II_ENG.pdf · DISCLAIMER "The terms used concerning the legal status of any country, territory,...

Local Government Finance:The Challenges of the 21st Century

Second Global Report on Decentralization and Local Democracy

United Cities and Local Governments

Cites et Gouvernements Locaux Unis

Ciudades y Gobiernos Locales Unidos

GOLD II2010

0w2010 00 PCB Portada DEFcarta ang 15/9/11 23:09 Página 1

Local Government Finance:The Challenges of the 21st Century

Second Global Report on Decentralization and Local Democracy

Edward ElgarCheltenham, UK • Northampton, MA, USA

GOLD II2 0 1 0

United Cities and Local Governments

Cités et Gouvernements Locaux Unis

Ciudades y Gobiernos Locales Unidos

0w2010 00 PCB Portadelles DEFcarta ang 8/9/11 14:14 Página 1

DISCLAIMER"The terms used concerning the legal status of any country, territory, city orarea, or of its authorities, or concerning delimitation of its frontiers orboundaries, or regarding its economic system or degree of development do notnecessarily reflect the opinion of United Cities and Local Governments. Theanalysis, conclusions and recommendations of this report do not necessaryreflect the views of the all members of United Cities and Local Governments."

United Cities and Local GovernmentsCités et Gouvernements Locaux UnisCiudades y Gobiernos Locales UnidosAvinyó 1508002 Barcelonawww.cities-localgovernments.org

Design and Layout MIA BCNProofreader Vivian YelaIllustrations and graphics Gérard Fagot


Foreword 5

Editorial project 7

Acknowledgments 9

1. Introduction 11

2. Africa 23

3. Asia-Pacific 61

4. Eurasia 113

5. Europe 149

6. Latin America 189

7. Middle East and Western Asia 231

8. North America 255

9. Financing Metropolitan Areas 285

10. Conclusion 309

Annex 331

Bibliography 357

3

CONTENTS


© United Cities and Local Governments 2011

All rights reserved. No part of this publication may be reproduced, stored in aretrieval system or transmitted in any form or by any means, electronic, mecha-nical or photocopying, recording, or otherwise without the prior permission of thepublisher.

Published byEdward Elgar Publishing LimitedThe Lypiatts15 Lansdown RoadCheltenhamGlos GL50 2JAUK

Edward Elgar Publishing, Inc.William Pratt House9 Dewey CourtNorthamptonMassachusetts 01060USA

A catalogue record for this bookis available from the British Library

Library of Congress Control Number: 2011936418

ISBN978 0 85793 976 0 (cased)

Printed and bound by MPG Books Group, UK


5

The 2nd Global Report on Decentralization and Local Democracy, that I havethe pleasure to present you with, is dedicated to local finance. It confirms thegrowing role of local governments in all regions of the world. It equallydemonstrates the important imbalances that can exist in the sharing ofresources and responsibilities between national and local governments. Theseimbalances have only been worsened with the impact of the global financialand economic crisis.

Thus, everywhere in the world, local authorities have more and more responsibilitiesin service provision, the putting into practice of social policy, environmentalmanagement, and local development. They ensure between two thirds to a half of thepublic investment in OECD countries as well as in certain emerging nations –China,South Africa and Brazil.

However, if the responsibilities of local government are growing, the share offunds available to ensure these responsibilities is often inadequate, inparticular in developing countries. This issue is made worse by the low level ofautonomy local governments have with regard to financial management in themajority of regions.

Without autonomy and resources local democracy is crippled. Its advances, whichinclude the free election of local representatives in the majority of countries, remainprecarious and can generate a profound disillusionment which threatens to ricochetback and fissure its own democratic foundation.

This divide between responsibilities and the sharing of resources specifically impactsthe attainment of the Millennium Development Goals (MDGs). In fact, it is insub-Saharan Africa and in the least advanced countries of Asia, where the means oflocal governments are the weakest, that the attainment of the MDGs is lagging themost. If local governments in the European Union spend nearly 3,250 € a year perinhabitant to meet the needs of their citizens, in sub-Saharan Africa and certain countries ofAsia only 24 € per inhabitant is available, and significantly less in the poorest countries.

The 2nd Global Report demonstrates that the financing of urban and local developmentis one of the weak links of development aid policies. With accelerating urbanization thecurrent level of available financing does not allow for a response to the existing andongoing “urbanization of poverty”.

FOREWORD

Bertrand DelanoëMayor of Paris, FrancePresident of UCLG


PPRREEFFAACCEEUnited Cities and Local Governments6

Our world organization, United Cities and Local Governments, has estimated that 200billion USD is needed annually, over the next 25 years, for investment in cities ofdeveloping countries to ensure that the most marginalized communities receiveessential services in order to reduce poverty and slum development.

For developed countries, the Report equally signals the constraints on local finance inresponding to structural changes –aging populations, migratory fluctuations,reductions in energy use and CO2 emissions, and risk prevention. But above all, itsignals against the tendency to unload on local governments where they are notdirectly responsible, a disproportionate weight of budgetary and financial deficitsthrough the assigning of new responsibilities without the necessary funding.

I can only support the conclusions of this Report on the need to establish new politicalregulations between central and local governments in each country, as well as at theworld level. A strengthened dialogue between the different levels of government istherefore indispensable so as to ensure a better sharing of means and competencies, abetter balance between democracy and solidarity, two pillars on which the future ofour countries, cities and populations is balanced, two principles based in the cardinalnotions of justice and responsibility.

Bertrand Delanoë


7

Publishers

• Director: Elisabeth Gateau, Secretary General, UCLG

• Deputy Director: Alberto Laplaine Guimarãis, Deputy Secretary General, UCLG

• Principal Adviser: Emilia Sáiz, Chief of Staff, Director of Statutory Issues and Institutional Relations, UCLG

• Coordination: Edgardo Bilsky, Director of Programmes and Research, UCLG

• Coordination Assistant: Claire Frost, Project Officer, UCLG

• Support team from UCLG: Dominique Arrestat, Pere Ballester, Mohamed Boussraoui,Jean-Baptiste Buffet, Sandra Climent, Alexis Demko, Sara Hoeflich, Ricardo Martínez,Mónica Mora, Carole Morillon, Thibaut Nancy, Natalène Poisson, Marie-Laure Roa,Maya Sawmy, Elisabeth Silva, Renske Steenbergen

• Interns: Justine Delefortrie, Eugénie Monasterio, Elsa Payan

Steering Committee

UCLG Regional and Metropolitan Sections• Don Borut, Secretary General, UCLG North America

• Jean-Pierre Elong Mbassi, Secretary General, UCLG Africa

• Josep Roig, Secretary General, Metropolis

• Rassikh Sagitov, Secretary General, UCLG Euro-Asia

• Guillermo Tapia, Secretary General, FLACMA, UCLG Latin America

• Frédéric Vallier, Secretary General, Council of European Municipalities and Regions (CEMR),UCLG Europe

• Peter Woods, Secretary General, UCLG Asia Pacific

• Selahattin Yildirim, Former Secretary General, UCLG Middle East & Western Asia

Experts Coordination• Jorge Martínez-Vázquez, Regents Professor of Economics and Director of International

Studies, Andrew Young School of Policy Studies, Georgia State University, U.S.A.

• Paul Smoke, Professor of Public Finance and Planning and Director of International Pro-grams, New York University / Robert F. Wagner Graduate School of Public Service, U.S.A.

EDITORIAL PROJECT


EEDDIITTOORRIIAALL PPRROOJJEECCTTUnited Cities and Local Governments8

Principal Partners• William Cobbett, Manager, Cities Alliance

• Àngel Cortadelles i Bacaria, Director General of International Relations, Generalitat deCatalunya, Spain

• Sandrine Delibiot, Director of International Relations and Marketing, Public and Whole-sale Banking, Dexia Bank

• Antoni Fogué, President of the Diputación de Barcelona and of the UCLG Committee onDecentralization and Local Self-Government, Spain

• Nathalie Le Denmat, Director of Local Governments and Urban Development, FrenchDevelopment Agency – Agence Française de Développement (AFD), France

Regional Advisers• Sandra Ceciarini, Director of Citizenship and International Cooperation, CEMR

• Rudi Hauter, Associate to the Secretary General, UCLG Asia Pacific

• Christopher Hoene, Director of Research, National League of Cities, U.S.A.

• Souleymane Maiga, Chief of Staff, UCLG Africa

• Néstor Vega, Academic Coordinator, FLACMA, UCLG Latin America

• Cenk Tikiz, General Coordinator, UCLG Middle East & Western Asia


9

Main authors by chapters

Introduction

• Jorge Martínez-Vázquez, Andrew Young School of Policy Studies, Georgia State University, U.S.A.

• Paul Smoke, New York University / Robert F. Wagner Graduate School of Public Service, U.S.A.

Africa

• François Yatta, Independent Researcher, Niger• François Vaillancourt, Université de Montréal, Canada

Asia Pacific

• Blane D. Lewis, Lee Kuan Yew School of Public Policy, National University of Singapore,Singapore

• Bob Searle, Independent Consultant and former head of Australia’s Grants Commission,Australia

Eurasia

• Natalia Golovanova, Center for Fiscal Policy (Moscow), Russia• Galina Kurlyandskaya, Center for Fiscal Policy (Moscow), Russia

Europe

• Luiz de Mello, Economics Department, Organisation for Economic Co-operation andDevelopment (OECD)

Latin America

• Jorge Martínez-Vázquez

Middle East & Western Asia

• Mehmet Tosun, University of Nevada, U.S.A.

ACKNOWLEDGMENTS


PPRREEFFAACCEEUnited Cities and Local Governments10

North America

• William F. Fox, University of Tennessee, U.S.A.• Enid Slack, University of Toronto, Canada

Financing Metropolitan Areas

• Roy Bahl, Andrew Young School of Policy Studies, Georgia State University, U.S.A.

Conclusions

• Jorge Martínez-Vázquez• Paul Smoke

Other Contributors

All UCLG sections, the Partenariat pour le Développement Municipal (PDM) based inCotonou, Benin for West and Central Africa and the Municipal Development Partnership(MDP) based in Harare, Zimbabwe for East and Southern Africa collaborated in thecollection of data and information.

Special acknowledgments for financial and advisory support

Generalitat de Catalunya

Diputació de Barcelona

Agence Française de Développement (AFD)

Cities Alliance

DEXIA Group


1. INTRODUCTION

JJOORRGGEE MMAARRTTÍÍNNEEZZ--VVÁÁZZQQUUEEZZ,,

GGEEOORRGGIIAA SSTTAATTEE UUNNIIVVEERRSSIITTYY,, UUSSAA

PPAAUULL SSMMOOKKEE,,

NNEEWW YYOORRKK UUNNIIVVEERRSSIITTYY,, UUSSAA

0w2010 01 Introduccio DEFcarta ang 15/9/11 07:08 Página 11

13

L ocal governments around the world todayplay a key role in facilitating development

and improving living standards. As morerobust governance mechanisms are adoptedand civil society develops even wherethey have historically been weak, localgovernments have grown to operate in anincreasingly open and responsive manner.Today they are considered by many centralgovernments to be important partners indealing with a range of public policy issuesand functions, including building moreefficient and equitable social service systemsand providing significant portions ofkey infrastructure that supports economicdevelopment and improves the quality of life.

The road to this point, however, has been farfrom smooth or easy, and many challengesto effective local governments persist tovarious degrees. Decentralization has beenuneven and has faced major obstacles. Inmany countries intergovernmental systemsremain problematic in various respects,and local governments are not adequatelyequipped to perform their functions welland to become more effective partners ofhigher level governments in meeting pressingcommon goals.

This report builds on the 2008 First GlobalReport on Decentralization and LocalDemocracy (GOLD I), which provides a broadbased overview of local government systemsaround the world. GOLD II focuses on aspecific aspect of decentralization —the fiscalarchitecture and performance of local governments.This topic was chosen for GOLD II becausefiscal architecture is fundamental to ensuringthat local governments can deliver public servicesand function successfully in meeting otheressential responsibilities.

Increasing fiscal decentralization (measuredas the subnational share of total nationalpublic expenditures) has been a globaltrend in recent decades. There are, however,

significant variations across and withinregions. Local budgets account on average for25 percent of public expenditures in theEuropean Union, for example, but less than 5percent in many developing countries. Iffiscal decentralization is evaluated in terms ofexpenditure and revenue autonomy, therehas been progress, but it has been unevenacross countries and generally less robust onthe revenue side. Global experiences alsodemonstrate that intergovernmental fiscalrelations are not fixed —they tend to evolvewith social, political, economic, demographicand technological forces that affect theoverall role of the public sector.

GOLD II takes the pulse of the current stateof the local public finances around the worldwith the main goal of identifying andanalyzing the principal challenges that localgovernments face in providing public servicesmore efficiently and equitably. The report alsooffers concrete recommendations for prioritypolicy reforms regionally and globally.

Why is Local Government FinanceImportant?

The potential importance of local governmentfinance is based on two main pillars. Thecore rationale is that local governments arewell positioned to improve how publicresources are used and the extent to whichdiverse citizen needs are satisfied. The secondjustification is the role that local governmentscould potentially play in dealing with severalsignificant contemporary global challengesthat broadly, although differentially, affectvirtually all countries.

The Core Rationale

The conventional case for decentralization isgrounded on two basic propositions. The firstis that local governments are closer to thepeople than the central governments, andthey have superior access to local information


IINNTTRROODDUUCCTTIIOONNUnited Cities and Local Governments14

that allows them to respond better to theneeds of citizens. The second is that they facestronger incentives to perform well on localmatters than the central government, so thatthey are in a better position to derive themost from public resources at their disposaland are more likely to seek innovative meansof doing so. These two propositions arerelated: access to local information andincentives to use it well must worksynergistically to produce better results.

Although the value of local governments inthis regard is well recognized, and there issome supporting if uneven empirical evi-dence, there are caveats. Close collaborationand innovative institutional arrangements areneeded between local governments andhigher levels of government to provideservices that involve economies of scale oraffect multiple local governments.

Equally important, the validity of the twobasic propositions regarding the benefits oflocal governments depends on meetingfundamental requirements. At a minimum,there must be sufficient autonomous localgovernment powers and resources, satis-factory local technical and managerial ca-pacity, and adequate incentives (electoraland beyond) for local governments to beheld accountable to their constituents andto behave in a fiscally responsible manner.Central governments can play a role inhelping local governments to meet theserequirements, but this requires time andongoing support in countries where theyare not in place. Thus, appropiateimplementation —the sequencing andpace—of intergovernmental fiscal reforms isjust as important as sound design.

Global Challenges and the Role of LocalGovernments

In addition to the general desirability ofdecentralization if appropriately designed and

implemented, a number of specific andinterrelated global trends present greatchallenges to individual countries and thebroader international community andreinforce the potentially important role oflocal governments. First, the world is facingmultiple environmental and resource crises,such as global warming, energy shortagesand food security concerns, which haveemerged prominently in domestic and globalpolicy circles in recent years. These crisesindividually and collectively impact localgovernments in very specific ways, but localgovernments may also be in a strong positionto help respond to them.

Second, increasing urbanization (see Figure1.1), which exacerbates the crises mentio-ned above and generates great publicservice needs, is a pervasive global trend,especially in developing countries. A ma-jority of the world’s residents now live inurban areas, and the share is expected toexceed 60 percent by 2030.1 According tothe United Nations (UN), 95 percent of theurban growth in the next two decades isexpected to be in Asia, Africa and to a lesserextent in Latin America, and it will befocused in small and medium sized cities.Rapid urban growth also implies anincreasing urbanization of poverty. Ifcurrent trends persist, one out of fivepersons will live in urban slums by 2030.The struggle to meet the MillenniumDevelopment Goals and advance the globalfight against poverty may be won or lostprimarily in the urban areas of developingcountries. Increasing urbanization alsocreates a need for innovative mechanismsto govern and serve metropolitan areasthat are growing in size, complexity andnumber. Developing sound intergovern-mental relations and an appropriate fiscalarchitecture in metropolitan areas presentdaunt ing cha l lenges because manydifferent governments and public enterprisesare typically involved in service provision in

1. United NationsDepartment ofEconomic and SocialAffairs PopulationDivision


15

a metropolitan area. Despite the challenges,some analysts believe that local governmentscan play an important role in meeting thedemands of urbanization and metropolitangovernance.

Third, many countries around the globeface a considerable backlog of infrastruc-ture demands and anticipate theemergence of new ones, in great partbecause of urbanisation. Addressing thechallenges of urbanization and the growthof large metropolitan areas lackingadequate basic infrastructure will requiresubstantial investments in the comingdecades, often in sectors for which localgovernments have major responsibility.According to one estimate, investment ininfrastructure and basic services in theorder of 200 billion USD annually will berequired over the next 25 years to meetthese shortfalls.2 The demand will begreatest in developing countries, but advancedindustrialized countries must also invest to

deal with their aging populations andinfrastructure. Special local investments inresilient infrastructure will be needed inmany countries that face a growing risk ofnatural disasters.

Finally, the global financial and economiccrisis that began in 2008 is deeper thananything experienced since WWII in termsof employment, income, and financial wealthlosses.3 The crisis has distressed practicallyall central governments around the worldand it has affected most local governmentsin some ways as well. At the same time,there is considerable diversity in how localgovernments across different countries havefared. While some local governments haveseen their funding cut and all types ofexpenditures reduced, others have actuallyexperienced a growth in funding and haveincreased certain types of expenditures. Insome countries, local governments may be ableto play a significant role in mitigating theeffects of the global financial crisis.

2. World Bank (2005)estimated theinvestment needs inpublic infrastructure indeveloping countries,amounting to 600billion USD per yearover the next 25 fiveyears. However, thesefigures include allpublic infrastructures,whether national(energy,communications andinformationtechnology, transport;water and sanitation,etc.) or urban (localroads, local watersupply, and sanitation,waste disposals,schools, streetlightning...).(WorldBank. 2005.“Infrastructure and theWorld Bank: Aprogress report”, TheWorld Bank). TheUCLG Committee onLocal Financeestimated one third ofthis amount, i.e. 0.4percent of World GDP,needs to be channeledto urban infrastructure(UCLG. 2007. UCLGPolicy Paper on LocalFinance, UCLG)

3. UCLG, The Impact ofthe Crisis on LocalGovernments. (UCLG.2009. “China”, TheImpact of the GlobalCrisis on LocalGovernment. UCLG).

Second Global Report on Decentralization and Local Democracy.GOLD 2010

Note: Territory size shows the proportion of all extra people that will start living in urban areas between 2002 and 2015, ineach territory.

Source: World Mapper; City Growth (2005)

Figure 1.1: Expected Urban Growth Between 2002-2015



The Structure and Requirements ofLocal Government Finance Systems

If local governments are to realize theirconsiderable potential in public service andhelp to effectively deal with prevailing andemerging challenges and crises, they mustoperate under a legal framework, institutionalstructures, and procedures that meet certainrequirements. Some of these are explicitlyfiscal in nature, while others relate to thelarger political and institutional context inwhich local governments operate.

Core Elements of the Fiscal System

Local governments are typically assigned arange of service delivery and other keyfunctions by constitutional or legal provisions.It is generally accepted that these functionsshould be appropriate in terms of theirrelevance for localities and their suitabilityfor local implementation. There is alsogeneral agreement that clarity of functionalassignment is important to ensure thatlocal governments and their constituentshave a consistent understanding of localresponsibilities. Sufficient expenditure autonomyis considered critical so local governmentscan respond to local needs.

Local governments also need access to funds todischarge their functions and to meet evolvingexpectations of their constituents. Centralgovernments have a comparative advantage inrevenue generation, so a major portion of localresources is often derived from shared taxesand intergovernmental transfers. Transfers canbe unconditional or conditional, and they maybe used for recurrent and capital spending.Transfers should be funded by a stable andpredictable pool of resources and allocated byappropriate criteria or formulae. The balancebetween conditional and unconditionaltransfers may vary in different contexts, butsome unrestricted resources allow localgovernments to exercise the autonomy that is

central to their own comparative advantage inservice delivery.

Beyond transfers, local governments need tohave dedicated sources of revenue overwhich they must have a degree ofdiscretionary control. This allows for thecreation of a tangible linkage between thecosts and benefits of local service delivery,and it also provides local governments with ameans to increase the amount of revenuethey can raise independently to finance therange and level of services demanded fromthem. Local own-source revenues may takethe form of taxes on appropriate bases, orthey may be non-tax revenues, such as userfees and charges, and license and registrationfees, among others.

Finally, as intergovernmental fiscal systemsmature, local governments need to haveadequate access to infrastructure finance.Some development spending can be fundedwith transfers, but eventually localgovernments, particularly in urban areas,need access to the capital market, whetherdirectly or, in less advanced systems,through intermediary institutions with somegovernment involvement. Local borrowing,however, needs to be governed by a suitableframework and adequate fiscal responsibilitysafeguards.

Non-Fiscal System Requirements

The focus of this report is on finance, butother aspects of intergovernmental systemscovered in GOLD I are critical to ensuringeffective local governments. As noted above,accountability is central to attaining thepotential benefits of decentralization. This isoften framed as the political dimension ofdecentralization, and the mainstream “goldstandard” for accountability is regulardemocratic elections. Not all countries haveor want free and competitive local elections,however, and other mechanisms that allow


17

for citizen engagement with local govern-ments —public access to information, feed-back, and complaint mechanisms, etc.— canimprove accountability. Moreover, local elec-tions alone are a rather blunt accountabilityinstrument, and non-electoral mechanismscan play a critical role in enhancing localaccountability even where elections are wellestablished.

Institutional dimensions of decentralizationare also extremely important. Local govern-ments need appropriate organizationalstructures, well-defined systems andprocedures for managing public resources,and suitable frameworks and mechanisms forengaging with other levels of government,private sector firms and nongovernmentalactors. Moreover, local governments mustpossess or be able to develop the capacityneeded to properly operate within theinstitutional framework.

Although these political and institutionalaspects of local government systems werecovered in GOLD I and are not given primaryattention in this report, their role in makingfiscal decentralization effective cannot beover-stated. Without adequate accountabilitymechanisms, appropriate operational systemsand sufficient capacity, autonomous localfiscal powers can lead to problematic ratherthan productive outcomes.

The Global Reality of Local GovernmentFinance Systems

Some countries have long had robust localfinance systems with strong development ofthe components outlined above, and manyothers have taken steps to develop systemsin recent years. At the same time, allcountries —from the most advanced industrialto the most fragile developing— face variouschallenges illustrated throughout GOLD II.Some challenges are related to weaksystem development and capacity constraints,

particularly in developing countries, ormore generally to resource shortfalls. Otherchallenges are external to the finance systembut affect demands placed on it and the way itfunctions.

System Challenges and Dilemmas

Many elements of local finance systemsoutlined above do not exist, are incomplete,or have been implemented inconsistentlywith the underlying framework, particularlyin developing countries. Fiscal frameworksrange from well to poorly designed (relativeto normative principles and contextualrealities) in terms of revenue andexpenditure assignments, correspondencebetween revenues and expenditures,transfers, subnational borrowing frameworks,etc. More broadly, overall constitutional andlegal frameworks for local government(with respect to legal status, politicalmechanisms, empowerment, administrativeand staffing structures, etc.) range frombeing well developed to barely havingbegun.

A common problem with fiscal systems isinsufficient clarity in the assignment of localgovernment expenditure responsibilities.Even where responsibilities are reasonablywell defined in more advanced systems,expenditure challenges may be created byunfunded mandates from higher levelgovernments and the lack of well developedmethodologies and practices to translateexpenditure assignment responsibilities intoquantifiable resource needs. Degrees ofautonomy in expenditure decisions alsovary widely.

An overarching challenge with service pro-vision in a multi-level government system iswhich functions should be undertaken ateach level and how levels should interact,including the metropolitan governanceissues outlined above. These are tough




decisions since there is a common trade-offbetween fiscal viability at higher levels andpolitical connectivity at lower levels. Instruggling to achieve a balance, countriesmust consider the benefits and pitfalls ofamalgamation versus division, as well asthe potential value of creating mechanismsto bridge jurisdictional fragmentation, suchas the use of special districts and/or frame-works for voluntary joint initiatives acrosslocal governments.

Progress has been made in developing taxsharing and intergovernmental transfers,but problems persist. Transfers may beinadequately or unreliably funded, and thecriteria used to allocate resources may beunclearly specified or inappropriate. Despitegrowing fiscal disparities across localities inmuch of the world, few countries usegenuine equalization grants to increaseparity in access to basic services acrosscommunities, some of which have lowrevenue capacity or high spending needsdue to demographics or other factorsbeyond their control. Where equalizationgrants exist, they may be poorly funded orundermine incentives for local tax efforts orexpenditure efficiency. Many countries alsostruggle with the right balance betweenunconditional grants, which promoteautonomy, and conditional grants, whichensure attention to national priorities.

Challenges to local revenue generation areparticularly pervasive. Although there ismore agreement about the need for strongexpenditure autonomy than there is forrevenue autonomy, some discretion is seenas necessary to promote local account-ability. Even where taxes that are widelyconsidered to be appropriate local sources,such as the property tax, are allowed, theymay not be well used. The property tax is adifficult and expensive tax to administer andtends to be especially unpopular amongtaxpayers. Even when it is relatively well

administered, its revenue potential may belimited, and other productive revenuesources have often not been assigned tolocal governments.

Only a few countries in more developed re-gions have robust systems of local govern-ment development finance. Many countriesimplement capital conditional grant pro-grams and local governments dedicate alarge share of resources to financing invest-ments, but the longer-term response to theneeds outlined above must includeenhancing responsible access to creditfor local governments. Some countrieshave successfully operated financialintermediaries for local governments, butthis approach has faced challenges and hasbeen undermined by political pressures inmany cases.

These challenges to developing robust localfinance systems, and in some countries poorlocal government performance, have led toinstances of backtracking on decentralization.Since the publication of GOLD I there hasbeen an emerging recentralization trend insome countries around the world.Disappointing performance in emergingsystems, however, may result from expectingtoo much too quickly from nascent localgovernments and failing to adequatelysupport building their capacity to fulfill theroles expected of them.

External Challenges

A number of major phenomena outlinedearlier —natural resource crises (environ-mental, energy, food security), urbaniza-tion, infrastructure shortfalls, and the globalfinancial crisis— were framed as problemsthat local governments could contribute toalleviating. At the same time, it is impor-tant to recognize that they also pose considera-ble challenges to local governments.Dealing with them effectively will require


19

more resources, greater technical expertise,and considerable ability to negotiate com-plex issues with a range of interestedparties with varying degrees of power.Thus, the extent to which local governmentscould take action to respond to theseserious global threats to developmentdepends on the extent to which they areproperly equipped and supported to do so.

In this regard, it is important to note that somecentral governments seem not to understandtheir own critical role in providing anenvironment conducive to local governmentaction intended to meet these challenges. Onthe contrary, the tendency towardsrecentralization in some countries seems tohave been exacerbated by the response to theworld financial and economic crisis. Centralgovernments in a number of African, LatinAmerican, and Eurasian countries haveadopted policies of unilaterally interrupting thedisbursement of revenue sharing and othertransfers. In other cases, central governmentshave increased control over funding allocationsor mandates over how local governments mustspend resources.

The Diversity of Experience

Although local government finance is im-portant in many countries and some basiccommonalities and challenges as outlinedabove can be identified, it is important tokeep in mind that there are also importantdifferences across regions and countries. Asreported in GOLD I, there is great variationaround the world in how local governmentsare structured and empowered. GOLD IIdemonstrates in more detail that there is alsoextraordinary variety in how the fiscalarchitecture of local government is organizedand performs both across different regionsand among countries within each region.

Historical roots and trajectories have a lotto do with how local government systems

are structured and the roles they play. Inmuch of Latin America, for example, theinfluence of centralized colonial regimescan be seen, particularly on fiscal matters.At the same time, some large countries,such as Argentina and Brazil, have longtraditions of provincial governance, andlocal governments have, with variousinterruptions, been more important inLatin America than in other non-OECDregions. In the Middle East and WesternAsia, the strong influence of the OttomanEmpire can be seen in still heavily cen-tralized systems, fiscal and beyond, thatprevail throughout countries in the region.

A number of regions exhibit considerableintraregional variation. Although centralizedsystems have dominated the Asia-Pacificregion, diversity is evident. Some countrieswere colonized by Britain (e.g. India andMalaysia) and others by France (e.g.Cambodia and Vietnam), with a few outlierinfluences (e.g. Spain/U.S. in the Philip-pines and the Netherlands in Indonesia).Some countries, such as China and Thai-land, were never colonized for extendedperiods. Their systems draw on their own aswell as borrowed traditions. Australia andNew Zealand differ from other Asia-Pacificcountries in that both were British colonieswhere descendants of colonists stayed andco-existed with indigenous people, insti-tutionalizing but adapting colonial gover-nance conventions. These variousinfluences have resulted in a great varietyof local government structures and fiscalsystems in the region.

The countries in Eurasia started in funda-mentally similar positions with the sameinitial system of administration and fiscalarchitectures inherited from the formerSoviet Union, but they have opted toreorganize their local governance systemsquite differently. In the Africa region, thereis a stark contrast between the centralized




local administration traditions of formerFrench colonies in West and Central Africaand the strong local government traditionsleft by the British in East and SouthernAfrica, although the latter were oftenweakened in the post colonial period.There has been considerable effort todecentralize and strengthen local govern-ments across the region, and in manycountries there is now a mixture of thelocal administration and local governmenttraditions.

In Europe, many countries show richdecentralization experiences with stronginstitutional underpinnings, but the systemsvary considerably and face significantpolicy challenges. In North America,Canadian and U.S. local governments playan important role in the public sector,but they are creatures of intermediategovernments (provinces or states) ratherthan the national government. This leads tointernal diversity since each province/statehas separate local government legislation, asituation which also occurs in someother countries in other regions, such asArgentina, Australia and India. In theU.S., there is a particularly complex localgovernment structure with thousands ofcounties, and tens of thousands of sub-county general-purpose governments andspecial-purpose districts.

These governance traditions across regions,of course, have been subject to evolvingpolitical and economic forces over the yearsthat have resulted in many changes to thesystems, including to the fiscal architecture.At the same time, the influence of thesetraditions persists in both obvious andsubtle ways. In moving forward with futurereforms, it is important to be aware of thenature and strength of this influence andwhat it implies for the pursuit of viableand sustainable local government financereforms.

Summary of the GOLD II Mission and Organization of the Report

It is not too dramatic to state that localgovernment finance systems around theworld are currently at a crossroads. Efforts todecentralize and more fully empower localgovernments have been prominent, but theyhave encountered a variety of challenges,both relatively universal and fairly specificto particular regions and countries. Theoverall situation has been exacerbated bythe emergence of a number of prominentand consequential crises –environmental,economic, and financial– in recent years.

Times of crisis present an opportunity toreflect on how local government finance sys-tems work and how they can be improved.There is, of course, the possibility ofoverreacting during crises and makingshortsighted and ultimately problematicdecisions, both small and large, for the sakeof getting through difficult times. Suchopportunistic reforms may alleviate immediateproblems but are likely to ultimately underminethe ability of local governments to meet theirresponsibilities in effective and sustainableways.

Moving forward with the reform of localgovernment finance requires systematicanalysis of the positive and negativeaspects of current intergovernmental sys-tems, as well as careful consideration of howlocal governments can be empowered andsupported to play a more productive role.This introduction has broadly outlined someof the key issues and options that need tobe considered on this front. The rest of thisreport considers these issues and optionsmore deeply at the regional and global levels.

The following chapters focus on specific UCLGregions: Africa (Chapter 2), Asia-Pacific(Chapter 3), Eurasia (Chapter 4), Europe(Chapter 5), Latin America (Chapter 6),


21

Middle East and Western Asia (Chapter 7) andNorth America (Chapter 8). Each of thesechapters reviews the local governmentfinance systems in the target region and thecontexts in which they are operating. Thechapters outline intergovernmental systems,with a focus on describing and analyzing thefiscal aspects. Positive and problematicfeatures of local government finance aresummarized, and specific opportunities andchallenges are highlighted. Finally, eachregional chapter closes with a summary of themain issues and regional specific policyrecommendations and issues for furtherinvestigation.

It is important to note that the regionscovered in these chapters vary in terms of the

number of countries included and the leveland quality of information available. Thus,some chapters cover a greater proportion ofcountries than others, and in some casesmore attention was given to countries forwhich better information was available.

Following the regional chapters, Chapter 9focuses on the special circumstances andchallenges of Metropolitan Areas acrossregions. Finally, the report concludes with anoverall summary of key findings in Chapter10. The final chapter also provides generaladvice for broad-based policy reforms andfuture work needed to more fully understandlocal government finance and to develop andsupport more detailed reforms both globallyand in specifications and countries.



0w2010 02 Africa DEFcarta ang 15/9/11 23:02 Página 22

2. AFRICA

FFRRAANN CC,,OOIISS YYAATTTTAA,,

IINNDDEEPPEENNDDEENNTT RREESSEEAARRCCHHEERR,, NNIIGGEERR

FFRRAANN CC,,OOIISS VVAAIILLLLAANNCCOOUURRTT,,

UUNNIIVVEERRSSIITTÉÉ DDEE MMOONNTTRRÉÉAALL,, CCAANNAADDAA


25

T he majority of African countries havemade progress with decentralization in

recent years, particularly at the policy and po-litical level. As indicated in the GOLD I report,local governments today are an importantpart of the institutional landscape. They aretaking on more and more responsibilities andtheir actions have progressively become anincreasingly important part of the day-to-daylives of citizens. Reforms to reinforce localgovernments continue to be undertaken andare expected in a number of different coun-tries; such as strengthening their financial re-sources and gaining freedom in raising andusing these resources.

However, countries have progressed atdifferent rates and rarely has the path beenlinear. In Gabon, for example, the decen-tralization process began in 1996, but transferof responsibilities did not begin until 2009. Inmany countries this process is also facingdifficulties. Globally, the share of publicexpenditure managed by local governmentremains low. The implementation ofdecentralization policies is half hearted andcertain national governments are taking stepsbackward. In Mali and Burkina Faso, theprinciple of the simultaneous transfer ofresponsibilities and resources, was recentlyquestioned. In Uganda, the management oflocal personnel has been recentralized andthe status of the capital city has beenmodified in order to create a metropolitanauthority, whose governing body is appointedby the central government. In Malawi andTogo, local elections have been suspended forlong periods and centrally appointed officialsmanage local governments (the practice since1999 in Malawi). In some countries, the shareof the revenues of local government comingfrom national resources has decreased inrecent years (Benin, Côte d’Ivoire, Mali,Senegal, Tanzania, Togo and Uganda).

The African continent is currently faced withrapid urbanisation rates (3.2 percent per year

on average for the period of 2005-2010,UNFPA 2007). However the development of alarge metropolis and medium size cities isaccompanied, to varying degrees, by seriousservice deficits and slum expansion. In 2007,the United Nations estimated that 72 percentof the urban population of sub-Saharan Africalived in slums (ibid). Improving the quality oflife for these populations, and specificallyimproving access to basic services isindispensable to ensure economic develop-ment and to reinforce social and politicalstability. These challenges must be met bylocal authorities, working hand in hand withnational leaders across Africa.

To respond to this challenge, local authoritiesneed appropriate and sufficient resources andresponsibilities. In this chapter, the situation oflocal finances is examined for a sample ofAfrican countries, including Algeria, Cameroon,Kenya, Mauritania, Morocco, South Africa,Zambia and Zimbabwe. The chapter is dividedinto five main sections: the macroeconomiccontext; the description of local financialsystems; an analysis of the main problems offiscal decentralization; the impact of the 2008-2009 financial crisis and finally recommenda-tions for improving the financial circumstancesof local governments in Africa.1

Macroeconomic and Financial Framework ofFiscal Decentralization

Fiscal decentralization in Africa, like anywhereelse in the world, is strongly conditioned bythe macroeconomic and financial situation ofcountries. The overall low level of publicspending has a direct negative impact on thefunding of local management and on fiscaldecentralization.

Africa has seen renewed high growth in thelast decade; an average GDP growth rate ofaround 5.5 percent in recent years, with areduction to 2 percent in 2009 due to theeconomic and financial crisis. According to the

1. The Africa chapter ofthe GOLD II reportdraws primarily ondata from 24 countriescollected by UCLG,completed with datafrom recent researchand publications.

Second Global Report on Decentralization and Local DemocracyGOLD 2010


AAFFRRIICCAAUnited Cities and Local Governments26

African Development Bank2, this growth–prior to the world financial crisis— wasmainly due to the increase in oil exports andthe oil price shock that pushed up the growthrates of oil-producing countries like Angola,Nigeria, Sudan, and Chad. Nevertheless,non-oil producing countries have alsoexperienced high median growth rates.

The economic recovery, however, continues tobe tenuous. Development and structuralgrowth is extremely vulnerable to natural andexternal shocks over which African countrieshave no control. In addition, Africa is strugglingto increase its share of international trade. Justbefore the implementation of the structural ad-justment programs, designed to bring Africa inline with the rest of the world, 12 percent ofworld trade was from the African region. Thisfigure slipped to 9 percent in the 1990s, and islikely lower at present.

In addition, the region continues to sufferfrom poverty. Recent projections suggest thepoverty rate in Africa will exceed 38 percent in2015, which is significantly higher than the22.3 percent Millennium Development Goals(MDGs) target. To achieve the MDGs goal ofhalving poverty by 2015, the continent needsto increase and sustain annual growth atbetween 7 percent and 8 percent.

Fiscal decentralization is suffering from thedifficult financial situation most Africancountries are in today. Although revenuegeneration by African countries has beengradually improving since the early 2000s,following four decades of stagnation, thefuture outlook is not bright. Trade liberaliza-tion has had a direct impact on public finan-ces, especially on tax revenues.3

Even with the completion of fiscal transition, ascountries move to less reliance on trade taxesand more reliance on internally generatedrevenue, analysts are questioning whetherthe expected incomes from an extended

implementation of the value added tax (VAT)can replace reductions in customs duties. Thepossible revenue loss would impact publicspending and consequently social programs,especially in health and education. In Congo,for example, the simulated loss of customsrevenue from application of the EconomicPartnership Agreement (EPA) with theEuropean Union is equivalent to total publicexpenditure on education.4

This change in public finances has occurredalong with the further deterioration of fiscalbalances.5 Despite their success in mobilizingresources, African countries are struggling tocover their primary expenditure, as well astheir contribution to investment and theinterest expenses of domestic and foreigndebt. In three regional groups out of five(Central Africa, West Africa and North Africa)the average fiscal balance has moved fromdeficit to surplus since 2005 as a result ofbetter management of public finances andtaxation. The improved situation in NorthAfrica was due to oil-exporting countries withhigher oil prices and increased production.These countries posted a surplus budgetbalance of approximately 5 percent of GDP.However, countries such as Egypt, Sudan,and Tunisia have seen their deficits increase.The budget position of Central Africa im-proved as a result of the recovery of itsprincipal economy, Cameroon, and the oilrevenue from Gabon, Congo, and EquatorialGuinea. In West Africa, the reportedimprovement corresponds to the implemen-tation of macroeconomic reforms, even if thesurplus is due to Nigeria's oil resources. In theother two regions (East and Austral Africa), arecovery seems to be taking more time.

Local Governments Structures andLocal Financial System in Africa

In this broader context, local finance systemshave received growing interest across Africa.In West and Central Africa piloting mechanisms

2. African DevelopmentBank (AfDB),Development Centre of the Organisation for Economic Co-Operation andDevelopment, andUnited NationsEconomic Commissionfor Africa: AfricanEconomic Outlook.

3. Free trade agreementshave generated majorconsequences: At thebilateral level, underpartnershipagreements betweenAfrican-Caribbean-Pacific countries andthe European Union,removal of duties onimports would reducetotal tax revenues inthe Gambia and CapeVerde by 20 percentand in Ghana, Rwanda,Mauritius and Senegalby 11 percent. In theWest African Economicand Monetary Union(WAEMU), gradualremoval of customsduties on intra-community trade hashad seriousconsequences.Between 1 July 1996and 30 September2006, member Stateswere paid an amountequal to 143 billionUSD as compensation.Lastly, global tradeliberalization hassharply reducedcustoms duties. Theseare the foundation ofpublic finance in Africa,accounting for anaverage of 14.7percent of taxrevenues in WestAfrica –from 5 percentfor Nigeria to 34percent for the Gambia(Busse et al, 2004) –and 16.4 percent inCentral Africancountries –from 2.4percent EquatorialGuinea to 24 percentfor the DemocraticRepublic of the Congo.Customs revenues forcertain countries wereoften as high as 30percent in the early2000s.

4. Source: BeyondEconomic PartnershipAgreements in Africa.Laurence Hinkle,Mombert Hoppe andRichard Newfarmer

5. African DevelopmentBank (AfDB),Development Centre ofthe Organisation forEconomic Co-Operation andDevelopment (OECD)and the United NationsEconomic Commissionfor Africa: AfricanEconomic Outlook.


27

for decentralization and local finance havebeen created or reactivated recently, forexample: the National Committee of LocalFinance (CONAFIL) in Benin and Côte D’Ivoire;and the National Council on Decentralizationand Inter-Ministerial Committee on LocalServices in Cameroon (2010); and the BurkinaFaso’s National Decentralization Conferencein 2010 which focused on financing decen-tralization. In 2009, Gabon initiated transfer ofresponsibilities 14 years after the firstdecentralization laws were passed. InMauritania, a new law for local government isbeing studied. The project would introduceregions as a new unit of local governance andexpand local government roles. However, inMali, the law promising that transfers ofresponsibility would be accompanied withcorresponding financial resources was revised.

In South Africa, the process of local financemodernisation has been underway since 2003,with the Municipal Finance Management Act,which revised budgetary and accountingprocesses, the internal auditing and controlsystem, and the procurement and biddingprocesses. In other countries of Southern andEastern Africa, it should also be noted, that theprocesses of legislative and constitutionalreform are underway. In Kenya, reformsapproved in August 2010 include: aconstitutional status for local governments(counties), direct election of mayors and theexecutive and giving them control over muni-cipal budgets, as well as clarification of thefunctions and resources of local governments,and an assignment of 15 percent of thenational revenues. However the future ofexisting local authorities in urban areas remainunclear. In Uganda, a new program of financialmanagement and transparency (FinancialManagement and Accountability Programme –FINMAP) has been in place for the last 3 yearsto strengthen the fiscal decentralizationstrategy (FDS). In 2010, Zambia approvedtheir Plan for Decentralisation Implementationand restructuring of the system of transfers to

make fund allocation more transparent andpredictable. In Malawi, the Parliament madeplans to reduce the number of wards in 2010.The government has also created a nationalcommission on local finance and increasedbetween 2004 and 2010 the sectors trans-ferred to local governments from three toseven (initially agriculture, education andhealth; now also forests, water and envi-ronment, and gender equality).

In North Africa, where decentralization isadvancing more slowly, some important re-forms are underway. In Morocco, the reform ofthe “Charte Communale” increased theflexibility of the fiscal and accounting controlsconstraining local government, reduced theveto power of central over local governmentand increased local responsibilities. Amulti–year program is also being introduced tostrengthen local government autonomy inpricing local services and to formalize theagreement between public decision makers andcentrally appointed accountants, includingmanagement control indicators fixed within thecontract. The King recently announced thatboth the role and autonomy of regions would bereformed. In Algeria, the government ispreparing a municipal code that would give localgovernments more flexibility to manage theirbudgets and revenues in the future. It alsoproposes both solidarity funds to supportequalization and systems of citizen consultation(local referendums).

The following section deepens the discussionof five important aspects of fiscaldecentralization: State structure and terri-torial organisation, local government respon-sibilities and expenditures, revenues,transfers, and access to loans.

States and Territorial Organisation

There are five federal countries in Africa:Comoros, Ethiopia, Nigeria, Sudan andTanzania (Tanzania is more of a bipartite



AAFFRRIICCAAUnited Cities and Local Governments28Ta

ble 2

.1:

Num

ber o

f Dec

entra

lised

Ent

ities b

y Lev

el - P

opula

tion

Indic

ator

s for

27 A

frica

n Co

untri

es, 2

008

Coun

tryLoc

al leve

lInt

ermed

iate

Interm

ediat

e Ave

rage lo

cal po

pulat

ionPop

ulatio

n of ca

pital

% urba

nTot

al coun

try po

pulatio

n lev

el 1lev

el 2(in

thousa

nds)

(in tho

usand

s)po

pulat

ion

(in tho

usand

s)

Alger

ia15

4148

wila

yas 1

212,

947

6534

,178

Beni

n77

120

223.

