in terms of efficiency and may even be equi-table (for example, by enabling public subsidies to betargeted to the poor, or by reducing the travel costs forthe poor to obtain quality services). Most public ser-vices, however, embody some degree of “publicness,”some relevant externalities, some broader political orsocial purpose (often redistribution), or some combina-tion of these features. Such services cannot be fully pri-

vatized or fully funded through user charges withoutbreaching either efficiency or equity concerns. Althoughregulation of private services can help deal with someissues—for example, content of educational curricu-lums—it cannot deal with others—for example, equalaccess for the poor—without continuing an explicit orimplicit subsidy. Consequently, although privatizationmay often be one way to deal with many problems ofpublic service delivery, in other ways it may make thetask of the institutional policy designer even more com-plex, since the number of agents has expanded and thedegree of control over agents has diminished.

Establishing Information Systems andCompetitive Governments

One of the theoretical arguments supporting decen-tralization is that local governments have better

Services can be produced (delivered) or provided(financed) by many different public institutional struc-tures, including: • Central government (department, decentralized

agency, or enterprise)• Regional government (department, decentralized

agency, or enterprise)• Local government (department, decentralized agency,

or enterprise)• Central-regional arrangement• Central-local arrangement• Central-regional-local arrangement• Regional-local arrangement• Association of local governments• Special-purpose local authority (encompassing more

than one local government, coterminous with a localgovernment, or covering less area than a local govern-ment)

In each of these cases the service may be produced byone structure (such as a regional enterprise) and provid-ed by another (such as a specific local governmentdepartment).

Services can also be provided in whole or part by thepublic sector but produced by the private sector. Again,there are many possible forms:

• Build-operate-transfer (BOT) or build-own-operate(BOO) arrangements

• Other forms of public-private partnerships• Development charges, exactions, and similar schemes• Franchise arrangements• Service contracts

Finally, public services can also be both produced andprovided by private agents, sometimes in response tocoercive legal requirements and sometimes even volun-tarily:• Compulsory provision by developers• Compulsory provision by individuals (through vouch-

ers or self-financed)• Voluntary provision (through formal or informal

arrangements)• Provision by nongovernmental organizations (church-

es, enterprises) Even this extensive list is incomplete. Many of the

arrangements listed have a number of possible variants,and there are ways to combine all these organizationalstructures. Moreover, different structures might apply forpolicymaking, regulation, financing, production, and soon. Finally, different structures may apply for different ser-vices and for local governments with different character-istics (size, financial capacity, and so on).

Box 5: Alternative Ways to Produce or Provide Local Public Services

Source: Bird, 1995.


information about local preferences; thus decentral-ization should enhance allocative efficiency.Information asymmetries make it difficult for centralgovernments to monitor local government agents.The decentralization literature developing this argu-ment has implications for the design of intergovern-mental grants and the earmarking of transfers andrevenues (Cremer, Estache, and Seabright, 1995;Boadway, Marceau, and Sato, 1997). But no seriousresearch has been done in developing or transitioneconomies to test such basic propositions as the pre-sumed greater knowledge of local preferences by localrather than central governments. There is, however,anecdotal evidence that decentralization increases theavailability of local information (Crook and Manor,1994). Recent work on Colombia is particularly inter-esting in this respect (World Bank, 1995b).Nonetheless, it is important to keep informationalconstraints in mind when analyzing and prescribingdecentralization strategies.

However the decentralization literature is inter-preted, it seems clear that the efficiency of any com-petition engendered by decentralization will largelydepend on the rules of the game that are set andenforced by the central government or perhaps by anindependent judiciary. As Inman and Rubinfeld(1996) show, electoral competition alone is unlikely toproduce efficiency—although as Breton (1996)shows, government competition can take many otherrelevant forms in addition to counting votes. Very lit-tle research has been done—even in industrial coun-tries—on the effect alternative institutional formshave on outcomes. This issue is particularly importantin developing countries, where institutional settingsvary dramatically and can strongly influence the out-comes of decentralization. Thus developing a betterunderstanding of this issue should be a top priority forthe World Bank.10

One conclusion that can be emphasized, however,is that the key to usefully competitive governments isto make relevant decisionmakers accountable for theirdecisions, and that the key to effective accountabilityis to make relevant comparative information publiclyavailable.11 Accountability and public information areeasier to achieve in industrial countries than in devel-oping countries. At its base, the ultimate mechanism

for competition is, on one hand, the ability of citizensto compare governments in terms of the services theyprovide and the taxes they levy (“exit”) and, on theother, the ability of citizens to affect and alter the deci-sions of government (“voice”). It is important to askhow relevant this mechanism is for developing coun-tries, where poor information, weak democracies andtraditions of participatory decisionmaking, and oftenweak rural labor markets constrain both popular par-ticipation and interjurisdictional mobility (voice andexit). Moreover, poor information and weak capacitymake hierarchical accountability particularly difficultin developing countries.

One key lesson relevant everywhere is thus that themore that is known, and the more publicly it isknown, the better the outcome of decentralizationefforts is likely to be, whatever their rationale andwhatever the circumstances in which they occur.12

Because the feasible and desirable implementation ofthis generalization may vary widely from place toplace and for different services, World Bank contribu-tions to supplying and supporting information rele-vant to understanding and evaluating the impacts ofdecentralization can play a crucial role in improvingoutcomes. As with the information-theoretic ap-proach to decentralization, the competitive approachsignals both the importance of improving knowl-edge—and public awareness of knowledge—in orderto improve outputs and the importance of consideringin detail the institutional structure of particular decen-tralization arrangements. It is not enough to discuss,for example, tax assignment or the case for or againstequalization transfers. Close attention must also bepaid to the details of how such fiscal institutions actu-ally work in the context of particular countries inorder to understand and analyze the incentives facingvarious decisionmakers. Indeed, assigning revenuesources to localities that lack channels for public par-ticipation may be counterproductive.

An unfortunate side effect of decentralization insome countries has been the virtual disappearancefrom the central government’s cognitive horizon of reli-able information on the provision of such services. Therole of the central government in ensuring and moni-toring effective and efficient decentralization is espe-cially critical when the main goal is to enhance service


delivery, particularly for services (such as health andeducation) that are important not only for nationaldevelopment but also for poverty alleviation and gen-eral welfare. Decentralization does not mean that thecentral government no longer has any responsibility inthese areas. What it means is that the nature of thisresponsibility has changed from delivering servicesdirectly to regulating and monitoring the efficiency andequity of services delivered by others—usually localgovernments. The essential tool for this task is an accu-rate information system—generated, for example, byrequiring local governments to file uniform and infor-mative budgets and financial reports. Uniform infor-mation is important so that comparisons can be madeacross communities and yardstick competition used.Fiscal transparency is fundamental to sound decen-tralization policy and, indeed, public policy in gener-al. All government accounts should be comprehensive,comprehensible, and widely available. Budgets shouldbe drawn up to display the real status of publicfinances, ideally including the long-run implicationsfor taxation of current spending decisions. This acuteneed for reliable information has been neglected inmost decentralizing countries (Box 6).

Some of the needed incentive for compliance withinformation demands can be created for local gov-ernments by making timely submission of suchreports a condition of receiving fiscal transfers. Thisrequirement may imply significant setup costs for sys-tem design, training, and implementation. To spendmoney wisely in a decentralized system may thusrequire some initial investment. While such concernsare less important when full responsibility for the effi-ciency and equity of services are devolved (as in a for-mal federal structure) rather than simply delegated tosubnational governments, the need for an adequateinformation system remains critical to macroeco-nomic management even in this instance, given theimportance of such services as education and healthin total public expenditure. Regardless of the form ofdecentralization, an important institutional problemis thus how to ensure that the relevant central agen-cies have adequate incentives to monitor subnationalactivity. And, as noted earlier, such information iseven more necessary when democratic institutionsare important.