5 41

8,792

Burk

ina Fa

so35

145

prov

ince

s13

regio

ns43

1,475

2015

,746

Cam

eroo

n36

0 mun

icipa

lities

and 1

4 citi

es1,

248

5718

,879

Côte

d’Ivo

ire71

8 mun

icipa

lities

and

56 de

partm

ents

19 re

gions

20

3,80

249

20,61

72 d

istric

ts (A

bidjan

and Y

amou

ssou

kro)

Egyp

t46

41 vi

llage

s and

220 t

owns

433 c

entre

s29

gove

rnor

ates

115

11,10

043

83,0

83Et

hiopia

556 w

ored

as68

9 + 2

city r

egion

s15

34,

585

1785

,237

Gabo

n52

49

depa

rtmen

ts29

540,

818

851,

515

Ghan

a13

816

91,

661 (

2001

)50

23,3

82Ke

nya

175 (

1 city

, 44 m

unici

paliti

es an

d 130

rural

coun

cils)

47 co

untie

s221

83,1

0022

39,0

03Ma

lawi

40 (2

8 rur

al dis

trict

s, 8 u

rban

town

s and

4 cit

ies)

330

732

1914

,269

Mali

703

49 ci

rcles

8 reg

ions

181,4

9432

12,70

0+ D

istric

t of B

amak

oMa

urita

nia21

614

673

413,

200

Moro

cco

1503

( 221

urba

n and

1282

rura

l) 16

regio

ns21

627

5634

,859

Moza

mbiq

ue43

477

1,09

937

21,6

69Na

mibi

a54

13 re

gions

3325

137

2,109

Nige

r26

536

855

215

1615

,306

Nige

ria77

436

stat

es +

fede

ral

115

778.

548

149,

229

distri

ct of

Abuja

Rwan

da31

(30 d

istric

ts an

d Kiga

li city

)31

080

018

10,47

3Se

nega

l54

314

regio

ns29

2,60

442

13,71

2So

uth A

frica

(SA)

226 l

ocal

mun

icipa

lities

8 m

etro

, 44 m

unici

pal d

istric

ts9 p

rovin

ces

172

2,34

561

49,0

52Ta

nzan

ia25

citie

s, 97

rura

l dist

ricts

2,

456

2541

,049

(108

town

ships

and 1

0571

villa

ges)

Togo

354 m

unici

palit

ies30

pref

ectu

res

5 reg

ions

1914

5242

6,02

0Tu

nisia

264

24 go

vern

ates

3999

267

10,4

86Ug

anda

1010

divis

ions,

97 di

stric

ts an

d mun

icipa

lities

301,

533

1332

,370

sub-

coun

ties a

nd to

wns

Zam

bia72

dist

ricts

35

3,00

035

11,8

63(4

citie

s, 14

mun

icipa

lities

and 5

4 dist

ricts

)Zim

babw

e90

129

1572

3711

,393

Sour

ce: Com

pila

tion

by a

utho

rs, n

atio

nal d

ata

colle

cted

by

UCLG

pro

vide

d by

the

loca

l gov

ernm

ent a

ssoc

iatio

ns. O

nly

gove

rnan

ce le

vels

with

a d

irect

ly o

r in

dire

ctly

ele

cted

cou

ncil

or a

ssem

bly

are

liste

d he

re. F

or th

e da

ta o

n po

pula

tion:

City

Pop

ulat

ion

(nat

iona

l pop

ulat

ion:

200

9 es

timat

ions

, urb

an p

opul

atio

n: 2

008

estim

atio

ns).

Tog

o: p

lann

ed u

nder

new

legi

slat

ion

1. T

he E

xecu

tive,

the

Wal

i or G

over

nors

, are

nom

inat

ed b

y th

e Cen

tral

Gov

ernm

ent.

2. A

s sp

ecifi

ed in

201

0 Con

stitu

tion.


29

confederation). South Africa and theDemocratic Republic of the Congo (DRC) arequasi-federal States. Other countries areunitary States. Table 2.1 presents an over-view of the data on the territorial organisationof the countries analysed in this chapter.

Great diversity can be observed in territorialorganization, as far as tiers and range of localauthorities are concerned. This diversitycomplicates comparison across countries.Since 2005, an increase in the number of localgovernments has occurred in some Africancountries (Burkina Faso, Côte d’Ivoire,Ghana, Kenya, Mozambique and Uganda).

Responsibilities and Expenditure

Overall, African local governments have li-mited spending autonomy. The division ofresponsibilities and expenditure between localauthorities and central governments is oftenunclear. The responsibilities of localgovernments are written into the constitutionor the laws of most of the African countriesanalyzed, except for Tunisia and Niger, butthere is a huge divergence between de jureand de facto practice. This is the case, forexample, in Côte d’Ivoire, Egypt, Ethiopia,Rwanda, and Algeria. The responsibilitiesgranted to local government are often sharedamong several local levels and the state.

In all countries, the higher levels of govern-ment (region/provinces or centre) exert acontrolling power over local government. Thisis exercised through budget rules andcirculars that specify the various percentagesto use when preparing the budget (forexample, not more than 50 percent on payrollexpenses or the earmarking of at least 40percent of a specific transfer for investment).Higher levels may also have veto powerduring budget review.

Budgetary controls, both prior to and postapproval, are not uncommon. Budget

approval is often dependent on legal com-pliance with mandates on revenues, ex-penditures, and budget balance. Localgovernments in countries such as Ghana,Gabon, Egypt, Tunisia, Rwanda, and Zambiaare strictly controlled by higher levels, withreviews of the budget structure as well ascomponent budget lines. For countries likeKenya, Uganda, and Burkina Faso, thereviews concern only the budget structure.In Malawi, South Africa, Tanzania, andAlgeria they focus principally on budgetlines. In Morocco, the recent organisationalreforms of local finance (2009) and localaccounting (2010) have reduced vetopowers. In Senegal, the individual decisionstaken by local governments are notsubmitted for prior approval to a higherauthority.

Finally, the timeframes for budget approvalscan be long in many countries. InFrancophone West Africa, for example, Maliand Niger, legal approval timeframes aregenerally exceeded, because it takes so longfor the oversight authority to providecomments.

Another indicator of the autonomy of localgovernments is whether or not they are freeto hire and dismiss personnel. Only about athird of the countries analyzed give their localgovernments full control over these decisions:Benin, Gabon, Ethiopia, Mozambique,Morocco, Rwanda, Tanzania and South Africa.In Ghana, Mali, Malawi and Algeria, decisionsto hire and dismiss local-level staff are takenjointly by central and local government. Forall other countries, local employees are eitherhired after approval of or directly by thecentral government, as is the case in bothEgypt and Tunisia. In Zambia localgovernments have full control over themanagement of their staff (includingsalaries), but central government caninterfere in these processes. In Malawi andKenya, local authorities can hire and fire their




subordinate staff, whereas only the centralgovernment can appoint professional staff. Insome countries, such as Mauritania, theauthority to hire staff is used as a pre-electionvote-catching technique, which increases thewage bill (Yahya Ould Kebd, 2010).

The procedure for assigning powers and res-ponsibilities to local governments varies fromone country to another. Most countries transferspecific blocks of responsibility (e.g. SouthAfrica and Morocco), while countries like Nigerand Cameroon grant general powers andresponsibilities to local governments. Somecountries, such as Tanzania, provide a detailedlist of mandatory and optional functions.

There are three groups of expenditure res-ponsibility in the countries analyzed here.The first group of basic tasks common toall countries comprises of the maintenanceof small equipment, roads, street lighting,waste collection, recreational activities(culture, leisure, and sports), as well asurban and residential planning. The secondgroup includes preschool and primaryschool education and basic health services,as well as transport and environmentalprotection. These occur in Algeria, Benin,Guinea, Morocco, Namibia, Niger andZambia or in some municipalities (largeurban municipalities in Kenya, for example,or regional administrative centres). In afew countries, such as South Africa, Ugan-da and Zimbabwe, the third moreadvanced functions, including thepromotion of local economic developmentand employment, energy distribution,police and security, and secondaryeducation may also be assigned.

The sharing of powers and responsibilitiesamong the various levels is very common incountries such as Mali, Tanzania, Zambia andZimbabwe. For example, local governmentsare generally responsible only formaintenance of school infrastructure. Sche-

duling, investment, and staff salaries aremanaged by different ministries. Otherfactors of note include:

• Responsibility for water and electricityremains significantly different betweenAnglophone and Francophone countries. InSouth Africa and Ghana, water boards orcompanies often provide them; inFrancophone countries they are more likelygoverned under a national enterprise. InKenya, former municipal enterprises haverecently formed regional groups.

• In countries with several subnational levels,the correct scale for service delivery isunder debate. For example, in Côte d’Ivoireand South Africa, the search for economiesof scale and externalities has led lawmakersto assign some services such as thedistribution of water and the production ofenergy to higher decentralized levelsand/or the State.

• Concerns for redistribution and social equityhave led lawmakers in most countries togive central government control over suchareas as welfare and housing. For othercountries, these are assigned to and sharedbetween the central and local levels. InCôte D’Ivoire, for example, socialassistance is a municipal responsibility.

Given the common divergence between dejure and de facto functional assignment, it ishelpful to use other indicators ofdecentralization. Figure 2.1 examines theratio of local to total public expenditures.6

The (non-weighted) average for all coun-tries analyzed is 8.3 percent, but thedifferences between countries areimportant. The highest ratio (17-23percent) is in Uganda, Rwanda, and SouthAfrica. However, the ratio in most countries(16 countries) is lower than 10 percent, oreven under 5 percent (11 countries). Inother words, the contribution of African local

6. Local publicexpenditure includesthe intermediate level(governorate, region,department, etc.).The data comes fromrecords compiled byUCLG and completedby data from otherworks and studies.


31Second Global Report on Decentralization and Local DemocracyGOLD 2010

0%

5%

10%

15%

20%

25%

30%

Benin

Burkin

a Fa

so

Cote d'Ivoire

Egypt

Gab

on

Gha

na

Kenya

Malaw

i Mali

Mau

ritan

ia

Mor

occo

Moz

ambiqu

e

Niger

ia

Rwanda

Seneg

al

South A

frica

Tanz

ania

Togo

Tunisia

Ugand

a

Zambia

Zimba

bwe

Africa

Niger

4%

2%

6%

11%

3%

10%

5%

9%

3%

1%

8%

2% 2%

15%

3%

18% 16%

2%

10%

4%

9% 8%

23% 23%

Source: Calculation by authors with country data collected by UCLG and completed by data from other studies (see bibliography).

Figure 2.1. Local Expenditure Share of Total Public Expenditure in 23 African countries, 2007

0%

5%

10%

15%

20%

25%

30%

Benin

Burkin

a Fa

so

Cote d'Ivoire

Egypt

Gab

on

Gha

na

Kenya

Malaw

i Mali

Mau

ritan

ia

Mor

occo

Moz

ambiqu

e

Niger

ia

Rwanda

Seneg

al

South A

frica

Tanz

ania

Togo

Tunisia

Ugand

a

Zambia

Zimba

bwe

Africa

4%

2%

7%

1%

4%

7%

3%

7%

4%

1%

11%

1%

15%

Niger

4%

21%

4%

18%

24%

2%

8%

4% 5%

10%

8%

Source: Calculation by authors with national country data collected by UCLG and completed by data from other works andstudies (see bibliography).

Figure 2.2. The Weight of Local Authorities in Total Public Revenue, Africa, 23 countries, 2007



authorities to public expenditure is generallyfar below that of those in Europe (where theaverage is 25 percent), and this does noteven consider the level of autonomyexercised in making these expenditures.

Total Resources and Own-Source Resources

In Africa, there has been less decentralizationof revenues than expenditures. Figure 2.2reports the local share of total governmentrevenues for 23 countries. In the majority ofcountries (17), local government revenuesrepresent less than 10 percent of publicrevenue and in 13 countries, less than 5percent. Only Tanzania and Rwanda exceeds20 percent, followed by South Africa andNigeria (15-18 percent).

The numbers in Figure 2.2 include autono-mous revenue as well intergovernmentaltransfers. Figure 2.3 breaks down localrevenues into sources, showing that in themajority of countries, about half of local

government revenues come from inter-governmental transfers while the other halfcome from local taxes. However, the situationdiffers substantially across countries. In somecases transfers provide the bulk of localrevenues, 85-90+ percent in extreme caseslike Tanzania and Uganda. In Egypt, localgovernments fully depend on shared taxesand transfers.

On the other side there are countries wheretaxes, fees, and charges dominate localbudgets: Niger, Senegal and Togo, followedby South Africa. Here own revenues provideapproximately two thirds of local revenue.South Africa falls in this range, if sharedtaxes are included. However in somecountries (Niger and Togo for example),the predominance of own revenue is notnecessarily a sign of autonomy. Morocco andKenya fall slightly below a 50/50 split in theirbalance between own revenue and transfers,and below this are Mauritania, Tunisia,Mozambique, and Malawi.

Source: Calculation by authors with national country data collected by UCLG and completed by data from other works andstudies.

Figure 2.3: Structure of Local Revenues, Africa, 15 countries, 2007


33

Own source incomes for the majority of localgovernments come from a limited number oftaxes, charges, and duties. Many countriesgenerate incomes from local services, butthese are often insufficient to cover costs.Principal sources include: property tax (bothbuilt and non-built land), residential tax,licenses and various charges (for business,markets, construction permits, bars,gambling, shows, advertising, hotels) andservice fees (water, sanitation, wastecollection). Vehicle taxes exist in certaincountries (Zimbabwe), but they are oftenshared with the center (Algeria, Côte D’Ivoire,Egypt, Morocco). Local governments in fewcountries may levy business taxes (Algeria,Morocco, Tunisia, Togo) or non-petroleumresource extraction (mining, forestry) shareswith the central governments of Morocco,Tanzania, and Gabon.

More unusual are taxes on the informal sector(Burkina Faso). The participation of localgovernments in personal revenue tax is

limited, but sometimes they are included inshared taxes, over which local governmentshave no control, or transfers (Gabon,Morocco, Kenya). This is also the case forshared portions of the VAT (Algeria, Egypt,Morocco, Nigeria, Senegal and Togo) andportions of the tax on the production or exportof petrol in the principal production countries(Algeria, Cameroon, Côte D’Ivoire, Egypt andNigeria).

Even where fiscal bases are already limited,the central government may weaken themfurther. Zambia, for example, abolished thetax on agriculture (crop levies) in 2009.Uganda suspended the graduated personaltax, the main local revenue source, in 2005.Tanzania has also eliminated local taxes andreduced tax rates, and Mozambique cut taxrates by 30 percent. On the other hand,Morocco has assigned a share of the VAT, aswell as the management of various taxes tolocal and regional governments. Some areshared between the different levels (for


0%

2%

4%

6%

8%

10%

Benin

Egyp

t

Keny

a

Malaw

i

Mali

Moroc

co

Moz

ambiqu

e

Niger

Sene

gal

South Af

rica

Tanz

ania

Togo

Tunisia

Uga

nda

Africa

3%

0%

1% 2% 2%

5%

1%

2%

3%

9%

2% 1%

2%

1%

3%

Source: Calculation by authors with national country data collected by UCLG and completed by data from other works andstudies (see bibliography).

Figure 2.4: Autonomy indicator, Africa, 14 countries, 2007



example: 5 percent of local service taxes forregional governments and 95 percent to thecity).

As noted above, however, the relativeimportance of own revenue in local budgets isnot an indicator of financial autonomy if sub-national entities have little or no power todetermine the tax base, fix tax rates or collecttaxes. Figure 2.4 presents a weightedindicator of autonomy: the ratio of the ownrevenues of local governments to total public

revenue. It was obtained by multiplying thesize of the local sector (Figure 2.2) by theshare of local taxes and fees (Figure 2.3). Thisindicator enables us to consider the impact offinancial autonomy depending on the size ofthe activity covered by this autonomy. Itshows that the country with the highest levelof fiscal decentralization is South Africa.

This national measure, however, does notshow the significant differences within eachcountry among the various types of local

Country Rate Basis Tax collection Fees collection Veto of central government

Algeria No No No No Yes

Benin Yes No No No No

Burkina Faso Yes No No Yes

Cameroon Yes No No Yes Yes

Côte D’Ivoire Yes No No Yes Yes

Egypt No No No No Yes

Ethiopia Yes Yes Yes Yes No

Gabon No No No Yes No

Ghana Yes No No Yes Yes

Kenya Yes Yes Yes Yes Yes

Malawi Yes No Yes Yes No

Mali Yes No No Yes No

Mozambique No No Yes Yes (Private) No

Niger Yes No No Yes No

Rwanda Yes No Yes Yes Yes

Senegal Yes No No Yes Yes

South Africa Yes Yes Yes Yes Yes

Tanzania Yes Yes Yes Yes Yes

Tunisia No No No No Yes

Uganda Yes No Yes Yes Yes

Zambia No No Yes Yes Yes

Zimbabwe Yes Yes Yes Yes Yes

Source: Compilation by authors, from data collected at national level by UCLG and from other studies (see bibliography).

Table 2.2: Taxation Power of Local Governments for 22 African countries, 2008


35

governments, and between local govern-ments of the same type. As a rule,decentralized, intermediary tiers in non-fe-deral countries –provinces, regions or de-partments– depend more on transfers thanmunicipalities. This is partly because theywere created more recently and the mainlocal tax bases were already taken bymunicipalities. Rural (or small)municipalities are also usually moredependent on transfers than larger urbanmunicipalities.7 This is because the ruralareas have less economic activity and thetaxation of rural properties, is problematicor even prohibited.

This analysis of local resources must also beconsidered within the legislative andregulatory framework of local taxes. In mostAfrican countries, regulations on decentra-lization and/or constitutions define a scopeof specific resources for local governments.The legislative and regulatory documentsgoverning decentralization define theconditions for exercising power by localgovernments with respect to taxation andresource collection.

Table 2.2 reviews the power of the variouscountries to raise taxes. It describes whetheror not local governments are free to set taxrates or change the base of some taxes. Italso specifies whether local governments

collect local taxes and fees. Lastly, the tableindicates whether higher-tier authorities havethe right to control or veto local governmentsbudgets.

Fiscal autonomy ranges from local govern-ment having full authority without anycentral veto on local resources at oneextreme, as in Ethiopia, to a situation inwhich the centre can veto local revenuesand/or budgetary decisions, as in Egypt.Between extremes, local governments incountries like Uganda and Gabon cannotcreate new taxes or fix tax bases, but theyhave significant authority to collect revenuewithout higher interference. Mozambique isinteresting because its local governmentshave no authority to create taxes or definebases, but have great latitude in collection,and local governments in this country haveinnovated to involve the private sector in taxcollection.8 In Tunisia, tax collection is, inprinciple, under the jurisdiction of the centrebut some sixty local authorities actuallycarry out their own tax collection (BenLetaieff, 2010).

There are important distinctions betweenFrancophone and Anglophone countries asshown in Figures 2.5 and 2.6. In theformer, local governments have no powerto raise taxes. Parliamentarians and thecentral government change local tax bases

7. In Mali, the share ofrevenues obtained fromtransfers is 36 percentfor all municipalities. Thisfigure ranges from 68percent for those withfewer municipalities ofless than 10,000inhabitants to 43 percentfor those withmunicipalities of 20 to50,000 inhabitants and16 percent for Bamako.The intermediate level ofgovernment (cercles)receives 42 percent of itsfunding from transfers,while 52 percent, ofregional governmentfunding is made up oftransfers. In the samevein, in Burkina Faso,transfers account for 29percent of municipalrevenues as a whole, but67 percent of revenuesfor municipalities with apopulation less than20,000 inhabitants and18 percent for those withpopulations of more than100,000. They alsoaccount for 47 percent ofthe funding of regions. InMozambique transfersaccount for 70 percent ofthe incomes of rural localgovernment incomes,but less than 40 percentfor large cities. In SouthAfrica, medium size citiescollect 44 percent of therevenues collected bylarge cities, while rurallocal governmentscannot even reach 5percent of this figure.

8. “Privatization of markettax collection has beenwidely promoted in EastAfrica but was found tobe rather complex andrisky. Based onunrealistic fiscal potentialanalysis lump sumagreements might leadto an eitherinappropriate profit ofthe private tax collectoror a high loss for thedistrict. Profit margins forprivate collectors werefound to be mostlyinadequate, resulting inno significant increase,but occasional decreasein public revenues.Monitoring for districts isapparently difficult” (Niña Boschmann, 2009,p. 9)


Figure 2.5: Own Tax Process: Model for Francophone Africa

Parliament

Tax Base and TaxRate (or range)Defined

Tax Census

Tax Collection Tax RevenueTransferred toNational Treasury

Funds Released toLocal Governments

Local GovernmentCentral Government



and rates. In some countries, local govern-ments have the right to define the tax rate.For example, in Côte d’Ivoire localgovernments can change rates for businesstax, licences and residence tax. In Benin,local governments can fix property taxrates within the range of 4 percent to 8percent that has been determined byParliament. In Senegal, local authoritieshave the power to determine in certaincases the basis for the revenues of publicproperties or services. They have thepower of the “additional cent”, a surcharge

that can be established by deliberation.This allows local authorities to chargelevies at the local level.

In Anglophone influenced countries, such asTanzania and Zimbabwe, local councils havemore room to set local taxes or change taxrates (Figure 2.6). In some cases, suchpowers are closely monitored by the centralgovernment before or after municipal counciladoption. For example, in Kenya, all changesin property tax rates must be approved by theMinister for Local Government. In South

Box 2.1: Local/Shared Taxes and Local Government in Nigeria

In Nigeria, three dimensions of local revenue include: power over local tax, responsibilityof tax collection, and who benefits from taxes. A tax can be determined at the federal orfederated state level, collected by federal, state, or local governments and spent by anylevel. Several examples illustrate: (1) Capital gains tax is set at the federal level, collectedjointly by the federal government and states and spent by states; (2) VAT is set by thefederal government, collected by the federal government and federated states and spentby the federal government, federated states and local governments; (3) television andradio licences are set by the federal government, but collected and spent by localgovernments; (4) property tax is set at the federated state level, but collected and spentby local governments. In all cases, local governments collect their taxes and fees, even ifthey are set by federal or state governments.

Source: Akindele S. T. and OLaopa (2002): Fiscal federalism and local government finance in Nigeria.Republic of Nigeria and Olisa Agbakoba SAN & Hilary Ogbonna (2004): Local Government Administration andDevelopment in Nigeria – A Capacity Building Manual by the Human Rights Law Service HURILAWS

Figure 2.6: Own Tax Process: Model for Anglophone Africa

Parliament

Tax Base defined (insome countries bylocal gov’t)

Tax Rate defined(under thesupervision ofcentral gov’t)

Tax Census Tax Collection Funds Controlled byLocal Government

Local GovernmentCentral Government


37

Africa, property tax rates are set by localgovernments within bounds.

With respect to local fees and non-taxrevenue, local government, in bothAnglophone and Francophone countries havea bit more flexibility. In most Francophonecountries, the National Assembly decideswhich fees can be levied, but localgovernments have the power to change taxrates. For example, in Ghana, daily variationscan be seen in market access fees,slaughterhouse fees, food for impoundedanimals, etc. Unlike with tax revenue, localgovernments are responsible for collection

and can even set up municipal fees collectionmanagement boards under certain conditionsand up to a certain amount.

Non-fiscal resources can be significant. InMorocco, for example, revenues from services(8 percent) and assets (17 percent) make upa quarter of local government incomes. InWest Africa, incomes from services and use ofpublic properties yielded nearly 20 percent oflocal incomes from 2004 to 2007.9 In SouthAfrica revenues from services (water,sanitation, and electricity) are responsible fornearly half of the income of local governments(around 44 percent in metropolitan and

9. Source: Partenariatpour leDéveloppementMunicipal (PDM).


Box 2.2: An Example of the Importance of Municipal Taxes Collection: The case of Burkina Faso and Morocco

In a Francophone context characterised by the principle of the unity of the statetreasury (unicité de caisse), it is useful to compare the proportions of local revenuesmanaged by the municipal revenue collection agency and the Treasury. Thisdistinction also covers the tax register, the tax bases and fees determined by localgovernments.

In Burkina Faso, on average, 43 percent of local government revenues were collected bymunicipal collection agencies in 2004. Thus, local governments in Burkina Faso havereal control over nearly half their tax and fees revenues. The 57 percent of local revenuefrom local taxes determined by law consists of resources that are not controlled by localgovernments.

Overall, the smaller municipalities have greater room for decision making. About 20local governments control nearly 50 percent of revenues, with six of them controllingmore than 80 percent. Major towns like Ouagadougou directly collect 37 percent ofrevenues, which means that a large proportion of their local revenue is managed bycentral authorities. This is due to the structure of the economic activity in areasdominated by a large modern sector.

In Morocco, the revenues managed and levied by local authorities are becoming moreimportant than revenues managed by the State on their behalf. Taxes and receipts,service products and property revenues account for 49 percent of local revenuescompared to 42 percent for revenues managed by the State; the remaining 9 percentcorrespond to the transferred revenues.

Source: Calculations by the authors based on data from local governments in Burkina Faso and Morocco.



medium size cities, but less than 10 percent insmall cities and rural areas). Such revenues,however, are limited in the majority of Africancountries.

Another issue is the management of the taxcollection process, which remains extremelycentralized in Francophone countries.Deconcentrated departments conduct thetax censuses to estimate the local tax base.This is recorded on the roll by tax authorities,who issue tax assessments and collect theselocal taxes. The deconcentrated departments,Treasury Department and/or RevenueDepartment, transfer the revenues tothe National Treasury. After this, revenuecollection information is released to localauthorities. Although local governments areformally excluded from tax collection, in somecountries they assist with certain activities,such as the distribution of tax notices (labourand fuel, for example).

In the Anglophone countries, centralizationis less significant. Local governmentsthemselves are the main agents of the taxcollection process. Generally, localgovernments collect their taxes and fees,even if central or state governments deter-mine the amounts. For example, in Kenya, lo-cal governments are in charge of updating theassessment roll (for which they can use theservices of private companies), determiningthe rates of taxes to pay (subject to centralreview), collecting taxes and following upoutstanding payments.

In some English-speaking countries, the taxprocess entails major costs for localauthorities, especially for collection. InTanzania, for example, the cost ofadministering local taxes can reach 80percent of yields and can exceed 100 percentin some districts. Zimbabwe experiencesthe same problem aforementioned. Inthese countries, negotiations are underway to “contract” central assistance in

local tax collection; in Tanzania, this will bedone for a fixed period of time by theTanzanian Revenue Authority. In French-speaking Africa, local authorities are seekingsome role in revenue collection. Thesedifferent tendencies reflect the diversity ofcentral-local fiscal relations across thecontinent. A balance must be found to ensurean appropriate role and better collaborationbetween both actors in the management oflocal finance.

In countries with a French administrativetradition, local authorities are usually legallyprohibited from opening private bankaccounts, although this is becoming morecommon in a growing number of countries.The public treasury is the bank of the publicsector and local authorities are obliged todeposit their liquid funds in it. Thiscompulsory deposit function is reinforced bythe fact that the treasury collects bothnational and local taxes. In Cameroon, localgovernments were allowed to hold accountsin commercial banks until 2010, but theymust now use the treasury system. InAnglophone countries, municipalities areoften free to open accounts and keep theirfunds in the banks of their choice.

State Transfers to Local Government

The systems of transfer to local authoritiesvary tremendously across the African region.A number of characteristics are noteworthy.

• The significant use of conditional transfersin the majority of countries. In SouthAfrica only one transfer in fifteen isunconditional, (the Equitable Sharegrant). This situation undermines localgovernment's autonomy, as the prioritiesof sectoral ministries prevail. In BurkinaFaso, on the other hand, only one in fourtransfers to local government isconditional (the global grant for equipmentwhich must be spent on investment).


39

• The use of formulas to distribute nationalfunds to local governments. Someformulas are simple while others are morecomplex. The Equitable Share in SouthAfrica, for example, has four differentcomponents (base service; institutionalcapacity; equalization; correction andstabilisation). For basic services forexample there are levels of support: onefull grant for poor households relying onmunicipal services and a second partialgrant for households not yet relying onmunicipal networks. The simplest formulasare based on population, and often ofgeographic size, urban or rural nature ofthe local government and level of poverty.

The analysis of intergovernmental transfersmust answer two questions: 1) How is thetotal amount transferred to be set? 2) How isit transferred? In the countries of Africa thereare diverse answers to these questions. Forthe first question, the total amount to betransferred is determined using one of thethree following methods:

• A fixed percentage, either of total centralrevenue or of revenue from one or morespecific taxes, assigned either frombudgeted or collected revenues. In Ghana,for example, 7.5 percent of nationalrevenues has been dedicated to transferssince 2008, while 2 percent of five taxes isdedicated to transfers in Côte d’Ivoire. InMorocco, 30 percent of the VAT fundstransfers, while in Kenya 5 percent of thepersonal and corporate income tax goesinto the Local Authority Transfer Fund.Senegal by law has established that theDecentralization Endowment Fundsreceives 3.5 percent of the total VATincome.

• A fixed amount determined yearly, as withany other national budget allocation, withor without consulting local governments.This is the case in Benin or Tunisia, where

fixed annual amounts were set in the1990s, but each year since has seen areduction, not only in the nominal amountof the transfer, but also its relativesignificance in municipal revenues.

• A variable amount determined by a for-mula (a percentage of the recipient's ex-penditures or an amount depending onpopulation characteristics: age, education,etc.). An example of this can be seen inSouth Africa's Equalization Grants, wherethe amount is determined using thenumber of poor households among otherindicators.

Each alternative has its advantages anddisadvantages. For example, the centralgovernment can become disinterested in thecollection of a tax if it retains only a smallpercentage. A fixed percentage enables localgovernments to project the amount oftransfers and therefore manage revenuesbetter. Annual determination of transferamounts, on the other hand, reduces theirpredictability to local governments, although itallows for better adaptation to changing needs.Thus, if the center transfers functions, it canensure their finance by increasing transfers.Also, funding budget deficits encourages localgovernment wastage, even when administra-tive controls are in place, as the experience inMorocco, until 1996, demonstrates.

For the second question on transfer me-chanisms, there are two main dimensions:equalization and conditionalities. The first,aims at guaranteeing a minimum level ofgovernment services across the country byensuring that all local governments have ade-quate revenues. The second aims at ensuringthat the behaviour of local governments,usually in terms of spending, aligns betterwith the priorities of the central government.

Equalization formulas are widely used. Asimple indicator such as population size or




target size, is often used (Burkina Faso,Malawi, and Mozambique), but more complexformulas are also used (South Africa). Condi-tional grants are paid, if the recipient meets anumber of criteria (submission of a budget,

financial accounts, etc.), or must be spentaccording to certain criteria (minimumpercent on investments, in a specific sector,etc.) or with some percentage financed bylocal revenues. Equalization transfers give

Box 2.3: Using “origin base” or “derivation” as a Rule for Shared Tax Transfers

Several countries use the derivation principle to share resources with local governmentsand to structure the sharing of national taxation revenues based on the place ofcollection.

In Nigeria, VAT is distributed to local governments based on the following criteria: 50percent of VAT is shared among local governments –30 percent by population and 20percent by origin of collection (derivation)–. Also, of the 13 percent of oil extractionresources due to the central government, 30 percent is returned to the local level basedon derivation.

In the Democratic Republic of Congo (DRC), the Constitution provides a derivationprinciple which allocates 40 percent of national taxes collected in their territory, toprovinces.

In Gabon, the national government transfers a percentage of personal income tax basedon what local authority collected it. The distribution is as follows 25 percent to cities, 7percent to the equalization funds of cities and 68 percent to the State. For intermediatetiers (departments), the distribution is: 65 percent to local government, 7 percent toequalization funds and 28 percent to the central government.

In Senegal, the principle of derivation is used to return the following percentages ofnationally collected taxes: 59 percent of the vehicle tax; 50 percent of the surplus valueof real estate; 60 percent of the Unique Global Contribution; 60 percent of courtimposed fines on the territory of cities and rural local governments.

The derivation principle can be useful if it is applied to taxes paid by the residents of theterritory that benefit from them and are not exported. This works for the individualincome tax but not taxation on imports (VAT for imported goods, customs tariff).Indeed, this taxation is often collected in a few specific locations (international airport,harbour, capital city, border town) and is then paid by consumers nationally via itsinclusion in the price of products. This is why a derivation revenue sharing formula wasreplaced by one based on population in Cameroon. The derivation revenue sharingformula was considered to be giving undue revenues to Douala (Cameroon’s mainharbour).

Source: Yatta François (2009): La décentralisation fiscale en Afrique; enjeux et perspectives Karthala, p. 324 and UCLGcountry profiles.


41

local governments more autonomy thanconditional transfers—they are usuallyunconditional.

There are many examples in Africa to illustratedifferent types of transfers (see Annex 2.2).For example, in Algeria and Benin, the purposeof transfers is to ensure fiscal balance, so theycover the differences between expenditure andrevenues. This, however, can discourage localrevenue efforts and encourage unnecessaryspending. Other systems can increase sometypes of spending, for example, salary costs.

The fear of counterproductive effects oftenleads central governments to impose variousforms of control (for example on recruitment),thereby reducing the autonomy ofmunicipalities in decision-making.

A well-designed transfer system should pro-mote municipal autonomy both directly (in thetransfer parameters) and indirectly (in the legaland regulatory environment). It may also beappropriate for transfers to have incentiveeffects on local revenues, for example, toreward efforts at tax roll registration or local


Box 2.4: An Effort at Transparency and Objectivity: Transfers in Kenya and Malawi

In Kenya, the Local Authorities Transfer Fund (LATF) was started in 1998 to improvepublic services, the management of local finance, and the reduction of localgovernments’ debt. LATF is funded by 5 percent of the national income tax. The amountmobilised by the fund increased from 12.9 million USD (1999) to 121.8 million USD(2009). It is redistributed between 175 local governments on the basis of a fixedamount per local authority. Sixty percent of the total is based on population and theremainder based on urban population (increasing with it). Some criteria must befollowed to benefit from the LATF: (1) spend a maximum of 45 percent of the totalbudget on staff and (2) devote at least 65 percent LATF transfer to the investmentbudget. Until 2010, withdrawals were based on the following rule: 60 percent of theallocated amount is released if the council submits its budget to the LATF committee;the additional 40 percent is released if the council submits its public accounts, arevenue enhancement plan, and a service delivery plan to the LATF committee.

In Malawi, the National Local Government Finance Committee (NLGFC) is in charge ofsupporting local authorities in preparation of their budgets and the distribution of nationalgrants to local authorities. Two thirds of these grants are distributed to urban authoritiesaccording to the following formula: 50 percent of the amount equally distributed betweenlocal authorities, 30 percent on the basis of the population, 13 percent to local authoritiesaffected by drought on the basis of their population and 7 percent to the less developedcities according to their population. The remaining third of the national amount of the grantis exclusively reserved to rural local authorities, of which 30 percent is equally sharedbetween local authorities and 70 percent is shared according to 4 factors: population,surface, illiteracy, and infant mortality. The District Assemblies must submit their draftbudget 90 days before the beginning of the fiscal year.

Source: Ministry of Local Government (2004) Status Report overview of the status of decentralization 1993 – 2004 andF. Yatta (2009) and Jamie Boex, Randson Mwadiwa, R. Kampanje (2001): Malawi Intergovernmental Fiscal TransfersStudy. Final Report. Government of Malawi, UNCDF, UNDP. Lilongwe and Atlanta, April 2001.



tax collection. Such transfers must not belinked to revenues as is the case in Tunisia,or in Ghana, with the District AssembliesCommon Fund (DACF). Transfers may also tryto facilitate efficient expenditure management,for example, by increasing school attendancerates. Some transfers reward compliance withadministrative requirements (budget, develop-ment plan, etc.). For example, in Madagascarmeeting such conditions allows access to theLocal Development Fund and in Ghana, it givesaccess to and distribution of the Assembly'sDistrict Development Facility (DDF). In Zambiathe central government is developing anew formula-based transfer system tomake the financing of local governmentmore predictable and transparent, whichwill include performance incentives.

Borrowing by Local Governments

Access for African local governments to debtand financial markets is limited. Even in thecases where legislation authorises access, thefinancial situation of local authorities and thelimited development of financial systems,makes access difficult.

The table below (Table 2.3) presents theprinciples that govern the access of localgovernments for borrowing by country. Insome countries, such as Kenya and Tunisia, theamount of the loan is restricted. In othercountries, such as Malawi, debt service levelsare taken into account; the supervisoryauthority cannot approve a loan beyond agiven debt service threshold. In most

Indicator Used Financial Market Specialized Financial Institution

Amount Kenya Kenya

Rwanda Morocco

Tunisia Rwanda

Tunisia

Debt service Malawi Morocco

Mozambique

Total resource threshold Benin Benin

Burkina Faso Burkina Faso

Côte d’Ivoire Côte D’Ivoire

Egypt Egypt

Gabon Gabon

Mali Mali

Niger Morocco

Senegal Niger

Togo Senegal

Uganda Togo

Source: Compiled by the authors

Table 2.3: Conditions of Access to Loans to the Local Governments of 16 African countries


43

countries, such as Burkina Faso, supervisoryauthorities use the criteria of a threshold oflocal government’s own resources to ensurethat they can pay back loans. Moroccoassesses multiple factors: the amount of theloan, debt service, and the resource threshold.

Access to loans, however, is generally notdirect. In most African countries, a financialintermediary manages funds earmarked forlocal governments. The lending systems forlocal governments use a wide range ofinstruments. There are currently 11 financialinstitutions for local authorities in Africa:Development Bank of Southern Africa (DBSA),Infrastructure Corporation Limited (INCA) inSouth Africa, National Agency for LocalGovernment Investment in Mali (AgenceNationale d’Investissement des Collectivitésterritoriales-ANICT), Municipal DevelopmentFund (Fonds de Développement Communal-FDC) in Rwanda, Loan and Support Fund forLocal Authorities in Tunisia (Caisse de Prêts etde Soutien aux collectivités locales-CPSCL),Special Council Support Fund (Fonds d’Équi-pement et d’Intervention Intercommunal-FEICOM) in Cameroon, Loan Fund for LocalAuthorities (Caisse de Prêts aux CollectivitésTerritoriales-CPCT) in Niger, MunicipalDevelopment Agency (Agence deDéveloppement Municipal-ADM) in Senegal,Loan Fund for Local Authorities (Fonds de Prêtaux Collectivités Locales-FPCL) in Côte d’Ivoire,Council Support Fund (Fonds d’EquipementCommunal-FEC) in Morocco and DevelopmentFund for Local Authorities (DFLA) in Malawi.Some institutions are more active than othersdue to lack of funds.

Three of these financial institutions are more orless specialized as intermediary institutions;that were created for the purpose of fundinglocal authorities. They are the DevelopmentBank of Southern Africa (DBSA), theInfrastructure Finance Corporation Limited10

(INCA) in South Africa, and the MunicipalInfrastructure Finance Fund (FEC) in Morocco.

Both the INCA and DBSA draw their re-sources from financial markets and lend tolocal authorities in the same way as a classicbank. Unlike South African bankinginstitutions, the FEC in Morocco borrow fromthe financial market, the government, andfrom various development partners whoprovide support for decentralization and/orlocal development.

Several other institutions are financiallyautonomous agencies with an administrativestatus: the National Agency for LocalGovernment Investment in Mali, the MunicipalDevelopment Fund in Rwanda, the SpecialCouncil Support Fund in Cameroon, the LoanFund for Local Authorities of Niger and the Loanand Transfer Fund for Local Authorities ofTunisia.

Lastly, there are some agencies created in the1990s, through international cooperationprojects such as the Municipal DevelopmentAgency in Senegal and the Loan Fund for LocalAuthorities in Côte d’Ivoire and theDevelopment Fund for Local Authorities inMalawi. These agencies are sustained bygovernment resources and internationalsupport.

The functions of specialized lending systems arediverse and varied, ranging from their essentialrole of lenders to grants and institutional support(Table 2.4). Few loan systems accept to serve asguarantees for loans taken out with otherinstitutions. In general, the absence of thisfunction is explained by the fact that no financialinstitution on the marketplace lends to localauthorities. It is sometimes also explained bythe fact that central government wishes to steerborrowing by local governments towards asingle institution.

Aside from Zimbabwe, which is a specialcase11, South Africa represents the mostsignificant regional experience. In South Africa,two-thirds of loans taken out by local

10. INCA was created in1996 by a consortiumof public and privateinstitutions, inparticular the FirstNational Bank of SouthAfrica, a few SouthAfrican insurancecompanies andfinancial institutionsand four internationalpartners: Proparco, asubsidiary of AgenceFrançaise deDéveloppement (AFD),Dexia internationaland two German andBritish financialinstitutions. The fourinternational partnerssit on the board ofdirectors of the INCA.




governments come from long-term marketloans. Already in the mid-1990s, South Africanlocal governments were using long-term loansfor capital investments in water, electricity,roads, education and other municipal services.The relatively sound finances of local govern-ments and the widespread belief that thegovernment would bail out those in trouble,enabled local governments, especially whiteones, to access loans through bank loansand municipal bonds. Bank loans are themost common form as only the richest localgovernments can access resources frommunicipal bonds. These were predominantlymade up of the local governments thatcurrently make up the city of Johannesburg.Lastly, it must be remembered that municipalbonds barely made up 1 percent of centralgovernment bonds.

In Francophone Africa, Douala implementedone of the rare experiences of bond issuancefor the capital market. In 2005, the city raisedaround 19 million dollars from the new stock

market, the Douala Stock Exchange. TheRegional Stock Exchange (BRVM) of Abidjan isin theory an opportunity for local authorities inWest Africa to access financial markets. Thisstock exchange, which was established underthe West African Economic and Monetary Union(WAEMU), is governed by the Regional Savingsand Capital Market Board (CREPMF) and hasbeen operational since 1988. In principle, themarket is open to all public (government andadministrations) or private (listed and unlistedcompanies) entities. In practice, however, forthe moment, no local government has beenable to access a BRVM loan as the centralgovernments refuse to provide approval, a pre-condition of all BRVM loans.

In spite of the limited access to borrowing,it remains possible to observe problemsof local authority debt. In Algeria,for example, problems of budgetmanagement brought on structural deficitsthat the government is attempting to getunder control. In 2007 approximately 980

11. The offering wasmainly made up ofinsurance companies,pension funds andPost Office SavingsBank reserves. In the1990s, the centralgovernment requiredinsurance companiesand other marketparticipants to keep apercentage of theirholdings in bondsissued by the centralgovernment, para-state organisationsand municipalities. Asfinance institutions(pension funds, PostOffice Savings Bank,etc.) were practicallyobliged to buy them,these bonds had lowinterest rates, makingthem a veritablewindfall for all bondissuers. Analysts admitthat the municipalbond market inZimbabwe is actually a"false" market that isunsustainable. Thereare two main reasonsfor this, the requiredauthorisation ofministries in charge oflocal governments andfinances and the Stateguarantee, andprescriptions fromcentral governmentwith respect to thefinancial market. Wewill therefore onlydiscuss the case ofSouth Africa.

Grants Loans Interest subsidy Guarantee Institutional support

Tunisia (CPSCL) X X X X

Morocco (FEC) X X

Senegal (ADM) X X X X

Rwanda (FDC) X X

South Africa (DBSA) X X

Côte D’Ivoire (FPCT) X

South Africa (INCA) X

Cameroon (FEICOM) X X X

Niger (FPCT) X

Mali (ANICT) X X X

Malawi X

Source: François Paul Yatta (2009). "La décentralisation fiscale en Afrique: Enjeux et perspectives", "Economie etDéveloppement" Collection, éditions Karthala and the UCLG country profiles.

Table 2.4: Loan System Services Provided to Local Authorities


45

local governments were in deficit; in 2008this number has increased to nearly1,200. Through the complementaryfinance law passed in 2008 thegovernment erased 22.3 billion DZD inlocal government debt and measures weretaken to reduce future deficits. In 2010,the number of local governments in debtdropped to 400. In Kenya the governmenthas conditioned local government accessto the LATF grant on progressive debtreduction.

Main Issues of Fiscal Decentralization in Africa

In this section, we will analyze the mainproblems linked to territorial organization,expenditure, total and own resources,transfers and local debt.

Territorial Organisation

Territorial organisation has a direct effect onthe decentralization process and the


Box 2.5: The Relationship between Decentralization and Deconcentration

A comparison of selected East Africa countries, Kenya, Malawi, Tanzania, Uganda andZambia reveals different models. Kenya has a deconcentration bias characterized by thesimultaneous presence of both a local government and deconcentrated administration.Districts and county councils, for example, cover the same territory. A centrallyappointed district commissioner heads the district while the county is headed by a clerk(appointed by central but officially accountable to the elected country councils).Consequently, there is very strong control and there are often cases of conflicts of tasksbetween local governments’ officials and deconcentrated services. Conversely, Ugandahas a low level of deconcentration (even if the latest tendencies are towardsrecentralization). In this country, there are no deconcentrated agencies that competewith local government departments. Local governments are therefore in a strongerposition, although central directives in recent years have to some extent made themexecuting agents of national policy. Tanzania is between the two extremes. In thiscountry, deconcentration means the transformation of a position of central control to aposition of assistance and facilitation of local governments.

This competition between deconcentrated departments and local governments can evenresult in resource wastage. In Zambia the decentralization process begun in 2002generated a parallel deconcentration of certain sectors to the district level, notablyeducation and health. This brought a duplication of human and financial resources and amultiplication of conflicts. In Malawi, ministries implement deconcentrated activities thatduplicate local government activities, including in health services, environmental andforest management, road maintenance, farming and irrigation. Business permits toprivate entrepreneurs are issued both by the Ministry of Trade and Industry and localauthorities. However, this duplication does not translate into a better level of serviceprovision.

Source: JICA (2006) Local Level service Delivery, Decentralization and good governance : A comparative study ofUganda, Kenya and Tanzania: education, health and agriculture sectors; Jamie Boex et al (2001) : MalawiIntergovernmental Fiscal transfer; F. Yatta (2009)



effectiveness of responsibility transfers tolocal government. The manner in whichdecentralization and deconcentration aresuperimposed and how they relate to eachother—independent, complementary orduplicative— has an important impact on howtransferred responsibilities are taken incharge by local governments and inconsequence on their budgets.

In the last ten years, there have been a seriesof reforms in several African countries that havecreated intermediate levels of governancebetween central government and municipalities(Burkina Faso, Côte D’Ivoire, Gabon, Mali,Morocco, Senegal, Togo and soon Mauritania).In Kenya, the new constitution approved byreferendum in August 2010 will transform the47 counties into an intermediate level ofgovernment, assuming the existing localauthorities remain in place (they will in theshort term). Mali has three local governmentlevels: communes, circles (the equivalent ofdepartments or provinces) and regions. InSenegal the regions, cities, and rural localgovernments coexist. Madagascar, a countrythat has been fully divided into local authoritiessince independence, has been oscillating betweenregions and provinces as the intermediate levelfrom one constitution to another.

There is an ongoing debate about the relevanceof the various levels and the availability of theresources to support them. A cost-benefitanalysis of these choices must take into accounttheir impact not only on public services and onthe representation of regions in nationaldecision-making, but also on internal tensionsbetween regional groups.

Expenditure Control of Local Authorities

There is commonly a lack of clarity in thesharing of responsibilities between the centralgovernment and local governments.Decentralization laws often do not providesufficient detail on the sharing of

responsibilities and even clear laws may notbe accompanied by modifications tocontradictory sectoral laws and policies, ascan be seen in Benin, Zambia and Tunisia. Forexample, if decentralization laws transferresponsibilities for education or transportationto local governments but sectoral laws remainunchanged, the stage is set for conflictbetween deconcentrated agencies which, forvarious reasons (prestige, tradition, humanand financial resources) do not wish to giveup mandates, and local governments whowish to take up their new functions. Since thedeconcentrated services are often moreequipped to act, in contrast to the localgovernments that are often lacking financialand human resources, decentralization isconsiderably slowed12.

This confusion regarding the division oftasks may be aggravated by internationaldonors. New development aid modalities –budgetary aid and sector wide support –often favour the recentralization of sectoralpolicies – such as education, health,water and sanitation, among others –through the concentration of financialmeans in the ministries, without taking thenew responsibilities of local governmentsinto consideration. In Mauritania, forexample, only 4 percent of theinvestments real ized in c it ies thatfell under the responsibilities of localgovernment were devolved and includedin local budgets.13 Thus, in practice andcontrary to the legislation, localgovernments are deprived of these tasksand their corresponding resources, whichsignificantly reduce fiscal decentralization.In addition, donors tend to introduceparallel aid mechanisms that do notsupport and may undermine governmentinstitutions. Several donors developalternative interventions with communities,social and professional associations, non-governmental organisations (NGOs), tradi-tional chiefs, etc.

12. Reforms are currentlyexpected in a numberof countries, notably inKenya and Zimbabwewhere development ofa decentralization plan(DecentralizationImplementation Planand NationalDecentralizationPolicy) will permit theclarification ofresponsibilitiesbetween the differentlevels of government.

13. See Yahya Ould Kebd(2010).


47

The transfer to local government of new res-ponsibilities, without the corresponding humanor financial resources (unfunded mandates)creates difficulties similar to those generated byimprecise laws. In Uganda, for example, newlaws on territorial organisations (Land Act) andon education have created new districts withoutfinancial resources, and needed humanresources and salaries are not assured. InSouth Africa and Zambia the local authoritieshave also experienced recent cases offunctional transfers without the correspondingresources (in the domain of local economicdevelopment, tourism and library creation). InZimbabwe local government received a grant tocover the transfer of education responsibilitiesin the 1990s, but this was progressivelyreduced and finally eliminated.

Inappropriate transfers of powers and res-ponsibilities to local governments continue to

be problematic. This is the case, for example, ofcountries like Mozambique and Malawi wheresocial security and public housing are financedby local authorities, often under sharedjurisdiction. Local governments generally do nothave the financial resources required to set upredistribution policies at the local level. Giventhe financial and economic disparities betweenlocal governments, these services can beharmonised only at the national level, if somelevel of equity is to be maintained among thevarious territories.

Another set of problems concerns the level ofautonomy local governments have to definetheir spending priorities and in some countriesto control their own resources. The controlexercised by central government on budgetdecisions weakens local authorities. As notedearlier, in the various countries, the draft budgetor budget proposal voted-in by the municipal


Box 2.6: Omnipresent Central Government Control: the example of Tanzania

In Tanzania, the Office of the Prime Minister in charge of Regional Administration andLocal Government, which replaced the Ministry of Regional Administration and LocalGovernment, exercises a tight control on the fiscal power of local governments.

In spite of the decentralization policy, both the Office of the Prime Minister and theMinistry of Finance (MoF) continue to exercise a great deal of influence on the fiscalactivities of local governments through regional secretariats. The Office of the PrimeMinister provides guidelines and budget standards and audits local governments toensure that they comply with these standards. It monitors local budgets and providesthe technical capacity required and technical training. However, local governments donot have a great deal of leeway due to central government restrictions implementedthrough the conditional grants that are used to finance nearly all local governmentexpenditure.

It is obvious that the promotion of decentralization by deconcentration is a difficult task.Even ten years on, the central government continues to take part in nearly every aspectof the fiscal activity of local governments.

Source: Odd-Helge Fjeldstad (2001) Fiscal decentralization in Tanzania: For better or for worse?; Jameson Boexand Jorge Martinez-Vazquez (2005) Local Government Finance Reform in Developing Countries: The Case ofTanzania



council is submitted for the approval of higherauthorities, such as the local representative ofthe national authority (prefect, governor, etc.)and/or ministries (Interior, Local Government,Finance, etc.) In this case, the local authoritymust wait for review by these higher authoritiesbefore final budget approval.

There may also be budget controls to ensureadherence to national policies, both before andafter local budget approval. In the event of adisagreement between the local governmentand the national authority, the final decisiontypically favours the latter. In Francophonecountries, if the deliberations of the municipalcouncil do not take the review of the higherauthority into account, the latter can declarethe budget unenforceable.

The “unified treasury” principle dominant inthe Francophone countries also weakens thelocal authority control over revenues. Aspreviously explained the centre can requirelocal governments to use their resources forcentral priorities, creating accounting problemsand reductions in available resources for localauthorities.

Own Resources of Local Government

The financial difficulties associated with povertyafflicting African countries partly explain whythey have problems transferring sufficientresources to local government. In Africa the taxratio as a proportion of GDP reached 27.4percent in 2007. For middle income countries(18 countries), the tax ratio as a proportion ofGDP varies between 25 percent and 34 percentthat is to say levels that are comparable withsimilar countries of other continents. But for lowincome countries the tax ratio as a proportion ofGDP is lower than 15 percent (32 countries)(OECD, AfDB and UNECA, 2010) This low levelof taxation, for the majority of countries,explains why the State was only able to partiallycover the demand of public services inmany areas even before the implementation

of decentralization. The decentralization ofresponsibilities did not lead to a relevant transferof resources because there were no resources totransfer, either at the central or at local level.

As noted in the first section, fiscal decentra-lization is currently suffering, in most countries,from an incomplete fiscal transition. Customduties are on the decline in most Africancountries, driven by regional integration andglobalisation, and Governments are hesitant toshare their resources, as currently they do nothave a replacement tax that is as easy to collectand has the same returns as customs duties.

Even if local governments could identify citizenpriorities and reflect them in local budgetsapproved by the supervisory authorities, theyhave very few resources for meeting the needsof their residents. Why is this so?

• Local governments have limited revenuegeneration powers.

• Central governments see local governmentresources as expendable and the burden offiscal adjustments often falls on them. Thereare many examples of this practice, whichreduces the predictability of local funding.Controls exerted over tax collection andthe practices of “unified treasury” systemsalso limit the level of local governmentscontrol over their resources, receiving onlypartial information on tax collection in theircommunities.

• Local tax administration, particularly for theproperty tax, requires significant technicalskill and is affected by the magnitude ofinformal urban settlements. As noted in theintroduction, 72 percent of the sub-Saharanpopulations live in slums. In this context,land registers are difficult to establish andeven more difficult to update.

• Other issues relate to the methods usedto calculate property tax. In general, urban


49

development and resulting increases inproperty value, are not well taken intoaccount. In Tanzania, for example, the lawestablishes two modalities for settingproperty values for local taxation: themarket value and the replacement cost. Butthe notion of property in Tanzania is limitedto the right of use, which converts propertyinto a non commercial good. Given the non-existence of a property market, the methodto evaluate property tax relies on the valueof the replacement costs of structures. Thesecond constraint is linked to delays in re-evaluation of the tax base: in Morocco, adecree has established a revaluation period,every three years a maximum and aminimum value is set for the square meterreference price; in Ghana, the period for thisre-evaluation should be five years but in

reality is between 10 and 15 years. Whenyou add the lack of land titles for the greatmajority of properties, it is fair to say that,land management issues are insufficientlyaccounted for when examining local financeissues. It is an area that deserves to beinvestigated as a possible lever of localresource mobilization.

• The lack of qualified human resources inlocal government in the majority ofcountries, and in particular in small localgovernments and rural areas, does not allowlocal governments to take on themanagement of local taxes. This constraintisolates local governments from taxmanagement, most notably in Francophonecountries, or leads local governments inAnglophone countries to delegate this


Box 2.7: Collaboration Between Central Governments and Municipal Services in theManagement of the Fiscal Chain: the example of Côte d’Ivoire.

To significantly improve the tax recovery rate (33 percent), decree No. 97-35 of 22January 1997, strengthened by law No. 2003-489 of 26 December 2003, regulates thecollaboration between the central role in local tax collection and municipal services. Thisincludes the identification of taxpayers of property tax, patents, licences andcomprehensive taxes, the keeping of files and their quarterly update (creation, editingand discontinuation of activities by the taxpayer), the acquisition of all the elementsnecessary for the setting up of roles and the control of recovery (distribution ofwarnings).

Teams were created in each municipality for tax collection. The were composed asfollows:

• Local employees (the number of agents dedicated to this depends on the extent ofthe municipality and its fiscal potential);

• Treasury employees (those responsible for municipal revenues); National FiscalDepartment employees, (responsible for property tax revenues).

All the above-mentioned operations are carried out by these teams.

Source: PDM (2002) and Yao Charles Kouassi (2010).



collection to national agencies (for examplein Zimbabwe) or the private sector (Kenya,Mozambique).

• Finally, one of the structural constraints thatgoes against the efforts of localgovernments to mobilize their own revenuesis often linked to the lack of trust on the partof the population. Problems of corruptionexposed in certain countries sometimes alsoaffect the local level where, due to proximitythese bad practices are directly felt by resi-dents. Even if measures have been taken toimprove the situation, African localgovernments still have to convince theircitizens that they are able to design andimplement local policies, improve the qualityof local expenditure and deliver adequateservices to them.

Transfers

Transfers represent around 50 percent of thefunding of local governments, and even morein many African countries. They make up the

difference between the cost of transferredpowers and responsibilities, own revenuesfrom local taxes and other sources, and theyalso offset differences in development andresource bases between local governments. Areview of mechanisms of transfer to localgovernments in Africa reveals the followingproblems.

• Fiscal decentralization suffers from a lackof knowledge about the cost of thetransferred responsibilities. Solving thisproblem is even more difficult whenfunctions are not clearly defined.

• Needs-oriented funding solutions are insome cases based on the expenditurefunding packages that existed beforedecentralization, as is the case in BurkinaFaso with the payment of civil servants’salaries. This means that localgovernments who were in advantageouspositions before decentralization (capitaltown with a proximity to power,national-level civil servants who enjoy

Box 2.8: Examples of the Assessment of the Cost of Transferred Powers and Responsibilities: Cameroon and Senegal

Cameroon is one of the few countries to carry out work to assess the cost of transferredresponsibilities. Two studies commissioned by the department in charge of decen-tralization show that the current cost of the traditional mandates of municipalitiesstands at FCFA 36.2 billion while the cost of the transferred tasks can been estimated atFCFA 115 billion (base 2000) or between FCFA 190 and 200 billion (updated in 2005).

In Senegal, the decentralization process has transferred nine responsibilities to localgovernment: environment and natural resource management, health, social welfare,youth, sports and recreation, culture, education, planning, land use planning, townplanning and housing. The government has decided to delegate to local government thesums that used to be earmarked for these nine responsibilities when they were providedby central government, representing an annual amount of approximately FCFA 6 billion(USD 8.5 million).

Source: F. Yatta (2009); and the Economic Commission for Africa (2004): Renforcer la gestion financière locale et ladécentralisation fiscale en Afrique : Etat des lieux et proposition d’un plan d’action.


51

local services, member of the ruling group,etc.) will continue to be advantaged. Thisis also the case in Gabon. For example, thefinancing of primary school teachers,benefits local governments who alreadyhave schools, while financing based on thenumber of school-going age childrentakes into account present and not pastneeds. If decentralization is implemented

by changing gradually from the first modeof financing to the second, it is essential tomake sure that local governments canraise the necessary funds.

Some formulas cover the differencebetween revenues and expenditures. In1996, Morocco replaced this type offormula with an equalization mechanism.


Box 2.9: Transfers that Never Arrive at Destination!

In African countries, the government does not always pay the monies due to localgovernment, which often represent very significant amounts.

For example, in Ghana one of the four payments planned under the District AssembliesCommon Fund (DACF) for 2009 was delayed, presumably to be paid in 2010. Thisrepresents 25 percent of the scheduled amount and approximately 20 percent of thetotal revenue of Ghanaian local governments.

In Côte d’Ivoire, Decree No. 98-05 of 14 January 1998 established the currentexpenditure block grant (Dotation Globale de Fonctionnement - DGF). This grant iswritten into the budget law each year on the basis of 2 percent of selected tax revenuescollected during the two years preceding grant payment. The levies include tax onbusiness profits, taxes on salaries, domestic value added tax and service tax. Inpractice, the total amount of grants has been below 10 billion FCFA, well below 2percent of the included taxes.

In Gabon, the government does not transfer the legally specified share of national taxes(business tax, tax on the revenue from securities, income tax from liberal professions,VAT) to local governments. This requires implementation decrees that have never beenissued.

In Cameroon, the budget law for the 2004 fiscal year applied 50 percent of themunicipal share of the VAT as a contribution to the national efforts under the HeavilyIndebted Poor Countries Initiative (HIPC). This diversion was supposed to stop at theend of the 2004 fiscal year, but it continued until the end of the 2009 fiscal year.

In Malawi decentralization policy recommends the transfer of 5 percent of the netnational revenues to local governments via unconditional grants, but the governmenthas never transferred more than 2 percent.

Source: F. Yatta (2009) ; Gérard Gagnon (2002): Le financement du développement local; Un état des lieux en Afriquede l’Ouest, Eléments de comparaison Ghana-UEMOA; contributions from the workshops organized in Rabat, 7 May2010.



This was done as the first methoddiscouraged tax effort, promoted un-necessary spending and created a dys-functional situation. The use of thistype of budget balancing transfershould be discouraged.

• Irregularity and unpredictability oftransfers is also a problem. Agreedupon transfers never arrive (fully) attheir intended destination despitehaving been voted in Parliament andwritten in the budget law. In Tanzania,for example, the formulas that exist tocalculate the transfers are simply notapplied. In these situations, localgovernments find themselves withoutthe necessary resources to implementpublic services. There is a risk of thissituation becoming more widespreadduring tax transition and even worsewith the global economic crisis creatinggeneral challenges for public finance.

• An even more serious problem occurswhen countries step back from goodprinciples on which their transfers arebased. In several countries (BurkinaFaso, Mali), the policy documents ondecentralization establish the principleof simultaneity between the transfer ofresponsibilities and the transfer ofresource, but the difficulties inimplementing this principle hasresulted in the postponement oftransfers.

Borrowing Capacity

Despite the high urbanisation rate andthe increase in the demand for publicinvestments, the use of debt by localgovernments is very limited, even wherepermitted by legislation. Only a few large localgovernments have access to borrowing. Thedebts taken out by local government are oftenshort-term loans (credit facilities) with retail

banks and in many countries they are usedprimarily to ease cash flow strains.

• Weak demand from local governments is amain reason for the under-development ofthe local credit market. The small share oflocal governments in public finance andthe limited local revenue system observedin most countries increases risks and theprice of credit.

• Local government access to the financialmarket remains very limited even incountries where there is a developed andefficient financial market.

• Specialized finance institutions haveyielded mixed results:

– Loans end up going only to localgovernments with institutionaldevelopment sufficient for them to meetloan requirements (application preparation,financial assessment of the project,etc.).

– The activity of specialized financeinstitutions has not succeeded inestablishing a culture of borrowing andrepayment, which has often beenundermined by the existence of severalother sources of "more easilyobtainable" funding, such asgovernment subsidies and grants fromdonor agencies. Loans frominternational donor agencies aregenerally cheaper than other sourcesregardless of the country'smacroeconomic and financial position.The situation has been made morecomplex with the recent development ofsub-sovereign loans at interest ratesranging between 0.75 percent and6 percent granted by internationaldonors to some African cities, such asOuagadougou and Dakar. Although thisis desirable in principle, it needs to be


53

coordinated with the financial market inorder not to affect its development.

– In some countries, specialized financeinstitutions have been a major factor inthe increase in the debt of local

governments often with seriousproblems of repayment. For example, inTunisia, more than half of the country'smunicipalities have overextendedthemselves after 10 years ofintervention on the part of the Loan and


Box 2.10: Financing: towards endogenous solutions

The FinAfricaCities programme, co-financed by the AFD and Cities Alliance is currentlyexamining how to finance investments in Sub-Saharan cities. These cities will haveapproximately 340 million additional inhabitants over the next twenty years, requiringinvestments of approximately 45 billion USD a year.

For this investment to occur, local governments must increase their relative share in thefinancing of local investments for two reasons. The first is that national budgets willfocus on increasing expenditure on their own core functions. Secondly, current trendsseem to indicate that development aid will not be able to play a much larger role than itdoes today.

The increasing need for investment implies a significant change in the scale and volumeof financing. In the end, both the modalities of urban financing and the financial systemwill need to be changed. The on-going changes in the African economies may indicatethat now is the time for a progressive evolution towards endogenous solutions.

The ideal endogenous solution is financing in local currency, supported by resourcescoming from local savings. These savings are gathered by financial institutions andbanks of the continent on the one hand and the financial markets on the other. Thecreation of specialized institutions for local governments on the Pan-African level or atthe regional level is neither necessary nor appropriate, given the number ofdevelopment banks already in existence across the continent. On the other hand, theefficiency of the systems will be supported by the development of structures such aslocal investment funds like those recently created in China, India, and South-East Asia.

The second endogenous solution, which is almost universally used, is financing via landdevelopment and investment. The reluctance of African governments to engage in thisprocess for a variety of reasons is currently untenable given the needs. Theimplementation of this solution supposes the existence of specialized operators toservice local governments. The institutional positioning of these operators is veryimportant. Partnerships between the public and private sectors allow a betterdistribution of risks across projects. They therefore seem well adapted to the Africansituation if their decision making autonomy is guaranteed.

Information facilitated by Thierry Paulais/FinAfricaCities/Cities Alliance.



Support Fund for Local Authorities(CPSCL).

– In South Africa, the Development Bankof Southern Africa (DBSA) and Infra-structure Finance Corporation Limited(INCA) seem to have turned away fromlocal governments because of the risksinvolved. Such difficulties have led tothe closing of some of these financeinstitutions (CPCT in Niger and FDC inRwanda) and the restructuring of manyothers (DBSA in South Africa, FPCL inCôte d’Ivoire and CPSCL in Tunisia). Ingeneral, recovery problems arise be-cause of the failure of localgovernments to budget the debtservice, the refusal by some electedrepresentatives to honour commit-ments of predecessors, politically mo-tivated government bailouts ofdelinquent local governments, and laxpolicies of the specialised financeinstitutions.

There is a need to learn from these lessonsand to review the role and institutional formatof specialised finance institutions in order toensure greater management autonomy andtransparency and to extend access toborrowing to local governments beyondmajor towns.

The Global Financial and Economic Crisis

In 2009, the impact of the financial crisistranslated into a reduction of growth in themajority of the economies of the region; theGDP growth rate dropped from 5.6% in 2008to 2.5% in 2009. All countries were affected,especially those with intermediate incomesand those oil exporter countries that weremore integrated into the global economy.Local governments were also affected.

According to international organisations,while the recovery began in Africa during

the first trimester of 2010, only a moderateimprovement in the growth rate is forcast(4%). The losses in employment that havetouched millions of households, will havelong term effects, most notably on humandevelopment. This will have implications formeeting the Millennium Development Goals(MDGs).

The financial crisis also reduced the transferof funds from expatriate workers. Afterhaving reached approximately 41 billion in200814, it is estimated that with the crisis,there has been a reduction of 6.6% in 2009.For some communities, this has had seriousimpacts as these remittances not only financea major part of local economic activity, butalso sustain household spending.

Another feared effect of the crisis concernsofficial development assistance. Is unlikely toreceive the additional resources foreseen inGleneagles, largely due to the reducedcontributions of certain European donors im-portant for Africa (OECD, 2010). Reductionsin ODA in 2007 already affected numerouslocal investment programs.

Even if it is too early to have quantifiedevidence of the impact on local govern-ments, they will be affected given theexisting low level of local resources. Tosupport the economy during the economicdownturn, two thirds of regional countrieshave cut their public spending: in ba-lancing their budgets they will be forced tomake difficult choices. Local governmentswill most likely suffer significant fundingreductions, as was the case in a number ofcountries in 2009.

The impact of the crisis on local revenues willin all likelihood be more important in 2010due to the behaviour of the principal revenuesof local governments, especially land andproperty tax. In many African countries, localtaxes are paid only when necessary, for

14. Source: African Centrefor Gender and SocialDevelopment (2009),African Perspectiveson the GlobalEconomic andFinancial crisis,including the impacton health.


55

example when the taxpayer needs a permit tobuild or sell property. The social impact of theeconomic downturn will limit newconstructions and transactions, andsubsequently reduce revenues.

Investment and credit constraints have al-ready led to the closure of some businessesand the curbing of the creation of newcompanies. This has affected and willcontinue to affect the flow of resources of lo-cal governments derived from local taxes.

On the whole, we are witnessing the effects ofa double edged sword, with a sharp increasein the demand for social expenditure for thepoor as a result of the crisis within the realeconomy on the one hand, and a significantdrop in own resources of local authorities onthe other.

Conclusion and Recommendations (presented by UCLG Africa)

The decentralization process observed inAfrica over the last two decades has had animportant impact. It has contributed to adeepening democracy and improvinggovernance in many African countries thatwere in search of legitimacy. Its objective hasalso been, naturally, the improving of servicedelivery and the quality of life for citizens.However, as discussed in previous sections,local governments in Africa are faced withserious funding problems that hamper theimplementation of their responsibilities.

In the light of these problems, the followingsix proposals aim at reinforcing financialdecentralization and making publiccommitments in favour of decentralizationmore credible. These proposals should alsocontribute to the creation of a necessarypolitical dialogue between the mainstakeholders concerned (governments, localgovernments, political parties, civil society,NGOs, etc.):

1) A necessity: the strengthening of the roleof local governments in public expenditure toincrease the credibility of decentralization.

Decentralization will be perceived as an un-realisable dream if the transfer of respon-sibilities to local governments does notcontribute to improving citizens’ quality oflife due to local elected officials not havingthe resources to meet the responsibilitiesassigned to them by law. The imple-mentation of the principle of subsidiaritydictates that local governments carry out,through the responsibilities devolved tothem, a significant share of public expen-diture. The first reasonable objective wouldbe to double African local governments’share in public expenditure over the nextfive years. In a 10-15 year horizon, localgovernments should be able to reach theworld average (around 20-25 percent) ofpublic expenditure implemented by localgovernments.

To ensure that this goal is met and to facili-tate their lobbying of national stakeholderssuch as the central government,Parliament and development partners,national associations of local governmentsmust have at their disposal preciseindicators to estimate their share in thepublic expenditure and follow its evolution.This is only possible through theimplementation and strengthening of amethod for institutionalizing standardizedand systematic study and analysis of localgovernments’ accounts.

2) An emergency: clarification of respon-sibilities between the different levels ofgovernment and estimation of the cost ofdecentralized responsibilities.

A recurrent observation in Africa is that thesharing of responsibilities between the diffe-rent levels of government lacks precision andthat the transfer of responsibilities to local




governments has not gone hand in hand witha simultaneous transfer of financial resourcesto implement them. There are numerous lawsthat have been waiting for years to be appliedor that have been too quickly ratified withoutdefining the means of their implementation,and without their harmonisation with theexisting laws and regulations. Maintaining acentral veto power which strips localgovernments of their substance andautonomy, barely having taken on their newresponsibilities, clouds the vision ofdecentralization in the field.

In fact, most local governments and theirassociations call for greater clarity as to theirresponsibilities and the means they shouldreceive to assume them. An urgent priorityin this regard is the launching of asystematic evaluation of the cost oftransferred responsibilities by examining,ministry by ministry, the budget allocatedwhen these were direct central responsibi-lities. This objective evaluation of the cost oftransferred responsibilities should also takeinto account their territorial dimension,which would allow for a better understandingof their coverage and the cost of transfer.Given that each ministry cannot solve theproblem of transferring the correspondingresources in isolation, a possible solutionwould be to establish a specific fund for thetransfer of responsibilities, jointly managedby the ministry responsible for localgovernments and by the national localgovernment association. This fund wouldhouse all resources for transferredresponsibilities and could eventually alsoinclude additional resources assigned inaccordance with the development of politicalwill to advance the decentralization process.Planning and programming transfers ofresponsibilities would allow transfers to beadapted, along with the necessary technicalresources, according to the national budgetcapacities to support them financially andeffectively.

3) A priority: reinforcing the financial resour-ces and autonomy of local governments.

Giving more room in financial matters forlocal governments is one of the objectivesof fiscal decentralization. As shown in theprevious sections, local taxes are the mainsource of revenues for local governments.This includes local fees and own taxes andthe fees and taxes shared with the centre.

To improve local government control ofand the efficiency of local tax systems,they should take part in the tax process,both in identifying the tax base and incollecting fees and taxes. In the countrieswhere the administrative culture assignsthe identification of taxpayers and thecollection of fees and taxes to the nationallevel, a first improvement could be toestablish performance contracts betweenthe centre and the local authorities inorder to identify taxpayers and collectlocal taxes.

It is also equally advisable that areas ofshared national taxes be discussed withthe national local government associationbefore the adoption of the annual budgetlaw, and that this law should undergo atransparent design process. This wouldhelp avoid unexpected situations, such asthe 2009 Budget Law withdrawal in Benin,of the industrial tools and licence tax base(outillage industriel de l’assiette de lapatente) which had a considerablenegative impact on the revenues of localauthorities. During this dialogue, localauthorities should pay particular attentionto the taxes applied to economic activityand to land and property, and to thepossible sharing with local governments,taxes on personal and corporate income,as is done in other continents.

As far as economic activity is concerned,the main questions to be discussed con-


57

cern the sharing of the value added tax(VAT) and of the taxation of informalsector activities. With regard to thispoint, it would undoubtedly be useful ifnational local government associationstake the initiative to set up specializedcommittees on these topics and presentproposals, rather than waiting for thegovernment to come up with unilateralsolutions. This recommendation is evenmore applicable with regard to theinformal sector: local authorities are bestpositioned to identify and control theoperators in this sector and negotiatewith them both the tax level that seemsappropriate and how to identify thecorresponding taxpayers.

The efficiency of land and property taxesdepends on knowledge of the tax base. Tobenefit from this knowledge, it is ne-cessary to have tools such as a land re-gister, addresses or cadastral database,the creation and management of whichoften remain beyond the currentcapacities of most African local authorities,in particular those of medium and smalltowns and rural centres. For this reason itis important to discuss with the nationalgovernment the possibility of technicalsupport for the creation and managementof a geographical information system thatwill allow the identification of the use,appropriation and exploitation of land andnatural resources. This support shouldalso include the creation of technicalteams guaranteeing the updating andmanagement of this system. The creationof these kinds of instruments must alsobenefit from decentralized cooperationprogrammes. This project must be apriority as it is decisive for the creation ofa mortgage market whose current non-existence represents a significant liabilityfor the development of a dynamic,efficient land and property market withinAfrica’s municipalities.

In order to effectively mobilize revenue, localauthorities need to have a good knowledge oflocal fiscal potential and of the best way ofmobilizing it. If there is one field in which localauthorities have to emphasize their addedvalue, it is precisely in their ability to improvethe collection of taxes and fees which willcontribute to improving their service provisionto citizens, while justifying the financial legiti-macy of decentralization policies, which intheory, should provide an increased abilityto act, drawn from their stronger publicapproval.

Local officials must firmly commit to mobi-lizing local resources as this is the onlysustainable way of reinforcing local autho-rities’ financial autonomy. In addition tomobilizing own resources, they should alsorun campaigns on general fiscal awareness inorder to inform citizens on the good use madeof local resources. This is why local budgetaryreforms, in particular performance reportingto populations and central governments, theimprovement of local public expenditure, andthe tradition of internal audits all play animportant part in reinforcing the legitimacy oflocal authorities to their citizens and inimproving civic fiscal behaviour.

The effort local authorities make towardsfinancial autonomy must not be underminedby central budgetary management or unifiedtreasuries. For this reason it is necessary toprogressively relax controls on budgeting andreplace them with legal controls andincreased transparency. It is also necessaryto seriously consider the relaxing or evenabandoning of unified treasury systems incountries that include local governments inthem. If this is to occur, local governmentsmust adhere to stricter financial managementprocesses and increase the regularity ofreporting to their constituents and the centralauthorities. Adopting instruments such as theparticipatory budgeting could help in thisprocess.




4) A major request: better defined rules fortransferring financial resources from centralgovernments to local authorities.

Given the important share of transfers withinlocal budgets, it is necessary to make therules that govern the organization of thesetransfers as transparent and legible aspossible, in order to ensure the traceability oftransfers for the State, as well aspredictability for local authorities.

The dialogue on the organization of financialtransfers from the State to local authoritiesmust be based in the most preciseinformation available on the amounts andorigins of funds, as well as the modalities ofallocation (equalization transfers orconditional transfers). This information shouldbe transmitted to each municipality and madepublic in an official document such as anofficial bulletin to facilitate civil societydebate. This dialogue would be even morefruitful if it were organized within a frameworkof equally consultative partnership, Thesecould follow the pattern used in the nationalcommittees on local finances created in someStates –in particular within the West AfricanEconomic and Monetary Union (UEMOA)–, inwhich representatives from the FinancesMinistry, the Ministry in charge of localgovernments, and the national association oflocal authorities are brought together. Itwould be advisable that such frameworks alsoinclude representatives from the Parliament,from economic stakeholders and from civilsociety. They could also, if necessary, be opento representatives from developmentpartners.

Transfers should, to the greatest extentpossible, be predictable. For this reason it isadvisable in the context of Africa to set aminimum level of public resources that mustbe transferred by the State to local authoritiesevery year, and to ensure that these follow anobjective formula negotiated between the two

parties. Such a formula should take intoaccount the need to encourage localauthorities involved in local economicdevelopment, as well as the need to addressequity and equalization between localauthorities with different economic levelsand/or facilities.

5) A necessity that cannot be ignored:encouraging access of local authorities toloans and financial markets.

Although most States’ legislations allow localauthorities the ability to borrow from financialestablishments or to access financial markets,in reality this ability is only used by a verylimited number of African local authorities.The often drastic conditions imposed on localauthorities when attempting to access loansprovides a first explanation of this state ofaffairs. A second reason is the lack ofadaptation of the financial tools andrequirements that local authorities must useand meet to access credit. Thirdly, there is acertain level of ignorance both nationally andon the part of local authorities both of thepotential offered by the financial market andof the true debt capacities of local authorities.

Nevertheless, the tremendous need forinfrastructures and facilities in African localauthorities is obvious, and cannot be met onlywith the mobilization of local authorities’savings or even with the mobilization of Stateresources. It is necessary that localauthorities have access to loans and financialmarkets to meet the challenges raised by thecontinent’s rapid urban growth.

Delegating some local public services toprivate operators is a first method that manylocal authorities are currently using torespond to the challenges of urban growth.This approach is particularly useful for thecommercial services, where the income fromtheir operation generally justifies theinvestment.


59

In order for local governments to accessinstitutional financing from the financialmarket, they must have clear books ofaccount and improve their local governanceto be able to inspire the confidence offinancial operators. However it should alsobe the responsibility of the State to promoteinstruments that could facilitate localauthorities’ access to loans. The develop-ment of financial institutions specialized ingranting loans to local authorities or in theintermediation for local authorities’ accessto the financial market is a first step,provided that the work of these institutionsis better evaluated.

The almost complete absence of Africanlocal authorities from the bond market isalso a sign of the poor development ofAfrican financial markets and, undoubtedlyof the lack of preparation on the part ofAfrican central and local authorities toaccess potential revenue source that couldeasily be exploited if developed. This diffi-culty could be caused by the problem mostlymedium and small local authorities have inestablishing acceptable applications forfunds. The idea of a mutualisation ofdemands, along with technical supportorganized in the framework of a develop-ment fund for cities, as suggested by theWorld Organization of Metropolises (METRO-POLIS), is worth studying. A study shouldbe launched across Africa to analyze theconditions for making this proposal a realitywithin the context of the continent.Launching this reflection is a priority agendafor United Cities and Local GovernmentsAfrica (UCLGA).

The improvement and harmonization ofbudgetary and accounting procedures oflocal authorities is a pre-condition foraccessing loan. These procedures do indeedvary greatly between local authorities bothwithin the same country and between localauthorities from different countries. In

addition to this external and internalheterogeneity, there are several constrainswhich remain important: local accountingwhich is poorly adapted to decentralization,non-consolidation of the expenditureswithin municipal budgets, lack of builtcapital, etc. This work will increase thetransparency of local accounts, an essentialelement for developing a sustainablefinancial market.

6) A pressing need: strengthening humanresources for African local authorities, inparticular in the field of management andfinancial engineering.

This strengthening typically comes fromwhat local authorities can do themselves,via programmes and projects ofdecentralized cooperation, both at thenational level and at the continental/inter-national level. The local governmentassociations must learn from each other toimprove negotiation procedures withnational government and the follow-up ofbudget negotiations at the parliamentarylevel. These exchanges can contribute tothe modernization of local financial systemsso as to facilitate in the long-term theaccess to financial markets and promotecitizen participation in budgetary decisionsat the local level. One of the first tasks willbe strengthening the capacity of localgovernments and their associations toregularly follow and analyse, in eachcountry, their finances and financing.

The idea of creating a fund that wouldencourage the strengthening of capacitiesvia the exchange of staff and experts wassuggested by the participants of the latestAfriCities Summit held in December 2009 inMarrakech. UCLGA is currently studying thebest way to implement this proposal asquickly as possible with the support of allpartners wishing to participate.



0w2010 03 AsiaPacific DEFcarta ang 15/9/11 23:07 Página 60

3. ASIA-PACIFICBL ANE D. LEWIS*

LEE KUAN

YEW SCHOOL OF PUBLIC POLICY

NATIONAL UNIVERSIT Y OF SINGAPORE

SINGAPORE

BOB SEARLE*

CONSULTANT

AUSTRALIA


* The authors thank participants in the UCLG GOLD II

ASPAC Workshop in Guangzhou, China, on 10

November 2009 and attendees at the UCLG ASPAC

Workshop on Local Finances in Batam, Indonesia, on

25-26 June 2010 for useful comments on drafts of

this Chapter, especially Sung Il Lim of the Korea

Research Institute for Local Administration. Blane

Lewis was principally responsible for the work on

South Asia, East Asia and South East Asia; and Bob

Searle for Australia and New Zealand.


63

I n most countries in the Asia-Pacific Region,the context in which fiscal decentralization

policymaking and implementation takes placedoes not usually change radically over shortperiods of time. Nevertheless, a number ofcountries in this region have recently seenquite significant changes to the structure anduse of local government.

In July 2009, the President of Pakistan post-poned local elections based on theunanimous decision of governors. In themeantime, governors appointed “non-political”administrators to replace elected mayorsand vice mayors; and the magistracysystem has been revived. As such, the veryexistence of local government in Pakistan isunder debate.

In China, the intergovernmental system hasin the recent past concentrated on promo-ting economic development, and this em-phasis has generated a number ofnoteworthy successes. Given these achieve-ments, Chinese leaders now appear poisedto re-focus their decentralization programon delivering quality local public services.The change will be operationalized as part ofthe government’s renewed attempts toaddress equity and poverty concerns in thecontext of its recently initiated program to“build a harmonious society”.

In 2003, Japan’s government launched abroad set of “Trinity Reforms”, which ithoped would ease many of the long-standingconstraints on local government operations.The overarching goal of the reforms is toprovide sub-national governments withmore fiscal autonomy: particular objectivesfocus on reducing sub-national governmentreliance on specific purpose transfers fromthe central government, increasing accessto own-source revenue and streamlining theuntied equalization grant. It is too early tojudge whether these objectives have beenachieved.

In Indonesia, recent changes to laws en-vision the eventual decentralization ofproperty tax to the local level. This changehas the potential to significantly increasethe amount of own-source revenuesavailable to local governments but alsoportends some daunting administrativechallenges. More broadly, the governmenthas begun to outline revisions to its lawson both administrative and fiscal decentra-lization with a view to again improve thelegal framework introduced in 1999.

Cambodia passed an Organic Law onDecentralization and DemocraticDevelopment in early 2009 and is now inthe process of formulating theimplementation plan to establish districtand provincial administrations asintermediate tiers between the centralgovernment and the communes. In Nepal,the writing of a new constitution isprogressing (with completion due in 2011)and will include, a clarification of the rolesand responsibilities for the tiers of localgovernment, provisions for a more securebase for local elected officials, and makefunds transfers more transparent andformula-based.

In general, it might be said that localgovernment in the countries with moredeveloped economies is undergoingreform to the structures and procedureswithin which it operates. On the otherhand, local government in the nations withdeveloping economies is at an earlierstage in its development.

This chapter describes and analyzes localgovernment finances in the Asia-Pacificregion. It is divided into separate sectionsfor South Asia, East Asia, South East Asia,Australia and New Zealand. Each sectionconcentrates on the experience of thecountries for which a reasonable amountof reliable primary and secondary



AASSIIAA--PPAACCIIFFIICCUnited Cities and Local Governments64

information is available. In South Asia,the local government finances of India,Pakistan, and Bangladesh are highlighted,with some comments also provided onNepal and Sri Lanka. In East Asia, thefiscal decentralization systems of China,Japan, and Korea are reviewed. Thesection on South East Asia examinesthe intergovernmental framework inIndonesia, Philippines, Vietnam, andThailand, and provides comments onCambodia and Malaysia. The countriescovered in Asia are among the mostprominent in their respective regions andaccount for over 50 percent of theworld’s population. The section on the Oceaniasub-region covers Australia and NewZealand. It has in some cases been difficultto differentiate between local governmentand other levels of sub-nationalgovernment. Generally, it can be assumedthat the lower tier of sub-nationalgovernment equates to local government.

In examining the structure andperformance of each country’sintergovernmental fiscal system, thechapter focuses on expenditures,revenues, intergovernmental transfers,and debt finance. A variety of input andoutput criteria are taken into account inevaluating performance, including theadequacy of local government resources,local tax and spending discretion, incen-tives in the intergovernmental transfersystem, local government administrativeand management capacity, and the qualityof local public services. In addition, thechapter highlights a number of pressingconstraints on local finances across sub-regions and countries, and provides obser-vations for improving fiscaldecentralization policy for selectedcountries where sufficient data wereavailable. A final section of the papermakes some broad conclusions thatmay be important for the reform of de-

centralization policy across countries inthe Asia-Pacific region.

South Asian Local GovernmentsFinances

Structure

Expenditure1

Sub-national governments in India contri-bute up to about two-thirds of consolidatedgovernment spending, one-third of whichcomes from states and local governments(i.e. “local bodies”). State governments areempowered to constitute lower levelgovernments, including urban and rurallocal bodies (ULBs and RLBs)2. Of total localexpenditure, 85 to 90 percent—is ULBspending concentrated in the largest urbancenters; the remainder is from RLBs, withvillage Panchayats responsible for about 85to 90 percent of rural local governmentexpenditure. There is, however, a great dealof variation across states and localgovernments. Approximately 60 percent oftotal local government spending goes to“core functions”, including water supply,street lighting, sanitation and roads. Notsurprisingly, ULB spending is more concen-trated than RLB spending on these corefunctions.

Legally, Pakistan has a three tier system ofsub-national governments although thesewere all dissolved as elected bodies inJuly 2009. Prior to that, sub-nationalgovernment spending amounted to aboutone-third of consolidated public sectorexpenditure, with most sub-nationalspending carried out by provinces asopposed to local governments (Districts,Tehsil Municipal Administrations—TMAs,and Unions3). In theory, districts have ex-penditure responsibility for elementaryand secondary education, primary healthcare, and the environment; in practice

1. See Table 3.1.

2. The 3682 Rural LocalBodies are eitherPanchayats (of whichthere are three tiers)or Autonomous RuralCouncils; while the247 033 Urban LocalBodies are knowneither as MunicipalCorporations,Municipalities or NagarPanchayats.

3. There are, in total,over 6600 localgovernment units inPakistan.


65

these functions appear, for the most part,to be carried out exclusively by or inconcurrence with provinces. TMAs areresponsible for basic municipal servicesand building control. Unions perform veryfew minor functions.

The sub-national government share of totalspending in Bangladesh is more limited thanthat of India and Pakistan, amounting to 15percent or less. Local governments(Union Parishads, Pourashavas and CityCorporations) account for an estimated two-thirds of the sub-national total. UnionParishads’ spending focuses on agriculture,road construction and maintenance, andhealth and sanitation; Pourashavas’expenditure is mostly devoted to wastemanagement, water supply, sanitation, anddrainage; City Corporations concentratetheir spending on the usual urbaninfrastructure services. Many of thesefunctions are carried out concurrently withdeconcentrated agencies of the centralgovernment.