Adopting Asymmetric Decentralization

Economic, demographic, and social diversity is oftenreflected in a multitude of government and deliverystructures even within a single country. For example,South Africa has provincial and regional governments,rural districts, one and two-tiered metropolitan gov-ernments and smaller urban municipalities, and spe-cial service delivery units such as water boards andhealth districts. Given such diversity, both in thenature of political jurisdiction and the characteristicsof households, “one size fits all” is definitely not truefor decentralization. Different instruments may havevery different effects in different circumstances, andvery different approaches may be needed to achievesimilar (or acceptable) results. For example, privatiz-ing water services may achieve efficiency and equityobjectives in a dense, urban setting, but it may fail toreach similar goals in a sparsely populated ruralregion. Private sector delivery and financing of waterservices may then have to be complemented with pub-lic sector and community delivery systems for specif-ic areas. To accommodate the need for such diverseapproaches, asymmetrical central policies—treatingdifferent units differently—may be required to achievesimilar outcomes.

An important element of such an approach is theprinciple of asymmetrical decentralization. For exam-ple, in many countries it may be feasible to decen-tralize political, economic, and administrative re-sponsibilities to large urban areas. Similarly, at theregional level, fiscal and administrative capacity maymake it easier to decentralize responsibilities only tosome provinces or states. In other cases it may be fea-sible to decentralize responsibilities directly fromcentral government to the private sector rather thanto local governments.

But asymmetrical decentralization may raise a fun-damental political problem: the perceived need tohave a law that treats all units similarly, in the face ofthe reality that there are wide and relevant differencesbetween them. In South Africa, for example, there isgrowing political debate about the capacity of manynewly formed provinces—especially those that haveinherited significant administrative problems from theapartheid system—to manage the delivery of social

services such as health and education. An option maybe to decentralize responsibilities only to thoseprovinces that show the capacity to manage such ser-vices. But this option may be politically difficult toimplement because it will require some elected repre-sentatives to agree to receive less autonomy than theircounterparts in other provinces. In such cases, anddepending on the country, solutions may be soughtthrough such means as making the application of aparticular decentralization provision conditional onthe satisfaction of a number of preconditions, on thevoluntary assumption of certain financing or otherobligations, or on the signing of a contract that per-mits individuation of both the timing and extent ofdecentralization. Such legislation assures the ultimate

decentralization of political and economic powers anddoes not leave it to the discretion of the center but,rather, to the actions of elected representatives of sub-national governments.

Synchronizing Policy

As noted earlier, a difficult but important design issueis how to match fiscal, political, and administrativearrangements to achieve the potential benefits of decen-tralization. Most countries do not get the mix right—they devolve decisionmaking to local levels withoutproviding budgets to make meaningful decisions, orthey decentralize finance without ensuring adequate


Good data on local finance are hard to find in developingcountries. In some countries careful examination ofnational data sources may disclose some relevant infor-mation, but data are often old, incomplete, budgetedrather than actual, aggregated (say, by region), or in someother way not easily usable for analytical purposes.Moreover, information is often hard to interpret. Fewdeveloping countries compile local fiscal data in a waythat is comparable over time, let alone across countries.What information exists is often the result of special andoccasional studies; the main international example is themassive work of Bahl and Linn (1992). A common reac-tion to this work by those with little experience in the fieldis, “Why do so many of the numbers refer to the early1980s or even the 1970s?” But those who have carried outsuch work can only applaud the massive data collectioneffort reported in this volume. No similar effort seems tohave been carried out anywhere since then.

Local finance data are poor for a number of reasons.First, few countries attach much importance to its sys-tematic collection. Second, collecting such data is difficultin countries with low administrative capacity and poorcommunications systems. Third, local government struc-tures are often complex and differentiated, which makesit difficult to design adequate data collection proceduresand to ensure that the data gathered is reliable. Fourth,local government finance is complicated: tracing fiscalflows among and within the array of local governancebodies found in most countries is not easy. Fifth, even ifinformation is obtained, how should it be interpreted?

Incidence and other problems of economic analysis arecritical in this connection, but surprisingly little empiri-cal progress has been made toward resolving such prob-lems even in the most developed, and statistically sophis-ticated, countries. Finally, and in the end perhaps mostimportant, local finance data are poor because no one isreally asking for improvement: other things seem muchmore important, and resources are scarce, so this omis-sion may seem understandable.

Such neglect is often a mistake, however. Local gov-ernments are among the most important service deliver-ers in most countries; unless they do better, the public sec-tor is not likely to do much better, and how can theirperformance be assessed in the absence of information?Moreover, governments everywhere are decentralizing;again, how can decentralization policies be properlydesigned, or adequately implemented, in the absence ofthe information needed to assess them? Finally, in theabsence of good information, policies and assessments aretoo often based on what central officials or visiting expertshappen to know, or think they know, about what is goingon in local governments. Given the variability of local real-ity in most countries, such biased judgments may be verymisleading. It is difficult to read accounts of local financein China, for example, without being reminded of the taleof the seven blind men describing an elephant, each inap-propriately generalizing from the bit they happen to know.All in all, paying more attention to getting the facts straightabout local (or, more broadly, subnational) finance wouldseem to be a worthwhile investment in most countries.

Box 6: Why Do We Know So Little about Local Government Finance?

Source: Bird, 1995.


local accountability through political decentralization.Even within the fiscal realm, coordination is extremelydifficult. Countries sometimes decentralize expenditureresponsibility without providing adequate revenues (asin Russia and other countries of the former SovietUnion), or decentralize revenue but mandate virtuallyall expenditures (as in Ecuador). Or the central gov-ernment devolves a service fiscally and politically, yetdetermines all wage and employment policies, or issuesthe locality with a long list of service delivery mandatesthat provides very little flexibility in local decisionmak-ing. Decentralization has been likened to a souffléwhere all ingredients must be present in the rightamounts and prepared in the right way to achieve suc-cess (Parker, 1995). Moreover, like a soufflé, the bestmethod of preparation will depend on the environmentand the best mix of ingredients is a matter of taste.

Policy synchronization is difficult not only becauseof the many fiscal, political, and administrative issuesthat require careful consideration, but also because eachservice and even each function within a service will dif-fer with regard to the appropriate form of decentraliza-tion. Depending on the nature of the service (external-ities and interjurisdictional spillovers), the politicallandscape, and possibly the administrative capacity, theamount of autonomy given to a local government willdiffer (Bossert, forthcoming).13 For example, certainpublic health programs (family planning, immuniza-tions) are of national concern, and thus local govern-ments may be given some autonomy for implementa-

tion but must be responsible to the central governmentfor achieving nationally specified outcomes. The whole-sale decentralization of the health sector led to the fail-ure of one state to undertake a family planning programdeemed important by the central government in thePhilippines, and to a drop in vaccination coverage inPapua New Guinea (Kolehmainen-Aitken andNewbrander, 1997).

The issue of providing flexibility to local levels indetermining wages and job requirements—decentraliz-ing the design of labor market conditions in the publicsector—provides an important example of the need forpolicy synchronization in the design of decentralization.In the Philippines, for example, decentralization of thehealth system resulted in health care workers doing thesame job being paid at three different rates based on theirprevious employment (central, provincial, or local). Thisgave rise to morale and budgetary problems. Inadequateattention to the financing of pension obligations whentransferring employees from one level of government toanother has complicated decentralization efforts inColombia and many other countries. The failure to givelocal governments any leeway in staff complements orsalaries largely vitiated initial decentralization efforts ina number of Eastern European countries (such asRomania). Such issues are especially important withrespect to large, labor-intensive activities such as healthand education, where matters are further complicatedby externalities (for example, with respect to publichealth) and basic welfare (accessibility) concerns.