In Nepal, the local governments’ share oftotal public sector expenditure is also verysmall, but perhaps surprisingly highconsidering the state of development oflocal government and the uncertaintysurrounding future levels and spheres ofa c t i v i t y u n d e r t h e c o n s t i t u t i o nnow being rewritten. The pressure fromlocal government is for the constitution tolimit the functions of the centralgovernment and, under the subsidiarityprinciple, give more responsibilities tosub-national government.

Own-Source Revenues

Indian state and local government own-sourcerevenues comprise just over one-third oftotal consolidated public revenues. Stategovernments assign revenue sources tolocal governments but only municipalities

have any real tax powers. The localgovernments’ share of total sub-nationalown-source is 10 percent or less, dominatedby ULBs, which collect over 90 percent oflocal government revenue. Property taxesare the most important urban governmentown-source revenues. The octroi (a tax ongoods brought into the locality fromoutside) was important in the past but hasnow been abolished in all but one state.RLBs derive most of their own-sourcerevenue from various minor licenses, fees,and charges. The small share of rural localgovernment own-source revenue isconcentrated at the gram parachat level,with the vast majority of all rural revenuebeing derived from transfers.

Revenue raised by the sub-nationalgovernment in Pakistan is only aboutseven percent of total public revenue.Again, the majority of sub-national own-sourcerevenue is from provinces. Local own-sourcerevenues were only 8 percent of localbudgets in 2008-09, suggesting that suchrevenue is less than one percent of totalpublic revenues (Bahl and Cyan, 2008).Most district revenue comes from a varietyof charges and fees. In theory, theproperty tax has been assigned to TMAsbut the provinces administer this tax inmost areas. Unions have no revenueraising authority.

Sub-national government own-sourcerevenues in Bangladesh are even morelimited than in India and Pakistan. Totalsub-national government own-sourcesamount to less than two percent ofconsolidated public revenues and there isno information on how the sub-nationaltotal is distributed across different levelsof governments. For Union Parishads themost important own-source revenue is theholding tax on the value of land; forPourashavas and city corporations, themajor own-source revenues are those




from the property tax, land transfer taxand building permits.

Intergovernmental Transfers4

As Table 3.1 shows, local governments inIndia experiences a very high level of verticalfiscal imbalance (revenue is only 3 percent oftotal public revenue while expenditure is 33percent of national total). Intergovernmentaltransfers to local governments in India makeup nearly 90 percent of their total revenues,with transfers being received from bothcentral and state governments but mostcoming from the latter as central transfers tolocal governments are mostly distributedthrough states. State transfers to localgovernments, including both revenue sharingand grants, comprise around one third of totalstate expenditures. Both types of transfersappear, in most cases, to be rather ad-hocand tied to particular kinds of spendingmandated by the grantor. On a per capitabasis, transfers to rural local governments aresignificantly larger than those to ULBs butthere are significant deviations from thesenorms (Rao, 2000b; Oommen, 2008). In partthese variations relate to the differential rolesplayed by the State Finance Commissions(SFCs), which are charged with determiningappropriate revenue sharing and grants-in-aid allocations.

Service delivery by Pakistani local govern-ments is also financed largely throughintergovernmental transfers. At thedistrict government level, transfers makeup 90 percent or more of total revenues(Kardar, 2006), but Karachi is theexception to this rule with about 20percent of its revenue coming from own-sources. TMAs are somewhat less relianton provincial transfers, with transferscomprising 60 to 80 percent of totalrevenues, depending on the province anddistrict of location. Unions are financedexclusively by a grant from provinces.Most intergovernmental transfers to thelocal level are general-purpose in nature,although provinces also provide specificproject grants.

Intergovernmental transfers also dominatelocal government revenues in Bangladesh,with transfers constituting 90 percent ormore of the total. There is considerablevariation, however, with rural local govern-ments (Union Parishads) being particularlydependent on grants and urban localgovernments (Pourashavas and CityCorporations) somewhat less so. The mostimportant transfer is the block allocation.While it is the largest grant, it is still verysmall from a national perspective, makingup less than two percent of the state4. See Table 3.2.

Share of Total Public Expenditure (%) Share of Total Public Revenue (%)

Country Sub-national Upper Tier Lower Tiers Sub-national Upper Tier Lower Tiers

India 66 33 33 33 30 3

Pakistan 33 28 5 7 6.5 0.5

Bangladesh 15 5 10 2 1 1

Nepal 9 -- 9 14 -- 14

Source: UCLG data files and other sources. Figures are approximations. Data is from most recent year available: India 2004, Pakistan 2004, Bangladesh 2005,Nepal 2008.

Table 3.1: Sub-national Government Share of Total Public Expenditure and Revenue – South Asia


67

budget. This general-purpose transfer isintended to fund local developmentactivities. Other development grants arespecific-purpose in nature, often focusedon particular projects (some of which arefunded by international donors). Localgovernments in Bangladesh also receivesmall grants to support the salaries andallowances of local civil servants.

Debt Finance5

Indian sub-national governments arepermitted to borrow and many do so.Resources borrowing include centralgovernment, donors (indirectly), public andprivate financial institutions, and capitalmarkets. Returns to bondholders are taxfree. Borrowing to cover budget deficits iscommon at all levels of government, andconsolidated borrowing to cover deficits hasrecently reached around 15 percent of GDP.About 80 percent of this borrowing is doneby state and local governments. Mostworryingly, much of the borrowing appearsto finance recurrent rather than capitalexpenditures. The majority of localborrowing is by urban local bodies, withmuch of it concentrated in the states ofAndhra Pradesh, Gujarat, Karnataka,Madhya Pradesh, Maharashtra, and Punjab,

and the widely disseminated experience ofTamil Nadu with the Tamil Nadu UrbanDevelopment Fund (see box 3.1). Statesformally guarantee the repayment of localgovernment loans. The totality ofoutstanding state guarantees extended tolocal governments may amount to as muchas 10 percent of GDP or slightly less(Pradhan, 2004).

Although Pakistani provinces may legallyassume debt, local governments may not.Local governments in Bangladesh arepermitted by law to borrow in order to financecapital development. There is no data onamounts borrowed by local governments ofany type, although it is clear that it has notamounted to much. It seems all localgovernment borrowing is done through thecentral government’s on-lending channels(Boex et al, 2002).

In Nepal, the Town Development Fundlends to municipalities for infrastructuredevelopment, but other tiers of localgovernment have no access to borrowing.In Sri Lanka, local government borrowingis from the Local Loans and DevelopmentFund of the central government but thereare insufficient funds to meet localgovernment demand. 5. See Table 3.3.


Relative Importance of Type of Transfer

Country Share of Transfers in Total Local Revenue (%) Revenue Sharing General Purpose Special Purpose

India 90 Medium Low High

Pakistan 90 Low High Medium

Bangladesh 90 Low High Medium

Nepal n.a. (but heavily dependent) n.a. n.a. n.a.

Sri Lanka n.a. (but heavily dependent) n.a. n.a. n.a.

Source: UCLG data files and other sources. Figures are approximations. Data are for most recent year available: India 2004, Pakistan 2004,Bangladesh 2005.

Table 3.2: Intergovernamental Transfers as Local Government Revenue Source – South Asia



Performance6

Current and Capital Resource Adequacy

In India, municipal governments generallyoperate under severe fiscal constraints. This

can be attributed both to their narrow andinelastic tax base and an indifferent attitudeto the use of funds. Within this context, localgovernments are not able to generatesufficient resources to meet even theoperating and maintenance costs of the6. See Table 3.4.

Box 3.1. Municipal Bonds in India

The Tamil Nadu Urban Development Fund (TNUDF) was established in 1996 fordevelopment of urban infrastructure at the state level. It was the first public-privatepartnership providing long-term debt for civic infrastructure and was not stateguaranteed. The unit capital is contributed by the government of Tamil Nadu (49percent) and three Indian Financial Institutions with international finance support(World Bank, Japan Bank for International Cooperation and the German KfW). TheTNUDF also raises funds from capital markets.

This fund has been an important first step towards linking domestic capital markets andthe financing need of small cities. The eligible borrowers are urban local bodies(Corporations, Municipalities and Town Panchayats) and private institutions creatingurban infrastructure.

Since the launch of the Jawaharlal Nehru National Urban Renewal Mission (JNNURM) in2005, the role of financial markets in financing sub-national governments has increasedsharply. The JNNURM has two main missions: a) the development of urbaninfrastructure and governance, and b) the development of basic services for the urbanpoor.

More generally, municipal issues are provided as revenue bonds, with fixed interest rateof around 12-14 percent, with or without government guarantee. Their maturity isbetween 7-15 years and they are structured obligations. Investor income can be eithertaxable or tax free.

Country Borrowing Legally Authorized Extent of Borrowing

India Yes Moderate

Pakistan No None

Bangladesh Yes Very Limited

Nepal Yes (but only by Municipalities) n.a.

Sri Lanka Yes limited

Source: UCLG data files, among others sources.

Table 3.3: Legal Authorization and Extent of Local Government Borrowing – South Asia


69

services they provide. Only a few largeurban local bodies in India (the million pluscities) appear to have sufficient revenuefor service delivery. Large urban centerscan have significant own-source revenues,making up well over 50 percent of theirtotal revenues, but some of this flows toparastatals they use to provide services on acost-recovery basis. Nevertheless, they areless dependent on transfers than other localgovernments, which are severely resourceconstrained, especially in rural areas.Rural bodies have few own-source revenuesand are almost completely dependenton transfers from the central and stategovernments. Most state governments alsofind themselves in precarious fiscalconditions and this in turn limits theextent to which they can transfer resourcesto local governments, thereby extendingfiscal constraints to the lowest levels (Rao,2000b; Oommen, 2008).

In 2005, the Indian government began alarge urban renewal program andextended grant assistance tomunicipalities for urban infrastructure andgovernance. Based on city developmentplans, tied grants will be made availablefor better basic services for the poor,urban land development, and majorinfrastructure developments.

Local governments in Pakistan also sufferfrom harsh resource constraints. Access totax revenue is very limited at the local level,as noted above, and transfers from provincesto local governments are not usually sufficientto meet residual funding needs. As expected,the more constrained a Pakistani province isdue to lack of transfers received from thecentral government, the more constrained itstransfers are to local governments. Localgovernment capital needs must be met out ofown-source revenues and transfers as theyare not allowed to borrow. This furtherexacerbates their capacity to produce

adequate levels of services (Manning et al,2003; Bahl and Cyan, 2008).

Bangladeshi local government resources arealso extremely constrained. The resourcesavailable are exceptionally limited to allowthem to satisfy the meager serviceresponsibilities that have been assigned tothem at minimally acceptable levels (Boexet al, 2002; Fox and Menon, 2008).

In Nepal, local governments cannot fund alltheir recurrent expenditure, so that revenuereassignment or grant increases must beachieved if services are to be provided. InSri Lanka, the provision of services andmaintenance of infrastructure are amongthe most important issues facing localgovernment.

Taxing and Spending Discretion

State governments in India determine thekinds of services that local governmentsshould deliver and the particular taxes thatthey may levy (based on recommendationsfrom the SFCs). The central governmenthas allowed for reasonably significantexpenditure responsibility and tax autono-my to be decentralized to local governmentsin recent amendments to the constitutions.Decisions regarding implementation ofthese provisions, however, rest with thestates, which generally have not assignedextensive fiscal powers to localgovernments. Of course there are ex-ceptions to this rule, including Karnataka,Madhya Pradesh and Kerala for example,which have been more proactive indevolving expenditure and revenue assign-ments (World Bank, 2007).

Indian local governments have limiteddiscretion over the taxes assigned to them.State governments exert much control overdefining bases and setting allowable rates oftax. Local governments do, however, have




reasonable spending autonomy over theirown source revenue where restrictions do notseem particularly onerous. This is not thecase with transfers. The vast bulk of revenuesharing and grants-in-aid is tied to particulartypes of spending or projects that the statesdeem important. In this regard, localgovernments are better viewed asimplementing agencies of the states thanautonomous local entities. Service deliverydiscretion is further constrained by theconcurrent nature of almost all assignedfunctions and the parallel institutionalarrangements (essentially local level para-statal bodies of the states) that have beendeveloped for the delivery of services such asslum improvement and water provision.

Law allows for significant fiscal respon-sibilities to be devolved to the local level inPakistan, but provinces determine preciseexpenditure and revenue assignments oflocal governments under their purview.Expenditure assignments could vary quitewidely across provinces but are in generalquite limited. Pakistan’s constitution allowsfor the complete separation of tax basesacross levels of government but provincialgovernments have access to very limitedsources of revenue. This in turn constrainsrevenue assignments at the local level.

As is the case in India, local governments inPakistan have little discretion over the taxesthat are assigned to them by provinces andcan only change taxation arrangements withthe approval of the provinces which definelocal tax bases (including exemptions) andset local tax rates. In addition, provinceshave authority to assess property valuesused in determining liabilities related to theproperty tax, ostensibly a TMA revenue.

Local governments in Pakistan also havelimited control over spending on servicedelivery. Provinces exercise significant controlsover both recurrent and development (capital)

spending. Provinces prescribe spendingpriorities in local budgets and sometimes evenrequire ex-ante approval for local governmentsto incur expenditures. Provinces specify indetail the expenditure objects targeted intransfers that they make to local governments,especially those for capital spending, and makebalanced budgeting a condition of grantreceipt. In extreme cases, provinces have usedassessments of poor performance to take backcontrol of functions originally decentralized.

Most employees of local governments inPakistan are provided by other tiers ofgovernment. Local governments have little orno authority over assigned staff and cannotfire civil servants appointed by other tiers.Federal and provincial pay increasesestablished in law are simply passed throughto local governments without regard to thelocal budget or to local government’s abilityto pay. This constitutes a significantunfunded mandate.

As is clear from previous discussion, publicexpenditures and revenues are verycentralized in Bangladesh. The centralgovernment dominates service delivery evenin education and health where relatedfunctions in other nations have beendevolved to local governments. Similarly thecentral government keeps a tight reign onalmost all sources of revenue.

Bangladeshi local governments have verylittle decision-making authority, even overtaxes and services under their responsibility.The district council (a deconcentrated unit atthe regional level) must approve all localbudgets and local tax and fee rates are set bythe central government. In addition, mostblock grants are earmarked. Central officialsare actively engaged in local governmentdecisions on spending priorities and thecentral government must also approvedevelopment projects prior to implementa-tion. Essentially all local government staff


71

members are employees of the centralgovernment so, while mayors and councilsare popularly elected, they have very littleauthority over their staff. Mayors cannotappoint independent (i.e. non centralgovernment) staff without the approval ofthe center and such approval is verydifficult to obtain. In many ways, localgovernments in Bangladesh are deconcen-trated units of other tiers of government.

In Nepal, local governments have very littlepower to set the taxes and fees for therevenue sources available to them, and lackof clarity of revenue assignments results ingovernment Ministries encroaching on localgovernment revenue mandates. There issimilar overlapping of responsibilities forservice delivery that creates confusion andnon-provision by the poorly-resourced localgovernment sector. In Sri Lanka, there is also

confusion in the allocation of revenue basesbetween the central and local spheres ofgovernment.

Incentives in the IntergovernmentalFramework

Many Indian states have integrated per-formance incentives into revenue sharingschemes for local governments. Theseincentives are operationalized as weights tovariables used to allocate revenue sharesand are intended to encourage good localgovernment behavior in revenue effort,service management, etc. However, theweights are developed exclusively, based onpast behavior and may therefore be seen asrewards for past behavior rather thanencouraging future performance. Generally,empirical evidence suggests thattransfers are negatively associated with


Table 3.4: Performance of Intergovernmental System – South Asia

Weak outcomes.Very limited outside largeMunicipalities.

n.a.Limited discretion andconfused mandates.

Inadequate.Sri Lanka

Weak outcomes.Very limited outside largeMunicipalities.

These have only recentlybeen introduced.

Limited discretion andconfused mandates.

Grossly inadequate.Nepal

Weak service outcomes.Overt focus on magisterialand police functions.

Weak and corruptdeconcentrated taxadministration.Management of servicedelivery mostly a CGconcern.

Some effective use ofincentives to encourage“local” revenue effort.

LG discretion extremelylimited. LGs essentiallydeconcentrated units of CG.

Resources of LGs severelyconstrained.

Bangladesh

Weak service outcomes. Reasonably sound taxadministration at districtlevel. Decent expendituremanagement but mostlycarried out by centralemployees.

No experience withperformance incentives.

Limited LG discretion overtaxes and spending.Provincial controlubiquitous.

Most local governmentssuffer harsh resourceconstraints.

Pakistan

Weak service outcomes.Weak tax administration andexpenditure management.

States have madesignificant use ofincentives but rewardsfocus on past not futurebehavior.

LGs have very little fiscaldiscretion. States controllocal taxing and spending tolarge extent.

Large urban LG resourcesreasonably adequate; smallurban and ruralgovernments severelyresource constrained.

India

Service OutcomesManagement CapacityPerformance IncentivesFiscal DiscretionResource Adequacy

Source: Author’s assessment based on UCLG data files, among other sources (see bibliography)



local own-source revenue generation (Mathurand Peterson, 2006).

Pakistani provinces have not, in general,made much use of performance incentivesin their dealings with local governments.Most provinces have not incorporatedincentives into the allocation of transfersthat they make to local governments(Peterson, 2002).

The Bangladeshi experience with incentiveshas been somewhat more positive. Thelocal revenue raising incentives createdby the intergovernmental transfers, forexample, appear to be relativelyconstructive, or at least benign. Transfersare unlikely to discourage local revenueefforts and may even encourage it. It’sclear at least that increasing grantlevels are positively associated with localown-source revenues. On the other hand,the limited discretion local governmentshave over spending, including those fromtheir own-sources, may well dampenenthusiasm for increasing collections (seeFox and Menon, 2008).

In Nepal, the grant system is supposed tobe formula based but is ad-hoc in practice,with no criteria specified in legislation orpolicy documents. They are, however,experimenting with minimum conditionsand performance measures in village anddistrict development committees andmunicipalities. In Sri Lanka, local govern-ments are overly dependent on transfers,but their lack of regularity and predict-ability is a problem for fiscal management.

Tax Administration and ExpenditureManagement

Indian local government tax administrationand expenditure management is quite weak:a prescribed budget process, which includesformulation of a plan, its approval by local

elected officials and implementation, is notgenerally operational (Bahl et al, 2005). Astudy in Rajasthan found that localgovernments there made no use of optionaltaxes under their authority. The localofficials’ explanation was that councilsand/or citizens would be reluctant to pay,even if they resulted in improved servicedelivery. This attitude could also explain localtax collection inefficiency. On the serviceside, local governments do not generallyseek additional responsibility, a function oftheir unwillingness to assume either greatermanagement burdens or the financial costsassociated with the increased responsibility.There are, however, some reforms andinnovative experiences (see Box 3.2) thatcould be used as examples by other statesand local governments (Mathur andPeterson, 2006).

The system of local tax administration inPakistan is unusual. District governments havetwo types of tax collection units, one to collectprovincial taxes and another that focuses onthe administration of district revenues. Thecapacity of both tax administration agencies isreasonably sound, although districtsthemselves only benefit to a certain extentfrom such capabilities. TMA tax administrativesystems have a long history but theiradministrative capacity is weaker than that ofdistrict governments. Lack of training amongtax collectors is a particular problem (Peterson,2002; Bahl and Cyan, 2008).

The quality of local expenditure manage-ment capacity is also reasonable at the localgovernment level, at least from a technicalperspective. Unfortunately, most civilservants working in district and TMAgovernments are federal or provincialemployees (but paid from local budgets)and can be transferred at the discretion offederal or provincial managers. As such,existing local management capacity can beillusory and temporary.


73

While the revenue authority of localgovernments in Bangladesh is limited, theycould be getting more out of their revenuesources than they do. Many municipalgovernments have made little effort to expandthe property tax database or conductvaluations in a timely manner, and collectionefficiency is low. Moreover, there arewidespread reports of corruption in taxcollection and an apparent culture of taxavoidance and evasion adds to the problem.Some estimates indicate that only about onequarter of the businesses in Pourashavasregularly pay their tax liabilities (Fox andMenon, 2008).

The management of spending and servicedelivery is the responsibility of higher tierrather than local governments in Bangladesh.Local governments seem to be heldresponsible for the local accountabilityfunction but not for the management anddelivery of services. This “implicit bifurcation”of regional and local public administration haslong been an issue and creates poor incentivesfor management performance and servicequality (Fox and Menon, 2008).

In Nepal, the very limited capacity of localgovernment staff is currently being addressedthrough donor funding, but much still needs tobe done. As might be expected after aprolonged period of civil unrest, Sri Lanka has

a particularly large need for capacitydevelopment within local government staff.

Service Delivery Outcomes

Local public service outcomes in India aregenerally inadequate (World Bank, 2004).Basic education, health, water and sanitationhave been judged to be some of the weakestin the world. Illiteracy is high, local publichealth systems are non-functional in manyparts of the country, quality of drinking wateris less than minimally adequate and generalliving conditions are often unsanitary. Accessto social services by the rural poor is par-ticularly problematic. There are, of course,exceptions.

Pakistani local public service outcomesare similarly sub-standard, but variationis extreme across provinces and localgovernments. A contributing factor topoor public services is the apparentindifference and corruption of the nation’sbureaucrats: a problem for all tiers ofgovernment. Citizen surveys in Pakistanrate the honesty and responsiveness oftheir civil servants much lower than dothe citizens of other countries in SouthAsia (Peterson, 2002).

Bangladeshi service outcomes are very weakacross the board. Local governments in


Box 3.2. Reform of Property Taxation in Andhra Pradesh

Prior to the changes, property taxes were only 13 percent of municipal revenue and 0.1percent of State GDP. The tax suffered from being based on difficult annual rentalvaluation provisions, a complex set of exemptions and high administrative costs.

The state clarified and simplified the valuation system, to make it more objective,standardized but widened the exemption scope to improve cost: benefit ratios, madepayment easier by involving the banking sector and provided grant incentives tocouncils to improve collection rates relative to demand notices. Increases in collectionswere generally over 25 percent.



Bangladesh are reputed to be inefficient andunaccountable providers of public services.Local bureaucracies tend to be isolated fromcitizens and are, in general, seen to be highlypoliticized and corrupt. Local administrationsare more given to political and social controlthan they are to the delivery of typical publicservices. To the extent that public services areemphasized at all, they tend to be morerelated to magisterial and police functionsrather than those that might be more usefulin raising the welfare of citizens (Boex et al,2002; Fox and Menon, 2008).

Selected Pressing Issues and BindingConstraints

The most pressing issue in India is the generallack of fiscal decentralization to urban andrural local bodies. Despite having theconstitutional power to devolve significantspending and revenue authority to localgovernments, the states have not done so inany meaningful way. While some localgovernments have unilaterally started takingon more responsibility for service delivery andtax collection a more proactive policy needs tobe adopted. Nevertheless citizens often seemto be reluctant to pay for more and/or betterlocal services (and where they may be morewilling they have few mechanisms throughwhich to register their willingness). Fordecentralization to proceed, improving thecooperation between local governments andtheir citizens must be a priority (Mathur andPeterson, 2006).

A second area of concern relates to theperformance of the State FinanceCommissions (SFC). As noted above, SFCs arecharged with a wide range of duties, includingdetermining local tax assignments, derivingappropriate local shares of state revenues,and establishing the appropriate amount ofgrants-in-aid to local bodies. The performanceof SFCs on all these counts has often been dis-appointing. Where SFCs have been proactive

and focused, their recommendations havelargely been ignored by the states.

Since July 2009 when the President ofPakistan dissolved elected local governments,reportedly based on the unanimous decision ofprovincial governors, the very existence oflocal governments in Pakistan has been underdebate. Beyond that most basic of issues, twoothers deserve to be highlighted. One is therole of the provinces in decentralization.Similar to their state counterparts in India,provinces in Pakistan determine the extent towhich decentralization proceeds within theirjurisdiction. So far, they have decided thatdecentralized public service provision andfinancing stops, for the most part, at theprovincial level; and the central governmenthas not stepped in to override this stance. Thesecond issue concerns the extremely low levelof local government own-source revenues andthe lack of any linkage it has to servicedelivery and cost recovery. Although limitedown-source revenue is often a problem indeveloping countries, it is especially acute inPakistan. Moreover, it has not been a priorityof central policy-makers in the past and it isnot clear that they see lack of tax authorityand discretion as a problem (Bahl et al, 2008).

The key issue in Bangladesh is the extent towhich government wishes to genuinelydecentralize its public sector to the local level.Currently, the system for the provision andfinancing of public services is more akin todeconcentration rather than decentralization.Local governments have very limitedresponsibility for delivering services and raisingrevenues and next to no discretion over theservices and taxes that have been assigned tothem. All civil servants at the local level areessentially central government employees.Significant reorganization would be necessaryto implement a more decentralized system.

A second set of important issues relates to thecomplexity of the Bangladeshi government


75

structure. The five separate levels ofoperation results in increased costs ofadministration, and an excessive number ofofficial levels of government. This is furthercomplicated by the existence of other, parallelinstitutional arrangements for service deliveryoperating alongside the official structure. Inaddition, the complex structure seems itselfto be constantly in flux, as each newgoverning administration seeks to use thesystem to further its own and differentpolitical ends. Again, the effective operationof any system of service delivery andgovernance, regardless of its deconcentratedor decentralized orientation, is made difficultby the consistent and sweeping changes overtime (Fox and Menon, 2008).

In both Nepal and Sri Lanka, the pressingissues relate to the current state ofdevelopment of local government and theneed to clarify legislation and improve thecapacity of local government to manage andprovide service delivery. Nepal particularlyneeds to give more responsibilities andrevenue sources to local government.

Conclusions and Observations on SouthAsia

India may wish to revisit the strategy ofletting states decide on the extent ofallowable decentralization to local bodies.The voluntary approach has so far resultedin little success. Policymakers mightconsider whether a mandatory approach,under the careful direction and guidance ofthe center, might be a more viable way toproceed.

In this regard, the Indian FederalGovernment should consider a number ofactions:

• Devolve specific functions to localgovernments to reinforce local demo-cracy: where concurrent responsibilities

are deemed necessary, respective res-ponsibilities of the states and localgovernments should be defined;

• Award more revenue autonomy and dis-cretion over the use of revenue instru-ments to local governments;

• Give exclusive authority to local govern-ments over the property tax and taxeson motor vehicles; and give access topersonal income tax, via a piggybackscheme, at least to some urban localgovernments;

• Develop a transparent and predictablesystem of formula based transfers tolocal governments, so as to provideadequate revenues, equalize fiscalcapacities to meet expenditure respon-sibilities and improve fund distribution;

• Provide local urban and rural bodies withcapital grants;

• Promote a more structured budget pro-cess, along with improved auditing;

• Facilitate access to loan funds for localgovernments to finance capital ex-penditure, and develop their credit-worthiness and administrative capacity.Existing regulations concerning muni-cipal access to borrowing should bereduced or streamlined.

In addition, the central government could bemore pro-active in supporting and guidingState Finance Commissions (SFCs). SFCs needincreased capacity to carry out their assignedroles more effectively and, with increasedcapacity, the government might considermaking SFC recommendations binding likethose of the Central Finance Commission.

In Pakistan, the first policy priority for thecountry should be to revive the system of




elected local governments. In particular,Pakistan should:

• Reconsider the practice of allowing pro-vincial governments to decide the extent ofdecentralization at sub-provincial levels,which has led to limited decentralization.

• Revisit the basic objectives on whichintergovernmental fiscal relations arebased.

• Devolve more expenditure responsibili-ties and tax authority to local govern-ments.

• Assure the development of a modernsystem of fiscal transfers.

Bangladesh first needs to decentralize theirsystem of public service provision andfinance. Then a number of policy questionsautomatically present themselves. Likeother nations in south Asia, these relate toexpenditure assignments, tax assignments,intergovernmental transfers, localborrowing and methods of vertical andhorizontal accountability. A first priority,however, should be to make local officialsmore accountable to their citizens by havingthem engaged by the local governmentsand giving elected officials more controlover the hiring and firing of these staff.

East Asian Local GovernmentsFinances

Structure

Expenditure7

Expenditures are extremely decentralized inChina, where sub-national governments8

are responsible for about 70 percent ofrecorded total public spending. Local (i.e.sub-provincial) governments account formost of this, contributing about 50 percent

of the recorded total, but there are alsosubstantial amounts that are not recorded.Local government expenditure budgetsfocus on education and health services: andthe bulk of public sector spending ineducation and health —as much as 85percent— is carried out by local govern-ments. Chinese local governments areinternationally different from most in thatthey also devote reasonably large shares oftheir budgets to social insurance schemes ofvarious kinds (including unemploymentcompensation, health insurance, andpensions), functions usually carried out bythe central government (Bahl and Martínez-Vázquez, 2009).

Expenditures are also quite decentralized inJapan. Japanese sub-national governments(47 prefectures and 1820 municipalities)comprise about 60 percent of total publicexpenditure, about two-thirds of which is doneby municipalities. Sub-national governmentsfocus on infrastructure, education and socialwelfare, which respectively constitute 25percent, 18 percent and 18 percent of localgovernment expenditure. These shares arereasonably similar for both prefectures andmunicipalities.

Following the strong move towards de-centralization in the 1990’s, spending is alsoquite decentralized in Korea. The relativesize of Korean sub-national government hasincreased over 5 fold since these reformsand sub-national governments (specialmetropolitan area of Seoul, 6 metropolitancities and 9 provinces) and lower-levelsub-national governments (75 cities, 86counties and 69 districts) now accountfor about 45 percent of total publicexpenditure. Adding expenditure by localeducation authorities, which areindependent of local governments, wouldincrease this to about 60 percent, such thatlocal governments dominate the publicsector. The upper level of sub-national

7. See Table 3.5.

8. There are 34provinces, 333prefectures, 2862counties or districtsand 41 636 townshipsin China.


77

government spends about 40 percent ofthe sub-national total. Local governmentexpenditure traditionally focused on publicworks and economic development but socialservices, including safety, health, housingand community development, have recentlybecome the priority.

Own-Source Revenues

Taxes are somewhat centralized in China,although not as much as in some othercountries with the same degree of expendituredecentralization. The sub-national portion oftotal public sector tax revenues amounts toabout 40 percent. The shares of total publicrevenues of provinces, prefectures andcounties/townships are approximately 15percent, 15 percent and 10 percentrespectively, although there seems to besome annual variations (Shen and Zou,2006). Many taxes are assigned to provinces,including land use, property, and vehicletaxes and stamp duties. Provinces havediscretion in assigning these taxes to lowerlevels and some do so, but lower level taxassignments vary widely. Four of the majortaxes in China, including Value Added Tax andCorporate Income Tax, are shared betweenthe central and local governments – theseconstitute substantial parts of local finance.

Sub-national governments in Japancollected approximately 40 percent of total

tax revenues in 2007, split about equallybetween prefectures and municipalities.This is expected to increase to 50 percentunder the 2003 Trinity Reforms. Noteworthyprefecture taxes include an enterprise,consumption and inhabitants’ (income) tax.The most important local taxes are theresidential and non-residential property(fixed asset) tax and the inhabitants’ tax.Own-source revenues comprise slightlymore than one-third of total revenues ofboth prefectures and municipalities. Thisrelatively low level persists even after theTrinity Reforms, which were a response toexcessive centralization in its public revenuecollections and over-reliance of localgovernment on grants. The reformincreased local government own-sourcerevenue and commensurately reduced thelocal tax allocation (untied revenue sharing)grants and to national government subsidies(tied grants) to local government.

Revenues are quite centralized in Koreacompared with other countries in thesub-region. Korean sub-national governmentsaccount for about 35 percent of nationalpublic revenues, with upper and lower tiersrespectively raising about 22 and 13percent. Property taxes are the mostimportant source of own revenues for localgovernments, making up over half of thetotal. Income and consumption taxes arealso important. Own-source revenues make




China 70 20 50 40 15 25

Japan 60 20 40 40 20 20

Korea 45 17 28 35 22 13

Sources: UCLG Data Files among others. Figures are approximations. Data are the most recent available: China 2005, Japan 2007 and Korea 2010.

Table 3.5: Sub-national Government Share of Total Public Expenditure and Revenue – East Asia



up about 60 percent of total localgovernment revenues. As part of the centralgovernment’s fiscal decentralization policy,a Local Consumption Tax was introduced in2010. As a result, the revenue share oflocal government is expected to increase byabout 3 percent.


Intergovernmental transfers of variouskinds make up around 60 percent of totalsub-national level revenues in Chinawhere the system is very complex andnon-transparent (Bahl and Martínez-Vázquez,2009). The most important transfers arerevenue sharing mechanisms (includingVAT, business tax, individual income tax andenterprise income tax), tax rebates andearmarked grants. These make up about 80percent of local government grant revenue,whereas equalization grants amount toonly about 5 percent. A myriad of otherintergovernmental transfers comprise theremainder. There is little official informationprovided about the allocation of transfersamong levels of sub-national government.

Japanese sub-national governments alsoreceive almost 40 percent of their revenuein the form of transfers from the centralgovernment – both local tax allocations andtied subsidies – or the prefectures. TheLocal Allocation Tax (LAT) is by far the most

important transfer, followed by variousspecial purpose grants. The LAT is a fiscalsupport and equalization transfer. Specialpurpose grants are used largely to funddelegated functions, although some addressspillovers or provide fiscal performanceincentives. Conditional grants target specificsectors and require compliance with tightoperational standards. The Trinity Reformsdiscussed earlier reduced the amount of thetied grants to local government as well aschanged (simplified) the basis on which theLAT is distributed. The Ministry of HomeAffairs, however, does not managedistribution transparently.

Intergovernmental fiscal transfers accountfor about 40 percent of Korean localgovernment revenue. The current transfersystem, established by structural changes in2005, consists of Local Shared Tax (LST),National Treasury Subsidies (NTS) and theSpecial Account for Region-wide andRegional Development (SARRD). It is overlycomplex and compromises the consistentmeasurement of the fiscal effects of grants.Under the LST, there are 4 types of grants:Ordinary Local Shared Tax (OLST), SpecialLocal Shared Tax (SLST), DecentralizationShared Tax (DST) and Real Estate SharedTax (REST). Prior to 2005, the LST consistedonly of OLST and SLST and operated asan instrument of fiscal support andequalization, similarly to the LAT of Japan.9. See Table 3.6.



China 60 High Low Low

Japan 40 High Low Medium

Korea 40 High Low Medium

Source: UCLG data files and other sources. Figures are approximations. Data are for most recent year: China 2005, Japan 2006, Korea 2010.

Table 3.6: Intergovernamental Transfers as Local Government Revenue Source – East Asia


79

The 2005 increases in DST and REST weredesigned to compensate for expendituresincurred for additional NTS projectstransferred to local government and for thefiscal loss to local governments caused bynational tax policy.

NTS are essentially special purpose grantsdesigned to support costs of delivering mostlydelegated services. A majority of NTS projectsare subject to matching and other conditionsstipulated by the central government. SARRDis a grant to support regional developmentprojects to enhance competitiveness.Transfers between upper and lower-localgovernments include Metropolitan andProvincial Subsidies (M&PS), ProvincialRevenue Sharing (PRS) and MetropolitanRevenue Sharing (MRS). M&PS is a conditionalgrant distributed to cities, counties, anddistricts together with NTS funds. PRS is ageneral grant to lower level local governmentsbased on population, tax collection costs andother factors. MRS is a general grant tosupport basic fiscal needs of districts and tomitigate fiscal inequality among them.

Debt Finance10

Under existing laws, sub-national govern-ments in China are not allowed to borrow.Nevertheless, many do borrow, both tocover operational deficits and to financecapital spending. Sub-national governmentsborrow directly from commercial banks andother financial institutions, or indirectly

through their locally owned Trusts andInvestment Companies (TICs). In turn, thelatter borrow from banks and throughcapital markets. It is estimated that in2004, total outstanding debt of sub-nationalgovernments was about $120 billion USD,about one-third of which was due toborrowing (Shen, Jin, and Zou, 2006).

More recently, economic commentatorshave estimated the outstanding debt ofChina’s local government investmentcompanies to be about 11.4 trillion yuan(about $1.7 trillion), with additionalcommitments of a further $1.9 trillion,much of it tied to infrastructure projects.Other sources talk about $880 billion, afigure approximately corroborated by ChinaInternational Capital, an investment bank.11

These loans are backed by assets (typicallyland) or supported by implicit governmentguarantees and have been used to developlocal transportation, energy, water, urbaninfrastructure, and public housing. Thereare concerns that the proceeds from landsales used to repay these bank loans mayfall short, especially if the country’sbubble-prone real estate market tumbles.

Japanese sub-national governments arelegally allowed to borrow from commercialbanks and to issue domestic bonds; thelatter are becoming increasingly morecommon.12 The traditional permissionsystem has been abolished but a “priorconsultation” system still gives the central

10. See Table 3.7.

11. “Shell Game”, TheEconomist, March 11th2010.

12. In 2003, the FinancialInvestment Local Fundcovered 60 percent oflocal bonds, privatefinancial institutionscovered 30 percentand the marketcovered the remaining10 percent.



China No Significant

Japan Yes Significant

Korea Yes Moderate

Table 3.7: Legal Authorization and Extent of Local Government Borrowing – East Asia

Source: UCLG data files and other sources.



government some control over sub-nationalborrowing. This could be because, whilelocal governments secure their borrowingwith own-source revenues, local borrowingis also implicitly guaranteed by the center(Mochida, 2008). Government claims to bemoving increasingly toward a system that isbased more on fiscal rules and marketdiscipline. Sub-national governments inJapan borrow significant amounts. The sub-national share of public capital spending isaround 70 percent of the consolidated publicsector total and almost all of this is financedby borrowing. Outstanding sub-nationaldebt was about one-third of GDP in 2005,which is very high by internationalstandards (Mochida, 2008).

Korean sub-national governments mayborrow to finance capital spending andrecovery from natural disasters. They cansource funds either via direct loans (publicloan funds or bank loans) or the issuance ofbonds. The former process (particularly,public loan funds) is more common. Inrecent years, local governments' access topublic loan funds, relative to the centralgovernment, has decreased sharply butaccess to the special account, the "RegionalDevelopment Fund" operated by upper-levellocal governments, is still possible.Traditionally, Korean local governmentshave tended not to borrow for capitalfinancing, except for some relatively richupper-level local governments and largecities. Korea maintained a very strict"permission system" on local borrowing.This was softened in 2005, by theintroduction of the "total volume controlsystem” under which local governmentsmay borrow, subject to quotas set by thecentral government. In 2007, totaloutstanding debt of sub-national govern-ments represented 2.1 percent of GDP,and local borrowing represented about 3percent of local annual revenue throughoutthe 2000s (Lim Sung Il, 2010).

Performance13


Given the vastness of China and thediversity of local governments, it is difficultto generalize about local resource adequacy.However, it does appear that resourceconstraints have become generally moreproblematic since the 1994 reforms underwhich revenues became more centralized,expenditure responsibilities remainedessentially the same and intergovernmentaltransfers failed to adequately replace thelost own-source revenue of many sub-national governments. On the other hand,this period coincided with a very rapidexpansion of the Chinese economy andimprovements in the standards of living,and the resultant additional demand forservices could be the cause of additionalpressure on local budgets. Whatever thecause, resources generally become moreconstrained as you move down thehierarchy of government tiers. Given theineffectiveness of equalization grants due totheir small volume, current and capitalresource constraints are particularly bindingfor rural areas and for less economicallydeveloped regions. Thus, there aresubstantial disparities between urban andrural areas in service provision levels bylocal government, especially in education,health and water supply (Bahl and Martínez-Vázquez, 2009).

Local governments in Japan appear to havesufficient current and capital resources todeliver services with which they have beencharged. Grants seem to adequately fill thevertical fiscal gap, although one objective ofthe Trinity Reforms is to make localgovernments more self sufficient andefficient by limiting access to transfers andforcing them to collect more revenue. Inprinciple, the support and equalization grant(the LAT) assures adequate current13. See Table 3.8


81

revenues for all local governments to deliverservices at basic or standard levels and thecentral government’s implicit guarantee oflocal borrowing allows local governments toraise sufficient capital to meet financingneeds, but this is not always the case. The"Trinity Reform" resulted in fiscal pressure,particularly on local governments withrelatively poor fiscal capacity. This pressure,worsened by the increasing demands of anaging society and the impacts of the globalfinancial crisis, has differentially affectedlocal governments.

In general, local governments in Korea alsoseem to have adequate resources to delivertheir assigned services, although there issome variation. While own-source revenuesare insufficient to fund routine servicedelivery, transfers appear to fill the verticalfiscal gap reasonably well, at least in theaggregate. The equalization transfer (LST)also seems to meet its intended purpose.Although borrowing has not been

substantial, intergovernmental transfershave, again, proved to be more thanadequate to compensate for any shortfallsin accessing capital finance. Nevertheless,recent developments, including a sharpincrease in demand for social welfareservices and other unfunded mandates,have increased fiscal pressures on localgovernments.


Formally, local governments in China havelittle tax and spending discretion. Local taxassignments are allowed by higher levelgovernments but all tax bases are definedand rates are officially set by the center. Localexpenditure responsibilities are also assignedby higher level governments and the centralgovernment has a variety of mechanisms tootherwise influence local spending, includingexpenditure laws and regulations, spendingmandates and overt political control.Nevertheless, in practice, many local


High quality serviceoutcomes.

Good quality taxadministration, althoughperhaps too little use of taxrate authority. Goodexpenditure management.

There are some incentivemechanisms to encouragebetter performance, butexperience is limited.

Authority over local taxbases and rates is limited.LG have reasonable controlover spending their own-source revenue but otherspending is controlled tolarge degree by centralgovernment.

LG current and capitalresources seem adequate,although increasing fiscalpressure from additionaldemand for social welfareservices.

Korea

High quality serviceoutcomes.

High quality taxadministration andexpenditure management,within narrowly assignedresponsibilities.

CG uses incentives in blockgrant to encourage localrevenue effort. Someperformance incentivesincorporated into specificpurpose grants, as well.

Tax authority over rates andbases limited. Spendingcontrolled to large degree byCG through delegation ofresponsibilities andmandates.

LG current and capitalresources adequate.

Japan

Significant variation acrossregions but generally low inrural areas.

Good quality taxadministration andexpenditure management,although significantvariation.

Performance incentiveshave focused on economicdevelopment.

Limited formal fiscaldiscretion. Off budgetactivity significant.

LG resource constraintsincreasingly problematic,especially at lowest levels ofgovernment.

China


Table 3.8: Performance of Intergovernmental Systems – East Asia




governments appear to have significantdiscretion, as indicated by the “illegal” use ofloan funds and the use of off-budget solutionsto overcome fiscal constraints and meetservice delivery ambitions. In general, localgovernments in China enjoy (or take) a highdegree of freedom in the use of non-taxrevenue items: with the application of landdevelopment and real estate business feesbeing particularly important sources ofrevenue in some areas.

Japanese local governments havereasonable control over most taxesassigned to them. Although the centralgovernment defines the tax bases, localgovernments set rates within rangesformulated by the center. Local govern-ments are also allowed to create their owntaxes under certain well-defined conditionscodified in law, but very few have utilizedthis capacity. Local governments have alsotraditionally used “extra-legal taxes” butthese have been in decline since the 1990sand are now insignificant. Examples ofthose that still exist include prefecturestaxes on the nuclear fuel industry andmunicipal taxes on vacant houses, smallliving spaces and to support the historicaland cultural environment.

Local governments in Japan have limited ex-penditure discretion. The center determinesspending levels and standards and canmandate local governments to assume other(nominally “autonomous”) functions, thuscreating unfunded mandates. In recent years,such requirements have increased sharply,particularly in the social welfare. Because ofthe level of control, sub-national expenditureresponsibilities in Japan may best be seen ascentrally delegated functions and not genuinedecentralization (Mochida, 2008).

Local governments in Korea havereasonable control over most of their ownrevenues (Lee, 2003; An, 2003). Central

government sets the types, base, and rangeof local tax, however local governmentsdetermine the tax rate applied with thecentrally set range. In practice, most localgovernments do not use this flexibility.Currently, a tax simplification process isunder way. In 2011, tax sources will bereduced from 16 to 11, with a furthersimplification of local taxes expected tooccur at a later date.

Korean local governments have both auto-nomous and delegated functions. In theorylocal governments have complete discretionover spending on autonomous tasks; butspending on delegated functions isinfluenced significantly by central govern-ment, as are the projects funded by theNational Treasury Subsidies. As such, localgovernments appear to have a reasonabledegree of discretion over their expenditures.The discretion is, however, constrainedsomewhat by the fact that the center setssub-national civil servant wages andpension benefits, which comprise a largeportion of autonomous spending.


The intergovernmental system in China has,in the recent past at least, provided significantincentives for economic development in theregions, taking precedence over equalization.The extent to which these incentives havestimulated economic growth is the subject ofgreat debate. On the fiscal side, the effect oflocal tax rate competition on economic de-velopment has been clearly documented.There has been a “race to the bottom” amonglocal governments in regions with favorableeconomic conditions and a “race to the top”for local governments in less well developedregions (Yao and Zhang, 2007).

Japan incorporates some incentives into itstransfer system. The support and equaliza-


83

tion grant (the LAT) ostensibly provides anincentive for local tax effort by increasingthe amount of the grant as fiscal capacitydeclines. The latter is estimated as afunction of “standard” rather than actualrevenues. Like many such calculations,standard own-source revenues are thosethat would be derived if the localgovernment used the standard tax rate. If asub-national government sets a higher rate,it keeps the associated extra revenue. Inpractice, the effects are unclear since mostsub-national governments use the standardrates. In addition, some analysts argue thatthe equalization grant provides an incentivefor local governments to over-borrow, sincegrant increases with expenditure needs,which are determined in part by the size ofstandard interest and principal repayments.But the effects of such an incentive are alsounclear since the central government stillcontrols the amounts that municipalities canborrow. Some special purpose grantsprovide performance incentives but there islittle information on the objectives, designor implementation of such incentives.

Korea has gone relatively far in attemptingto establish incentives in its intergovern-mental system. For instance, some of thespecific purpose grants in Korea havematching components that are designed tostimulate spending of certain kinds. In2010, a new block grant is set to beimplemented with incentive mechanism forown-source revenues.


The quality of tax administration and ex-penditure management varies widely acrosslocal governments in China. The overallsystem of tax administration appears tohave improved somewhat since the 1994reforms. Prior to that time China operated aunified central tax administration system

that was reasonably efficient in collectingtaxes, but created significant rent seekingbehavior and divided loyalties among taxadministrators vis-à-vis their central andlocal “clients” (Shen, Jin, and Zou, 2006).Expenditure management policies in recenttimes have tended to focus on thestimulation of economic growth to a muchgreater extent than on the delivery of publicservices, but this does seem to be changing.

Japanese local government tax administra-tion and expenditure management appearto be of reasonably high quality, at leastwithin the limited bounds within whichmunicipalities are allowed to operate. Muchlocal government taxation and spending is“co-managed” by or “integrated” with thecentral government.

Local government tax administration andexpenditure management also appear to besound in Korea. Administration of taxes andexpenditures makes extensive use ofcomputerized systems, and a simplificationof the tax system commenced in 2011 willfurther assist in efficient tax administration.One remaining problem may be that localtax authorities seem to make insufficientuse of the control they have over the ratesof their assigned taxes, but this may be dueto the disincentives imposed by the transfersystem and not because they lack capacityof making such adjustments (Lee, 2006).


The quality of local public services in Chinavaries a great deal across the country. Inthe east and southeast regions, servicesappear to be of good quality: elsewhere andespecially in rural regions, service qualitylags. Broadly speaking, those regions thatare growing well economically have accessto relatively more resources, at least someof which are devoted to service delivery; inpoorer regions, particularly in rural and




remote areas where own-source resourcesare much more limited, local public servicedelivery is very weak.

The quality of local public services inJapan is very high. Public service outcomesfor education, health and infrastructureare among the best in the world.Access to services is fairly equitable acrosssub-national jurisdictions, indicating thatthe intergovernmental fiscal system designedwith this goal in mind has been working well(although very recently it has been undersome strain).

The quality of public services in Korea is verygood. The country’s infrastructure and socialservice outcomes are consistently rankedamong the best in the world for countries witha similar level of income. Like Japan, Koreahas a strong tradition of assuring equitableaccess to public services and the geographicdistribution of service outcomes is quite even,notwithstanding some concentration ofquality public services in Seoul. In addition,local governments appear to have becomeincreasingly accountable to citizens in therecent years and this helps to assure qualityservices.


China’s current system of intergovern-mental fiscal relations needs to be improvedto balance economic development andequalization, and a higher average quality ofpublic services. The system has recentlyfocused more on economic developmentand less on public service outcomes. Thismay now have changed with government’srenewed focus on addressing equity andpoverty concerns, and the “building of aharmonious society”.

Several issues stand out as China moves toconcentrate more on local public service

delivery. First, the high level of debt ofsub-national government in China must becorrected as it poses threats to both thenational and international economies.Second, while the assignment ofexpenditures needs clarifying, central andlocal governments need to find a moreadequate equilibrium in the responsibilitiesassigned to sub-national governments. Forexample, social insurance schemes(especially pensions) should probably not bea local responsibility (Bahl and Martínez-Vázquez, 2009). Third, the lower levels ofgovernment lack sufficient revenues, ingeneral, to cover their expenditureassignments. They have very few taxes attheir disposal and very little official controlover those available to them: and thetransfer system fails to address verticalimbalances and is ad-hoc and overlycomplex. Fourth, the distribution of grantsshould be more transparent and predictableas between the levels of sub-nationalgovernment and between units of localgovernment. A final concern relates to thetight accountability to central officials, whichmust be reduced if autonomy of localgovernment is a real objective.

The critical fiscal decentralization issue inJapan concerns the degree of controlexerted by the central government overlocal governments. For a mature, developedcountry, municipalities still have very littlediscretion over their fiscal affairs (Mochida,2008). The second and related issueconcerns the “one-size fits all” approach toservice provision and financing, and there isgrowing concern that the varied preferencesof residents across the country are notbeing addressed by the overly centralizedsystem (Aoki, 2008). Although localgovernments would appear to be wellplaced to respond to local demand, theyhave not been given sufficient authority todo so. As Japan increases fiscaldecentralization under the Trinity Reforms,


85

the fiscal relationships between the centraland local government will need to bereassessed. Widening fiscal capacities oflocal government will mean changes to thedistribution of LAT funding if equality ofservice provision is to be maintained.Japan’s high dependency on debt fundingwill also need critical evaluation.

A key issue in Korea concerns the assignmentof service responsibilities across levels ofgovernment. As in many countries, there issome lack of clarity regarding serviceassignment, but the major problem appearsto be some mis-assignment of functions: thepractice of “revenues are owned by thecentral government, but executions are doneby local governments” needs to be changed togive better fiscal balance. The centralgovernment seems to play a relatively largerole in the delivery of health, environmentaland police services compared to some othercountries. A second concern, even after the2011 tax simplification initiatives, may be theproliferation of taxes at the local level. A finalproblem concerns the system of inter-governmental transfers where the issue is theexistence of multiple transfers with the sameobjectives, some of which are allocated by thecenter and others by regional governments(Lee, 2003; Lee, 2004).

Conclusions and Observations on East Asia

China appears ready to re-orient itsintergovernmental system to focus more onthe delivery of public services. Policymakerscould begin by reducing higher levelsubnational government authority to assignexpenditures and revenues to lower levelgovernments (Shen and Zou, 2007). Inaddition the following objectives should beconsidered:

• Building a stronger consensus betweencentral and sub-national governments onthe assignment of local level service res-

ponsibilities and tax authorities to thedifferent levels of local government,particularly concerning the service res-ponsibilities of local governments for theprovision of social insurance;

• Assigning more substantial tax and non-tax revenues to local governments andconsidering allowing at least somediscretion for tax rates and feeschedules;

• Restructuring the system of intergovern-mental transfers with a view to making itsimpler and more fiscally equalizing bothbetween different levels of governmentand across the country;

• Formalizing local government borrowingauthorities (and introducing rules for thesize and use of loan funds) and moregenerally bringing all fiscal and financialoperations on budget;

• Developing improved measures ofhorizontal accountability (such as theLocal Congress of People).

In Japan, the results of the Trinity Reforms(Doi, 2004) have so far been mixed.Transfers to local governments werereduced and own-source revenues haverisen by a commensurate amount, but thereduction in special purpose grants, relativeto the support and equalization grant, hasbeen somewhat disappointing. In addition,while the overall vertical fiscal gap didnot increase with the reforms, somesub-national governments did experience anet loss in revenue. As such, inequality inresources and levels of service delivery has,for the first time, become an issue. Finallythere is some evidence that the reduction inblock transfers has led to increasedborrowing by sub-national governments.This may be worrisome given the alreadyhigh level of local debt, especially at a time




when the central government is turningaway from administrative supervision tomarket discipline as the main mechanism ofcontrol. This suggests that the future policyagenda in Japan may wish to increaseefforts in:

• Providing sub-national governments withmore fiscal autonomy by replacingspecific purpose grant with own-sourcerevenues,

• Clarifying the twin roles of the LATdistribution and improving itsequalization performance to improveequity in the distribution of resources forservice delivery as own-source revenuesincrease and LAT transfers decrease,

• Taking steps to ensure that localgovernments do not borrow excessively,and

• Eliminating unfunded mandates associa-ted with social welfare services.

Although Korea’s system of public serviceprovision and financing appears to functionreasonably well, there are several im-provements that the government mightconsider:

• Realigning its assignment of serviceresponsibilities across levels of govern-ment, withdrawing itself from theprovision of some services while stillmaintaining its responsibility for policy.

• Balancing functional and fiscal distribu-tions to guarantee full coverage of thecosts in the case of transfers of functionsfrom central to local governments.

• Further restructuring of the National Taxand Local Tax System and improving thetaxing power of local government byeliminating nuisance taxes (i.e. those that

do not generate much revenue net ofcollection costs) at the local level. Aspectsof revenue assignment, tax base structure,tax payer's rights and convenience shouldbe reconsidered and property, income andconsumption taxes need to be harmonized.The fiscal accountability to local residentsshould also be considered.

• Revisiting the system of intergovern-mental transfers. Here the focus wouldbe on consolidating the number oftransfers that focus on similar goals,streamlining the design of individualtransfers to reduce conditions and thenumber and complexity of tied grants,increasing the equalization attributes ofthe Local Services Tax distribution andclarifying the relationship between thisand other transfers.

• Enhancing the links between the taxsystem and the grants system with theaim of improving local autonomy,accountability and efficiency.

• Introducing differentiated policy on fiscaldecentralization to allow for theapplication of differentiated fiscal auto-nomy, considering fiscal capacity, eco-nomic conditions, population and otherfactors across local governments.

• More autonomy in local borrowings andits linkage with the citizen's burden.

South-East Asian Local Governments Finances

Structure

Expenditure14

In Indonesia, expenditures have been de-volved to sub-national levels to a significantextent. Sub-national government (of whichthere are 33 provinces, 349 regencies14. See Table 3.9


87

(kabupaten) and 91 cities (kota)) spendingmakes up about 35 percent of total publicexpenditure; and spending by localgovernments (kabupaten/kota) is about 80percent of the sub-national total. The mostimportant local service in Indonesia iseducation: local governments spend about35 percent of their budgets on primary andjunior secondary school education.Infrastructure is next most important with abudget share of around 15 percent.

In the Philippines, sub-national spendingmakes up about 25 percent of total publicexpenditure (net of debt service). Localgovernment (i.e. 1,496 municipalities, 138cities, and 42,025 barangays) expenditurecontributes about 55 percent of Philippinesub-national totals. Health and infrastructureare the most important services amonglocal governments in the Philippines, eachmaking up about 11 percent of totalexpenditure budgets. In education, theyhave responsibilities only for theconstruction and maintenance of schools.

Local budgets in Vietnam increased fromabout 26 to more than 45 percent of totalpublic expenditure from 1992 to 2008. Data

are not available nationally to disaggregatesub-national spending among provinces andcentral cities (63), provincial cities andtowns (93), rural districts (550) and thecommunes (more than 10,600 communesand wards). Information for two provinces(Ninh Thuan and Thai Binh), however,suggests that in 2008, provincial, district,and commune spending made up about 75,15, and 10 percent of the sub-national totalrespectively (Hanai and Hugen, 2008). Sinceprovinces are responsible, in large measure,for assigning particular expenditures todistrict and commune levels, significantvariation across provinces is likely, but itseems that provinces are dominant.Available information suggests that districtsand communes focus their spending in theeducation and health sectors.

Sub-national spending in Thailand15 makes upabout 26 percent of total public sectorexpenditure (including state ownedenterprises) in the 2011 budgets. There areno precise data on the local government(i.e. municipal and sub-district) portion ofthe sub-national total (Amornvivat, 2004).Most local spending is focused on educationand health, which account for about 20

15. In Thailand, the sub-national governmentsector is made up of75 provincialadministrativeorganizations, 2006municipalities and5,742 tambonadministrativeorganizations orspecial types of localgovernments.




Indonesia 35 7 28 8 5.5 2.5

Philippines 25 11 14 10 2.5 7.5

Vietnam 45 30 15 35 25 10

Thailand 26 n.a. n.a. 15 n.a. n.a.

Cambodia Less than 5 n.a Less than 5 Less than 1 n.a. Less than 1

Malaysia n.a n.a Less than 5 n.a. n.a. Less than 1

Source: UCLG data files and other sources. Figures are approximations. Data are for most recent year available: Indonesia 2006, Philippines 2005, Vietnam2008, Thailand 2010 actual and 2011 budget.

Table 3.9: Sub-national Government Share of Total Public Expenditure and Revenue – South-East Asia



percent of local budgets: the composition oflocal expenditure varies greatly, however,depending on the degree of urbanization ofthe jurisdiction.

In Cambodia, provincial and district ad-ministrations above the 1,630 electedcommune and sangkat councils (urbancommune) have long been deconcentratedentities of the central government. Communesaccount for a very small share (less than 5percent) of total public expenditures. Theymostly finance councilor allowances and smallinfrastructure development projects fundedthrough transfers from the centralgovernment and, indirectly, from donors. The2009 Organic Law on Decentralization andDemocratic Development created indirectlyelected councils at the provincial and districtlevel and foresees specific functions andresources being given to all sub-nationallevels, including the communes.

In Malaysia, the application of the “OneMalaysia” policy has resulted in a morecentralized form of government within theFederation and a very small local govern-ment sector. Local government is seen as aservice delivery arm of the central govern-ment (deconcentration). There are, how-ever, some calls from local government formore autonomy and specific expenditureassignments being provided.

Own-Source Revenues

As in many developing countries, Indonesiantax revenues are very centralized,especially given the level of expendituredecentralization. Provincial and localgovernment own-source revenues compriseonly about eight percent of governmentrevenues; and local government own-sourcescontribute just under a third of the sub-nationaltotal. The tax on electricity sales is the mostimportant own-source revenue in Indonesia,making up about 20 percent of the total. Local

taxes on hotel and restaurant sales and publichealth center charges are also noteworthy;each comprises about 10 percent of totalown-source revenues.

Philippine sub-national government own-sourcerevenues contribute around 10 percent ofconsolidated public revenue. UnlikeIndonesia, however, the bulk of sub-nationalrevenue generation —about 75 percent—derives from local (municipal, city, andbarangay) government collections. Thereal property tax is the most significantown-source revenue, respectively accountingfor about 30 and 40 percent of own-sourcerevenues of municipalities and cities. Thebusiness tax is also important for some localgovernments. One concern is that electedcouncilors often lack knowledge on sources offinance for local government.

Sub-national revenue in Vietnam isreasonably important. Own-source revenuesat the provincial, district, and commune levelhave been increasing rapidly, from about 35to 44 percent of total public revenue from2008 to closer to the present. At theprovincial level, the most important taxes arethe VAT (excluding imports), corporate profittax, personal income tax on high incomes,gasoline and oil fees. Data from the twoprovinces noted above suggest that most,perhaps just under 85 percent, are collectedby provinces with the remainder coming fromdistricts (Hanai and Hugen, 2008). Revenuesof communes are negligible. Again, givenprovincial discretion to assign local taxes,some variation across provinces is likely.Overall, land-based taxes dominate localown-source revenues. The current tension inpublic revenue collection in Vietnam is thatthe central government wants large cities tobe more self-reliant while they see a need forgreater financial assistance.

Sub-national revenues in Thailand comprisedabout 15 percent of the consolidated total in


89

2010. The most important local taxes are thebuilding and land tax, the land developmenttax and the signboard tax (a localadvertisement tax). Local governments alsolevy license fees and fines but these are notimportant quantitatively. There isconsiderable variation between localgovernments, with major cities obtaining asignificantly larger proportion of their totalfunding from own source revenues.

In Cambodia, there is currently very littlerevenue collected by the commune/sangkatcouncils and the future of revenue collectionby them and the newly established districtcouncils is yet to be determined. InMalaysia, in line with the nation’s “OneMalaysia” policy, there is heavy centralizationof revenue management. There are,however, some calls from local governmentfor greater revenue autonomy to be givenand local government being able to fix taxrates without state approval.


Intergovernmental transfers dominate localgovernment revenues in Indonesia, makingup at least 90 percent of the total. The

general-purpose equalization grant (DAU) islegislated to be at least 25 percent ofnational general revenue and is the largesttransfer, comprising about two thirds of localrevenues. Central government tax (propertyand income) and non-tax (natural resource)sharing with local governments is alsoimportant and comprises the bulk of theremainder. There is also a small specialpurpose capital grant (DAK).

In the Philippines, transfers contribute up to70 percent of local government revenue,although there is significant variationregionally and between the two lower tiers.The formula-based Internal RevenueAllotment (IRA) is by far the most importantrevenue source17. The share of the IRA intotal revenues of cities and municipalities isabout 50 percent and 75 percent,respectively. National revenue sharing andcategorical (tied) grants of various kinds areconsiderably less important to localbudgets.

Intergovernmental transfers in Vietnamcomprise half of total local revenues. Sharedtaxes distributed to provinces include VAT(except on imports) and those on corporate

16. See Table 3.10.

17. At each level of localgovernment, thedistribution is basedon population (50%),area (25%) and anequal share (25%).Due to the size of thegrants pool and thedistribution formula, itdoes not cover thecosts of devolvedfunctions, has resultedin greater disparitiesbetween rich and poorlocal governments andhas been incapable ofencouraging localrevenue collections.




Indonesia 90 Medium High Low

Philippines 70 Low High Low

Vietnam 50 High Medium Medium

Thailand 85 Low Low High

Cambodia Very high n.a. High Low

Malaysia Very close to 100 n.a. n.a. High

Source: UCLG data files and other sources. Figures are approximations. Data are for most recent year available: Indonesia 2006, Philippines 2005, Vietnam2008, Thailand 2011.

Table 3.10: Intergovernmental Transfers as Local Government Revenue Source – South-East Asia



and personal income, profits, consumption,and gas and oil. The sharing rate is the samefor all taxes but varies across provinces,depending on differential expenditure needs andfiscal capacities. Other transfers to provincesinclude an equalization grant and variouscategorical transfers (e.g. for national programsand emergency needs). Provinces, in turn, maketransfers to district and commune governmentswithin their jurisdictions. Transfer objectivesand mechanisms are often unclear, althoughprovinces claim to allocate grants to lowerlevels for both general and specific purposes.

Intergovernmental grants make up thebulk of local government revenues inThailand, comprising nearly 85 percent oftotal budgets. Three main types oftransfers are made to local governments:general-purpose grants, specific purposegrants, and subsidies. Nearly all grants aretied to particular types of expenditures(even so-called general-purpose grants).The total grants allocated to localgovernment in Thailand fell by nearly 30percent between 2008 and 2010.

In Cambodia, commune revenue assignmentsdefined in the original 2001 decentralizationlegislation have never been implemented.Most funding for the councils derives from theCommune/Sangkat Fund as untied transfers.

In turn, that Fund receives resources fromboth the central government budget and frominternational donor agencies, though donorfunding is declining and is now only about 30percent of the total. Although the funds arereceived as untied transfers, about one third isneeded for administration costs and theremainder mostly finances small infrastructureprojects. The future design and scale of transfersto commune/sangkat and district councils iscurrently under discussion as part of the plan toimplement the 2009 law noted above.

In Malaysia, local government receives veryclose to all its funding in the form of grantsand contributions from the federal and stategovernments. These enable bothinfrastructure development and recurrentservice provision. That the federal grants passthrough the state governments causes aproblem of funds flow to local governments.

Debt Finance18

By law, Indonesian local governments areallowed to borrow for capital developmentfrom government sources, private financialinstitutions or capital markets. In practice,local borrowing to finance capital spendinghas been limited and has been exclusivelyfrom the government or from internationaldonors through the government’s on-lending18. See Table 3.11


Indonesia Yes Limited

Philippines Yes Limited

Vietnam Yes Limited

Thailand Yes Very Limited

Cambodia No None

Malaysia Yes n.a.


Table 3.11: Legal Authorization and Extent of Local Government Borrowing – South-East Asia


91

mechanisms. Over the past twenty fiveyears kabupaten/kota and their waterenterprises have borrowed, in total, lessthan one percent of GDP, and currentoutstanding debt is considerably less thanthis (Lewis and Woodward, 2010).

Philippine local governments have access toa similar set of loan mechanisms but appearto have borrowed from a wider array ofsources than their counterparts in

Indonesia. Some local governments haveissued bonds and many have borrowed fromeither government or non-governmentinstitutions, with the bulk of localborrowing in the Philippines —about 80percent— coming from government sources.One interesting loan mechanism is outlinedin Box 3.3. Outstanding debt of Philippinelocal governments has recently beenestimated at about 0.6 percent of GDP.Elected councilors are said to be wary of


Box 3.3: The Local Government Unit Guarantee Corporation (LGUGC) in the Philippines

Established in 1998, the LGU Guarantee Corporation (LGUGC) is a private financialcredit guarantee institution owned by the Bankers Association of the Philippines (38percent), the Development Bank of the Philippines (37 percent) and the AsianDevelopment Bank (25 percent).

Its primary goal is to make private financial resources available to creditworthy localgovernments (LGs) through its credit guarantee. The LGUGC’s credit enhancementfacilitates LGs infrastructure development projects in the capital market. Over the pastdecade, the LGUGC has supported LGs to mobilize capital from a range of banks andbond investors for all types of infrastructures projects.

Three main services provided are:

• A guarantee mechanism. The LGUGC guarantees the debt of LGs, water districts,electric cooperatives, renewable energy technology providers and state universitiesand colleges. The guarantee fees range from 1 percent to 2 percent per annum.

• Credit rating services. In the absence of an entity specialized in LG risk evaluation,the LGUGC has established an internal LG credit screening and rating system. Thesystem evaluates the LG’s capacity to pay and willingness to honor contractualobligations.

• Program management. The LGUGC manages the guarantee fund for the ElectricCooperative System Loss Reduction Project of the World Bank – Global EnvironmentFacility and is the program manager for the Capacity Building to Remove Barriers toRenewable Energy Development – a Loan Guarantee Fund of the UNDP.

Out of the 26 projects that the LGUGC has guaranteed, nine have been related to water(for example: Indag Water District for about $315,000 USD). Of the municipal bondsfloated in the Philippines market since 1998, increasingly more have had an LGUGCbond guarantee. As of January 2009, closed deals totaled about $60.4 million USD.



taking on debt that extends beyond theirperiod of office and this fear needs tobe overcome if more appropriate use ofdebt financing is to be achieved (AsianDevelopment Bank, 2008).

Provincial governments in Vietnam areallowed to borrow in domestic markets(either bank loans or bonds) to financecapital expenditures, and some have doneso. District and commune governments arenot allowed to take on debt.

Municipalities and sub-districts in Thailandneed higher-level government approval toborrow, with borrowing restricted tofinancing local capital development. Inpractice local government borrowing hasbeen very limited and most infrastructuredevelopment has been funded throughtransfers, donors or from local fiscalreserves. There are no data available on thelevel of borrowing by Thai local governments(Amornvivat, 2004).

In Cambodia, the Organic Law on Decentralizationand democratic development prevents localgovernment from accessing debt. In Malaysia,local governments may borrow from either thefederal or state governments, as eitheroverdrafts or mortgages, but only with theapproval of the state government.

Performance19


The distribution of resources across localgovernments in Indonesia is quite skewed.As such, it appears that at least some localgovernments in Indonesia have insufficientresources with which to discharge theirfunctions (World Bank, 2006). However, asa whole, local governments do not spend allthe operating funds that are available tothem and are building up large surpluses.This is partly a capacity problem.

Access to local government capital financingin Indonesia is insufficient relative toinfrastructure needs at the local level.Central government allocates only a smallcapital grant (DAK) to kabupaten/kota.Currently the DAK makes up less than fivepercent of total kabupaten/kota revenuesand local governments have not borrowedsignificantly to finance capital development.Operation and maintenance ofinfrastructure and public investments bylocal government is a particular concern inIndonesia (Lewis, 2003a; Lewis, 2007a).

In the Philippines studies suggest thatcurrent local government funding isinsufficient to cover the cost of deliveringservices under their authority (Manasan,2004; World Bank and Asian DevelopmentBank, 2005). Inadequacy of funds seems tobe a particular problem for municipalities,although perhaps less so for cities(Manasan, 2004).

Resources for local capital development inPhilippines are perhaps slightly lessconstrained than in Indonesia, althoughsome of this apparent adequacy may bebeing forced on local governments, whichare required to use 20 percent of their IRAfor capital development (Manasan, 2004).Philippine local governments also receive avariety of capital transfers from the centralgovernment, but these are mostly small.Other capital grants are allocated as part ofa grant-loan mixed distribution packagefrom the Municipal Development Fund(World Bank, 2009). The larger variety offinancing sources notwithstanding, localgovernment debt finance in Philippines hasbeen similar in size to that in Indonesia.

It is difficult to judge current and capitalresource adequacy of local governments inVietnam. On the current side, themaintenance of infrastructure appears to besomewhat of a concern but this may be19. See Table 3.12


93

either a problem of resource constraints ora difficulty with expenditure assignments.In terms of capital requirements, there areno capital transfers in the Vietnameseintergovernmental system and, as noted,local governments are not permitted toborrow to finance capital development; assuch any local capital spending needshave to be met out of general resources(Martínez-Vázquez, 2005). In any case,since district and commune expenditure andrevenue assignments vary across provinces,resource adequacy probably also variesacross local governments.

As local government’s functions in Thailandhave not yet been very clearly identified, it isdifficult to assert the adequacy of resources toassure local responsibilities are fulfilled, butsome sources say local governments have runquite substantial budget surpluses in recentyears (Amornvivat, 2004; JICA, 2008).However the National Municipal League ofThailand contest this and claim that localgovernments have had recurrent revenueproblems as central government has notprovided the amount of shared taxes definedin the Decentralization Act of 1999. This wasoriginally set at 35 percent, although this levelhas never been reached, and it was only 25percent in 2009. If surpluses are beingexperienced they could be indicative ofcapacity constraints as much as over-funding.Unfortunately there is insufficient informationwith which to make a judgment on this.

In Cambodia, the central government’sCommune/Sangkat Fund makes up nearlyall local government revenue. It is receivedas untied funding and is used nearly entire-ly for administrative expenses and smallinfrastructure projects. There is effectivelyno recurrent service delivery undertaken bylocal governments, partly a result of thelack of clarity in functional assignments aswell as resource and capacity constraints.In Malaysia as in most nations, finance is

said to be a major constraint of localgovernment performance.


Indonesian local governments have littlediscretion over their tax resources. Centralgovernment enumerates allowable taxesand charges and defines tax bases,although local governments have someflexibility in setting tax rates up to limits setby the center. In the early years ofdecentralization, local governments wereallowed to create their own taxes andcharges, but this authority was rescinded bylegislative change. Sub-national govern-ments need to have their budgets approvedby the central government but recentchanges to legislation seem to have madethis much more of a formality than it hasbeen.

Once budgets are approved, Indonesianlocal governments have completeexpenditure authority over their own-sourcerevenues and their DAU transfer revenue.Even for the small DAK capital grant, whichis tied to specific spending sectors, localgovernments have quite wide latitude inchoosing specific spending activities withinthe designated sectors. They are requiredby law to spend 20 percent of their totalbudgets on education, but this is rathereasily achieved now that teacher salariesare included in the calculation under arecent change in the interpretation of thelaw. Unfunded mandates are not a problemfor local governments. The central govern-ment is, however, developing “minimumservice standards” which, if enforced ascurrently designed, would cost more toachieve than the resources available tomany kabupaten/kota (World Bank, 2006).

In the Philippines also, central governmentstrictly defines allowable local governmenttaxes and their bases, and local




governments are also quite constrained insetting tax rates. The tariffs for some localtaxes are fixed by the center and localgovernment rate flexibility for other taxes isbound at the upper end to limits whichappear to be quite low by internationalstandards. Finally, local governments areonly allowed to change rates of taxes overwhich they have authority once every fiveyears and then by no more than 10 percent.On the other hand, local governmentsappear to have quite wide purview in

establishing and developing fees andcharges, but these do not result in muchrevenue (Manasan, 2004).

The central government in Philippines exertscontrol over local spending in several ways.In some cases, it limits the amounts thatlocal governments can spend on certainclasses of expenditures and in otherinstances it prescribes minimum levels onparticular types of spending. For example,spending on personnel services is limited to

Table 3.12: Performance of Intergovernmental Systems – South-East Asia

Capacity is limited to thatnecessary to enable them tobe arms of the centralgovernment.

Some relationship of grantsto financial position andrevenue collection, but notstrong.

Very little.LGs have very few resourcesthat are not grants fromother spheres ofgovernment.

Malaysia

Generally poor serviceoutcomes.

Generally very limited.Incentives within the grantsystem are currently beingconsidered.

LGs have very little revenueor expenditure discretion.

LGs are extremely poorlyresourced

Cambodia

Generally weak outcomesbut improvingresponsiveness.

Local efforts to improve taxadministration andexpenditure management inevidence—tax systemcomputerization andcontracting out services.


Taxing discretion limited tominor charges/fees;spending heavily influencedby CG.

LGs resources need to belinked to responsibilitiesand the legislated amount ofthe shared tax transfershould reflect theseresponsibilities.

Thailand

Significant improvementsneeded. Little horizontalaccountability in mostprovinces.

Tax administrationcompletely centralized.Spending reasonably wellmanaged—low wage bills.


LGs have very limited taxauthority. Moderatediscretion over managementof service delivery.

LG resource adequacyquestionable, especially fordevelopment andmaintenance ofinfrastructure.

Vietnam

Service delivery needs to beimproved.

LGs need to improve taxesand spending management.Significant cash build-ups.

Explicit use of incentiveslimited to employment oftransfer intercept in casesof non-repayment of loans.

LGs set some tax rates butcan make changes onlyevery 3 years. CG heavilyinfluences LG spending;Unfunded mandates aparticular concern.

Inadequate resources forsignificant number of LGs,particularly municipalities.

Philippines

Service outcomes need to beimproved.

LGs need to reinforce theircapacities to administertaxes and spend resources.As significant accumulationof cash reservesdemonstrate.

Minor and ad-hoc experiencewith performanceincentives. Formalencouragement of localrevenue effort through blocktransfers discontinued.

LGs set tax rates undercentrally imposed ceilings;LGs have (almost) completediscretion over spending.

Some LGs have insufficientfunds.

Indonesia




95

45 percent or 55 percent of budgets,depending on the income class of the localgovernment and, as noted above, 20percent of a local government’s IRA must beset aside for development spending and fivepercent of total local revenue must bereserved for a ”calamity fund” (World Bank,2009).

Philippine local governments also face anumber of significant unfunded mandates.The most binding of these include local civilservant salaries in general, and benefits forhealth sector employees in particular. Localgovernments have also recently beenrequired to fund health insurance premiumsfor their impoverished residents. Finally,local governments are expected to offerfinancial support for many centralgovernment agencies operating in theirjurisdictions, such as for police, fireprotection, and the courts.

Vietnamese local governments have verylimited tax autonomy. The central govern-ment determines all allowable sub-nationaltaxes, defines tax bases, sets tax rates andhigher level governments in the hierarchicalstructure approve the budgets of levels ofgovernment “below” them. Provinces assigntaxes to district and commune levels at theirown discretion, with few exceptions, passingon the centrally defined bases and rates.Such local revenue assignments are fixedfor “stability periods” of three years but canbe changed thereafter. The only revenueauthority local governments have relates totheir ability to establish relatively minor tollroad, school, and health fees (Hanai andHuyen, 2008).

Local governments in Vietnam havesignificantly more expenditure autonomythan either Indonesia or the Philippines.Although provinces establish the serviceresponsibilities of districts and communes,they do not micro-manage service delivery.

Somewhat unusually in the region, allsub-national governments, including districtsand communes, have discretion to makedecisions regarding staffing and remuneration(above minimum levels set by the center).Local unfunded mandates are illegal undercurrent law (Martínez-Vázquez, 2005).

Thai local governments have limited taxingand spending discretion. The centralgovernment assigns allowable taxes,defines their bases and sets rates (orceilings). As in Vietnam, the only revenueauthority that local governments have is indeveloping user fees and charges. Localgovernments in Thailand have limiteddiscretion in setting spending priorities andallocating funds across expenditure types.All local government budgets must beapproved by higher level governments,which often insist on significant changes.Almost all grants are tied to particular typesof spending. On the other hand, localgovernments do have some discretion indeciding how particular services can bedelivered. Many local governments have, forexample, contracted out delivery of someservices (Weerasak, 2006; JICA, 2008).

In Cambodia, the limited resources of thecommunes/sangkats and their responsibilityfor councilors’ remuneration means thateven though their transfers are untiedand their expenditure decisions are notinterfered with, they have limitedresources with which to exercise theirexpenditure autonomy. At this stage in thedevelopment of decentralization, they haveno revenue-raising autonomy.

In Malaysia, the discretion of local govern-ment is limited by the fact that theexecutive is an appointed arm of the state(intermediate level) government in thefederation rather than an elected sphere ofgovernment. The main own-sourcerevenues of local government are property




taxes, rents, license fees and fees forservices. Local government needs theapproval of the state before changing taxrates or introducing new taxes.


Indonesian has had only limited experiencewith structuring performance incentives intothe intergovernmental fiscal system.However, the central government doesallocate 10 percent of the national propertytax to local governments based on theirprevious year’s tax collection performanceand the DAU grant was increased in line withthe extent to which own-source revenuesexceed “potential revenues”. However, themechanism was dropped, at least in partdue to local government complaints (Lewisand Smoke, 2009).

Interestingly, research has shown that thesize of the DAU is, in any case, positivelyassociated with local own-source revenues,even without the incentive (Lewis, 2005). Alikely explanation for this outcome is thatincreased transfers and under-spendinglead to increased reserve funds, whichin turn lead to increased interestearnings, a form of own-source revenue.Finally, a more positive experience withincentives is illustrated by a recent(partial) withholding of DAU untiedtransfers to local governments that did notcomply with budget reporting requirements.After delaying transfer payment, thenoncompliant local governments almostimmediately conformed to the relevantregulations.

In the Philippines, experimentation withperformance incentives has been even lesscommon than that in Indonesia. Evidencesuggests, however, that larger block grantallocations are associated with less localrevenue generation (Manasan, 2004). Thus,

larger transfers provide a disincentive tolocal tax effort. More positively, asmentioned above, the central government’sprerogative of withholding a localgovernment’s IRA if it fails to repay a loanon time or in full has encouraged gooddebt repayment performance by localgovernments.

No specific performance incentives havebeen integrated into the intergovernmentalframework in Vietnam. On the other hand,the lack of clear expenditure and taxassignments may create disincentives toimprove spending efficiency and revenuemobilization. The centralization of taxadministration may also providedisincentives for local tax collection, andnegotiations involved in determiningmany transfers may provide incentives forlobbying behavior at the expense of localrevenue effort. Regarding the equalizationtransfer, the use of outdated expenditurenorms to determine spending needsmay create negative incentives in serviceplanning and decision-making; and the factthat equalization transfers are reduced iflocal own-sources increase sends the wrongsignal regarding local revenue effort(Martínez-Vázquez, 2005).

The central government in Thailand has notyet used explicit performance incentives inthe intergovernmental transfer system andlittle information exists on the generalincentive effects of the central-localdecentralization framework. However, it hasbeen hypothesized that excess slack in thefiscal system has induced reasonablysignificant innovation among some localgovernments in developing their fiscalstrategies (Weerasak, 2006).

Cambodia does not yet have any experiencewith performance incentives within itsfiscal transfer system but is consideringthis development. Malaysia measures local


97

government performance through a “LocalGovernment Star Rating” classificationintroduced in 2008 and some of the grantsto local government rely on a council’sfinancial position and revenue collectionperformance, but incentive structures are nota feature of grant distribution.


Indonesian local government revenueplanning and tax administration are in-sufficiently developed and these processes arenot standardized to one set of financial reportsand a fixed timetable for their production. Adesirable annual level of tax revenues is notestimated by local officials and most localgovernments simply target a fixed percentageincrease in own-source revenues each year,then meet the target with relative ease (Lewisand Oosterman, 2009). Another issue is theexcessive cost of tax administration. Currently,the ratio of the cost of administering taxes torevenue yield, across all local governments inIndonesia, is well over 50 percent and morethan 10 percent of local governments havecost-to-yield ratios exceeding 100 percent(Lewis, 2006).

Similarly expenditure management suffersfrom weaknesses. As noted above, manyIndonesian local governments do not fullyspend the resources to which they haveaccess and have accumulated a surprisinglylarge stock of fiscal reserves sincedecentralization began in 2001. Currently,kabupaten/kota fiscal reserves are equal toabout 25 percent of annual localgovernment expenditure. As is the case forrevenues, the distribution of fiscal slackacross Indonesian local governments isskewed: the fiscal reserves ranging fromabout five percent to nearly 80 percent ofannual expenditure. Even so, only a fewIndonesian local governments have drawndown on these reserves to finance needed

capital spending (Lewis and Oosterman,2009). On the expenditure side of theirbudgets, Indonesian local governmentsspend a significant amount on governmentadministration (25 percent of local ex-penditures) but also spend about 35 percentof total budgets on education, often morethan the pre-determined optimum amount(Lewis, and Pattinasarany, 2009b).

While Indonesian local governmentborrowing has been limited, as noted above,repayment of loans has been weak overall(Lewis and Woodward, 2010). Over fortypercent of total principal and interestrepayments that have fallen due are inarrears. Most outstanding loans to localgovernments and their water enterprises(especially) would be considered non-performingby normal definitions (although such loansare referred to as merely “stuck” inIndonesia). The Indonesian centralgovernment appears to be addressing theproblem, however, with an extensive programof local debt restructuring.

Local tax administration in the Philippineshas seen improvements made but thereremain important challenges. As little as 60percent of total own-source assessedrevenue is collected by local governmentsand collection efficiency appears to havedeclined in recent years. Many local officialsdo not use the tax authority that is availableto them. In addition, local governmentscollect many nuisance taxes and, althoughhard data are lacking, tax administrationcost efficiency is low (Manasan, 2004).

In terms of expenditure management,Philippine local governments spend, onaverage, about 25 percent of their budgetson administration (Manasan, 2004) and likethose in Indonesia, have accumulatedquite large fiscal reserves. The stock ofsub-national reserves in the Philippines isbetween 35 percent and 40 percent of




annual expenditure. However, unlike thesituation in Indonesia, local governmentreserves in the Philippines are not fullyavailable for appropriation, owing torequirements regarding the maintenance ofliquidity sufficient to cover long-termliabilities. Finally, as already mentioned, thecentral government can cut the IRAallocations due to non-payment of debt andthis has undoubtedly encouraged timely andfull loan repayment (World Bank, 2009).

Tax administration is centralized in Vietnam.Thus tax administrative effectivenessand efficiency are not issues for localgovernments. Local officials do sometimesattempt to improve the performance ofcentral tax administrators by limitingtheir salaries and benefits but this haslittle overall impact in improving taxadministration. On the other hand,Vietnamese local governments do havesome control over spending and seem to doa reasonably good job in terms of managingtheir expenditures. By internationalstandards, local governments spend less onwages and salaries and significantly moreon capital development (Martínez-Váz-quez, 2005).

Despite having limited discretion over theirrevenues, local government taxadministration in Thailand appears to bequite competent and many localgovernments have modernized their taxadministrative systems with a view toimproving collection efficiency (Weerasak,2006; JICA, 2008). Others have embarkedon successful civic education schemes toencourage residents to pay their taxliabilities (Weerasak, 2006). Thai localgovernments have also been innovative onthe expenditure side, with many nowcontracting-out the delivery of someservices and improving cost efficiencies.Some local governments have also adoptedmedium-term expenditure frameworks and

others have engaged in genuineparticipatory budgeting, becoming moreresponsive in the process (Weerasak,2006).

In Cambodia, tax administration by localgovernment is restricted by the delay inpassing local government finance legislationto clarify assignments and give legalcapacity to collect. Until this happens, localgovernments’ in Cambodia have littleauthority and capacity. On the expenditureside, the small projects undertaken bythe communes are included in theirdevelopment plans and budgets, but theseCouncils generally have only one staffmember so administrative capacity is verylimited.

In Malaysia, the “One Malaysia” policy hasled to local governments having very fewduties in revenue collection, so they havevery little capacity. On expendituremanagement, local government’s majortask is to deliver functions that arecontrolled by the central government sowhile they have some service deliveryexperience, their overall capacity inservice delivery management, includingpolicy determination and the setting ofstandards, is limited.


As in other developing nations, local publicservices in Indonesia need improvement,although since decentralization a decadeago, citizens have become more satisfiedwith the level of service provided bylocal government, (for example a recentstudy estimated that over 85 percent ofcitizens are at least somewhat satisfiedwith local public school education),social service outcomes rank disappointinglylow from both regional and internationalperspectives (Lewis and Pattinasarany, 2009a).Infrastructure quality is, on average, sub-standard


99

and deteriorating, but citizen satisfactionwith low quality services, which may inpart be related to improvements theyperceive in the role of local governmentsand governance, creates few incentives forlocal governments to improve servicedelivery performance.

Philippine residents, like their counterpartsin Indonesia, appear quite content with thequality of public services. A recent survey ofhealth services indicates that just less than90 percent of Filipinos are at leastsomewhat satisfied with quality (Azfar andMeagher, 2001), but a recent World BankReport qualified the situation of publicservice delivery in Philippines as quite poor(World Bank, 2009). As with Indonesia, it islikely that low expectations explain highlevels of satisfaction, but these lowexpectations engender little support forservice improvements.

Information on the quality of local publicservices is lacking for Vietnam but sourcessuggest that significant improvements areneeded (Martínez-Vázquez, 2005). Improvingservice delivery at the local level, however, isconstrained by the almost total lack ofhorizontal accountability. Vietnam has startedto experiment with local service score-cardsand this is a useful beginning in establishinglocal government accountability to citizens,but much remains to be done.

There is a similar lack of information aboutlocal public service outcomes in Thailand.Performance of local governments is notmonitored and outcome data are notregularly collected. It does seem, however,that local governments are becoming moreaccountable to their citizens. Local govern-ments, for example, are increasingly usingpublic hearings to seek opinions on andexperience with particular developmentprojects. While public participation inbudget decisions is still somewhat limited,

the positive experience of some localgovernments gives cause for someoptimism (Weerasak, 2006).

In Cambodia, the level of resourcing of thecommunes, the lack of clarity in theirexpenditure assignments, and limitedcapacity have led to ad hoc provision of arather limited range of small-scale services.Some of what is delivered is of poor qualityand has limited impact.


The most critical local government issues inIndonesia are related to the distribution ofrevenues across local governments and tothe quantity and quality of spending by localgovernments. The max-min ratio of percapita own-source to shared revenuesamong local governments is nearly 400:1,quite large by international standards(Lewis, 2002). This overstates the inequityproblem since such revenues make up arelatively small portion of total revenues,but even after accounting for the equalizingDAU grants, there is a great disparity inlocal government access to revenues(Lewis, 2003).

At the same time, Indonesian local govern-ment’s management capacity is weak, andwhile some municipalities have madesignificant improvements in this domain,across the board this translates into lowspending efficiency. There is the potentialfor reaching higher levels of service deliverywith existing resources if spending isrefocused (especially away from administra-tive tasks) and efficiency is improved. Weakintergovernmental performance incentivesand inadequate efforts to build capacityhave contributed to this shortcoming.