To debate whether decentralization is good or bad isunproductive and misleading since the impact ofdecentralization depends on design. Yet it is essentialto consider the wide range of issues that influencedecentralization. The traditional fiscal federalismapproach is a starting point, but the need for a strongerfocus on institutions—both the rules that influence thebehavior of actors at different levels of government, inthe private sector and in civil society, and the organi-zations that implement those rules—is increasinglyevident in the World Bank’s operational work. Thisbroader agenda has led to an enhanced focus onaccountability and capacity, has strong implications forthe Bank’s project design and policy dialogue, and callsfor a reinvigorated research effort focused on develop-ing countries.

Incentives for Accountability

Decentralization is dispersing fiscal, political, andadministrative responsibilities across different tiers ofgovernment and between the public and private sec-tors. Responsibility for delivering services, for exam-ple, may lie with all or some tiers of government, withcommunity groups, or with the private sector. Thechallenge is to design decentralization so that it createsincentives that hold each entity accountable for itsresponsibilities and makes explicit the institutionalrelations between each entity.

How can decentralization provide such incentives? First, in democratic settings political decentraliza-

tion and elections provide direct political accountabil-ity. More broadly, as Breton (1996) demonstrates, in allfunctioning political structures there are a variety ofmeans by which indirect political accountability is

attained. Many developing countries, though, haveweak representative decisionmaking processes andlocal elites are often deeply entrenched. Local elitesoften take leadership roles, and although not necessar-ily bad, this can result in the hijacking of resourcesunless transparency and accountability are somehowenforced (Narayan, 1998). One way to improve localparticipation and accountability is through transparentbudgeting processes and public procurement proce-dures. Experience in Brazil and Mexico has shown thatparticipatory budgeting (that is, where all citizens areinvited to a public meeting to discuss budgetary prior-ities) can provide a critical link between communitiesand government.14 Governments should actively seekcommunity participation and use performance-basedbudgeting so that their constituents know not onlywhat inputs were used but what outcomes wereachieved. Similarly, publicizing procurement bids,proposals, and selection enhances community aware-ness and ultimately accountability. Stressing theimportance of community participation in local deci-sionmaking should be an important component of thedialogue on decentralization.

Second, by diffusing responsibilities across differententities, decentralization provides a basis for compari-son and competition (even if indirect). In developingcountries, where interjurisdictional mobility may beconstrained, competition among service providers ofgoods that can be privately consumed within a jurisdic-tion increases options for residents.15 (Decentralizationand private sector development have an interdependentrelationship whereby the former enables the latter, andthe latter strengthens the former.16)

Third, in distributing fiscal instruments to all levels ofgovernment, with the right to set rates, decentralizationcreates incentives for fiscal accountability. Being forced

26

5. Rethinking Decentralization in Developing Countries

to tax one’s constituency—either to deliver additionalservices or to pay for policy mistakes—is an importantrestraint on political decisionmaking. But in developingcountries, where it may not be easy for people to moveto the jurisdiction where the tax and service bundlematches their preferences or to vote the incumbent deci-sionmaker out of power, maximizing accountabilitymeans selecting tax instruments that closely match taxesand services. In this respect, user charges are particular-ly important for accountability because they create aclose link between the delivery agent and the client.17

Finally, access to well-functioning markets by gov-ernments and households may provide additionalcheck on politicians and administrators. For example,decentralizing borrowing powers offers a mechanismfor using capital markets to provide accountability bysending market signals on the performance of govern-ments and private firms. Well-functioning land mar-kets make “voting with one’s feet” a credible threatbecause policy decisions get capitalized in land values.In both cases information helps ensure that marketsfunction efficiently. Strengthening the regulatoryframework to improve the functioning of markets isessential for establishing the self-correcting methods ofaccountability that will lead to successful and sustain-able decentralized decisionmaking.

Properly designed and sequenced decentralizationmay deliver each of these forms of accountability. Theframework of incentives and accountability is especial-ly important given the uncertainty that generally per-vades the decentralization process. We may not be ableto say exactly what the “correct” form of decentraliza-tion is for a particular country, but we do know that cor-rect institutional incentives are essential both to revealmistakes and to provide a self-regulatory mechanism.

The Role of Capacity

Increasing fiscal, political, and administrative respon-sibilities and setting clear rules holding each tier of gov-ernment accountable for those responsibilities lead toan important set of questions: Does the public sector—that is, central and local governments—have thecapacity to support and manage decentralized respon-sibilities? Does it have the political capacity to identify

and respond to individual preferences? Does the cen-tral government have the administrative capacity toprovide technical and financial support where appro-priate? Do local governments have the capacity todeliver promised services at low cost? Political capaci-ty has to do with the channels for local participationand administrative capacity with the processes re-quired of central and local administrators and withtheir training. The evidence on these issues is sparse.

Local administrative capacity is sometimes consid-ered inadequate because bureaucratic requirementsimposed by the center are inappropriate for local deci-sionmakers. In reality, if the appropriate requirementswere assigned to each level of government according tothe information required for them to perform theirfunctions, local capacity would probably not be asproblematic. In other cases local administrative capac-ity is identified as a problem by the central governmentwhen in fact the central government may also lack thecapacity to manage local affairs. In yet other cases thedesign of intergovernmental fiscal relations does notprovide guidelines, resources, and incentives thatwould lead to strong local capacity. All these factorsmay be in play at the same time. For example, in SouthAfrica limited administrative capacity at the provinciallevel has constrained the decentralization of educationand welfare payments. There has been less debate onwhether the central government has the capacity tomanage the delivery process or whether the failure isdriven by the design of the intergovernmental system(Ahmad, 1995). Local capacity is a complicated issue,and the appropriate way to improve it may not simplybe through increased training of local officials.

Colombia provides an example of how local gov-ernments have relied on demand-driven processes forboth infrastructure investment and capacity building.This approach has increased allocative efficiency(matching investments to local preferences) and direct-ed resources to the poor (Box 7; World Bank, 1996b).

The experience with developing capacity for localdecisionmaking and administration is encouraging forbelievers in the potential allocative and democraticvirtues of decentralization. As in such well-knownAsian cases as the Orangi project in Karachi, Pakistan(Bird, 1995), local participation appears both to revealpreferences and to keep costs down. Depending on the

Rethinking Decentralization in Developing Countries 27

nature of the project, community involvement mayalso enhance targeting of the poor, since participantswho self-select communal work projects are willing tovolunteer their labor without remuneration.

In assessing the link between decentralization andcapacity, therefore, several issues and competinghypotheses need to be raised. The initial impressionthat capacity constraints may inhibit decentralizationseems to be an offshoot of a paradigm that looks atcapacity building in a top-down, supply-driven frame-work. The policy implication of this view is that capac-ity building should precede decentralization. A com-peting hypothesis suggests a more dynamic anddemand-driven relationship between decentralizationand capacity building: shifting responsibilities tolower-tier governments may provide the incentive forpublic officials to invest in capacity building or seekcreative ways to tap into existing sources of capacity.The latter could include not only outsourcing to the

private sector and nongovernmental organizations butintergovernmental contracts and agreements as well.

In addition, the discussion of capacity constraintsmay incorrectly assume that all subnational govern-ments are similar. In reality, urban governments (par-ticularly in large cities) have more capacity to manageand finance service delivery than do rural govern-ments. Similarly, different regional governments—provinces and states—have different fiscal and man-agement capacity. In this context the principle ofasymmetric decentralization offers an approach todecentralizing responsibilities where feasible ratherthan relying on an all or nothing approach.

Policy Dialogue and Project Design

The institutional perspective makes clear the impor-tance of understanding the concrete and specific cir-


Under Colombia’s “coparticipation” system, local com-munities provide labor and local materials for new pro-jects, and municipal governments contribute a portionof the cost. This approach not only fosters communityinvolvement in identifying needs and choosing projects,it also promotes community participation in the execu-tion, operation, and maintenance of the works.Municipalities have to prepare projects that are thenappraised using technical and environmental criteria.Project beneficiaries should be low-income rural fami-lies, and projects can be carried out by any contractor(private contractors, nongovernmental organizations,state agencies, universities) that competes to supply theworks and services.

Preliminary evidence is surprisingly encouraging onlocal capacity to carry out such functions. A recent studyof 16 municipalities found that decentralization hasenhanced local capacity in terms of labor, capital, andtechnology (World Bank, 1995b). Colombian municipal-ities are, for example, increasing the skills of local bureau-cracies through such means as competitive hiring, sharingthe services of professionals among municipalities, train-ing municipal employees, and rotating personnel acrossdifferent departments in the same municipality. Capitalspending capacity has also increased. One municipality

has totally privatized road maintenance; another has putprivate developers in charge of constructing urban roads.In other localities computers have been introduced tomonitor water and sanitation services. In addition, munic-ipalities have started to share equipment and they haveimproved their technological capability for internal orga-nization, planning, and monitoring to ensure better pro-ject management.

Underlying these improvements is a more basicchange: Colombian municipalities have been movingfrom a supply-driven (top down) to a demand-driven(bottom up) approach to public services. Increasingly,reflecting the new liveliness of local politics and (with sub-stantial variations from area to area) more extensive com-munity participation, people are getting what they wantrather than what someone in the capital thinks theyshould want. Emphasis has been put on roads, education,and water works; those are the needs that people perceive,and those are the needs that newly empowered andresponsive local governments are attempting to satisfy.Opinion surveys suggest that the resulting sectoral alloca-tion of resources is consistent with community prefer-ences, with most respondents indicating that they trustthe local government more than the national governmentto deliver goods and services.

Box 7: Developing Capacity in Colombia

Source: World Bank, 1995b.

cumstances of the country concerned. What can be donein terms of decentralization largely depends on where,when, and how it is done. Initial conditions determinethe level of trust, the reputation of the various actors, theexistence (and rigidity) of constraints to institutionalchange, and so on. The interests of the main stakehold-ers in the existing system must be understood to ascer-tain the existing or potential political support for changeand the resources available, and to determine the beststrategy (objectives) and tactics (timing, sequence, dura-tion) for change. Decentralization is political, and ade-quate political and institutional analysis is needed toundertake it successfully and appraise its success—or toassist or intervene in the process effectively.

As noted, decentralization has strong implicationsfor resource mobilization and use. Because these issuesaffect all of the World Bank’s main policy concerns—growth, macroeconomic stability, service delivery, and,most important, poverty reduction—the Bank mustengage countries in a dialogue on the design of decen-tralization. The potential benefits of decentralizationcan only be achieved—and the potential pitfalls canonly be avoided—if policy design focuses on creatingthe appropriate institutional arrangements in whichdecentralization can occur.

Even if decentralization efforts are far from optimallydesigned, the Bank may still be able to improve theirallocative and distributive outcomes—for example, byenhancing local capacity and working directly with localofficials (as in Colombia) or by using adjustment lend-ing to support a productive decentralization agenda (asin the Kyrgyz Republic). Useful dialogues on decentral-ization have been held in countries as diverse asVietnam—where focusing on rural service deliveryrather than central-provincial relations proved helpful—and South Africa—where a low-key but potentially highpayoff dialogue on decentralization has been sustainedthrough a pivotal period in the country’s history.

Tools of public policy exist in all countries to bringabout some desired consequences of decentralization.Yet institutions differ dramatically between industrialcountries and developing countries, and these differ-ences have implications for the extent and type ofdecentralization that may be appropriate. In develop-ing countries weak democracies, factor markets, andregulatory frameworks affect the accountability

imposed on local decisionmakers by citizens (“voice”),by competition (“exit”), and by the public sector (“hier-archy”; World Bank, 1997). Governments are general-ly not as responsive, transparent, or predictable as theyare in industrial countries. Thus the same policiesimplemented in an industrial country may have verydifferent consequences in a developing country (andthis outcome will vary between developing countriesdepending on their institutions).

Since the outcome of decentralization depends heav-ily on the institutional arrangements with which decen-tralization policy interacts, successful decentralization(that is, a process that achieves efficiency objectiveswithout jeopardizing equity or macroeconomic stabili-ty) requires a careful examination of each country’sinstitutions. A framework for assessing institutionalweaknesses and factoring them into the design ofdecentralization policy is shown in Table 2. This list ismeant only as a starting point. We must identify whichinstitutional structures are most important to achievethe benefits of decentralization and avoid the pitfalls,encourage their development, and compensate forinstitutional weaknesses (at least in the short run) whenadvising decentralizing governments on policy design.In the longer run, institutional strengthening is essen-tial. But given that decentralization is happening now,and institution building takes time, our advice shouldtake these institutional weaknesses into account.

In countries that seem to lack most forms ofaccountability, aspirations for decentralization shouldbe modest. In these cases decentralization can do moreharm than good, and the best technical advice the Bankcan provide is how to go forward with some benignforms of decentralization (such as local trash collec-tion) while encouraging the institutional developmentnecessary for further advances.

More Case Studies, Data, and Research

Because decentralization is an inherently country-spe-cific (and sometimes activity-specific) process, consid-erable contextual information is required to analyzeand assess, let alone assist. Detailed, information-intensive case studies are the essential building blocksto knowledge in this field.


Detailed data should be collected from country casestudies and a cross-country dataset developed to pro-vide a context against which to assess country experi-ences. Only those who have tried to do cross-countrycomparisons of decentralization can appreciate howbad, inadequate, and incomparable is the existing data-base on almost anything one might wish to examine. Itis not much of an exaggeration to say that one canprove, or disprove, almost any proposition aboutdecentralization by throwing together some set of casesor data. More comparable and meaningful cross-coun-try data are needed on various measures of decentral-ization (Huther and Shah, 1998), effects of alternativetransfer designs (Ahmad, 1996), and so on. Though inrecent years much work has been done on this issue byacademics and international agencies, much moreremains to be done. The World Bank’s DecentralizationThematic Group has undertaken a project to mergeseveral cross-country data collection efforts from dif-ferent parts of the Bank and from several external part-ners in order to form a comprehensive database thatincludes fiscal and institutional country variables. Aneffort will be made to regularize the collection of sub-national data in all client countries.

Given the importance of accurate, timely, and reli-able information with respect to what is going on andits effects on various policy outputs, the Bank shouldconsider fostering and supporting not only the devel-opment of information systems but also, and perhapsmore important, the creation of governmental andespecially nongovernmental forums and centers ofexpertise (such as the Center for Local Government inHungary or the official but still independent Financialand Fiscal Commission in South Africa). Unless anduntil the citizens of decentralizing countries have bet-ter access to information and analysis, they, the gov-ernments involved, and outside agencies (such as theBank) will continue to stumble along with respect tomany of the issues that are most critical to designingand assessing decentralization policies.