Another related set of constraints in Indonesiarelates to local own-source revenues. Like




most developing countries, the level ofIndonesian local government own-sourcerevenues is very low and this constrainsaccountability. Just as important is the factthat Indonesian local governments lackaccess to any good source of own revenuethat can be used to respond to citizendemand (Lewis, 2009). A new Indonesianlaw on sub-national taxation asserts thatauthority over the urban and rural propertytax will be decentralized to local govern-ment within five years, but implementationduring the indicated time frame is far fromcertain. If it does eventuate, it will benecessary to reinforce the capacity ofIndonesian local governments to use theproperty tax revenue effectively (Lewis,2009).

In the Philippines, the most pressing issue isarguably the under-funding of localgovernments relative to their service res-ponsibilities, especially in municipalities.Local tax administration also appears to bethe binding constraint. A particular problemin this regard concerns the property tax,which is not used as a local revenue source.Unfunded mandates by the center furtherconstrain resources.

A second major concern in the Philippines islocal governments’ lack of fiscal autonomy.As mentioned above, the centralgovernment maintains significant controlover both local revenues and expenditures.Over-control of local fiscal operationsconstrains the achievement of efficiency andaccountability objectives.

Perhaps the most important issue inVietnam relates to the structure of theintergovernmental system, particularly therole of provinces. Currently, Vietnam isallowing provinces almost completediscretion in setting the fiscal agenda fordistrict and commune governments. Arelated issue is the hierarchical relations

among levels of government, particularlywith regard to the budget preparation andapproval process. This model, under whichthe budget at each level must be approvedby the next higher level government (inaddition to the People’s Council), limitsplanning and budgeting autonomy, which inturn constrains the attainment of efficiencyand accountability objectives. Finally, therelatively large number of sub-national tiersincreases administrative costs and distortsimplementation of central, regional andlocal policies and programs (Martínez-Vázquez, 2005).

A particular constraint to effective de-centralization in Thailand is the number andsize of local governments. Thailand has over7,800 local governments at the bottom tier,making the average size of a local unit littlemore than 8,000 residents, with many beingmuch smaller. This constrains the properdecentralization of many services andpossibly increases the costs of servicedelivery. Another problem relates to theparallel structure of deconcentrated officesof the central government operatingalongside decentralized governments.Such a structure creates confusion inservice delivery and fiscal operations andconstrains the development of horizontalaccountability. Besides these institutionalproblems, some technical concerns alsoexist. There is a particular problem with theassignment of functions, which have notbeen clarified since the basic legislationwas passed in 1999. Lack of revenuedecentralization and discretion is alsoproblematic. Finally, the system of transfersis complicated and fragmented andconstrains local decision-making; transfersalso lack transparency, predictability, andtimeliness in their allocation (Martínez-Vázquez, 2005).

In Cambodia, the most pressing issue isto get the 2009 Organic Law on


101

Decentralization and Democratic Developmentimplemented. A plan to achieve this wasadopted in June 2010 and the period to2013 should see significant change. Thenewly created district councils will, however,remain indirectly elected by the electedcouncilors of the communes and sangkats, asituation that raises questions about theiraccountability to the people.

In Malaysia, clearly the most pressingmatter is to reintroduce elected localgovernments. Once the councils are back inplace, they need to be given sufficientautonomy to function effectively as re-presentatives of the people and providers ofservices.

Conclusions and Observations on South-East Asia

The foregoing discussion suggests a numberof different priority areas for policy reformacross the countries of South-East Asia.

In Indonesia, government might consider:

• Revising allocation procedures of themain DAU untied grants to improveequalization (including removal of thecounter equalizing wage paymentcomponent);

• Revising and eliminating the manyinefficient local taxes and levies(estimated to number 9,715 and accountfor 36 percent of local taxes and levies);

• Prioritizing expenditure managementthrough increasing awareness ofoperation and maintenance needs,better identification of investmentopportunities, and improving servicedelivery outcomes;

• Requiring relevant central ministries towork with local government to build lo-

cal administrative capacity (for generalpurposes and on the assumption theproperty tax will be decentralized andneed to be well managed);

• Opening a dialogue with local govern-ments to design and integrate aconsistent set of performance incenti-ves into the intergovernmental fiscaltransfer framework and working withlocal governments to further develophorizontal accountability.

In the Philippines, the central governmentcould consider ways to improve thefunding situation of local governments,especially cities. This might include:

• Increasing the IRA allocations;

• Decentralizing additional tax bases,including a piggy back scheme on thenational income tax;

• Building capacity of local governmentsin tax administration and other ownrevenue generation;

• Discontinuing its practice of makingunfunded mandates;

• Removing some controls over localgovernment revenue generation andspending and replacing them withperformance incentives.

In Vietnam, the government may want torethink the controlling role provinces playin district and commune fiscal affairs.Possible reform to consider might be:

• Explicitly assigning expenditure and re-venue responsibilities to sub-provinciallevels;

• Providing more autonomy in expenditureplanning and decision-making, and in



20. Unless the contextindicates otherwise,the term ‘State’ isused in the Australiancontext to includeeither or both of theNorthern Territory andthe Australian CapitalTerritory.


revenue generation to districts andcommunes;

• Implementing a modern property taxover which some rate authority would begranted to sub-national government.

Vietnam may also like to revisit the multi-level and nested national budgetpreparation and approval process as well asthe need for four levels of government.

In Thailand, the government might considerfollowing the Indonesian example andeliminate its deconcentrated offices. Apriority policy concern is clarifying theservice responsibilities of local governments,which are poorly defined. Currently, servicesare assigned in individual local governmentacts and sectoral laws rather than inconsolidated national legislation. If thatwere done, a more appropriate assignmentof revenues could be made and a systemof intergovernmental transfers developed.Local governance and accountabilityshould be developed, strengtheninghorizontal structures and avoidingexcessive bureaucracy. In addition, it will benecessary to improve the efficiency oflocal government operations, possiblythrough the amalgamation of some smallermunicipalities.

In Cambodia, the most important actionrequired is the implementation of the 2009Organic Law, which is underway. Thatprocess should involve clarification of thestatus of provincial, district and communecouncils and definition of revenue andexpenditure assignments for each. Thefiscal transfer system should then bedeveloped so as to ensure each sphere ofgovernment in Cambodia is equitably(though probably not fully) funded.

In Malaysia, the priority would be to providefor elected local governments with genuine

autonomy. Following this, the governmentcould increase the clarity in the division ofresponsibilities between local, provincialand central government and work towardseliminating overlaps to improve efficiency.

Australia and New Zealand LocalGovernment Finances

Australia and New Zealand differ from othernations in the Asia-Pacific Region in thatonly a small percentage of their populationis descended from the indigenous people. Inboth nations, the general governmentsector is about 35 percent of GDP but localgovernment is small, making up only about6 percent of total government expenditurein Australia and about 9 percent in NewZealand (2 percent and 3 percent of GDPrespectively).

Australia is a federation in which localgovernment is not mentioned in the nationalConstitution. In each of the six states andthe northern territory, local governmentexists under State20 legislation and eachcouncil is subject to the same legislation,irrespective of its population size, area,degree of urbanization and range ofservices. In the Australian Capital Territory,municipal functions are performed bythe territory government. The number ofcouncils in Australia remained fairly stablebetween 2000 and 2008 at about 700, butchanges in Queensland and the NorthernTerritory in that year resulted in a reductionof over 100 in the number of councils.

In New Zealand, there are two tiers of localgovernment, both of which operate underthe same national legislation. Most of thenation is covered by both regional councils,of which there are 12, and territorialauthorities, of which there are 73. Five ofthe territorial authorities also have thepowers of a regional council and areknown as “unitary authorities”. The structure



of local government has been stable sincethe late 1980s but there is now somepressure to reduce the number of regionsand form more unitary authorities, withmajor changes already agreed for theAuckland area. The unitary nature ofgovernment makes this process of overallreform much easier than in the Australianfederation where each state is responsiblefor local government.

The separation of powers between the twospheres of local authorities in New Zealandis, in theory, that the regions are theregulatory authorities and the territorialauthorities provide the services. Below theterritorial authorities is a tier of localorganization that exists at the discretion ofthose authorities – community boards.About half the territorial authorities havechosen to establish these and in 2006 therewere about 150 community boards. Somecouncils also have formal relationshipswith local indigenous Maori Liaisoncommittees. At both the regional and territorialauthority tiers of local government, thereis considerable variation in area andpopulation size.

All councils in Australia and New Zealand aremade up of elected officials, and with theexception of setting rate levels in New SouthWales, councils in both nations have a veryhigh level of autonomy within the activitieslegislated for local government. In termsof supervision, New Zealand councils are

largely left to the scrutiny of citizens andthe electoral process, the state of SouthAustralia operates a self-regulatory systemthrough the local government associationwhile other Australian states use a morelegislative supervisory model.

Expenditure21

Traditionally, the major responsibilities oflocal governments in Australia have beenlocal roads, recreational facilities, andcommunity amenities (including wasteremoval). Recent times have seen anexpansion into a wider range of activities inthe social security and welfare areas butthese are still a minor task overall. Theatypical structure of the population in thenorthern territory –high proportion ofpoor indigenous people in the territory’ssmall and scattered population– results incouncils in that jurisdiction having moreroles than elsewhere. The increasingproportion of local government expenditurespent on general administration in Australiais of some note and seems to be relatedat least in part to greater regulatoryrequirements being imposed on councilsby state legislation. Possibly resulting fromthese extra administrative costs, the shareof councils’ expenditure spent on roads –abasic municipal service– has been reducingover the last decade.

As in Australia, the main functions of localgovernment in New Zealand are roads and 21. See Table 3.13



Australia 36 30 6 20 17 3

New Zealand 9 n.a. 9 8 n.a. 8

Table 3.13: Sub-national Government Share of Total Public Expenditure and Revenue – Oceania

Source: UCLG data files and other sources. Figures are approximations. Data are for most recent year available: Australia 2006-7; New Zealand 2007-8.



transportation; activities associated withculture, recreation and sport; water andwaste water, and solid waste management.General governance is about 17 percent oftotal local expenditure and furtheramalgamation of regional and territorialcouncils may result in this being reduced. Inboth countries, there has been some costshifting to local government by otherspheres of government in recent years, butthese have been incremental and regulatoryin nature rather than a major unfundedmandate being deliberately transferred tothe local sector.

Under the principle of subsidiarity, it seemsunlikely that either nation has gone as far asit could to get service delivery down to thelowest level of government that couldreasonably provide it. However, bothnations have taken the view that servicessuch as education and health should beprovided on a more uniform basis than canbe achieved if they were transferred to localgovernment. In Australia, the greateruniformity the community expects isachieved through agreements between thecommonwealth and state governments andthrough conditions the commonwealthattaches to tied grants it provides thestates. There is no move towards givingthese types of services to local governmentand, if any change were to be made, itwould probably involve a greaterconcentration of power at the centralgovernment level. In New Zealand, thenational uniformity of education and healthservices is achieved though their assignmentto the national government and there is nopressure to change these arrangements.

Own-Source Revenues

The revenue sources of local government inAustralia and New Zealand are shown in Table3.14 and 3.15. In both cases, the tax base isthe value of real property but, in Australia, the

state governments also use this revenue baseand their absolute revenue flow from propertyis greater than that of the local governmentsector. Councils generally decide on the size ofthe rate levies and user fees, and collect theirown revenue. The only exceptions being thatcouncils in the Australian state of New SouthWales have a limit imposed by the stategovernment on the size of annual rateincreases22 and some regional councils in NewZealand arrange to have their revenuecollected by the territorial authorities. Theflexibility given to councils in New Zealand touse different valuation bases and createdifferent property-based taxes is wider thanthat available in Australia, and greaterflexibility given to some Australian statescoincides with higher levels of rate revenue inthose states.

Current reviews of taxation in both nationsmay impact local government. In NewZealand, there is some talk of the nationalgovernment levying a valuation-based taxon rural properties, and in Australia there isa quite radical proposal that all public sectortaxation might be collected by one agency,whether the revenue flow is for the central,state or local government sector.

Although the New South Wales StateGovernment imposes a limit on the annualincrease in rates, councils can apply forspecial exemption and the great majority ofsuch applications are successful.Nevertheless, it does appear that therestriction has some impact as the growth inrate revenue in New South Wales over thelast decade has been significantly lowerthan in all other states. The use of theproperty value as a revenue source by thestate governments may be a much largerand more widespread incursion.

In Western Australia, the legislativerequirement to use unimproved value as thebasis of valuation in all rural areas, results in

22. This is also the case inthe Northern Territoryuntil 2013.


105

very productive mining properties on poorquality rural land being grossly under-valuedand under-taxed. Notwithstanding thisrestriction, Western Australia has had one ofthe highest rates of property tax growth sincethe mid 1990s.

The pattern of revenue sources for councilsin the northern territory differs from otherAustralian states because of the highproportion of poor indigenous people in the

territory’s small and scattered population,and the way other levels of government usecouncils to provide non-municipal serviceson an agency basis, financed throughsubsidies. Associated with the revisedstructure of local government, a three-yearperiod of limiting rate increases is currentlyin operation in the northern territory butthere are, as yet, no data available tomeasure the impact of this state policydecision.


State Source of Revenue

Property Rates Sale of Goods and Services Interest and Dividends Other Revenue Self Funding Ratio Grants and Subsidies

New South Wales 35.7 32.0 4.3 12.0 84.0 16.0

Victoria 47.7 17.9 1.3 12.9 79.8 20.2

Queensland 27.4 39.0 2.4 19.3 88.1 11.9

Western Australia 40.2 20.7 3.2 15.7 79.8 20.2

South Australia 59.2 16.3 1.2 5.7 82.4 17.6

Tasmania 31.4 39.0 3.8 8.3 82.4 17.6

Northern Territory 17.2 16.5 1.9 12.9 48.4 51.6

Total 37.4 28.9 2.8 14.1 83.1 16.9


Table 3.14: Australian Local Government Revenue Sources, by State, 2006-07 (%)

Source of Revenue Regional Councils District/City Councils

Municipal Rates 43.9 61.0

Regulatory Income 4.3 6.4

Investment Income 7.9 5.1

Sale of goods and services, and other income 14.4 17.6

Self-funding ratio 70.5 90.1

Grants and subsidies 29.5 9.9

Total 100.0 100.0


Table 3.15: New Zealand Local Government Revenue Sources, 2007-08 (%)



Revenue from the sale of goods and servicesis an important supplement to rate revenue inboth nations, but is more important inAustralia where it is close to 30 percent oftotal revenue. Part of the difference betweenthe two nations in the importance of thisrevenue (and between States in Australia) isdue to some utility services, such as waterand sewerage, being funded through raterevenues in some jurisdictions and byconsumption-based charges in others. It hadbeen anticipated that, in New Zealand,revenue from sales of goods and servicesmight have been higher for regional councilsbecause of the enterprises such as ports andpublic transport they provide, but these seemto be managed outside the generalgovernment accounts of the councils unlessthey need supplementation or providesupport funding for other council activities.

Income from interest and dividends isrelatively high in both nations, but higher inNew Zealand where councils haveaccumulated relatively large surpluses overtime and have considerable investments.Interest and dividend revenue is over 10percent of the own-source revenue of theNew Zealand regional councils, indicating thatthey have been able to accumulatesubstantial investments – possible resultingfrom their trading enterprise activities. Forthis reason, the size and importance of thegrants provided to the regional councils inNew Zealand is somewhat surprising.

The impacts of the recent global financialcrisis on councils’ capacity to raise revenuein both Australia and New Zealandhave been relatively small. Some councilssuffered a temporary reduction in residents’capacity to pay rates and charges, but thegreater impact has been the lower interestrate received on investments and, in a fewAustralian cases, their investment portfolioswere grossly reduced in value through thecollapse of international banks. The longer-

term impact, at least in Australia, may be areduction in the spread of values betweenthe lower and higher ends of the propertymarket that will result in a redistribution of acouncil’s total rate impost away from thosewith greater capacity to pay, towards thosewith less.

Intergovernmental Transfers

Councils in Australia receive grants from boththe central and state governments, with thelargest program being the untied fundingfrom the central government. This programprovides about a third of all grants to councilsand is distributed by state grant commissionsbased on principles of fiscal and servicedelivery capacity equalization specified by thecentral government. This structure does notwork particularly well as the different statecommissions interpret the principles differentlyand similar councils in different statessometimes receive very different levels of percapita funding. The annual increases in thesize of the grants pool is determined byformula but the original amount was basedon a self-assessment of the Australiangovernment’s fiscal capacity many years ago,so the pool has no direct relationship to levelsof current fiscal stress (if any) within the threespheres of government. Another 15 percentof Australian grants are provided by thecentral government as tied grants forrecurrent service delivery –particularly roadmaintenance. These grants are managed byline agencies within the central bureaucracywhich at least partly follow the distributionalpatterns decided by the state grantscommissions, i.e. limited equalization.

At the state level, the largest program is theuntied subsidy to councils for reducing orexempting those in need from council rates,although these have been declining in realvalue in some states and no longer cover thefull cost of the states’ policies. Two states alsoprovide untied funding for more needy councils


107

with large indigenous populations. Tied grantsfrom the state governments are mostfrequently provided for capital infrastructuredevelopment and for services provided ona cost recovery basis. The distribution ismost frequently based on relative needsas expressed in submissions to stategovernment agencies responsible for programmanagement.

There are very few incentive mechanismsincluded in the grant arrangements inAustralia. The federal government usesagreements rather than strict conditions toensure that its objectives are achieved by grantrecipients and the state funding is rarelytied to local government achieving greaterefficiencies.

For councils in New Zealand, almost all grantsto local government are provided as tied grantsfor local road maintenance and construction.The grants are funded from national fuelexcise, road user charges, and vehicleregistration fees, and there is someequalization of service capacity within thedistribution so that poorer councils receive ahigher percentage of their total roadexpenditure budgets from grant funding. Theoverall size of the pool of grant fundingprovided to councils in New Zealand varies overtime as expenditure priorities of the nationalgovernment change.

In fact, changes in central government policy inthe eight years after 1999-2000 saw grants toregional councils increased from 15 to nearly30 percent of their total revenue while grantsto territorial councils remained stable at about10 percent of revenue. As a result, rates havereduced in importance from over 50 percent toless than 44 percent in the regional councils,while remaining much more stable atabout 57 to 60 percent in the territorialcouncils. While these movements areassociated with the different functions performedby the regional and territorial councils, it is an

issue requiring further analysis and negotiationwith the government of New Zealand.

The extent to which councils in Australia andNew Zealand are self-funded acts against thelikelihood that municipal services are ofroughly the same standard within each ofthese nations. Self-funding must result inricher councils being able to provide betterservices and the equalization mechanismsoperating through the grants systems areinsufficient to overcome the differences in fiscalcapacity. Clarity is needed in both nations as towhere the policy objective of equalization fits inrelation to the provision of municipal services,and changes may be necessary to expenditureassignments between spheres of government,revenue assignments, or both. In bothnations, the local government structureand management capacity do seem to beunderutilized and diseconomies of small scaleare obvious in many councils.

Debt Finance

All local authorities in both Australia and NewZealand have capacity to borrow, but thelevel of debt is exceptionally low in Australiawhere gross debt is about 25 percent of theannual operating revenue, and also low inNew Zealand where it is about 70 percent ofthe annual operating revenue. In both cases,borrowing is much more likely to be used tofund infrastructure for cost-recovery ratherthan social activities.

Overall, with self-funding ratios being veryhigh, with large investment portfolios and verylow debt levels, it must be assumed that thelocal government sectors in Australia and NewZealand are under little stress in funding theirservice delivery responsibilities. At a minimum,a different approach to intergenerationalfunding of infrastructures would seegreater use of debt and higher rate imposts(or lower investments) to fund more socialdevelopments.




Conclusions and Observations on Australia and New Zealand

Australia claims that the most importantnegative feature of the framework within whichlocal government operates is its lack of nationalconstitutional status. If this is the case, itdoes not seem to be associated with fiscalmanagement, as councils have a high level ofautonomy and the grants system has beendesigned to by-pass any constitutionaldifficulties. More work is needed by localgovernments to identify the implications of thelack of constitutional status. Such status may,for example, make changes to expenditureassignments easier to achieve and this wouldresult in a more active and influential localgovernment sector.

Local government representatives highlight thefiscal difficulties that councils are facing, beingsqueezed on the one side by rising communityexpectation, increasing responsibilities andcompliance requirements and on the otherby only modest revenue growth. Localgovernments in larger urban and regionalcouncils have far fewer economic constraintsthan small rural and remote councils;especially indigenous councils in the moreremote regions.

Work is needed to improve the long-termplanning in local government in Australia. Thesystem of fiscal transfers for local governmentis too cumbersome and creates difficulties forlocal government’s acceptance of grantdistribution results, especially for the untiedgrants. More work is also needed by thecentral government to see that the principlesfor distribution that it has agreed to with thestates are interpreted uniformly and appliedsimilarly by the seven state grantscommissions.

Along with the development of policy on theconstitutional status of local government, workis necessary in Australia on examination and

clarification of the vertical imbalance within thefederation to see that the revenue andexpenditure assignments to each of the threespheres of government are those that bestserve the nation. The rate pegging actions ofthe New South Wales government also need tobe abolished so that true autonomy is providedto local government in that State. Clarificationis needed on where equalization of serviceprovision by local councils fits into nationalobjectives as this will impact on grant systems,revenue assignments and the spread ofrevenue between rates and fees from the saleof goods and services.

In New Zealand, the major policy developmentmust relate to the future role of regionalcouncils and their amalgamation into unitaryauthorities. These councils currently appear tohave too limiting a range of functions andsuffer substantial diseconomies of scale. Theextent to which equalization of the capacity oflocal governments is an objective of the centralgovernment also needs policy clarification asthe current system of grants cannot achieveequalization and greater inequities aredeveloping over time: to the detriment of themore needy sectors of society. Policy mayalso be necessary on the accumulation ofinvestment assets by councils to see thatintergenerational equity of citizens is betterachieved, but this should result fromagreement of councils through theirassociation rather than being imposed by thenational government.

Conclusions

The above review demonstrates the significantvariation in fiscal decentralization frameworksand outcomes across and within sub-regionsof the Asia Pacific region. The differencesextend across standard variables such aslocal governments’ shares of consolidatedpublic expenditure and revenue, local dependenceon intergovernmental transfers, and localgovernment access to and use of capital


109

finance. Differences also span across anassortment of performance indicators relatedto resource adequacy, fiscal discretion,incentives, capacity and service delivery. Thisdiversity is at least in part dependent on thelevel of development of the nations’ economiesand their historical backgrounds, making itnearly impossible to generalize about fiscaldecentralization in the Asia Pacific.

Difficulties in making generalizations not-withstanding, findings are summarized inthree broad conclusions that may beimportant for policymakers in the region toconsider as they develop theirdecentralization reform agendas goingforward. The first relates to the appropriateeconomic objectives of decentralization. Oneof the principal arguments in support ofdecentralization is that it can improve publicservice delivery to citizens. As such, itmakes sense for countries to design andimplement their decentralization programswith an explicit view to makingimprovements in basic public services. China,for example, has tied its recentdecentralization agenda more towardseconomic development goals (and this policy isalso being emulated in Cambodia), thoughpublic services are now becoming a moresignificant concern. Bangladeshi policymakershave ignored service delivery in the design andimplementation of their decentralizationprogram, using decentralization more as aninstrument of control than as a mechanism forthe improvement of local services. Orienting adecentralization program toward the objectivesof improving services would appear to be anecessary starting point in the achievement ofthe objective.

A second conclusion also relates todecentralization strategy. Here the issue isthe extent to which it is the central govern-ment or middle tier governments that leadthe design and/or implementation ofdecentralization programs. In unitary na-

tions, it appears that strong central govern-ment guidance of decentralization is neededfor its more successful implementation. Infederal nations and elsewhere wheredecentralization decision-making is a mixof central and middle-tier governmentpolicy implementation, decentralization oftenseems to get stuck at the middle tier.However Australia has proved an exceptionto this rule and Vietnam seems to be makingefforts to move beyond this dynamic. In anycase, the general point is that whendecentralization stops at the middle tiers, theachievement of service delivery outcomes isconstrained. As such, central governmentsneed to develop mechanisms to ensure thatprovision and financing of public services isdevolved to the lowest levels feasible.

A final conclusion concerns accountability.This review has shown that where countriesin the region have paid attention toaccountability in the design and execution oftheir fiscal decentralization, they have mostfrequently stressed vertical accountability toother governments. Horizontal accountabilityis somewhat weak throughout the Asia sub-regions, although it is stronger in the olderdeveloped democracies of Australia and NewZealand. It is difficult for decentralizationprograms to deliver quality local publicservices in the absence of strongaccountability to the people. Part of theproblem in Asia may be technical in that thelink between service delivery and taxpayment is not strong; and this certainlyconstrains accountability. The larger issuelies outside the realm of the strictly fiscal,however, and relates more to the politicalenvironment in which local governmentsoperate. It is often argued that democraticelections at the local level are necessary forthe establishment of horizontal accountabilityand to create a robust citizen demand forquality services. To deepen local democracy,more active citizen engagement andgovernment responsiveness is also needed.




UCLG ASPAC Policy Recommendations for Fiscal Decentralization

After discussion at the Workshop on Local Government Finances held at Batam,Indonesia, on 25-26 June, the members of the UCLG-ASPAC Region agree that the basisof its policy on Fiscal Decentralization is as follows.

1. That local government be given greater autonomy by having its powers widened todecide all tax bases and tax rates, and the level of fees for service, within itsmandate.

2. That there be a review of the current allocation of revenue sources and servicedelivery responsibilities to each sphere of government and that they be reassignedso as to:

a. Reduce vertical fiscal imbalance;

b. Better match expenditure responsibilities with own-source revenues;

c. Ensure all spheres of government share in natural resource revenues; and

d. Reduce the impact of economic cycles on any one sphere of government.

3. That local government be provided with an increased and fixed share of nationalpublic sector revenue through grants and transfers, but with a reduced number ofgrants and a larger proportion being provided as untied funding.

4. The grants distribution to local government be managed by an institution that isindependent of government, has membership acceptable to grant recipients andoperates in a transparent manner to provide funds directly to the bank accounts ofthe recipients.

5. That there be more widespread application of horizontal fiscal equalization principleswithin the distribution of grant funding for local government.

6. That all local governments have access to loan funds, under conditions and limitsthat are acceptable to both the national and local governments.

7. That where one does not already exist, there be a local government financecorporation or equivalent established to ensure local government has access to loansat lower than commercial rates of interest.



UCLG ASPAC Policy Recommendations for Fiscal Decentralization (continued)

8. That all unfunded mandates of local government be removed by increased fundingor reallocation of responsibilities, and that future unfunded mandates be madeillegal.

9. That “finance follows function” in any future reallocation of responsibilities betweenspheres of government.

10. That consultative mechanisms be established to ensure that all spheres ofgovernment work together to coordinate development and infrastructure planningand renewal, and budgeting, and that local government has an increased role indetermining national service delivery objectives.

11. That other spheres of government review and amend their policies and proceduresto remove unnecessary restrictions on local government fiscal autonomy.

12. That where not already done, standard accounting and financial managementpractices be introduced for local government; that there be clear proceduralinstructions for practitioners implementing these standards and that both electedofficials and local government staff be trained in their application.

13. That national constitutions be amended where necessary to give local governmentappropriate status, that all sectoral legislation be examined to ensure itsconsistency with local government legal requirements and that local governmentrevenue management be firmly based in the law.

14. That local government associations be recognized by other spheres of governmentas the appropriate bodies to represent local government in intergovernmentaldiscussions on policy and fiscal arrangements.

For its part, local government will work with other spheres of government to achievethese objectives, will re-examine its own policies, by-laws, regulations, procedures andpractices to see that they best meet the overall objectives of government in providingservices to the people in an efficient manner.

Local government will also take steps to improve the transparency of its operations,involve the people more in its planning, budgeting and other activities, and willintroduce programs to encourage the payment of local taxes and charges.


0w2010 04 Eurasia DEFcarta ang 9/9/11 11:59 Página 112

4. EURASIA

NNAATTAALLIIAA GGOOLLOOVVAANNOOVVAA

AANNDD

GGAALLIINNAA KKUURRLLYYAANNDDSSKKAAYYAA

CCEENNTTEERR FFOORR FFIISSCCAALL PPOOLLIICCYY ((MMOOSSCCOOWW)),, RRUUSSSSIIAA


115

T he countries reviewed in this chapterinclude those belonging to the

Caucasus region (Armenia and Georgia),Central Asia (Kazakhstan and Kyrgyzstan),Eastern Europe (Moldova, Belarus andUkraine), and Russia, which coversEastern Europe and Northern Asia.

Local governments in the Eurasia regionhave faced major challenges over the last20 years due to the rapid adjustment of adiminished size and role of government.This problem has been encumbered bythe concurrent attempt on the part ofmost national governments to retainresponsibility for the large amounts ofsocial services that usually were providedby local governments during the Sovietperiod.

Concurrently local governments across theregion have been coping with a dramaticincrease in migration, both domestic andinternational. This has resulted in a redis-tribution of service needs in relation tothese positive and negative net-migrations, which no longer match theexisting infrastructure. At the same time,the social infrastructure (schools,hospitals etc.) in localities that have lostpopulation has been preserved in order toreduce social tensions, as budget-supported institutions have becomealmost the only area of employment.

In the early 1990s, despite the fact thatEurasian countries developed local self-governments from a common framework,each of the eight countries examined inthis chapter has responded to thosechallenges in its own way and has selectedits own model of local self-governance andintergovernmental fiscal relations. Forinstance, Kazakhstan has adopted a sys-tem of deconcentrated state bodies forlocal administration; a rather centralizedhierarchical system of public authorities

was selected in Belarus; and a two-tiersystem of local self-government wasimplemented in Russia and Moldova. Inbetween, there are states where local self-governments exist autonomously (Arme-nia, Georgia) or alongside state bodies oflocal administration (in Ukraine at theoblast and raion level, in Kyrgyzstan - atthe raion level).

Positive trends towards greater decentralizationcan be traced in some of the Eurasiancountries: Local taxes were adopted in2010 in Armenia and in Belarus a law OnLocal Government and Self-governmentwas adopted.

However, in other countries an inclinationtowards centralization is being observed;local tax reduction in Russia, Kyrgyzstan,and Georgia; nomination by the center oflocal self-government heads inKyrgyzstan; reorganization of localgovernments in Georgia; and limitations tolocal government autonomy in Russia.

The recent financial crisis revealedweaknesses within the local finance sys-tem of some Eurasian countries, while inother countries local budgets faired muchbetter than central/regional ones due tothe more stable revenues assigned to locallevel. For example while in Moldova, trans-fers to local governments were propor-tionally cut by 20 percent, in Kazakhstan,where grants were set up for a three-yearperiod, they remained stable. In Russia,the revenue of settlement budgets(poselenie) grew, in nominal terms, in2009 by 22 percent, however some largeindustrial cities (e.g., Yaroslavl, Orel,Kursk) faced shortcomings due to thereduction of Personal Income Tax revenueresulting from growing unemployment.

It should be noted that despite local financeproblems in Eurasian countries, significant



EEUURRAASSIIAAUnited Cities and Local Governments116

progress has been made since the early1990s in a number of areas:

• Explicit assignment of expenditureshave been put into law in all thesecountries;

• Meaningful revenue decentralization hastaken place;

• Transparent equalization transfer me-chanisms have been developed in mostof these countries.

In this chapter we review how particularcountries in Eurasia have addressed theirlocal finances in different ways. We alsotake into account the common difficultiesthat local governments in the region con-tinue to face and are hopeful that effectivesolutions can be found for improving thelocal finance systems in Eurasia.

Territorial Organization and the Size ofLocal Governments

Across the region there is a wide variety inthe size of the public sector as well as in

the division of spending across the levelsof government. The territorial andeconomic features of the countries underreview in this chapter are presented inTable 4.1.

As shown in Table 4.2, the size of thepublic sector ranges from 23 percent ofGDP in Kyrgyzstan to 45 percent of GDP inUkraine. Across the Eurasian governmentsector, local government spending makeup a relatively small percentage of totalgovernment spending, with Armeniahaving the lowest share at 6 percent.1 Thistrend reflects the low level responsibilitiesassigned to the local level across all Eura-sian countries. Countries with budgetsallocated to local levels, cities, and raionscan be united into one group for compa-rison. The category of local budgets inUkraine, Belarus and Kazakhstan includeoblast budgets thus making the expendi-ture share of these combined localbudgets the highest: 31 percent, 36percent and 53 percent respectively.Oblast executive authorities in this groupof countries are subordinates to thecentral government which retain control

1. The share of localgovernment in thegeneral governmentbudget according tothe IMF data differsfrom, figures found inselected publications(e.g. Tumanyan, 2009[2]). The differencecould appear due to thecountry’s accountingfeatures. Besides,researchers often usethe share of localbudgets in theconsolidated budgetarygovernment but not inthe generalgovernment.

Countries Population Urban population Capital City Total Area Population Density GDP per capita GDP per capita, PPP(thousands) (% of total) Population (thousands) (km2) (persons per km2) (2008 US$) (current international $)

Armenia 3,234.2 63.9 1,113.3 29.8 109.1 3,872.7 6,074.7

Belarus 9,671.9 73.5 1,829.1 207.6 47.7 6,230.1 12,278.2

Georgia 4,382.1 52.7 1,106.7 69.7 62.0 2,969.9 4,965.5

Kazakhstan 15,766.5 57.9 633.7 2,724.9 5.8 8,513.1 11,323.2

Kyrgyzstan 5,276.0 36.3 785.1 198.5 27.5 958.4 2,192.6

Moldova 3,572.7 41.8 592.9 33.8 110.5 1,693.8 2,979.4

Russia 142,008.8 72.8 10,470.3 17,098.2 8.7 11,831.5 15,922.5

Ukraine 46,143.7 68.0 2,765.5 603.7 79.8 3,898.9 7,276.8

Data source: WDI, capital city population and territory - countries’ statistical offices

Table 4.1: General Country Statistics, 2008


117

over the execution of functions even if ittransfers many of them to the lower level.

All countries under review became inde-pendent after the disintegration of the So-viet Union. With the exception of theRussian Federation, all of them are unitary

states2, and adopted their constitutionsbetween 1993 and 1997. These consti-tutions laid the foundation of local self-governance to be further developed inlaws on local self-governments or statelocal governments. Belarus was the lastcountry to introduce the law On Local

2. Although Ukraine,Georgia and Moldovaare unitary states, theasymmetricassignment of powerswith the special rightsfor the autonomousterritorial units(republics) allowsregarding them as anintermediate stepbetween the unitaryand federal state. Fordiscussion onasymmetricdecentralization inGeorgia see Kirn andKhokrishvili (2008).The idea offederalization ofUkraine is discussedwidely between theexperts and politicians.


Countries Armenia (2008) Belarus (2008) Georgia (2007) Kazakhstan (2007) Kyrgyzstan (2006) Moldova (2008) Russia (2008) Ukraine (2008)

Total Expenditure of General 23.9 48.7 29.0 20.6 23.3 41.4 42.9 45.4

Government (% of GDP)

Expenditure of Local Government 1.3 17.3 6.1 11.0 5.5 10.2 7.8 14.2

(% of GDP)

Local Expenditures as a Share 5.5 35.5 21.2 53.4 23.8 24.6 18.3 31.2

of General Government (%)

NOTE: Data for Belarus, Kazakhstan and Ukraine include oblast level figures. For Kazakhstan general government figures are not available, figures refer tothe budget of the government.Data source: for Kyrgyzstan – Ministry of Finance data, for other countries – Government Finance Statistics Yearbook 2009 (IMF)

Data source: for Russia for 2003-2004 author’s calculations based on the RF Ministry of Finance data, for other countries– IFS database (IMF).

Table 4.2: The Size of Government Expenditures

Figure 4.1: Local Government Outlays as a Share of General Government Outlays (%)

Armenia

Belarus

Georgia

Kazakhstan

Moldova

Ukraine

0%

10%

20%

30%

40%

50%

60%

2003 2004 2005 2006 2007 2008

Russian Federation



Government and Self-Governance in2010.3

As former republics of the USSR, thesecountries had, at one point, a commonsystem of administrative and territorial divi-sion. The USSR’s republics were divided into

regions (oblasts, krais, and autonomousrepublics). Regions were further divided intoraions (the equivalent of counties or districts)and cities of regional subordination. Raions,in their turn, included cities of raionsubordination, towns, and villages, rural andurban settlements. This administrative

3. Belarus regulated itslocal governments bya normative act(1991) that wasenacted before itsconstitution wasadopted.

Table 4.3: Territorial Organization of Local Governments 4(2009)

Source: compiled by authors (see bibliography).

Countries Regional level Intermediate level Local level

Armenia

Belarus

Georgia

Kazakhstan

Kyrgyzstan

Moldova

Russia

Ukraine 243 towns657 settlements10,222 villages

488 raions176 cities

25 regions:24 oblasts and Autonomous Republic of Crimeria2 special status territories: Kiev (capital) and Sevastopol

21,595 (1,734 urban and 19,861 rural)settlements

1,799 raions521 cities (“gorodskoi okrug”)

83 subjects of RF:21 republics, 46 oblasts, 9 krais, 4 autonomous okrugs,1 autonomous oblast, 2 federal cities: Moscow (capital) and St. Petersburg

3 municipium (Comrat, Bender and Tiraspol)51 towns847 villages (communes)

32 raions2 municipium (Chisinau and Balti)

1 autonomous territorial unit Gagauzia(Gagauz Eri)6

11 cities of raion subordination472 settlements (ayls)

40 raions12 cities of oblast subordination

7 oblasts(administrative regions without budget rights)2 special status territories:Bishkek (capital) and Osh

45 cities and townsabout 2,500 rural settlements without budget rights

160 raions39 cities

14 oblasts2 special status territories: Astana (capital) and Amaty

5 cities: Tbilisi (capital), Batumi, Rustavi, Poti, Kutaisi4 communities (Eredvi, Kurta, Tighvi and Azhara) 60 raions

9 administrative regions (administrative regions without budget rights)Autonomous republics of Abkhazia and Adjara5

1,524 rural municipal units (cities/towns of raionsubordination, villages, settlements)

118 raions23 cities

6 oblastsCity of Minsk (capital)

City of Yerevan (capital)48 urban and866 rural communities

10 marzers(administrative regions without budget rights)


119

structure became the starting point forestablishing local governments throughoutthe region and remained relevant during thefirst years of independence.

Although the countries originally had a uni-form administrative and territorial division,later, in the course of decentralization, theypreferred different options for territorial or-ganization of local authorities (see Table 4.3).

Local authorities in Eurasia do not alwaystake the form of countries with a longtradition of local self-government. Inmany countries (Belarus, Kazakhstan,Kyrgyzstan and Ukraine) the executivebodies of local governments are subjectto the executive bodies of the higherlevel (embedded in the “verticalpower structure”). At the same level ofadministrative-territorial division, thereare elected councils, who adopt the budgetprepared by the executive authorities, andtake other decisions according to thepowers assigned to them. The degree ofaccountability of the executive local bodyto the higher level of government and thedegree of influence of elected councils atthe execut ive local body var ies indifferent countries of Eurasia. Such anorganization of local self-government isa compromise between deconcentrationand the devolution of powers. Undersuch arrangement local authoritieshave the greatest attributes of localself-government: an elected local council,local ownership and local budgets whileunder a deconcentrated system the localbudget is usually a part of the centralbudget (as the budgets of marzers in Armeniaor the budgets of oblasts in Kyrgyzstan).

In Kazakhstan, local governments donot have the official status of localself-governments. According to the law they arecalled local bodies of state government,though they do have elected local councils

and expenditure responsibilities andrevenue sources assigned to them. Localauthorities include oblasts and cities ofrepublic subordination (Astana and Almaty)as well as raions and cities of oblastsubordination. At the same time, the centralgovernment directly interacts only with theoblast level, while oblasts buildintergovernmental fiscal relations withraions and cities of oblast subordination.There are local councils at the sub-raionlevel but as they have no budgets of their ownthey cannot be considered a full-fledged levelof public administration.

In Armenia, the raion level was abolished. Theterritory of the country is divided intoadministrative provinces (marzers), andeach province includes several formerraions. Marzers are deconcentratedsubdivisions of the central government.Local self-government units have beenestablished in the City of Yerevan (lawadopted in 2009, before which theterritory of Yerevan was divided into 12neighborhood communities), and in localcommunities at the level of rural and urbansettlements (first law adopted in 1996, newlaw in 2002).

Georgia has administrative subdivisions ofcentral government at the regional level(krais, autonomous republics and the city ofTbilisi). Initially, Georgia had two subnationaladministrative levels, the raion and the localself-governments at the settlement level.Only the settlement level had some degree ofautonomy.7 Since 2006 the new Organic Lawon Local Self-Government introduced aone-tier system of local self-government.Today the municipalities at the raion level(including large cities) constitute the onlyterritorial level of local self-governance.However the dissatisfaction with thereform results, calls for the necessity offurther local self-governance reform sincethe number of residents in newly

4. In some countries (e.g.Russia and Ukraine)the large cities(Moscow and St.Petersburg, Kiev) couldbe divided further intomunicipalitiesaccording to the speciallaws but thesemunicipal units havevery limited powers sowe do not considerthem in the Chapter.

5. There is also onebreakaway Republic ofSouth Ossetia on theterritory of Georgia.According to theOrganic Law “On LocalSelf-Government” “Onthe territory of Georgiawhich is not underGeorgia’s jurisdiction,the authority and rulesof establishing the localself-government bodieswill be determinedafter restoration ofGeorgia’s jurisdiction incompliance withconditions set by thislaw”

6. Territorial unit StingaNistrului (Transnistria)is not included into theintergovernmentalrelations system ofRepublic of Moldova.

7. See Shergelashvili andNarmania (2006).




formed municipalities is too large toallow populations to feel represented(Losaberidze, 2009).

In the Republic of Kyrgyzstan the system oflocal government was introduced in 1995 forBishkek city. By 2001, all cities, townsand villages received the right to localself-governance. Initially Kyrgyzstan had asystem in which oblasts and raions had localbodies of state executive power while cities,towns and settlements (ayls) set up localself-governments with their own respon-sibilities and revenue sources. Later oblastsbecame administrative subdivisions of thecentral governments without budget rightsand without their own representative bodies.According to the legislation raioncouncils (kenesh) consist of representativesof sub-raion municipalities and,therefore, express the interests of localself-governments. Heads of the executiveauthorities of the cities (mayors) are “elected”by the deputies of the local keneshes afternomination by the President of the Republic.Since 2009 the heads of the cities of raionsubordination, villages and settlements areapproved by the deputies of the localkeneshes on the proposal of the head of raionadministrations.8 However with the recentgovernment overthrow (2009), new processesor restrictions may be on the horizon.

In 1998 the territory of the Republic ofMoldova was organized into 11 units, of which9 were regions (judets), one was the autono-mous territorial unit Gagauzia, and the otherwas Chisinau municipality. In 2001 the regio-nal level (judets) was abolished and the raionsystem was adopted. The raions wereendowed with effectual power after the localelections of 2003 (Veverita, 2006).9 As aresult, the central government directly in-teracts with local self-governments in theraions, the two municipalities outside raions(Chisinau and Balti) and in the one auto-nomous territorial unit of Gagauzia. Raions, in

turn, interact with local governments insettlements and sub-raion cities. Transnistria,a breakaway territory on the border ofUkraine, though formally a territory of theRepublic, is not currently included into thebudgetary system of the Republic of Moldova.

Initially, the subjects of the Russian Federation,in compliance with the principles of federalismand on the basis of the old Law on GeneralPrinciples of the Organization of LocalSelf-Government (1995), set up variousmodels of local self-government. In most regions,local governments were established mainly atthe level of large cities and raions (the so-called“raion model” of local self-government). Inother regions, local self-governments were, forthe most part, set up at the level of large citiesand settlements (the “subraion model”), whileyet other regions had a combination of raionand subraion types of local self-government(the “two-tier model”).10 Starting from 2006, atwo-tier model of local self-governance hasbeen introduced in the country in accordancewith the new Law on General Principles of theOrganization of Local Self-Government (2003):local self-governments at the level oflarge cities (“gorodskoj okrugs”),11 raions andsettlements.