More generally, decentralization is an issue aboutwhich it may truly be said that “the devil is in thedetails.” Much research needs to be done—primarilyin case studies—on a variety of important issues, par-ticularly with respect to the outcome of alternativeinstitutional changes. These include but are by nomeans limited to the effect of intergovernmental trans-fers on local tax efforts and service expenditures, the


Table 2: Moving toward Better Design of Decentralization in Developing Countries

Institutional Weakness Impact Design Feature to Compensate

Weak democratic institutions • Less local accountability through “voice” mechanisms • Create channels for community and processes • Greater chance of local elites capturing benefits participation

• Decisionmaking less transparent, predictable, • Initiate process of participatory budgetingresponsive • Legislate open, public procurement

procedures

Weak legal and regulatory • Less interjurisdictional mobility and less • Diversify local service providers to give systems accountability through “exit” mechanisms more choice (provide technical assistance

Weak markets for land, • Decisionmaking less transparent, predictable, to facilitate public-private partnerships)labor, and capital responsive • Consider earmarking user fees to improve

accountability

Weak information systems • Less “hierarchical” accountability • Create more local incentives for complianceWeak regulatory systems • Decisionmaking less transparent, predictable, with central objectives

responsive • Improve information systems• Prioritize key regulations

Weak information systems • Moral hazard • Establish transparent and internally Weak regulatory systems • Soft budget constraint between levels of government consistent multitiered budgeting systemsWeak financial systems • Potential central government bailout • Establish subnational debt reporting,Nontransparent fiscal monitoring, and rules for central systems government intervention

• Legislate independence of central bank

All of the above • All of the above • Encourage only modest decentralization of certain local services

effect of fiscal decentralization on macroeconomicindicators and capital market development, alternativeforms and methods of strengthening local institution-al and service delivery capacity, assessments of the

extent to which different decentralization initiativesimprove (or worsen) governance, and potential usesand limitations of user charges as a means of financingdecentralized services.


32

Annex World Bank Economic and SectorWork on Decentralization, 1988–Present

Report ReportNumber Type Date Country Title

General13399 ER 10/26/94 — Stability in a decentralized economy11620 SR 10/1/94 — Health priorities and options in the World Bank’s Pacific member countries8634 ER 5/1/90 — Public sector decentralization: economic policy reform and sector

investment programs

Sub-Saharan Africa16593 ER 11/26/97 Ethiopia Public expenditure review13011 SR 11/30/94 Madagascar Decentralization and local government reform12930 SR 11/10/94 Eritrea Options and strategies for growth9643 SR 3/1/93 Malawi Public sector management review: selected issues8997 SR 4/23/92 Kenya Local government finance study9101 ER 6/1/91 Madagascar Beyond stabilization to sustainable growth8995 SR 6/1/91 Zaire A review of government expenditure8974 SR 2/1/91 Nigeria Urban transport in crisis7844 SR 1/1/91 Nigeria Road sector strategy paper8221 SR 12/1/89 Ghana Fiscal decentralization7963 SR 10/1/89 Mozambique Food security study7495 SR 10/1/89 Tanzania Population, health and nutrition sector review6930 SR 8/1/89 Zaire Urban sector development in Zaire: a new approach to poverty alleviation

East Asia and the Pacific16737 SR 6/1/97 Papua New Guinea Accelerating agricultural growth: an action plan16047 SR 2/25/97 Lao PDR Priorities for rural infrastructure development15745 ER 10/31/96 Vietnam Fiscal decentralization and the delivery of rural services: an economic

report15573 SR 6/27/96 China Higher education reform15898 SR 6/11/96 Philippines Education financing and social equity: a reform agenda14933 SR 11/13/95 Philippines A strategy to fight poverty14540 ER 10/18/95 China Public investment and finance14680 SR 9/5/95 Philippines Public expenditure management for sustained and equitable growth13847 SR 4/3/95 Korea, Rep. of Transport sector: resource mobilization challenges and opportunities13313 SR 11/30/94 Philippines Power sector study: structural framework for the power sector13143 ER 9/26/94 Vietnam Public sector management and private sector incentives12929 SR 9/15/94 China Power sector reform: toward competition and improved performance12343 SR 5/23/94 Philippines Devolution and health services: managing risks and opportunities11764 ER 11/1/93 Malaysia Managing costs of urban pollution: country economic report12407 ER 10/1/93 Indonesia Fiscal decentralization: towards a new partnership for progress11094 SR 7/28/93 China Budgetary policy and intergovernmental fiscal relations10716 SR 1/1/93 Philippines Fiscal decentralization study10050 ER 11/27/91 China Reforming intergovernmental fiscal relations9354 ER 6/1/91 China Economic development in Jiangsu Province8532 ER 8/1/90 Lao PDR Issues in public economics8074 ER 6/1/90 China Reforming social security in a socialist economy7927 SR 12/1/89 Indonesia Power sector institutional development review

7605 SR 6/1/89 China Revenue mobilization and tax policy7483 ER 6/1/89 China Macroeconomic stability and industrial growth under decentralized

socialism7098 SR 3/31/88 Philippines Transport sector review7007 ER 3/1/88 Indonesia Selected issues of public resource management

Europe and Central Asia16420 ER 7/3/97 Estonia Public expenditure review update16112 SR 6/25/97 Ukraine Public expenditure review: restructuring government expenditures15601 ER 2/24/97 Azerbaijan Poverty assessment15353 ER 7/15/96 Kazakhstan Transition of the state14546 SR 6/28/96 Bulgaria Private sector assessment14925 SR 12/19/95 Estonia Financing local governments14862 SR 11/29/95 Russian Federation Fiscal management in the Russian Federation14929 SR 8/1/95 Russian Federation Housing reform and privatization: strategy and transition issues14470 ER 7/20/95 Latvia Local government expenditures and resource transfers14110 SR 6/13/95 Russian Federation Poverty in Russia: an assessment13187 SR 12/12/94 Russian Federation Restructuring the coal industry: putting people first13039 ER 9/28/94 Poland Growth with equity: policies for the 1990s12273 ER 2/23/94 Bulgaria Public finance reforms in the transition11302 ER 12/1/92 Russian Federation Intergovernmental fiscal relations in the Russian Federation11190 SR 9/1/92 Romania Decentralization and local government reform10446 ER 7/1/92 Poland Decentralization and reform of the state10061 ER 5/1/92 Hungary Reform and decentralization of the public sector

Latin America and the Caribbean15543 SR 5/30/96 Belize Environmental report15298 ER 4/29/96 Colombia Reforming the decentralization law: incentives for an effective

delivery of services15044 ER 4/26/96 Paraguay The role of the state15272 ER 2/22/96 Bolivia Poverty, equity and income: selected policies for expanding earning

opportunities for the poor14085 SR 7/7/95 Colombia Local government capacity: beyond technical assistance13304 SR 1/20/95 Guatemala Basic education strategy: equity and efficiency in education12673 SR 8/8/94 Colombia Poverty assessment report12293 SR 6/29/94 Paraguay Poverty and the social sectors in Paraguay: a poverty assessment11933 SR 3/2/94 Colombia Toward increased efficiency and equity in the health sector:

can decentralization help?11130 SR 12/23/93 Venezuela Venezuela 2000: education for growth and social equity11160 ER 12/23/92 Venezuela Decentralization and fiscal issues8972 ER 1/21/92 Venezuela Public administration study8924 SR 7/1/91 Mexico Decentralization and urban management: urban sector study8918 ER 1/1/91 Ecuador Public sector finances: reforms for growth in the era of declining oil

output9093 SR 11/1/90 Honduras Social sector programs8994 SR 8/1/90 Colombia Decentralization reform: a review of political and administrative aspects7870 SR 10/1/89 Colombia Decentralizing revenues and the provision of services: a review of

recent experience

Middle East and North Africa16522 SR 5/8/97 Tunisia Higher education: challenges and opportunities13233 SR 7/26/96 Iran Education, training and the labor markets14043 SR 5/1/95 Islamic Intergovernmental fiscal relations and municipal finance

Rep. of Jordan management sector study10157 SR 6/1/92 Morocco Issues and prospects in the public sector8782 ER 8/1/90 Morocco Decentralization in local government expenditures and management7150 SR 3/1/88 Tunisia Municipal finance and management7115 SR 2/1/88 Morocco Municipal finance sector report

Annex World Bank Economic and Sector Work on Decentralization, 1988–Present 33


South Asia15677 SR 7/31/97 India The Indian oilseed complex: capturing market opportunities16506 ER 5/30/97 India 1997 economic update: sustaining rapid growth16558 SR 5/1/97 Bangladesh Municipal finance management sector study16113 ER 12/20/96 Bhutan Country economic memorandum15905 ER 7/31/96 Bangladesh Public expenditure review14960 SR 6/6/96 Pakistan Improving basic education: community participation, system

accountability, and efficiency15092 SR 11/20/95 Pakistan Supporting fiscal decentralization in Pakistan9518 SR 12/20/91 India Irrigation sector review7946 ER 1/1/90 Bangladesh Poverty and public expenditures: an evaluation of the impact

of selected government programs

Note: ER is Economic Report; SR is Sector Report.