In Ukraine local self-government is exercisedby territorial communities of villages, settle-ments, and cities, both directly throughvillage, settlement, and city councils, andtheir executive bodies, and through raion andoblast councils that represent the commoninterests of the territorial communities ofvillages, settlements, and cities. Bodies ofexecutive power in oblasts and raions,similarly to those in Kazakhstan and Belarus,are embedded in the state “vertical powerstructure”. The central government directlyinteracts with oblasts, raions and cities whilebasic levels of local self-governments (towns,settlements and villages) interact only withthe raion bodies of state power. In Ukraine adraft Concept on Local Self-Government

8. Initially head of the citiesof raion subordination,villages and settlementswere elected by directelections.

9. This story has evolvedsince: under the Law onRegional Development(2006), 6 developmentregions were established.These regions are notadministrative territorialunits and have no legalauthority, however in thecontext of actual initiativeto reform the Moldovanconstitution there havebeen proposals toreorganize raions intolarge regions more viablefrom an economic pointof view.

10. See for additional details:Kurlyandskaya at al.(2001), Martínez-Vázquez at al. (2006) orDe Silva at al. (2009).

11. Some regions with thelow population densityhave tried to avoidnegative consequences ofthe fragmentation of localgovernments. They haveorganized “gorodskojokrugs” across the wholeterritory of the formerraions preserving one-tierlocal governmentstructure on the territory.An example of this can beseen in the Far-Eastregions with lowpopulation density (e.g.Sakhalin andKamchatka).


121

Reform in 2009 envisaged the creation ofelected councils in the oblast and raions,giving them new responsibilities andresources. However this project wasabandoned in 2010, and a program ofmergers between small municipalities wasbegun. The new budget code (adopted July 8,2010), which entered into force on 1 January2011, introduced no major changes to thesystem of state funding. As long as citiesreceive direct resources from the Statebudget, the local government framework isunlikely to change. Concurrently there isincreased encouragement for joint fundinginitiatives between local governments forinvestments.

The Republic of Belarus has preserved the oldadministrative division with regard to localauthorities. Although de jure local councilscalled “bodies of local self-governance”12

exist, de facto they are included in the overallpublic administration system. Decisions onterritorial development, such as theconstruction of schools, care centers, or the

infrastructure of facilities (roads, transport)etc., flow down from the upper tiers ofgovernment. The same administrativepractice applies to the financial resources thatsupport these decisions. Local councils haveneither real authority to make decisions northe resources to execute them, though thelanguage used in local administration leadsone to assume the existence of localself-governance is plausible based on suchterms as “local councils”, “executive committee”or “the head of the city”. The new law OnLocal Government and Self-Governance(introduced in January 4, 2010, and cominginto force six months after its officialpublication) has not made any changes to thissituation. The new law regulates theresponsibilities of local councils and theresponsibilities of the executive committeesof oblast, basic (raions) and primary(settlements) levels. According to the new law“The oblast councils are superior to thecouncils of the basic and primary levels” and“basic level councils are superior to thecouncils of the primary level”. Executive

12. See the Constitution ofthe Republic of Belarusand the law On LocalGovernment and Self-government in theRepublic of Belarus.


Countries Regional level Intermediate (raion) level Settlement level

Armenia D LSG

Belarus LG/LSG LG/LSG LG/LSG

Georgia D LSG +

Autonomous Republics (LSG)

Kazakhstan LG LG LG (without budget rights)

Kyrgyzstan D LG/LSG LSG

Moldova LSG (include Autonomous Republics) LSG

Russia Subjects of Federation LSG LSG

Ukraine LG/LSG LG/LSG LSG

D – deconcentrated units of central governmentLG – local state government bodiesLSG – local self-government bodiesLG/LSG – local executive bodies included into hierarchical “vertical power structure” and local representative bodies

(councils) with a status of local self-government

Source: compiled by authors.

Table 4.4: Legal Status of Local Governments



committees could cancel the decisions oflower executive and administrative bodies,and the decrees of their heads, if they do notcomply with other acts of legislation or thedecisions of the council or the executivecommittee of the regional level.

Most Eurasian countries considersettlements and cities the main level oflocal self-governance closest to citizens.However, the question of the territorial levelsuitable for setting up local self-governanceremains open for discussion. On the one hand,settlement authorities are closer to thepopulation and therefore have a betterunderstanding of popular concerns; on the otherhand, raion self-governments can ensuremore effective services due to economies ofscale, higher levels of staff competency, andthe remains of the traditional raion-levelorientated budgetary system in former USSRcountries. At the same time, raions are oftencalled on to fulfill many mandates delegatedby the regions or central government (e.g., thesocial security function). Compromiseshave been found: such as two-tier systems oflocal self-governance or in combiningexecutive bodies to include both representativesof higher levels of government (from thehierarchical “vertical power structure”) and localcouncil members (from local self-government)at the raion level (see Table 4.4).

Local Finance

Expenditure Assignment

Initially, the assignment of expenditure res-ponsibilities between the levels of govern-ment in the countries of the former USSRcorresponded to the ownership of infra-structure and other properties: the centralbudget funded institutions belonging tocentral government, and local budgets wereresponsible for local institutions irrespectiveof the services rendered by them. Often lawswere amended to include provisions that

assigned the same responsibilities to differentlevels of government. It was thereforedifficult, when the new states were dividingcentral and local responsibilities, to decidewhat to allocate to each level of governmentand, accordingly, to estimate expenditureneeds in building a system of equalizationtransfers.

Gradually, however, the Eurasian countrieshave come to recognize the need for a clearerassignment of expenditure responsibilities inthe creation of an effective system ofintergovernmental fiscal relations. Most of thecountries under review regulate expenditureassignments by means of the Budget code orthe Law on Local Self-Governance. The onlyexception to this is Moldova where issues oflocal responsibilities are determined in theLaw of Local Public Administration and theLaw on Administrative Decentralization (bothenacted in December, 2006), while the Lawon Local Public Finance sets the publicexpenditure assignments.

In Russia, the laws establishing a new systemof expenditure assignment were adopted in2003. The law On Local Self-Government thatformally outlined the responsibilities of cities,raions and settlements came into force in2006.

In Armenia and Georgia local self-governmentresponsibilities are listed in the laws On LocalSelf-Government. For Armenia this law wasadopted in 2002 and replaced an earlier 1996law, in Georgia the Organic Law was signed inDecember, 2005.

In Ukraine, the Budget Act of 2010, likethat of 2001, is based on the distinctionbetween on the one hand, income andexpenses for services consideredguaranteed by the State, which are basedon a national equalization system (first"fund"), and on the other expenditureand revenue for local needs, which is



Expenditure/Countries Armenia Belarus Georgia Kazakhstan Kyrgyzstan Moldova Russia Ukraine

Public OrderSecurity, police C C, R C C,R C C, L2 C, (L2) C, R, L2Fire protection C C, R, L2 (L2) C C, L2, L1 L1 R R, L1Civil protection L1 C C C C, L1EducationKindergartens or pre-school education L1 * L2 L2, L1 L1 L2–>, L1 L2 L1Primary and secondary education C * C L2 L1 L2–>, L1 R–>, L2 L2, L1Vocational and technical education C * C R C L2 R C, RHigher education C * C C C C C ,(R), (L2) CSpecialized or additional/off-school education L1 * C L2 C L1 L2 L1Public HealthEmergency assistance C * C L1 C C R(aviation), L2 L2, L1Primary care C, (L1) * C, (L2) R C C L2, L1 L1Hospitals C * C C, R C, L2, L1 C R, L2 R, L2Social WelfareFamily welfare services C L2 C L2, L1 C L2, L1 C, R R, L2Welfare homes C R C R R R, L2Social security C, L1 C C C, L2, L1 C, L2, L1 C R, L2Housing and town planningHousing L1 L2 L2 L1 L1 L1 L1Town planning L1 L2 L2 L2 L1 L1 L1 L2, L1Cemeteries & crematoria L1 L2 L2, L1 L1 L1 L1 L1UtilitiesGas L1 L1 R, L2 L1 C, L2, L1 L1 R, L2, L1 District heating L1 L2 L1 L1 L1 L1Electricity L1 L1 L2 L1 C, P L1 L1Water & sewage L1 L2, L1 L2 R, L2, L1 C, L2, L1 C, L1 L1 L1Refuse collection L1 L2, L1 L2 L2 L2, L1 L1 L1 L2, L1Refuse disposal (L1) L2, L1 L2 L2 L2, L1 L1 L2 L2Culture, leisure & sportsTheatres & concert halls C, L1 C, R, L2 C, L2 R, L2 C, L2, L1 C, L1 R, L2, L1 R, L2Museums C, L1 C, L2 C, L2 R, L2 C, L2, L1 C, L1 R, L2, L1 R, L2Libraries C, L1 C, L2 C, L2 R, L2 C, L2, L1 C, L2–>, L1 R, L2, L1 R, L2, L1Parks & open spaces C, L1 C, L2 C, L2 C, L2, L1 L1 L1 L1Sports & leisure L1 L2 C, L2 R, L2 C, L2, L1 L2, L1 R, L2, L1 R, L2Other cultural facilities (clubs) L1 L2 L2 R, L2 C, L2, L1 L2–>, L1 L1 R, L2, L1Roads, transportLocal roads L1 L2, L1 L2 L2, L1 L1 L2, L1 R, L2, L1 L1Transport L1 C, R, L2 L2 L2 L1 L1 L2, L1 R, L2, L1

L1 – settlements; L2 – raions; R – regions; C – central government.(…) – voluntary/discretionary authority;* – expenditures are assigned according to the allocation of institutions (property); –>– responsibility for financing salary and some other expenses through earmarked grants to other level of government.

In countries with a multi-tier system of local government, the list of responsibilities of cities of oblast subordinations usually combines the lists ofresponsibilities for raions and settlements, cities of oblast level subordination fulfill the lists of responsibilities for raions, settlements and oblasts unlessotherwise stated.

Source: analysis of country legislation, UCLG country data files, publications listed in bibliography and comments from the UCLG workshop participants.

Table 4.5: Expenditure Assignment Among Levels of Government



not part of the matching system (thisincludes housing and municipal services,infrastructure and transport).

In Belarus, prior to the new 2010 law On LocalGovernment and Self-Government, expen-diture responsibilities were not delineated(see Bakanova and Martínez-Vázquez, 2007),and the assignment of functions were made inaccordance with assignment of institutions.The new law places more attention on thedivision of powers between executive autho-rities and local councils than on the verticalassignment of responsibilities.

In Kyrgyzstan, the responsibilities of localself-governments were outlined in 2008in the law “On Local Self-government andLocal State Administrations”. Meanwhile inKazakhstan, the expenditure assignments ofall levels of government were assigned in theBudget Code (2007). Sub-raion municipalitiesin Kazakhstan have no budgets of their own;however according to the Budget Code,raions and cities of oblast subordinationprovide in their budgets for the expenses ofcity districts, cities of raion subordination,settlements and auls (villages).

Across the Eurasian countries the wording ofexpenditure responsibilities in legislationcan be quite vague, and it is sometimesvery difficult to determine which level ofgovernment is responsible for the provision ofa certain service. In addition, in manycountries, there are exceptions to the generalrules of expenditure responsibility allocation,where an authority can take on additional(discretionary) obligations, usually financedfrom the budget of another level.Nevertheless, according to the legislation, thegeneral scheme of allocation of expendituresis as follows (see Table. 4.5).

The Ukraine budget code delineates “own”and “delegated” responsibilities. The latterincludes education, health care and social

protection. “Delegated” responsibilities arefinanced through the general budget fundsand are taken into account in the process ofthe allocation of equalization grants.

In some countries responsibilities are unclear:it is difficult to find out which level ofgovernment is responsible for what service.For example, in Georgia local government“Performs social-cultural activities andsupports the activities of the relevant objects(educational and educative institutions, etc.)having local importance” and “establish pre-school and other education institutions”;following this arrangement general educationcould be provided by local or centralgovernment (de facto it is provided by centralgovernment). In Moldova the legislation onlocal self-government assigns the samepowers to local authorities of the first andsecond levels (see Popa, 2007).

In Eurasian countries there are also somecases of local governments financing issuesthat do not belong to their exclusiveauthority. In Georgia local budgets financelocal police departments but according tothe Organic Law on Local Self-Governmentmaintenance of the public order does notbelong to the exclusive authorities of localgovernment.13 According to the same law,local governments could also have voluntaryauthority, within which authority “the localself-governing unit shall be entitled as itsown initiative to make decision on funding insocial, cultural and educational spheres”. Itis questionable though, if financing thepolice could be considered the socialspheres. Before 2009 the definition waseven stricter: voluntary authority couldinclude issues “not belonging to its exclusiveauthorities or to the authorities of the statebody and those not prohibited for the self-governing unit”.14 According to this formerdefinition financing the police by the localself-governing units could not be consideredas an execution of voluntary authority, be-

13. In Georgia according tothe legislation localgovernments couldhave “own”,“voluntary” and“delegated”responsibilities.

14. The similar restrictionsare pointed out in thelaw “On generalprinciples of theorganization of localself-governance in theRussian Federation”.But in the Russian lawsome exceptions aremade.


125

cause according to the GeorgianConstitution (Article 3), criminal police andinvestigation is an exclusive responsibility ofGeorgia’s higher state bodies (see CIESR,2009 for detailed discussion on the topic).

The issue of local police funding is a compli-cated one for more countries than simplyGeorgia. According to the RussianConstitution (Article 132) “the local self-government bodies … ensure the protection ofpublic order”, and according to the law OnLocal Self-Government in the RussianFederation (RF) local government issuesinclude “providing municipal police”, however,implementation of this issue requiredamending the existing law enforcementsystem as public order and safety was, priorto these legislative changes, provided by theFederal Ministry of the Interior. Somelocalities (usually large cities) historicallyfinance their “own” municipal police but thepolice force follows the instructions of andreports to the Ministry of the Interior. Nowinstead of separating police into the federaland local (as it was the intention at the timeof the reform of responsibility assignment)federal government is planning to simplyrecentralize the police force.15

Another problem of expenditure assignmentin Eurasia arises from unfunded mandates.16

Unfunded mandates exist in Armenia,Kyrgyzstan and Moldova. In Armenia, forexample, of the 19 delegated responsibilitiesgiven to local governments, only 2 arefinanced from the central budget. InKyrgyzstan, central government insteadof transferring funds for delegatedresponsibilities simply converted theunfunded mandates into official local govern-ment responsibilities without making anymodification to the revenue assignment oflocal government. The central government’sdecision on the growth of salaries in Moldovacould be taken as an example of unfundedmandate.

In Kazakhstan, Russia, Ukraine, Belarus17

and Georgia unfunded mandates areforbidden but de facto some legislativedecisions of higher levels of government doresult in additional financial burdens to thelocal level.

In Russia, for example, if federal governmentincreases the salary of federal police, regionaland local government must also increase theirexpenditures on public order and safety. In theyear that this change is implemented thefederal government provides a special grant forincreasing salaries but in the following yearssubnational budgets must find their ownrevenue sources to cover this additionalexpense. This practice hits local budgetsespecially hard during times of economic crisis(a recent example is Yaroslavl city).18 InKazakhstan, hot breakfasts at school wereintroduced by the central government withoutadditional funding assigned to the localgovernments providing the services.

Despite the formal delineation of authority,central governments often seek to retaincontrol over local spending. One way to do so isto allow the central treasury to manage(meaning control the allocation of) publicexpenditures (Armenia, Russia) even though,on paper, the treasury is simply supposed tomanage the internal control of funds and neverdecide on the priorities for the allocation ofresources, it can, and often does, exert quitesignificant influence over these. For example,Armenian observers have noted that thetreasury often intervenes in the financialoperations of the local communities.19 InRussia, the Federal Treasury asks for detailedassignments of budgeted expenses andrequires exact execution of those expenseswhich reduces flexibility within the budget exe-cution process.

Another method used by Federal or Centralgovernments to exercise control over localexpenditure is to regulate by law the number of

15. Regional governmentsco-finance the federalpolice. In the initialversion of the legislationon assignment ofexpenditureresponsibilities thecross-subsidizing offunctions of other levelsof government wasforbidden. However anexception was latermade for policefinancing.

16. The obligation for localgovernments to providea service or perform atask without full oradequate funding fromthe upper-levelgovernment thatintroduces therequirement orobligation.

17. It is not easy todetermine if theresponsibility could beconsidered as amandate in Belarus dueto unclear expenditureassignments andvertical subordination ofthe executiveauthorities.

18. Another example ofunfunded mandates inRussia can be found inthe laws devoted to thequality of services andsafety standards. Whilerendering publicservices, localgovernments fall withinthese federal normativeacts. However, theyoften do not have thefunds to comply with allthe requirements of thefederal legislation. If thefederal prosecutor'soffice (the mainauthority that overseesthe compliance with thefederal legislation) findsviolations of federalnorms and standardsby a municipality, thelocal government musteliminate the provisionof those services. Butother agencies couldcontinue to render thesame services, with thesame violatedstandards of federallaws unless or until theyare themselvesevaluated byrepresentatives of theprosecutor's office.

19. UCLG Country data fileon Armenia.




local government and public sector employees.This can be found in all the countries withexecutive bodies with vertical power structuressuch as Belarus, Kazakhstan and Ukraine, andthis is also true for Kyrgyzstan. Theexpenditure autonomy in terms of personnel atthe settlement government level in Moldova islimited due to the high share of expenditurefinanced through earmarked grants comingfrom the raion level (specifically for wagepayments in the social sphere).

In Georgia, according to the law on StateSupervision over Activities of LocalSelf-government Bodies, central governmentbodies that execute control over the localgovernments have the right to cancel localnormative acts (see Losaberidze, 2009).

In Russia, the Budget Code imposes constraintson municipalities depending on their share ofintergovernmental transfers in total revenues,discounting any grants for financing delegatedresponsibilities (see Box 4.1). As localgovernment budgets strictly depend onintergovernmental transfers it is little wonderthat local governments are under strict control.20

The expenditure structure of local govern-ments depends on the assignment of res-ponsibilities between the different tiers ofgovernment and between the public and pri-vate sectors.21 Expenditure outlays includeboth the amounts related to own anddelegated responsibilities and budgetary re-ports do not distinguish between spending onthese (see Figure 4.2).

20. The same norms existsfor the regionalgovernments (see DeSilva at al. 2009).

21. The share of healthcare expenditures alsodepends on theexistence of theMedical InsuranceFunds, while socialprotectionexpenditures dependon the existence of thePension Fund and otherfunds of accumulatedresources for socialprotection.

22. In transitional countriesthe low qualifications ofthe public sectoremployees at thesettlement level couldalso contribute or be aprincipal reason for thesmall number offunctions assigned tothe local level.

Box 4.1. Restrictions to Local Government’s Autonomy in Russia

Share of grants(including tax transfers) in total revenues of a local government Restrictions

0 ≤10% None

> 10% for two consecutive Salary of local officials shall not exceed the ceilingsreporting years established by the regional council.

> 30% for two consecutive Local spending is limited to matters explicitly reporting years listed as local expenditure responsibilities

in federal and regional legislation.

> 70% for two consecutive • The local government shall sign an reporting years agreement with the regional government to

increase the efficiency of local spending and enhance collections.

• The local government’s budget is submitted to the regional government for approval.

• The regional government audits end of year local budget execution reports at least every other year.

Source: De Silva at al. 2009, based on the Budget Code of RF.


127

The list of local expenditure responsibilitiesis significantly impacted by theadministrative-territorial division. Thus, thenumber of functions assigned to settlements isusually not large due to the diseconomies ofscale and spillover effects.22 Typicallysettlements are assigned responsibility forhousing and community amenities, pre-schooleducation and recreation. If raion budgets areincluded in the local government category,the list of responsibilities can include,in accordance with expenditure assignmentcriteria, general education and primary healthcare. If oblast budgets are regarded as local,expenditures on health care and education willconstitute a large portion of total outlays, andwill support many related responsibilities. Dueto economies of scale, services provided at thislevel will be less expensive and will reach agreater portion of the population.

In the presence of a large number ofthinly populated municipalities, per capitamanagement costs will grow, together with theshare of expenditures on general public

services in local budgets. This is illustrated bythe case of Armenia where the mainexpenditure line items are general publicservices (26 percent)23 and housing andcommunity amenities (26 percent). The shareof education is also important at 18 percent(see Figure 4.2).

In Georgia, where there are no settlementgovernments, local budgets spend much lesson general public services – 8 percent ofexpenditures.24 The main share of expendi-tures falls on Housing and communityamenities (40 percent). 16 percent ofexpenditures in Georgia go to economic affairs.In comparison, expenditures on culture (8percent) and education (11 percent) lookrelatively small.

In spite of the fact that both Russia andMoldova have a two-tier system of localself-government, the difference betweenthese two countries in responsibilities assignedto the local level results in different expenditurestructures. In Russia, education (general and

23. This share becomeseven greater if theadditional 11.9 percentfrom "Expenditures notclassified elsewhere" istaken intoconsideration, as mostof these expenditurescan be attributed toGeneral PublicServices.

24. It is difficult to comparerevenue andexpenditure structuresin different countriesespecially when dealingwith transitionalcountries. As a result ofdifferences in budgetclassifications,expenditures on thesame function can beassigned to differentfunctional groups inbudgets of variouscountries. The IMFdatabase could be agood source ascountries provide dataaccording to commonstandards incomparable budgetclassifications.However, excessiveconsolidation of thedata in the GFS causesother problem. Thedata gives one figurefor all localgovernments. This is isnot a significantproblem when we lookat a country with one-tier subnationalgovernment (such asArmenia or Georgia) ora country with a two-tier local self-government structure(Russia, Moldova).However, for countrieswhere the state localgovernment system issupplemented withlocal self-governmentbodies (Belarus,Ukraine) it is better todistinguish betweenthe different levels ofthe budgetary system.


0%

20%

40%

60%

80%

100%

Armenia Belarus Georgia Kazakhstan Kyrgyzstan Moldova Russia Ukraine

General public services Education Health

Housing and community amenities Economic affairs Recreation, culture, and religion

Social protection Other

Source: Authors’ calculations based on IFS database, for Armenia – on "Local Self-Government Reforms in Armenia (2007and 2008) Book 3", for Russia – on RF Ministry of Finance reports. Data for Kyrgyzstan is for 2006 (cf. Annex 4.1).

Figure 4.2: The Structure of Local Government Outlays (2008)



preschool) accounts for 37 percent of allexpenditures, in Moldova more than 50percent, while Russia spends 12 percent oflocal budgets on health care and Moldova onlyabout 2 percent. In Russia housing andcommunity amenities accounts for 21 percentof total outlays while in Moldova this is only 9percent.

In Kyrgyzstan, where raion budgets belong tothe local level, local spending on health carewas not high (only 5 percent of expenditures).At the same time, the share of expenditureson education reached 59 percent. Later theMedical Insurance Fund was created inKyrgyzstan and health care expenditure werecentralized.

Expenditures on general public servicesamount to between 11-13 percent inKazakhstan, Kyrgyzstan and Ukraine, whilein Belarus they are as high as 20 percent,this can be attributed both to a highnumber of administrative staff and also todifferent approaches to the classification ofexpenditures.

In Belarus, Kazakhstan and Ukraine, wherethe expenditure of local government includesthose of oblasts, the local level supports notonly pre-school and general education but insome countries even vocational training; thisadds to spending on education and the shareof this line item in expenditures is larger (25percent - 29 percent). Healthcare is anotherimportant expenditure item in this group ofcountries (19 percent - 20 percent). Economicaffairs remain at 11 percent of expendituresin Belarus and Ukraine and reaches up to 20percent in Kazakhstan.

One should also note the tendency towardslow spending on social protection: incountries with one-tier local governments(Armenia, Georgia) and countries with oblastbudgets (with the exception of Ukraine),relevant expenditures constitute 2-5

percent.25 The low share of this line item canbe explained by the fact that social support isfinanced through special funds. In Russia andMoldova, the share of social protection isabout 8-10 percent; while social protection isnot assigned to local governments, inRussia, however, regional governments oftendelegate social protection functions to thelocal level along with earmarked grantsfor its funding. On the other hand, localgovernments in Ukraine spend 17 percent onsocial protection.

The ideal assignment of expenditure res-ponsibilities does not exist. The scope ofresponsibilities in each of the countries underobservation depends, among other things, onthe territorial organization of authorities, andon the readiness of the central governmentto transfer them to the subnational level. Incountries with local bodies of state power(Belarus, Kazakhstan), more responsibilitiesare assigned to the local level through thedecisions of executive bodies of power. Also,raion governments have more responsibilitiesassigned to them than settlementgovernments. This approach can probably beexplained by economies of scale or low staffqualifications at the settlement level.

Revenue Assignment

In discussing effective intergovernmentalfiscal relations the next issue to examine isrevenue assignment. Some countries (forexample in Kyrgyztan) even began theassignment of revenues before that ofexpenditure responsibilities of localgovernments had been determined. In theRussia of the 1990s, regions and localgovernments first competed for revenuesources before addressing the question ofwhat should be funded by these moneys. TheRF Budget Code provided for the federationand regions to share tax revenues in a 50:50division,26 while the Law on FinancialFoundations of Local Self-Government

25. At the same time, oneshould remember thatper capita spending inabsolute terms ishigher in countries withmulti-tier budgetarysystems.

26. In practice, however,the normative hadnever been observed.


129

assigned shares of federal taxes to localgovernments (e.g., 50 percent of regionalenterprise property taxes were to betransferred to all local budgets).27

To assess the actual level of revenueautonomy, one has to distinguish betweenown taxes (taxes for which localgovernments have authority/discretion tochange tax rates and/or tax bases) andshared taxes (taxes shared between differentlevels of government whose rates and basescannot be changed by local government).From the point of view of local governments’

fiscal autonomy, the difference betweentaxes that are allocated to local governmentson a permanent basis and those where thepercentage received is set annually by a lawon budget of a higher level of government isimportant. However, not all countries havelegislation that clearly distinguishes betweenown and shared taxes. Many view all taxesthat go to local governments as own taxes,and distinct from transfers and grantsallocated as a sum of money. For example,under the RF Budget Code own revenuesinclude not only tax and non-tax revenuesbut also budget revenues in the form of

27. Some regions, such asRostov Oblast brokethis mandate, leavingRostov on the Don City,for example, with 36percent to 40 percentof collections in1997–2000.


Box 4.2. Tax Reform in Russia: Reduction of the tax burden at the expense of sub-national governments

The history of the tax system reform in Russia has amounted to a gradual reduction of thetax burden for taxpayers at the expense of regional and local government incomes.

Before 2004, the list of local taxes, according to the law On the Foundation Principles ofthe Tax System in the Russian Federation, included 22 taxes and fees. Fifteen taxes andfees from this list were cancelled for a Subject of the Federation in the case of itintroducing the sales tax of which 60 percent was due to local governments. However, in2004, the regional sales tax and the fifteen local taxes and fees that it substituted wereabolished by federal law.

In 2001, local governments were due 5 percent of the federal profits tax (which could bereduced to 0 percent if the local government so choses), in 2002-2004 this was reducedto 2 percent.

Starting in 2004, only four taxes were remitted to the local budgets: the land tax, thepersonal property tax, the registration fee paid by individuals engaged inentrepreneurship and the tax on advertising. From 2005 on, the Tax Code only assignedlocal government the land tax and the personal property tax.

As a result of the almost total lack of own revenue sources and administrative rightsover their collection, local governments are deprived of the flexibility to manage therevenue side of their budgets; moreover, the raion level of local government is leftwithout any own revenue sources.

The desire of the central government to cut the list of local taxes was partly a result of thecentralized collection of local taxes: tax authorities were not interested in improving localtax compliance and local taxes were abolished due to expensive tax administration.



grants, i.e. all revenues except forsubventions (transfers to support delegatedexpenditure responsibilities).

The search for stable revenue sources with anevenly distributed tax base represents a greatchallenge for transitional countries. However,due to spatial disparities across the countriesexamined, and their resulting differences inability to raise taxes, regional inequalities aregrowing. Also, the smaller the size of amunicipality (settlement, village), the moredifficult it is to find suitable revenue sources.

There is no doubt that property taxes are thebest alternative for local taxes. In the USSR,the group of property taxes included thepersonal property tax, the legal entityproperty tax and the land tax. Currently, someEurasian countries are still in the process ofseparating property taxes from land tax28

(Russia29, Armenia, Kyrgyzstan), others havecombined these revenue sources into a singletax. Also, in Moldova, Russia, Kyrgyzstan localauthorities can vary the tax rate within thelimits established by law, while in Armenia andUkraine they have no control over property taxrevenues. In all of these countries the taxbase for property taxes is set by Centralgovernment.

In Moldova and Ukraine local authorities, apartfrom property taxes, may impose local taxesand fees from a list, established by law by thecentral government. Usually, they can changethe rates of such taxes and fees within thelimits established by the Central govern-ment. In Armenia and some other countrieslocal governments can introduce and regulatelocal duties.

There has been a recent trend towards theabolishing of local taxes in some countries ofEurasia as some local taxes were considerednuisance taxes due to their high compliancecosts. For example, in Russia, the list of localtaxes was reduced from 22 to 4 in 2004 and,

finally, to 2 taxes in 2005 (see Box 4.2); inKyrgyzstan the number of taxes was reducedfrom 1630 to 8, and then in 2009, to 2, thelocal tax list was also cut in Georgia in 2004.31

Armenia provides an opposite example to thistrend: both land and property taxes are localas of 2011 and two additional local taxes(hotel tax and parking tax) are to beintroduced for 2011 (see Tumanyan, 2010).

Despite the obvious advantages of real estate(property) taxes as sources of local budgetrevenues, these taxes are not sufficient tocompensate the vertical imbalance in thebudget system of Eurasian countries. Thereare two reasons which contribute to this:(1) the significant amount of expenditureresponsibilities legally assigned to localgovernments and (2) the impossibility ofgathering high tax revenues from property,owing to the poverty of many jurisdictions oras a result of ineffective assessments of thetax base (for instance, in Russia the tax basefor the personal property tax is assessedaccording to the booked32 rather than marketvalue). Additional local taxes and feescollected in Ukraine, Moldova and some othercountries cannot solve this problem of verticalimbalance. For this reason the countries underreview also assign percentages of sharedtaxes to local governments.

The personal income tax (PIT) is often used asthe second-best choice to provide revenueautonomy to local governments. If paid at theplace of residence the PIT can be viewed as abenefit tax, or as a payment for public servicesfunded by local governments. Unfortunately,in some countries (e.g., Belarus, Russia,Ukraine) the tax, calculated on wages, is stillpaid at the place of employer’s registrationrather than at the place of employee’sresidence. This practice results in an unevenlydistributed tax base and a disrupted relationbetween the amount of taxes paid and theamount of services received.

28. This distinction has itsroots in the formerUSSR’s system oftaxation, where beforethe transition land waspublicly owned whileproperty could beprivate. This was thereason for creation oftwo separate taxes.

29. In Russia real estatetax was introduced asan experiment in 2cities: Novgorod andTver. According to theRF Government,property tax in Russiamay be introducedafter the presidentialelections in 2013.

30. Local taxes includeddog licenses; fees forthe organization of localcommercial auctions,lotteries, competitionsand exhibitions; tax onthe use of localsymbols; fee ontransactions made oncommodity and rawmaterials exchanges;tax collected ongrowing flowers ingreenhouses and thesale of them to thepublic; hunting andfishing licenses; tax ontourists travellingabroad; tax on videosalons and concertperformances.

31. The list of taxesincluded taxes oneconomic activity,gambling business,resorts, hotels andadvertisements. Toreplace these incomeslocal government wasassigned shares fromthe central tax.

32. Booked value is thevalue estimated at themoment of the creationof an asset, thencorrected throughamortization and theuse of some coefficientsto update it. So themarket value of theproperty often is farfrom the booked valueestimated for thepurpose of taxation.


131

Small business taxes (special tax regimes) areanother good candidate for the source of localrevenue. But successful applying of this taxdepends on the establishment of a propertax base. In situations where the localgovernment can administer the tax base forsmall business this tax could be a good sourceof local income.

Many countries also assign the enterpriseprofits tax to local governments. However, thisleads to a growth of revenue disparity due touneven distribution of the tax bases. Theassignment of the profit tax at the local levelalso exposes local governments to greaterrevenue volatility; this has resulted in ashortfall of local revenues during the recenteconomic crisis.

In addition to the central government, regionalor raion authorities also can assign shared taxrates to local governments. In Russia,subjects of the Federation can assign localgovernments shares of any taxes in theregional budgets. This means that localgovernment across the country have differenttaxes assigned to them. The RF Budget Codeonly requires that municipalities of the sametype must be assigned the same sharing ratesof shared taxes and that taxes must beimposed by a long-term normative act ratherthan a law on budget. In Kazakhstan, oblastscan set sharing rates of the PIT and the socialtax for raions and cities of oblastsubordination. In Moldova, authorities of thesecond level (raions) can assign taxes toauthorities of the first level. But the mostcomplicated system of shared taxes is inBelarus: some shared rates are established bythe budget code, others by the Republic’s andoblasts’ annual budget laws.

Of the countries under consideration, inArmenia, Moldova and Kyrgyzstan collection oflocal taxes (duties) has been transferred ordelegated to the local level. In Armenia localgovernment monitors the tax base and

controls the collection of land and propertytaxes. 75 inter-municipal inspections werecreated to fulfill this function. The delegationof property tax collection to the local level in2003-2005 resulted in a 38 percent increase incollected tax revenue. The delegation of landtax collection in 2006 resulted in a 36 percentincrease in collected tax revenue.33

In Kyrgyzstan, after the introduction of the TaxCode in 2009, local taxes and fees have beencollected by the National Tax Service. However,according to the Law on Local Self-Governmentand Local State Administration the collection oftaxes and duties is considered a responsibilitydelegated from the state level and before 2009the local government collected its own taxrevenues. The centralization of sales tax(which was converted from a local to a statetax) in 2009 resulted in a decrease in theamount of collected tax.

Financing local services through shared taxesand grants often results in a blurring of therelationship between taxes paid to andservices rendered by local governments. Localauthorities blame higher governments forfailure to allocate funds and residents do notdemand qualitative services in return for thetaxes they have paid.34

From the point of view of fiscal autonomy, weare interested in the estimation of localgovernment own revenues. However, theavailable reports do not allow revenues to beidentified separately. At first sight, it appearsthat local authorities can have more influenceon revenues than on grants. In the case ofshared taxes, however, local governmentshave little to no control or impact, and if theirfiscal autonomy is analyzed on the basis oftheir share of tax and non-tax revenues in thebudget structure it can be easily overes-timated.

Unfortunately in Eurasia local taxes are notreally “local”; central government can grant

33. For both taxesestimation was madefor 2008 in real terms(Tumanyan, 2009).

34. It is also important toremember thementality of those whogrew up under theSoviet system, whichcannot be changedeasily. There is atendency to think thatbudgetary servicesmean servicesprovided free of chargeand the fact that theyhave paid for municipalservices by paying theirtaxes is not wellrecognized (this isespecially true in thecase of indirect taxesthat are passed on toconsumers of goodsand services).




tax benefits at the cost of local taxes,thus reducing the local tax base and fiscalautonomy. For example, in Georgia andRussia, the central (federal) governmentlegislation establishes tax exemptions on localproperty taxes.

In Russia, for instance, before 2005 subjectsof the Federation were able to setdifferentiated sharing rates for taxes due tolocal government (different rates wereadopted for various municipalities in thejurisdiction of the same region). As a result, ifa region assigned sharing rates, the portion oftaxes in the structure of its local budgets washigher than in a region where grants-in-aidwere allocated. But it would be anexaggeration to assume that municipalitiesin the first region enjoyed greater fiscalautonomy than in the second.

It is also difficult to distinguish between taxrevenues and grants, especially if we deal withtax revenue in the place of grants. In Russia,starting from 2005, subjects of the Federation

could substitute the equalization grant by ashare in the personal income tax (and prior to2009, by shares of any federal and regionaltaxes due to a subject of the Federation).In this case, the grant amounts due to amunicipality is substituted by the taxcalculated on the basis of estimated taxreceipts according to the sharing rateestablished in the regional budget law. Ifactual tax receipts in the jurisdiction of amunicipality are higher than estimatedreceipts, the income of the municipality willnot be withdrawn into the regional budget.However, if a collected amount of a tax isbelow estimations, the municipality’s losseswill not be compensated. Also, and morerelevant to our discussion here, such incomeswill be shown as tax revenues, not grants.

Income taxes provide the greatest share oflocal government revenues in most Eurasiancountries (with the exception of Armeniawhere only taxes on land property wereassigned to local budgets). In Moldova,Russia, and Ukraine income taxes account

Source: Authors’ calculations based on IFS database. (cf. Annex 4.2)

Figure 4.3: The Structure of Local Tax Revenues (2008)

0%

20%

40%

60%

80%

100%

Taxes on income, profits, and capital gains Taxes on payroll and workforce

Taxes on property (incl. land tax) Taxes on goods and services

Other taxes

Armenia Belarus Georgia Kazakhstan Kyrgyzstan Moldova Russia Ukraine


133

for 75-87 percent of all local tax revenues, ofwhich the largest part is paid by individuals(see figure 4.3). Income taxes paid by legalentities include the small business tax andtaxes under the simplified taxation schemeand the profits tax. The latter does notoccur very frequently owing to its unevenlydistributed tax base. In Kyrgyzstan andBelarus income taxes amount to 47-48percent, but taxes on goods and servicesalso constitute a significant proportion of taxrevenues in these countries (about 35percent). In Kazakhstan, where incometaxes account for 36 percent of taxrevenues, there is also a social tax inaddition to the income taxes; actually, this isa wage tax, i.e. an income tax paid by theemployer.

Property taxes account for only 4 percent oftax revenues in Ukraine (where real-estatetax has not yet been introduced), 11 per-cent in Moldova, 14 percent in Kazakhstan,13 percent in both Russia and Belarus, 63

percent in Georgia, and 18 percent inKyrgyzstan. In Armenia, the proportion ofthese taxes in local budgets is 100 percentas they are the only taxes assigned to thelocal level in the country.

Intergovernmental transfers

Intergovernmental transfers are used inorder to reduce vertical and horizontalimbalances, to stimulate local governmentsto follow the central (or regional)government’s policies, or to ensure equalaccess to public services for all citizens.Local governments in all the countries underreview receive general purpose equalizationgrants. In Russia, half of the regions’ localgovernments, in addition to equalizationgrants, receive compensatory transfers forbalancing the budgetary system. Localgovernments in all the countries underreview also receive earmarked grants(capital grants, transfers for financingdelegated responsibilities, etc.).


Source: Authors’ calculations based for Armenia – on "Local Self-Government Reforms in Armenia (2007 and 2008)Book 3", for Georgia – on “Local Self-Government in Georgia” for other countries – on the countries’ Ministry’s ofFinance reports. (cf. Annex 4.3)

Figure 4.4: Structure of Local Transfers

0%

10%

20%

30%

40%

50%

60%

Armenia (2008)

Belarus (2005)

Georgia (2008)

Kazakhstan (2008)

Kyrgyzstan (2008)

Moldova (2009)

Russia(2008)

Ukraine (2008)

General purpose transfers Targeted transfers



Box 4.3. Main Characteristics of the Equalization Grants to Local Governments in Russia

In Russia, different formulas are used in different regions to determine equalizationgrants, however according to the Budget Code they should be based on either:

• A tax capacity index (where objective factors affecting the per capita cost ofmunicipal services are considered);

or

• A per capita equalization.

Revenue

The Tax Capacity Index (TCI) is used to estimate local government revenue. Generalformulas are used where the tax capacity for a given municipality, divided by its popu-lation, is compared to the same equation for the region.

For specific taxes, tax capacity is calculated by multiplying the specific tax rate by themunicipalities’ tax base.

Expenditure

Estimations of expenditure are usually based on the Expenditure Needs Index (ENI).Different formulas are used in different regions to calculate expenditure needs. Usuallythe formula is based on one (or more of the following):

• the number of consumers of the state services in a given municipality;

• the weighted sum of the expenditure factors (most common);

• the expenditure norms, established by the regional legislation.

Different factors can be taken into account for different expenditure items, the mostcommon are: size; proximity; level of urbanization; utility cost; population age structure;and remote area wage differences.

Transfer allocation formula

The most common transfer formulas are:

• Proportional equalization

• Full equalization

• Per-capita allocations

Usually the formulas applied to raions and cities (“gorodskoy okrugs”) are different fromthose applied to settlements (where per-capita allocations are most frequent). Inaddition, the responsibility for settlement budgets equalization is often delegated toraions.

Source: Author’s analysis of the RF regions’ legislation.


135

The low proportion of transfers in local bud-gets of Belarus (20.9 percent, see Figure4.4) is not surprising as, given the politicalregime in this highly centralized structure,the center is likely to be much less afraidthat the decentralization of revenues willlead to a loss of control over vertically alig-ned bodies of executive power. In addition,the central government has a legislativeopportunity to withdraw revenue surplusesin the form of negative transfers.

At the same time in Kazakhstan the propor-tion of grants is much higher than in Be-larus (56 percent of revenues) althoughKazakhstan also uses sharing rates of taxesand negative transfers. In Moldova, theproportion of grants in the structure of bud-gets is also high (57 percent).

In Russia, grants account for 46 percent ofconsolidated local government budgets; butthe share would be higher if the amounts ofshared taxes that replaced grants wereincluded.35

Equalization grants

The equalization grant methodology used bymost countries is based on the estimationof revenue (or tax) capacity and theestimation of expenditure needs. In somecountries (Belarus, Moldova, Kazakhstan,Georgia), grants are allocated to cover thedifference between estimated revenuesand expenditure assessments. Anotherpossible approach is to compare the indi-cator of revenues (adjusted on the cost ofexpenditures) with the target value (thisapproach was used in Georgia and is stillused in 2010 in Armenia). A more sophisti-cated approach is based on using a dimen-sionless fiscal capacity index based on percapita tax capacity of local governmentand expenditure needs index. This latterapproach is used in Russia (see Box 4.3); asimilar formula is used in Kyrgyzstan andsuggested in the draft Law on FinancialEqualization in Armenia.36

Since the countries of Eurasia shared thesame budgetary system in the past and face

35. Figures are notavailable due toreporting standards.

36. According to the draftLaw on FinancialEqualization (Armenia)part of the fund isallocated according tothe scale coefficient,another part– according to the fiscalcapacity index, fordetails see Movsisyan(2007). For moredetails on equalizationin Russia seeSlavgorodskaya et al. (2008).


Countries Formula to Determine Excess % of Excess Reassigned Funds Assigned to

Belarus Revj / Expj > 1 100% Central Budget

Kazakhstan Revj / Expj > 1 100% Central Budget

Moldova Revj / Expj >1.2 100% Equalization Fund

Russia Revj / Expj average > 2 ≤50%

(established by the law of the RF subject, (established by the law of the RF Equalization Fund

but not less than 2) subject but not more than 50%)

Ukraine Revj / Expj > 1 93% for Kiev Central Budget

95% for other donors

60% for the capitals of the oblasts

with population more than 950 thousand

Revj – revenue estimation used for grant allocation, Expj – expenditure estimation used for grant allocation

Sources: Authors’ analysis of the countries’ legislation

Table 4.6: Characteristics of Negative Transfers



similar challenges in the sphere of inter-governmental fiscal relations, it is no won-der that best practice and experience in thefield of equalization grants is shared amongthe former USSR neighbors. Most of thesecountries have also had methodologicalsupport from international organizations.

In some countries, municipal governmentspay negative grants in cases where theirrevenues exceed predetermined level (seeTable 4.6). The Russian Budget Codepermits the withdrawing of negative grantsinto the local government equalization fundif per capita tax revenues of a municipalityare in excess of the per capita doubleaverage tax revenues in the latest reportingyear. The amount of grants should notexceed 50 percent of the difference betweenthe per capita local tax revenues and thedouble average tax revenues. Though manyregions have included such provision in theirlaws, no more than 6 subjects of the RFcurrently employ it.