The authors are grateful for helpful comments provided by

Thomas Bossert, Shanta Devarajan, William Dillinger, Cheryl

Gray, Robert Inman, Christine Wallich, and Dana Weist.

1. It is interesting to note that Africa, the region with the high-

est proportion of decentralization projects, has completed the

least formal economic and sector work on the broader topic of

decentralization or intergovernmental fiscal relations. By con-

trast, Europe and Central Asia, the region that has focused its

analytical work most on intergovernmental fiscal relations, has

the lowest proportion of decentralization projects. The experi-

ence from these regions suggests that there may be inadequate

connection between sectoral and project issues on one hand and

the macroeconomic and broader policy dialogue on the other.

2. See, for example, recent studies by the International

Monetary Fund (Ter-Minassian, 1997), Inter-American

Development Bank (1994, 1997; Lopez-Murphy, 1996), and

CEPAL/GTZ (Aghon and Krause-Junk, 1996), as well as the

recent conferences sponsored by the U.S. Agency for Inter-

national Development (Washington D.C., October 1997) and the

Food and Agriculture Organization (Rome, December 1997).

3. The traditional methods used by the Bank to promote

transparency and accountability—procurement rules and

financial audits—are inadequate for improving the quality of

government at the local level because of the lack of basic insti-

tutions to ensure political and economic accountability.

4. Dillinger explored the importance of institutional issues

for Bank work in urban areas in his discussion of urban finance

in 1994. This paper furthers that analysis.

5. Alternatively, as in the traditional economic theory of “fis-

cal federalism” (Oates, 1972), both levels of government may

be seen as agents of different groups of citizens, with the cen-

tral government encompassing the subsets included within the

purview of the different local governments. So long as local gov-

ernments are hierarchically subordinated to the central govern-

ment, both approaches lead to the same conclusion: the central

government’s preferences—whether seen as representative of all

citizens or simply as the superior jurisdiction—should domi-

nate, and the design problem is as just stated (Seabright, 1996).

6. The tax and expenditure decisions made by local govern-

ments can also have a major impact on redistribution.

7. Bird and Gendron (1998) suggest that surcharges on a

value-added tax may also be feasible in some circumstances,

provided administrative capabilities are high enough.

8. An alternative view is the “federalist approach,” where the

primary objective is to ensure that all regions of the country are

able to provide such services at acceptable minimum standards

and have adequate resources to do so. In this case simple lump-

sum transfers, with no conditionality other than the usual

requirements for financial auditing, are called for (Shah, 1994).

In this approach it is essentially assumed that the fact that the

funds flow to locally responsible political bodies will ensure suf-

ficient accountability and that it is neither necessary nor desirable

for the central government to attempt to interfere with, or influ-

ence, local expenditure choices.

9. The success of this option would depend, of course, on

the rules that guide the budget process.

10. For example, recent World Bank research on decentral-

ization of education has indicated that the most important ele-

ment in improving outcomes (as part of education reform) is

parent involvement in teacher supervision (King and Ozler,

1998). Knowing which aspects of decentralization are most

important can help in creating the appropriate institutional

structures to benefit from decentralization.

11. Democracy without good information is not enough;

nor, of course, is information without democracy. Nonetheless,

even in countries without truly democratic institutions in which

decentralization is simply another instrument of the central

government, good information is essential to improving service

outcomes.

12. Toward this end, the Economic Development Institute is

working in Africa (Zimbabwe) and Latin America (Venezuela)

to develop pilot experiences with municipalities to increase

35

Notes

transparency through participatory budgeting and public infor-

mation campaigns.

13. Since some functions are of national importance, the

principal-agent framework (whereby the center provides man-

dates and incentives so that local governments will achieve

desired results) is appropriate. Yet decentralization provides

myriad opportunities for independent local decisionmaking

that affects efficiency, equity, and quality of services. How much

local discretion (or “decision space”) is provided to local gov-

ernments will differ depending on the sector and the particular

function. (For more discussion, see Bossert, forthcoming.)

14. Porto Allegro, Brazil, offers a fascinating case of trans-

parent budgeting and community participation in local govern-

ment fiscal decisionmaking. Indeed, Porto Allegro and a dozen

municipalities in Brazil publish their budget on the Internet (see

http://www.prefpoa.com.br/opart/default.htm). Regrettably,

most participatory budgeting in Latin America has focused only

on capital works, not on recurrent revenue and expenditures

(see also Paul, 1995, on India).

15. Pure public goods will be underprovided and under-

consumed if left to the market and thus must still be provided

by government. Thus accountability for local public goods must

still rely on electoral processes or interjurisdictional mobility.

16. Technical assistance can be provided to local govern-

ments to facilitate public-private partnerships. For example, risk

management templates for solid waste have been designed and

used in the Philippines to help local governments analyze the

structure of financing and the risks faced for different aspects of

service delivery under different contractual agreements.

17. If indeed mobility is constrained, then general local

taxes—such as a sales tax or a piggybacked income tax—are

appealing from an efficiency standpoint. (The problem of

interjurisdictional tax competition would not exist.) Yet in

most developing countries the administrative capacity is weak

for such tax collections. If local democracy is also weak, then

it may be more likely that funds raised through earmarked

user fees will be spent more efficiently to better reflect local

preferences.


37

References

Aghon, Gabriel, and Gerold Krause-Junk. 1996.Descentralización Fiscal en America Latina: Balance, Pricipios,Desafiós. Santiago, Chile: Proyecto Regional deDescentralización Fiscal CEPAL/GTZ.

Ahmad, Ehtisham. 1996. Financing Decentralized Expenditures:An International Comparison of Grants. Cheltenham, U.K.:Edward Elgar.

Ahmad, Junaid. 1995. “Funding the Metropolitan Areas ofSouth Africa.” Finance and Development 32 (September):50–53.

Alderman, H. 1998. “Decentralization and Targeted Transfers:Social Assistance in Albania.” Living Standards MeasurementStudy Working Paper 134. World Bank, Washington, D.C.

Bahl, Roy W. 1998. “China: Evaluating the Impact ofIntergovernmental Fiscal Reform.” In Richard Bird andFrancois Vaillancourt, eds., Fiscal Decentralization inDeveloping Countries. Cambridge: Cambridge UniversityPress.

Bahl, Roy W., and Johannes Linn. 1992. Urban Public Finance inDeveloping Countries. New York: Oxford University Press.

Besley, Timothy, and Anne Case. 1995. “Incumbent Behavior:Vote-Seeking, Tax-Setting, and Yardstick Competition.”American Economic Review 85 (March): 25–45.

Bird, Richard M. 1993. “Threading the Fiscal Labyrinth: SomeIssues in Fiscal Decentralization.” National Tax Journal 46 (2):207–27.

———. 1994. “Decentralizing Infrastructure: For Good or ForIll?” Policy Research Working Paper 1258. World Bank,Washington, D.C.

———. 1995. “Financing Local Services: Patterns, Problems,and Possibilities.” Major Report 31. University of Toronto,Centre for Urban and Community Services.

Bird, Richard M., and Ariel Fiszbein. 1998. “Colombia: TheCentral Role of the Central Government in FiscalDecentralization.” In Richard Bird and Francois Vaillancourt,eds., Fiscal Decentralization in Developing Countries.Cambridge: Cambridge University Press.

Bird, Richard M., and Pierre-Pascal Gendron. 1998. “Dual VATsand Cross-Border Trade: Two Problems, One Solution?”International Tax and Public Finance 5 (3): 429–42.

Bird, Richard M., and Francois Vaillancourt, eds. 1998. FiscalDecentralization in Developing Countries. Cambridge:Cambridge University Press.

Bird, Richard M., Robert D. Ebel, and Christine I. Wallich, eds.1995. Decentralization of the Socialist State. A Regional andSectoral Study. Washington, D.C.: World Bank.

Boadway, Robin. 1997. “Public Economics and the Theory ofPublic Policy.” Canadian Journal of Economics 30 (4a):753–72.

Boadway, Robin, Nicolas Marceau, and Motohiro Sato. 1997.“Subsidies versus Public Provision of Private Goods asInstruments for Redistribution.” Discussion Paper 942.Queen’s University, Institute for Economic Research, Canada.

Bossert, Thomas. Forthcoming. “Decentralization of HealthSystems: Decision Space, Innovation and Performance.”Social Science and Medicine.

Brennan, Geoffrey, and James M. Buchanan. 1980. The Power toTax. Cambridge: Cambridge University Press.

Breton, Albert. 1996. Competitive Governments. Cambridge:Cambridge University Press.

Breton, Albert, and Anthony Scott. 1978. The EconomicConstitution of Federal States. Toronto: University of TorontoPress.

Breton, Albert, Alberto Cassone, and Angela Fraschini. 1998.“Decentralization and Subsidiarity: Toward a TheoreticalReconciliation.” University of Pennsylvania Journal ofInternational Economic Law 19 (1): 21–51.

Buchanan, James, and Richard Wagner. 1971. “An EfficiencyBasis for Federal Fiscal Equalization.” In J. Margolis, ed.,Analysis of Public Output. Princeton, N.J.: PrincetonUniversity Press for National Bureau for Economic Research.

Cremer, Jacques, Antonio Estache, and Paul Seabright. 1995.“The Decentralization of Public Services: Lessons from theTheory of the Firm.” In Antonio Estache, ed., DecentralizingInfrastructure: Advantages and Limitations. Washington, D.C.:World Bank.

Crook, Richard, and James Manor. 1994. EnhancingParticipation and Institutional Performance: DemocraticDecentralisation in South Asia and West Africa. A Report toESCOR, the Overseas Development Administration.


Davoodi, Hamid, and Heng-fu Zou. 1998. “FiscalDecentralization and Economic Growth: A Cross-CountryCase Study.” Journal of Urban Economics 43: 244–57.

Dillinger, William. 1994. Decentralization and Its Implications forUrban Service Delivery. Urban Management Programme Series16. World Bank, Washington, D.C.

———. 1997. “Brazil’s State Debt Crisis: Lessons Learned.”Economic Notes 14. World Bank, Country Department 1,Latin America and the Caribbean Region, Washington,D.C.

Ehdaie, Jaber. 1994. “Fiscal Decentralization and the Size ofGovernment,” Policy Research Working Paper 1387. WorldBank, Washington, D.C.

Fiszbein, Ariel. 1997. “Local Government Capacity inColombia.” World Development 25 (July): 1029–43.

Fox, Jonathan, and Josefina Aranda. 1996. “Decentralizationand Rural Development in Mexico: Community Participationin Oaxaca’s Municipal Funds Program.” Monograph Series.University of California, Center for U.S.-Mexican Studies,San Diego.

Glaessner, Philip J., Lee Kye Woo, Anna Marian Sant’anna,Jean-Jacques de St. Antoine, and others. 1994. PovertyAlleviation and Social Investment Funds: The Latin AmericanExperience. World Bank Discussion Paper 261. Washington,D.C.

Grossman, David. 1996. “Strengthening the InstitutionalFramework.” In “Colombia: Reforming the DecentralizationLaw: Incentives for an Effective Delivery of Services.” WorldBank, Washington, D.C.

Hines, J., and R. Thaler. 1995. “The Flypaper Effect.” Journal ofEconomic Perspectives 9 (4): 217–26.

Huther, Jeff, and Anwar Shah. 1998. “Applying a SimpleMeasure of Good Governance to the Debate on FiscalDecentralization.” Policy Research Working Paper 1894.World Bank, Washington, D.C.

Inman, Robert P., and Daniel L. Rubinfeld. 1996. “The PoliticalEconomy of Federalism.” In Dennis C. Mueller, ed.,Perspectives on Public Choice. Cambridge: CambridgeUniversity Press.

———. 1997. “Rethinking Federalism.” Journal of EconomicPerspectives 11 ( 4): 43–64.

Inter-American Development Bank. 1994. “Fiscal De-centralization: The Search for Equity and Efficiency.” InEconomic and Social Progress in Latin America: 1994 Report.Baltimore, Md.: Johns Hopkins University Press for the Inter-American Development Bank.

———. 1997. “Fiscal Decision-making in DecentralizedDemocracies.” In Economic and Social Progress in LatinAmerica: 1997 Report. Baltimore, Md.: Johns HopkinsUniversity Press for the Inter-American Development Bank.

Kenyon, Daphne. 1991. Interjurisdictional Tax and PolicyCompetition: Good or Bad for the Federal System? Washington,D.C.: Advisory Commission on Intergovernmental Relations.

Kessides, Christine. 1993. Institutional Options for the Provisionof Infrastructure. World Bank Discussion Paper 212.Washington, D.C.

King, Elizabeth M., and Berk Ozler. 1998. “What’sDecentralization Got to Do with Learning? The Case ofNicaragua’s School Autonomy Reform.” Impact Evaluation ofEducation Working Paper Series 9. World Bank, Washington,D.C.

Kolehmainen-Aitken, Riitta-Liisa, and William Newbrander.1997. Decentralizing the Management of Health and FamilyPlanning Programs. Boston, Mass.: Management Sciences forHealth.

Kornai, Janos. 1992. The Socialist System. Princeton, N.J.:Princeton University Press.

Lall, Rajiv, and Bert Hofman. 1995. “Decentralization and theGovernment Deficit in China.” In MacroeconomicManagement and Fiscal Decentralization. EDI Seminar Series.World Bank, Washington, D.C.

Litvack, Jennie. 1994. “Regional Demands and FiscalFederalism.” In Christine Wallich, ed., Russia and theChallenge of Fiscal Federalism. A Regional and Sectoral Study.Washington, D.C.: World Bank.

Litvack, Jennie, and Claude Bodart. 1993. “User Fees PlusQuality Equals Improved Access to Health Care: Results of aField Experiment in Cameroon.” Social Science and Medicine37 (3): 369–83.

Lopez-Murphy, Ricardo, ed. 1996. Fiscal Decentralization inLatin America. Washington, D.C.: Center for Research inApplied Economics, Inter-American Development Bank.

Martinez-Vazquez, Jorge, and Robert McNab. 1997. “FiscalDecentralization, Economic Growth, and DemocraticGovernance.” Paper prepared for USAID Conference onEconomic Growth and Democratic Governance,Washington, D.C., October 9–10.

May, Ronald James. 1971. Financing the Small States in AustralianFederalism. New York: Oxford University Press.

McLure, Charles E., Jr. 1995. “Comment on Prud’homme.” TheWorld Bank Research Observer 10 (2): 221–26.

Mueller, Dennis C., ed. 1996. Perspectives on Public Choice.Cambridge: Cambridge University Press.

———. 1997. “Breton Competitive Governments: AnEconomic Theory of Politics and Public Finance.” A BookReview. Canadian Journal of Economics 30 (4a): 997–1003.

Musgrave, Richard A. 1983. “Who Should Tax, Where, andWhat?” In Charles McLure, ed., Tax Assignment in FederalCountries. Canberra: Centre for Research on Federal FinancialRelations, Australian National University.

References 39

Narayan, Deepa. 1998. “Participatory Rural Development.” InErnst Lutz, ed., Agriculture and the Environment: Perspectives onSustainable Rural Development. Washington, D.C.: World Bank.

Oates, Wallace. 1972. Fiscal Federalism. New York: HarcourtBrace Jovanovich.

Parker, Andrew N. 1995. “Decentralization: The Way Forwardfor Rural Development?” Policy Research Working Paper1475. World Bank, Washington, D.C.

Paul, Samuel. 1995. A Report Card on Public Services in IndianCities: A View from Below. Bangalore, India: Public AffairsCentre.

Pauly, Mark. 1973. “Income Distribution as a Local PublicGood,” Journal of Public Economics (2): 35–58.

Peterson, George E. 1997. “Building Local Credit Systems.”Urban Management Program Policy Paper. World Bank,Washington, D.C.

Prud’homme, Remy. 1995. “The Dangers of Decentralization.”The World Bank Research Observer 10 (2): 201–20.

Ravallion, Martin. 1993. “Poverty Alleviation through RegionalTargeting: A Case Study for Indonesia.” In Karla Hoff, AviBraverman, and Joseph Stiglitz, eds., The Economics of RuralOrganization. Oxford: Oxford University Press.

———. 1998. “Reaching Poor Areas in a Federal System.”World Bank, Development Research Group, Washington,D.C.

Rondinelli, Dennis. 1981. “Government Decentralization inComparative Perspective: Theory and Practice in DevelopingCountries.” International Review of Administrative Sciences 47(2): 133–45.

———. 1989. “Decentralizing Urban Development Programs:A Framework for Analyzing Policy.” U.S. Agency forInternational Development, Office of Housing and UrbanPrograms, Washington, D.C.

———. 1998. “What Is Decentralization?” World Bank PovertyReduction and Economic Management KnowledgeManagement System, Washington, D.C.

Salmon, Pierre. 1987. “Decentralization as an IncentiveScheme.” Oxford Review of Economic Policy 3 (2): 24–43.

Seabright, Paul. 1996. “Accountability and Decentralization inGovernment: An Incomplete Contracts Model.” EuropeanEconomic Review 40 (1): 61–89.

Shah, Anwar. 1994. “The Reform of Intergovernmental FiscalRelations in Developing Countries and Emerging MarketEconomies.” Policy and Research Series 23. World Bank,Washington, D.C.

Shah, Anwar, and Zia Qureshi. 1994. Intergovernmental FiscalRelations in Indonesia. World Bank Discussion Paper 239.Washington, D.C.

Spahn, P. Bernd. 1997a. “Decentralization in TransitionEconomies.” Paper prepared for the World Bank Seminar on

Fiscal Decentralization in Developing Countries,Washington, D.C., May 7.

———. 1997b. “Decentralized Government and Macro-economic Control.” Paper prepared for the InternationalInstitute of Public Finance 53rd Congress, Kyoto, Japan.

Tanzi, Vito. 1996. “Fiscal Federalism and Decentralization: AReview of Some Efficiency and Macroeconomic Aspects.” InMichael Bruno and Boris Pleskovic, eds., Annual World BankConference on Development Economics 1995. Washington,D.C.: World Bank.

Ter-Minassian, Teresa, ed. 1997. Fiscal Federalism in Theory andPractice. Washington, D.C.: International Monetary Fund.

Tiebout, Charles. 1956. “A Pure Theory of Local Expenditures.”Journal of Political Economy 64 (5): 416–24.

Treisman, Daniel. 1998. After the Deluge: Regional Crisis andPolitical Consolidation in Russia. Ann Arbor: University ofMichigan Press.

Tresch, Richard. 1981. Public Finance. Plano, Texas: BusinessPublications.

Triche, Thelma. 1993. “The Institutional and RegulatoryFramework for Water Supply and Sewerage: Public and PrivateRoles.” TWURD, WS-9. World Bank, Washington, D.C.

Triche, Thelma, Abel Mejia, and Emanueal Idelovitch. 1993.“Arranging Concessions for Water Supply and SewerageServices: Lessons from Buenos Aires and Caracas.” TWURD,WS-10. World Bank, Washington, D.C.

Vaillancourt, Francois. 1998. “Morocco and Tunisia: FinancingLocal Governments—the Impact of Infrastructure Finance.”In Richard Bird and Francois Vaillancourt, eds., FiscalDecentralization in Developing Countries. Cambridge:Cambridge University Press.

Van de Walle, Dominique, and Kimberly Nead, eds. 1995.Public Spending and the Poor: Theory and Evidence. Baltimore,Md.: Johns Hopkins University Press.

Wallich, Christine I., ed. 1994. Russia and the Challenge of FiscalFederalism. A Regional and Sectoral Study. Washington, D.C.:World Bank.

Walsh, Cliff. 1992. Fiscal Federalism: An Overview of Issues anda Discussion of their Relevance to the European Community.Canberra: Federalism Research Centre, Australian NationalUniversity.

Watts, Ronald. 1996. Comparing Federal Systems in the 1990s.Kingston, Canada: Institute of Intergovernmental Relations,Queen’s University.

Weingast, Barry. 1995. “The Economic Role of PoliticalInstitutions: Market-Preserving Federalism and EconomicDevelopment.” Journal of Law, Economics, and Organization 11(1): 1–31.

Wheare, Kenneth C. 1969. Federal Government. 4th ed.London: Oxford University Press.


Wildasin, David. 1997. “Externalities and Bailouts: Hard andSoft Budget Constraints in Intergovernmental FiscalRelations.” Policy Research Working Paper 1843. WorldBank, Washington, D.C.

Wilensky, Harold. 1974. The Welfare State and Equality:Structural and Ideological Roots of Public Expenditures.Berkeley: University of California Press.

World Bank. 1988. World Development Report 1988: Op-portunities and Risks in Managing the World Economy. NewYork: Oxford University Press for the World Bank.

———. 1994. World Development Report 1994: Infrastructure forDevelopment. New York: Oxford University Press for theWorld Bank.

———. 1995a. Better Urban Services: Finding the Right Incentives.A Development in Practice book. Washington, D.C.

———. 1995b. Local Government Capacity in Colombia. AWorld Bank Country Study. Washington, D.C.

———. 1996a. “Argentina: Provincial Finances Study.” Report15487-AR. Washington, D.C.

———. 1996b. “Colombia: Reforming the DecentralizationLaw: Incentives for an Effective Delivery of Services.” Report15298-CO. Washington, D.C.

———. 1996c. “Vietnam: Fiscal Decentralization and theDelivery of Rural Services.” Report 15745-VN. Washington,D.C.

———. 1997. World Development Report: The State in a ChangingWorld. New York: Oxford University Press for the World Bank.

Zhang, Tao, and Heng-fu Zou. 1998. “Fiscal Decentralization,Public Spending and Economic Growth in China.” Journal ofPublic Economics 67: 221–40.

rethinking decentralization in developing countries

Documents