Belarus, Kazakhstan, Ukraine and Moldova allalso use different variants of negative grantswith regard to relatively better off localgovernments, within a stricter framework ofrules. For example, in Belarus andKazakhstan, with their centralized budgetarysystem, local governments remit to thecentral budget any sums of estimatedrevenues in excess of estimated expenditu-res. In Moldova local government revenueestimations can exceed expenditure es-timations by 20 percent; any local revenues inexcess of that figure are to be transferred tothe central budget. In Russia and Ukrainethere are limits established for the amount ofexcess revenue that can be taken, while inother countries this is not the case and allsurpluses are collected. In Russia andMoldova revenues are taken into equalizationfund and redistributed between other localgovernments, while in other countries theserevenues go directly to the central budget.

Local experts have observed that negativetransfers reduce the incentive for own taxbase development in Eurasian countries,while advocates of the negative transfersystem note that in the countries with theuneven distribution of tax base negativetransfers with moderate withdrawals canallow for a greater over all assignment oftax revenue to the local budgets.

In countries with a multi-tier system of localgovernment, the hierarchical model37 oftransfer allocation is predominant (Belarus,Kazakhstan, Moldova). However, starting from2008, transfers to settlements in Kyrgyzstanhave been allocated directly from the budget ofthe Republic, not from the budget of the raion.In Ukraine, the central government allocatesequalization grants to oblasts and raions andraions, in turn, allocate grants to municipalitiesat the sub-raion level. In Russia, the subjectsof the Federation use various models: bothregions and raions have the right to equalizebudgets of settlements and a subject of thefederation may delegate responsibility for theequalization of local settlement budgets toraions. As a result, Russian settlements canreceive equalization grants only from regionalgovernments, or only from raion governments,or from both of them (in which case the grantsfrom the raion budget will be allocated takinginto account equalization grants from theregional government).

Countries also differ in their treatment ofdifferent types of local governments.Kyrgyzstan has three separate funds: forthe equalization of cities, raions and settle-ments, respectively, while in Russia citiescan receive grants from the equalizationfund for raions and gorodskoy okrugs andfrom the equalization fund for settlements.

Other transfers

As distinct from general-purpose equalizationgrants, special-purpose grants often have no

37. When localgovernment of lowestlevel receive transfersfrom the localgovernments of theintermediate level.


137

transparent allocation methodologies. Inall Eurasian countries under observation,special-purpose grants are provided forcapital investments, for salary payments andfor specific policy stimulation.

At best, capital grants are allocated on acompetitive basis, but clear provisions con-cerning capital grant allocation proceduresare absent in budget legislation. At worst,their distribution is a result of lobbying onthe part of individual municipalities.

In Moldova, special-purpose grants areprovided to local governments of first tier(settlements) from administrative-territorialunits of second tiers (raions) to coverwages, state obligatory social insurancecontributions to the budget of the statesocial insurance institution, and medicalinsurance contributions of staff in pre-schools, primary schools, general educationschools, lyceums, clubs and libraries.

Regional governments in Russia and thecentral government in Kyrgyzstan38 allocatecategorical grants to local budgets to fundgeneral education services.39 Grant moneyis used to pay wages and some otheroperating costs in educational institutions.Grants are calculated upon a minimumstandard of educational expenditures perstudent. Some regions in Russia use similargrants to co-finance expenditures on pre-school education (kindergartens).

Russia also uses the co-financing mechanismto support current expenses in the socialsphere. The Federal center and subjects ofthe federation can provide financial support tolocal self-governments for the execution oftheir expenditure responsibilities. While doingthis, they do not necessarily stipulate a co-financing input from local governments;the federal or regional government oftensimply assumes the funding of a certainresponsibility. Examples of such grants

include additional payments to teachers forclassroom management under the NationalEducation Project, additional payments togeneral practitioners under the Health CareNational Project, subsidies from a number ofregional budgets to pay for school meals, andso on. It should be noted that in Russia thefinancial crisis has reduced the number ofearmarked grants (above all for capital pur-poses) coming from federal and regionalbudgets. Other examples of specific grantsin Russia are transfers from regional andmunicipal finance reform funds (see Box 4.4).

Grants for financing delegated responsibilities(mandates) should be viewed separately.As a rule, such transfers are calculatedbased on estimated responsibility costs. Theresponsibility for social protection regardingcertain groups of citizens or theresponsibility to fund the privileges ofvarious populations, are often delegatedas mandates. For instance, in Russia andUkraine regions often delegate suchresponsibilities to raions and cities.

All countries in question exercise strict con-trol over public expenditures funded byearmarked transfers.

Local Borrowing

Along with analyzing revenue and ex-penditures structures it is significant to discussthe general rules of sub-national borrowing.Two main types of external funding areavailable for local governments: borrowingfrom financial organizations and budget loansfrom upper level budgets. Short termborrowing to cover cash gaps should usually berepaid in the current fiscal year.

In Kazakhstan local authorities can borrow inthe form of budget loans or issue securities.The debt limits of local executive bodies for therespective financial year is set by the CentralGovernment taking into consideration the

38. There used to be agrant for salarypayment in health carebut since 2008 healthcare is financedthrough the MedicalInsurance Fund.

39. Uzbekistan uses thesame approach tofunding schooleducation frombudgets of differentlevels through special-purpose grants.




Box 4.4. Russian Regional and Municipal Finance Reform Funds: Success or Failure?

The Regional Finance Reform Fund was established in 2000 with the support of a loan issuedby the World Bank to the RF Government. The Fund’s grants were awarded to competingregions for the chance to co-finance and implement reform programs. Subsequently, havingrecognized the success of the regional finance reform program, the Ministry of Finance startedto finance the regional Fund from its own resources and founded a Municipal Finance ReformFund in addition to it.

There was a special selection procedure for regions and municipalities. To participate in theprograms, they had to comply with certain criteria of financial management and develop aprogram for reforming a regional or a municipal finance management system includingmeasures to increase expenditure effectiveness. Any RF subject and/or municipality with apopulation of over 100,000 as well as capitals of RF subjects were eligible for taking part in thefederal Program of Finance Reform. During the time of the Fund existence, 54 RF subjectswere selected for funding, though later 5 were disqualified for failure to comply with criteria.

Not all the activities that were to be carried out under the grant program contributed to theeffective performance of the authority (participation required a huge number of reports andfollowing the completion of the reform program some regions annulled normative actsadopted for the program, demonstrating the ineffectiveness of part of the activities). Still, theFund stimulated implementation of good management practices including creatingtransparent systems of intergovernmental fiscal relations and implementation ofperformance-based budgeting. Non-participating regions were able to copy successfulbehavior models of the program participants and use their normative acts as examples fordeveloping own legislation. Thus, one of the most important advantages of federalism, thepossibility to copy innovations, was realized.

Some efficient mechanisms developed under the Program of Regional Finance Reform (suchas transparent methods of intergovernmental transfer allocation, local responsibilitiesinventory list, standards for local services quality, etc) were included into the Budget Code andbecame obligatory for all subjects of the Federation.

In response to the success of the Federal Program of Municipal Finance Reform, RF subjectsstarted their own municipal finance reforms. For instance, Stavropol Krai has demonstratedgood results in the regional municipal finance reform program.

Unfortunately, during the recent financial crisis the federal government decided to cut fundsallocated under the program, and in 2009 competitors received less funding in comparison toprevious years. The federal government did not provide grant financing in 2010 but in 2011the Fund will probably be restored.


139

purposes and tasks defined by the mid-termfiscal policy (three-year period). The amountspent on debt repayment and services shouldnot exceed 10 percent of the local budgetrevenue in the corresponding financial year.

In Belarus borrowing is allowed only in theform of budget loans from higher-level budgetsand loans, implemented through the issuance

of securities. The provision of loans by banks tolocal executive and administrative bodies is notallowed. The maximum local government debtsize for the next fiscal year should not exceed30 percent of forecasted local budget revenueswithout intergovernmental transfers.

Armenian legislation allows municipalities toborrow from the budget of higher level


Box 4.5. Municipal Development Fund of Georgia

The Municipal Development Fund of Georgia (MDF) was established in 1997 in thecourse of the implementation of the World Bank project aimed at reconstructingmunicipal infrastructure.

The MDF is governed by the Supervisory Board approved by presidential decree andincludes representatives of the Government of Georgia, relevant delegates of thePresident of Georgia and local self-governments. The Supervisory Board makes finaldecisions on funding municipal investment projects.

As a result of successful MDF project implementations, the government decided to placeother capital projects under the MDF management. Thus, the activity of the MDF wasfurther diversified.

The MDF had now been institutionalized; which has resulted in its own expert opiniontaken into account due to the sound relationship created with central and regionalauthorities of Georgia and donors, as well as being able to implement projects ofdifferent categories.

The main objective of the MDF is to mobilize funds from international financialinstitutions, donor agencies, central and local governments, and to make themavailable to local governments and self-governments for investments in municipalinfrastructure and services. The MDF also assists local self-governments inmanagement optimization and introduces international practices in the field of localself-government. The MDF established an additional source of funding for projects tobe implemented under the supervision of municipalities. Its structure perfectly meetsthe requirements of local self-governments that lack access to managementexpertise and financial markets.

The activity of the MDF is a considerable step towards improving municipal infrastructureinvestment in Georgia, particularly in view of the actual output being seen in thedevelopment of city road infrastructure, improved operation in water supply systems,enhanced quality of services provided by the educational and health institutions andgeneration of employment opportunities.



governments, as well as from commercialbanks. But in practice, municipalities do nothave access to the credit market.40 Instead,in Armenia, there is a practice of inter-municipal lending, where one municipalitycan receive an interest free loan fromanother. To date, there has been no case ofdefault on these debts. While this type ofloan is generally not a large amount, thiscould be an interesting case of best practiceand a model for other countries; thoughunfortunately it may be hard to replicate.

In Georgia local governments canborrow from the Georgian MunicipalDevelopment Fund (see Box 4.5). Theseloans are limited to infrastructuredevelopment: repair of existing serviceinfrastructure and roads, replacement ofequipment necessary for water supply,etc. Borrowing levels are limited by theamount of discretionary budgetaryresources available. Local governmentfunds 20 percent of the cost of a project,40 percent is provided as a grant, and 40percent is financed by a loan at theinterest rate of 15 percent annually with10 years maturity.

In Kyrgyzstan local government can re-ceive loans under agreement of theMinistry of Economics and Finance, andunder government guarantee. Short-termloans should be repaid during thefiscal year. Middle-term and long-termborrowing can only be used fordevelopment projects. Local self-governmentbodies are prohibited from borrowing incases where the total debt serviceobligations exceed 20 percent of the localgovernment annual revenue.

In Moldova local governments (both tiers)can borrow funds for currentexpenditures from financial organizationsand from upper-levels budgets, under thecondition that these should be repaid

during the current financial year. Thetotal amount of loans should not exceed 5percent of the total revenues of theadministrative unit receiving the loan. Forcapital expenditures loans can be takenfor short-term and long-term period. Inthis case the amount of outstanding loansalready received (together with interest)as well as the amount owed on the loansto be obtained (together with interest)can not be in excess of 20 percent of totalannual revenues of the respectivebudgets.

In Russia only domestic borrowing is allowedfor local governments. The total amountof the municipal debt should not exceedthe total amount of local revenue notconsidering intergovernmental transfers.For the municipalities that have a shareof grants (including tax transfers) as partof their total revenue less than 70percent for two consecutive reportingyears the total amount of the debtshould not exceed 50 percent of thetotal amount of local revenue lessintergovernmental transfers (see Box4.1). Municipalities can also receivebudget loans from higher level budgets.Normally, budget loans should be repaidduring the same financial period they aretaken, but as a result of the financialcrisis this term has been extended to upto three years.

In Ukraine local governments can borrowfrom financial and banking organizationsto cover cash gaps for a period of no morethan 3 months and within the currentfiscal year. Local governments (cities andCrimea) can also take loans for capitalpurposes (or development budgets) inthe domestic market. The total amount ofthese loans should not exceed 10 percentof the administrative unit’s totalexpenditures. All borrowing must beapproved by the Ministry of Finance. Only

40. Tumanyan (2009)wrote on this issue“The Article 59 of law on Local Self-government andArticle 30 of the law onBudgetary System ofArmenia state thatmunicipal securities areissued in accordancewith relevant law.Clause 6, Article 57 ofthe law on Local Self-government statesalso that theprocedures of issueand allocation ofmunicipal securities areestablished by thegovernment, but thereis no law or procedureadopted… and thecommunities did notissue municipalsecurities…”


141

cities with a population over 800,000 areallowed to borrow abroad. In the casethat a regional or local government bodydefaults on a loan, they are prohibitedfrom borrowing for the next five years.The maximum amount of debt is fixedannually in the Law on Budget. The totaldebt of central, regional and localgovernments must not exceed 60 percentof the GDP of Ukraine.

In addition to the shortage of ownrevenue sources, limited borrowing rightshamper the ability of local authorities tofund capital expenditures. While thetheory behind these limitations isintended to protect local governmentsfrom harmful debt practices, when debtlimits are set up as a share of annualbudgets it is difficult if not impossible tofinance long-term investment project. Asa result, local governments find it easierto access funds for capital expenditurefrom higher level government budgetthan from the financial sector. PublicPrivate Partnerships (PPPs), analternative that has been used with somesuccess in other regions as a way offinancing investment, is not widely usedin Eurasia.

Main Positive and Negative Featuresof Local Finance in Eurasian Countries

Decentralization in the countries ofEurasia was a spontaneous process basedon the trial-and-error that led to changesin the status of administrative-territorialentities, municipality borderlines, andassignment of expenditure responsibilitiesand revenue sources. Although formallocal self-governance in some countries isnearly 20 years old, the decentralizationprocess has not been completed. Thereare however common elements that canbe found throughout the region withregard to local finance.

In some countries positive trends can betraced in the field of decentralization. InArmenia the list of local taxes andduties will be enlarged as of 2011. In2010 Belarus adopted the law On LocalGovernment and Local Self-government.

However, in other countries recentraliza-tion tendencies have been observed,specifically in recent years in Kyrgyzstan(establishing of appointed heads of localself-governments) and in Georgia (annul-ment of settlement level of government).In Russia the nomination process forregional heads has affected the regionallevel, but has yet not reached the localself-government level.41 In all three coun-tries local tax lists were cut over the lastyears.

A huge impact on the pace of localgovernment establishment has beenmade by international assistance projects.External support for decentralization andlocal government reform provided by theWorld Bank, USAID, UNDP, EBRD, theCouncil of Europe EU and other inter-national organizations included bothconsultative and financial assistance.Although the formation of local self-governmentis not possible without an informedpopulation and the support of centralgovernment (one may refer toKazakhstan, a centralized state withoutlocal self-governments but with clearexpenditure assignment between levels ofgovernments), the assistance in thedevelopment of transparent legislation,creation of democratic institutions andsupport to local initiatives have providedan invaluable contribution to the decen-tralization processes in Eurasia.

In all the countries under review expendi-ture responsibilities are assigned todifferent levels of government bylegislation but the assignment is not

41. Unification of the localself-government modelin Russia resulted incentralization ofpowers in some regionsbecause former ownresponsibilities of localgovernments becamedelegated ones andlocal tax lists were cutwhile in other regionslocal self-governmentwere first establishedonly.




always clear. Sometimes the samedefinitions occur among theresponsibilities of authorities of differentlevels (e.g., Moldova), also someresponsibilities are still assignedaccording to the old method ofinstitutional delineation (propertyownership) (e.g., in Belarus). Sometimesde facto assignment differs from thatoutlined in the legislation: localgovernments finance responsibilities thatdo not belong to the local governmentauthority (e.g., Georgia). Co-financing ofexpenditure from different levels ofgovernment is rare but still exists (e.g. inRussia, Moldova).

Unfunded mandates have been significantlycut. In many Eurasian countries unfundedmandates are forbidden by the legislation.However there remain responsibilitiesthat, while not legally consideredunfunded mandates, place significantadditional burdens on local governmentbudgets and reduce their expenditureautonomy.

The central government has tried to exer-cise control over local spending. In somecountries (e.g., Kazakhstan, Belarus,Kyrgyzstan, Moldova) the number of localstaff is established by the centralgovernment through law or by imposinglimits on local budgets for certainexpenditure line items (e.g., Russia). InRussia, according to the President’sdecree, regional governments mustevaluate the performance of municipalraions and city governments.

Fiscal autonomy of local governments isrestricted in most countries of the region.All countries under review have a closedlist of local taxes, some countries(Belarus, Kazakhstan, Kyrgyzstan) set thesharing rates of shared taxes in thebudget law. In Russia, raion governments

have a significant amount of expenditureresponsibilities assigned to them but notaxes of their own. Besides, central(federal) governments can establish taxbenefits (exemptions, etc.) for local taxes(e.g. Georgia, Russia).

Delegation of tax collection powers is nota common. Only in Armenia, Kyrgyzstanand Moldova can local governmentscollect local taxes. The experience ofthese countries has shown that taxcollections at local level increase taxcompliance.

There has been significant improvementin the design of equalization grants. Allthe countries under review have legallyadopted formula-driven methodologies forallocating equalization grants from thecentral budget. The ability of coefficientsto adequately reflect capacities or needsof local governments can be discussed butthe rules of the game are the same for allgrant recipients. However local expertsnote that negative transfers in Belarus,Kazakhstan, Moldova and Ukraine createdisincentives for the development of thelocal tax base.

The design of conditional grants has notmoved along equally well. At best, capitalgrants are allocated on a competitive ba-sis, but clear provisions with regard to ca-pital grant allocation procedures areabsent in budget legislation. At worst,their distribution is a result of lobbying onthe part of individual municipalities. Allcountries in question exercise strictcontrol over public expenditures fundedby earmarked transfers.

Borrowing powers are limited in allEurasia countries. For some countrieslimitations are set as a percentage ofannual local revenue, for others – as ashare of local expenditure. In most


143

countries (Russia, Kazakhstan etc), thecapital expenditures of local governmentsare often co-financed by higher levelgovernments.

Despite a number of unresolved problemsin local finance in Eurasian countries,significant progress has been made overthe last twenty years.

Conclusions

Expenditure responsibilities have beenassigned to different levels ofgovernment in Eurasia, however,decentralization of revenue sources is stilla great challenge. Not only have thetransitional economies faced a difficultproblem of “ideal” local taxes, but theuneven distribution of tax bases andadministrative costs associated with taxcollections result everywhere in greaterlevels of decentralization of expenditureresponsibilities than that of revenuepowers. In addition to this, transitionaleconomies are confronted with effectivelyinfluencing their national tax systems,tax base estimation, tax evasion by largetaxpayers, and uncoordinated reformpolicies.

For local self-governments to becomemore autonomous it is necessary toenlarge the list of local taxes (specifically,in Russia, Kyrgyzstan, Georgia) or toassign for long periods those sharedtaxes whose sharing rates are currentlyestablished in annual budget laws(Kazakhstan, Belarus, Moldova). Localreal estate tax should be introducedinstead of separate taxes on land andproperty. Personal income tax is a goodsource for local revenue if combined witha flat rate and payments according to theresidence principle. Small business taxesare another good instrument if localgovernment could establish the tax base

and the tax rates. Central governmentshould not set tax exemptions on localtaxes.

The administration and collection of localtaxes and fees should be carried out atthe local level, delegation of the taxcollection to the local level increases taxcompliance and resulting revenues.

The recent financial crisis has revealedweaknesses in the local finance systemsof some Eurasian countries, while inother countries local governments fairedmuch better then central ones. Duringan economic downturn, the situation ofpublic sector revenue assignment ishighlighted, and the necessity of stablesources of income for public sectorrevenue becomes crucial for localgovernments. Widely used property andland taxes are the most stable sources ofbudget revenues, especially if the taxbase is not estimated on the basis of themarket value (as is the case for Russiaand some other Eurasian countries). Oneof the main sources of local governmentrevenues in Eurasia (except for Armenia),personal income tax also remained stablein Russia (except for large industrialcities) and Kazakhstan, however Ukrainereports significant downturns in revenuecollection. Small shares of stable localtaxes in the total amount of local revenueand strong dependence on the decisionsof the higher level authorities endangerthe stability of local finance.

Another important issue is the stability ofthe rules for allocations of intergovern-mental transfers that prevent localbudgets from unexpected changes in theamount of transfers. In Moldova, forexample, transfers to local governmentswere cut by approximately 20 percentdue to the crisis (Expert-Grup, 2009),so local governments have had to




significantly cut expenditures. At thesame time in Kazakhstan the amount oftransfers, having been set for a 3-yearperiod, was not reduced, and additionalgrants were transferred to the oblastlevel. In Ukraine in 2009 “local budgetsmissed about 3.7 percent of their ex-pected grant revenues” (Slukhai and Tse-khanovskiy, 2009), however localbudgets in the Ukraine were also con-fronted by an additional problem:“permanent delays in payments to localgovernments by the State Treasury …which almost paralyzed local governmentfunctioning”.

In Russia the crisis mainly influenced federaland regional budgets, while local self-government budgets (especially those ofsettlements) were not much affected by thecrisis; only a number of large industrialcities faced budget problems due to falls inrevenue. In most of the region’s countriesequalization transfers were not cut in 2009,capital transfers were reduced in the firsthalf of the year but were restored by theend of 2009.

Commercial credit shortages should notcause significant problems for localgovernments in most Eurasian countriesbecause local borrowing powers werealready restricted by the national legisla-tion. In Russia federal governmentsprovided budget loans to regionalgovernments, and regions, in turn, canprovide loans to local governments.Budget loans were also enlarged inUkraine.

It should be noted that prior to the crisisEurasian countries have created effectivemechanisms to address pressing issues oflocal self-governance. For example, theMunicipal Development Fund of Georgiabrings together funds from varioussources to be used for municipal

infrastructure development. A similarmechanism can be recommended to othertransitional economies provided it issupported by the central governmentsand international financial organizations.

The Russian Fund for Regional andMunicipal Finance Reform is also a goodexample of best practices that allows newmechanisms of public financial manage-ment to be tested in pilot jurisdictionsand then includes successful instrumentsinto common practice. Another usefulexample is the experience of inter-municipal loans in Armenia. Differentmunicipal best practice programs inEurasian countries (e.g. in Moldova) addto the improvement of local governmentperformance.

Inter-municipal cooperation is aneffective instrument to reduce costsassociated with the provision of municipalservices taking advantage of the availableeconomies of scale, whi le at thesame time allowing individual localself-governments to retain control overfunctions. This approach can be usedeffectively in other countries, especiallythose where local self-governments existonly at the settlement level and aretherefore not able to realize economies ofscale. Inter-municipal tax inspections inArmenia could be mentioned as anexample of these types of economies ofscale. In Russia, in some regions, raionsand cities finance ambulance servicestogether but the Russian Budget Codedoes not provide a mechanism for alloca-tion transfers between budgets of thesame level of government, so some mu-nicipalities are afraid that the joint pro-vision of services will cause difficultiesdue to its co-financing.

Of all countries under consideration, Belarushas the least effective system of


145

intergovernmental fiscal relations in theregion. First, the republic should decide onfurther developing its local authorities.Belarus has recently started a set ofreforms, including the development of anew law on local self-government, theadoption of the second part of the tax code,and starting the process of developing anew equalization grants mechanism. Givenpolitical realities, one can assume thatBelarus would benefit from the experienceof other Eurasian countries where bodies oflocal power, albeit subordinate to the centralgovernment, nevertheless, have legallyassigned responsibilities and own revenuesources.

Kazakhstan does not plan to set up localself-governments in the near future. In spiteof their absence, however, the centralizedsystem of the country promotes otherreforms to enhance performance of thepublic sector (performance-basedbudgeting, PPP mechanisms develop-ment). It is possible that in the future, uponcompletion of the budgetary reform, at thenext stage of development, the centralgovernment will reconsider the issue of localself-government.

Moldova needs to make intergovernmentalfiscal relations between raions and settlementsin their territory more transparent andformula-driven. Settlements should receivemore expenditure and revenue autonomyfor the provision of effective services.

In Russia, it is necessary to assign owntaxes to raion governments (the transporttax was the most suitable candidate fordecentralization but federal government isplanning to cancel it) and introduces thereal estate tax. As in other Eurasiancountries, the Russian central governmentshould entrust tax administration to thelocal level.

Armenia needs clear legislation on financingdelegated responsibilities. Municipalitiesneed additional support in accessingfinancial markets; as while they are allowedaccess by law, in practice this has yet to beeffectively implemented.

Ukraine seems to have renounced radicalreforms. Instead of an overall reform ofterritorial organization, the nationalgovernment is currently supportingvoluntary association to improve theefficiency of local management. The focus ison certain key sectors such as themanagement of housing and theimprovement of municipal services ratherthan a full institutional reform. The newOrganic Budget Law, adopted in 2010, doesnot introduce any changes in the principlesthat govern local government, but it doesreopen the debate on the formula forcalculating equalization transfers.

It is difficult to predict the local self-govern-ment development in Kyrgyzstan, but withthe current political instability centralgovernment can hardly be interested instrengthening the local government level.The internal problems of Georgia are alsonot a good environment for building stronglocal governments.

The countries of Eurasia cannot avoid pro-blems while developing local self-governments, however, the success ofneighboring countries facing similar challen-ges and the analysis of best practices incountries with a long history of localself-government should help in findingeffective solutions. The main recom-mendations for the Eurasian countries havebeen summarized in the box which follows.




UCLG Euro-Asia Regional Policy Recommendation

• Administrative-territorial division and the status of local government should beclarified: local executive bodies should be accountable to the local representativebodies and thus to the citizens;

• Clear assignment of expenditure responsibilities should be established;

• Unfunded mandates should be eliminated not only on paper but in the everydaypractice, delegated functions must be supported with adequate financial resources;

• Local governments should have autonomy over local budgets: expenditure autonomyshould be supported by revenue autonomy with sufficient revenue sources;

• Local government responsibilities at the lowest level (that closest to citizens) shouldbe increased and matched with adequate funds;

• Transfer allocations should be transparent and should not contain negativeincentives;

• The administration and collection of local taxes should be transferred to the locallevel;

• Inter-municipal cooperation should be developed: local governments should worktogether for better services provision;

• Inter-municipal short-term loans can be recommended as a good system of mutualassistance;

• Evaluation of the quality of local services and sufficiency of funding should beprovided by external non-partisan entity;

• Access to capital markets should be provided for local governments.

Recommendations developed by the participants of the UCLG GOLDII Eurasia Workshopin Moscow (February 2010).


0w2010 05 EUROPE DEFcarta ang 9/9/11 12:02 Página 148

5. EUROPE

LLUUIIZZ DDEE MMEELLLLOO **

EECCOONNOOMMIICCSS DDEEPPAARRTTMMEENNTT,, OOEECCDD


* The opinions and analyses presented in this chapterare the author’s own and do not necessarily reflectthat of the OECD or the organisation’s Membercountries. The author is indebted to DEXIA forcompiling the data set used in this chapter,providing background information on recentdecentralization initiatives in Europe and updatingthe country-specific information reported in DEXIA(2008) on revenue-sharing coefficients and fiscalrules.


151

T his chapter deals with local governmentfinances in Europe. The set of countries

under examination includes the EuropeanUnion Member States, and to further ex-tent Iceland, Norway, and Switzerland.These countries offer a rich array ofexperiences with fiscal decentralization butface common policy challenges, which thechapter aims to highlight. The institutionalunderpinnings of fiscal decentralization arestrong in Europe. For several decades,the continent has seen a continuing,though sometimes uneven, trend towardsgreater democratic decentralization tothe local and regional levels, asevidenced by the European Charter forLocal Self-Government of 1985, whichcame into effect in 1989 and has beenratified by all European Union (EU), andalmost all Council of Europe, memberStates.

The GOLD I report provides a compre-hensive review of the history of fiscaldecentralization in Europe, reported recentcountry experiences and described thepolitical empowerment of local governmentsin the region. The Report recognisesthat, to a large extent, the institutionalarrangements currently in place in differentcountries – such as the tax bases assignedto the local authorities, the expenditurefunctions under their remit and the financialrelations among the local jurisdictionsand between them and higher levels ofgovernment – are rooted in history, cultureand local traditions. But intergovernmentalfiscal institutions are far from immutable,because the role of local governments isalso shaped by evolving social, economicand demographic forces that affect citizens’demands for goods and services providedlocally.

A number of broad trends are shaping theoutlook for local government finances inEurope. First, European local governments

have been hit hard by the global crisis thaterupted in 2009 and the subsequentrecession.1 Several EU countries have beendeveloping fiscal consolidation plans torestore longer-term fiscal sustainability,which affect local and regional governmentsthrough increased pressure for tax hikesand expenditure restraint. For example, thefiscal consolidation programs announcedduring 2009-2010 have included tax hikes,especially the value added tax (as inCroatia, Czech Republic, Estonia, Hungary,Ireland, Latvia, Lithuania, Poland, Portugal,Romania, Spain and United Kingdom), payfreezes in the public sector (as in Italy,Latvia, Romania, the United Kingdom) andmore general expenditure cuts (as in theUnited Kingdom). Intergovernmental fiscalrelations have also been affected through areduction in central government grants andtransfers to the local authorities.2 Fiscalconsolidation has also implied a curtailmentby the central government of localgovernment financial autonomy in somecountries, such as Spain.3

Second, several territorial and tax reformshave been announced or implemented inEurope over the last years with a directbearing on local governments. They includeGreece’s administrative and territorialreform (the Kallikratis Project) aiming toreduce the number of municipalities throughmergers and to replace the prefectures byregional authorities with extended powers;Latvia’s abolition of regional governments;and France’s plans to merge thedépartements and régions to reduce thenumber of territorial counsellors and to setup “community of municipalities” for themajor cities with a broader set ofcompetencies transferred to them from theregional and departmental authorities.Municipal merger programs are also inplace in Finland, Luxembourg, Swedenand Norway. In the same vein, the newgovernment in the United Kingdom has

1. According to theCouncil of EuropeanMunicipalities andRegions (CEMR) andCouncil of Europe’ssurveys, Europe’slocal governmentshave, for the mostpart, beenpessimistic aboutthe outlook for theirfinancial positionsince the eruption ofthe global crisis.

2. This is the case ofBulgaria, forexample, wherecentral governmenttransfers to themunicipalities will becut by 2015 and theVAT rate may beraised if the planedexpenditure cuts arenot enough toreduce the publicdeficit.

3. To limit the publicdeficit, the centralgovernment of Spainhas restrictedaccess to credit forlocal authorities andtheir dependentpublic agencies tofinance investmentsas of 2011.


EEUURROOPPEEUnited Cities and Local Governments152

abolished the Regional DevelopmentAgencies and the Government Offices andall the regional strategies for economicdevelopment and housing, with the centraland council governments, as well asbusiness-led partnerships, picking up manyof the functions. In Norway, competenciesand corresponding revenue have beentransferred from the national governmenttowards the county councils to reinforcedecentralization and local democracy,as well as to streamline the share ofcompetencies between the different levelsof authorities.

Finally, European local governments will needto deal with emerging challenges. Theyinclude in some countries a reduction inlocal revenue sources imposed by the centralgovernment (for example, in Greece, as partof the fiscal consolidation program in France,as a result of phasing-out of the taxeprofessionelle; and in Germany as areduction in intergovernmental transfers). Atthe same time, there is an increased pressurefor value for money from local governments –a centralising trend of the last UK governmentin particular, but also in different forms seenin Scandinavia and the Netherlands –which may also lead to more pressuretowards outsourcing or privatization. Localgovernments will also have to deal with theemergence of new claims on local budgets,including in the areas of social protection intimes of crisis and in a context of rapidlyageing populations. Demands associated withthe need to integrate immigrants into thesocial fabric of the recipient countries alsotake a toll on local governments.

The focus of the chapter will be on the locallevel of government (municipalities andcities), rather than the middle-tier jurisdic-tions (such as, states, provinces, auto-nomous regions, etc.), although theanalysis of intergovernmental fiscal rela-tions will sometimes take into account other

levels of government. Different countryexperiences in the areas of expenditurecommitments, revenue assignment,including intergovernmental revenuesharing, and financing arrangements, es-pecially local government borrowing, will beanalysed with a view to highlightingcommon policy challenges, while recogni-sing that generalizations are often difficult.This is because common trends maysometimes mask very different underlyingforces shaping the public finances at thelocal level of government.

The paper is organised as follows. Thefollowing section builds on the GOLD IReport and provides an overview of thestructure of local government finances inEurope. The third section is related to theexpenditure side of local budgets, andattention is focused on the composition oflocal government outlays by functional andeconomic classification, as well as the mainpolicy issues related to local governmentexpenditure policy. The fourth sectiondelves into the revenue sources of localgovernments, including taxes, non-taxinstruments and intergovernmental revenuesharing, highlighting the main policy issuesrelated to the assignment of revenuesources to local governments. The fifthsection deals with the design of inter-governmental grants and transfer systemsneeded to bridge the gap between locallyraised (or shared) revenue and theexpenditure mandates assigned to localgovernments. The sixth section focuses onlocal government budget outcomes andindebtedness, with special emphasis onthe fiscal rules governing local governmentfinancial management. The seventhsection emphasises the short-term policychallenges facing European local governments,including those related to the recovery fromthe global financial and economic crisis, aswell as longer-term challenges related topopulation ageing and immigration. The



Population Population Urban Pop. in the Old Net GDP GDP Structure of Average (millions, growth rate population largest city dependency migration (billions per capita government population 1st January (%, 1999- (% of total (% of urban ratio stock (net of current (thousands of (number of of 2009) 2009 pop., 2005) pop., 2005) (pop. 65 + migration as % euros in PPP current euros sub-national municipalities

average) to pop. 15 of pop., 2008) terms, 2008) in PPP layers, 2009) (2008-2009)to 64, 2008) terms, 2008)

Austria 8.4 0.5 66.0 41.6 25.4 0.4 258.2 31.0 Federal (2) 3,540

Belgium 10.8 0.5 97.2 9.9 25.8 0.5 309.1 28.9 Federal (3) 18,180

Bulgaria 7.6 -0.8 70.0 20.2 25 0.0 78.8 10.4 Unitary (1) 29,015

Cyprus 0.8 1.6 69.3 34.7 17.8 0.1 19.1 24.0 Unitary (1) 1,510

Czech Rep. 10.5 0.2 73.5 15.6 20.5 0.7 210.2 20.2 Unitary (2) 1,670

Denmark 5.5 0.4 85.6 23.5 23.6 0.5 165.4 30.1 Unitary (2) 56,040

Estonia 1.3 -0.3 69.1 43.6 25.3 .. 22.6 16.9 Unitary (1) 5,910

Finland 5.3 0.3 61.1 34.0 24.8 0.3 155.7 29.3 Unitary (1) 15,265

France 64.4 0.7 76.7 21.0 25.1 0.1 1,735.8 27.1 Unitary (3) 1,750

Germany 82.0 0.0 75.2 5.5 30 -0.1 2,380.4 29.0 Federal (3) 6,655

Greece 11.3 0.4 59.0 49.3 27.8 0.3 265.6 23.6 Unitary (2) 10,870

Hungary 10.0 -0.2 66.3 25.3 23.5 0.2 162.0 16.1 Unitary (2) 3,160

Ireland 4.5 1.8 60.5 41.2 15.9 0.4 150.8 33.9 Unitary (2) 38,975

Italy 60.0 0.5 67.6 8.5 30.4 0.8 1,526.9 25.5 Unitary (3) 7,395

Latvia 2.3 -0.6 67.8 46.1 24.9 -0.1 32.6 14.4 Unitary (1) 19,205

Lithuania 3.3 -0.5 66.6 19.1 23 -0.2 52.1 15.5 Unitary (1) 55,965

Luxembourg 0.5 1.4 82.8 21.3 20.6 1.6 33.9 69.3 Unitary (1) 4,215

Malta 0.4 0.9 95.3 5.8 19.3 0.6 7.9 19.1 Unitary (1) 6,060

Netherlands 16.5 0.5 80.2 8.8 21.8 0.2 552.3 33.6 Unitary (2) 37,280

Poland 38.1 -0.1 62.1 7.1 18.9 0.0 539.3 14.1 Unitary (3) 15,380

Portugal 10.6 0.5 57.6 45.4 23.4 0.1 202.4 19.1 Unitary (2) 34,485

Romania 21.5 -0.4 53.7 16.6 21.3 0.0 259.0 .. Unitary (2) 6,760

Slovak Rep. 5.4 0.0 56.2 14.4 16.6 0.1 97.9 18.1 Unitary (2) 1,870

Slovenia 2.0 0.3 51.0 32.4 23.3 1.4 46.1 22.8 Unitary (1) 9,630

Spain 45.8 1.4 76.7 16.8 24.1 0.6 1,173.3 25.7 Quasi-federal (3) 5,620

Sweden 9.3 0.4 84.2 22.5 26.7 0.6 278.6 30.1 Unitary (2) 31,790

United Kingdom 61.6 0.5 89.7 15.7 24.3 0.4 1,788.7 29.1 Unitary (3) 151,110

EEUU 2277 449999..77 00..44 .... 2255..22 .... 1122,,550011..77 2255..11

Iceland 0.3 1.5 92.8 39.0 17.1 0.4 9.7 30.3 Unitary (1) 4,150

Norway 4.8 0.8 77.4 22.4 22.1 0.9 228.6 47.9 Unitary (2) 11,020

Switzerland 7.7 0.8 75.2 20.5 24.1 1.2 269.9 35.3 Federal (2) 2,950

Sources: Eurostat, World Bank (World Development Indicators, 2009), Dexia and CEMR (2009a).

Table 5.1: Socio-economic and Political Indicators



eighth section summarises the main policyconsiderations raised in the paper andhighlights key policy recommendations. Thefinal section concludes.

Structure of Local Governments inEurope: an overview

Institutional Arrangements

The GOLD I report noted that institutionalenvironments differ among the Europeancountries. First and foremost, variations inthe definition of local authority dependessentially on the territorial organisation ofdifferent countries and their institutionalsettings, which are shaped by history.4

Institutional arrangements depend on whetherthe country is a federation (Austria, Belgium,Germany and Switzerland), a quasi-federation(Italy and Spain) in the sense of havingregional authorities enjoying comprehensive

legislative, financial, and executive autonomywhile not being defined de jure as a federalcountry, or a unitary State (Table 5.1). Evenamong the unitary States, there are countrieswith a long tradition of decentralisedfiscal and financial management andempowerment of local authorities, such asthe Nordic countries. The United Kingdomhas a particularly interesting arrangementthat could be described as a case ofasymmetric decentralization by combining aquasi-federal setting for Scotland, Wales andNorthern Ireland (since 1998) with a unitarystructure in England.

There is in addition considerable hetero-geneity in Europe in the number of layersof administration within each country.This makes it difficult to put forward anall-encompassing definition of localgovernment that could be applied to theentire continent.5 In some cases, there are

4. See also DEXIA(2008) and CEMR(2009a) for moreinformation oninstitutionalarrangements in theEuropean Unioncountries.

5. To deal with thisdifficulty, this chapteruses to the extentpossible theclassification used byEurostat.

Box 5.1. Middle-tier Jurisdiction in Countries with Three Levels of Subnational Government

In 2009, there were approximately 980 middle-tier jurisdictions in Europe. They aresituated between the local and regional (or federated) levels of administration in the 7European countries that have three sub-national layers of government (Belgium,France, Germany, Italy, Poland, Spain and United Kingdom).

The institutional underpinning of middle-tier jurisdictions are also changing. In Belgium, adebate was launched in 2009 on the transformation of provinces into “territorialcommunities” in Wallonia and in “urban regions” in Flanders, both with redefinedmandates. In Italy, the reform of local governments approved by the Council of Ministersin July 2009 foresees a rationalisation of the provincial map. In England, the policy ofmainstreaming "unitary authorities" should have led to the abolition of county councils;however the Cameron government decided not to proceed with this process (law ofDecember 16, 2010). In France, the law of December 16, 2010 provides for the creation ofnew regional directors, elected at the cantonal level, but who will participate in bothdepartmental and regional level councils, which implies a redefinition of the role of thesetwo levels of government. It also allows for mergers between departments and betweenregions and between a region and the departments that are part of its territory.

Source: Dexia



Box 5.2. Municipal Restructuring in Europe: Ongoing and planned reforms

Recent initiatives

In Latvia, a local authority merger programme, which had been in place for more than 10 years,came to an end in July 2009, resulting in the replacement of 524 local authorities by 118 newentities. In Finland, the PARAS project for restructuring municipalities and services was launchedin 2007. At the end of 2012, when the project is due to come to completion, Finland is expectedto have 337 local authorities instead of 416 in 2007. In Northern Ireland, a local governmentrestructuring programme adopted in 2008 will come into effect in May 2011 with the replacementof the 26 current districts by 11 new districts with a broader array of responsibilities. In England,restructuring of the remaining two-tier system in some rural regions (county and district councils)resulted in the replacement of 44 local councils by 9 new authorities in April 2009. In February2010, new mergers were agreed for Norfolk, Suffolk and Devon.

Ambitious municipal merger programmes were launched in Greece and Luxemburg. InGreece, the Kallikratis institutional reform voted in 2010 plans to reduce the number of localauthorities from 1,034 to 370, as well as to strengthen their responsibilties. In Luxemburg, aproject launched in 2008 aims to reduce the number of municipalities from 116 to 71.

In Iceland, the number of local authorities is planned to be reduced from 78 in 2009 to less than40 in the next 5 years. Finally, in France the law of December 16, 2010 renewed themechanisms for municipal fusion, allowing for the creation of "new municipalities", based on thecurrent metropolitan areas. However this process can be blocked if the majority of the electedofficials of the municipalities concerned vote against it. The "new" local authority will have theability to create "devolved administrations" whose bodies issue from the new municipalities.

Inter-municipal cooperation

In Hungary, the government is reinforcing the system of “micro-regions with multiplepurposes” in its public services reform. In other countries, inter-municipal cooperation can befostered as a preliminary step towards municipal mergers. This is the case in Luxemburg andFinland, where the municipalities that are not ready for merging are required to create“partnership areas” to implement basic services, such as health care and social services.

In countries where inter-municipal cooperation is already well developed, there is a tendencyto rationalise these ventures whilst reinforcing their democratic legitimacy. In Italy, thereform of local government approved by the Council of Ministers in July 2009 plans thesuppression of Mountain Communities. In France, the law of December 16, 2010 plans toconsolidate the inter-municipal cooperation (90% of the local governments and 95% of thepopulation are already parts of metropolitan areas), to ensure that by 2013 no localgovernment is working alone (with the exception of Paris). The members of the deliberativeassemblies of the municipal associations with own revenue will be elected by direct universalsuffrage from the same list of candidates presented for municipal elections.

Source: Dexia



two easily identifiable sub-national levelsof government, such as middle-tier juris-dictions (Box 5.1), including the comu-nidades autonomas in Spain and therégions in Belgium, and local authorities,which in turn encompasses municipalities,communes, local councils or districts, andprovinces in some countries (such asSpain). In other cases, there are countrieswith local governments which have similarlegal status, such as France (municipalities,departments and regions), but which inother countries may have specialcharacteristics, particularly in the differentlaws applicable to urban and ruralmunicipalities, as in Cyprus and Poland, orin the duality of the county-district andunitary councils in England (but also inother regions of the United Kingdom). Insome countries, cities can gain additionalmunicipal powers and those of otherintermediate levels of government (eg thecities of Germany, Poland, which are alsodistricts, and in Hungary where cities canassume the rank of department).

There are other complexities too, related tothe special status of local authorities in anumber of countries. Capital cities mayenjoy a different status among localgovernments, as in France, Hungary, Italy,Poland and the Slovak Republic. Capitalcities can also take on the role of a middle-tier jurisdiction, as in Austria, Belgium,Czech Republic and Germany (regionalstatus). In addition, other non-capitalcities or urban areas may be defined asmiddle-tier jurisdictions. Cases in pointare Bremen and Hamburg in Germany,which have a double status of local andmiddle-tier jurisdictions. This is not to beconfused with particular administrativearrangements, including those of a supra-municipal nature, such as metropolitanareas, which are not considered to be aseparate stratum of administration.Geographically isolated areas, such as

islands and mountainous regions, are oftentreated differently in some countries(Greece or Italy, for example), on accountof their special needs and economicstructure.

The boundaries of local governments haveevolved over time in line with changes in theterritorial organisation of a number ofcountries, as well as evolving demands forlocal goods and services. In some countries,where local authorities are deemed toosmall, there have been efforts towards theamalgamation of local jurisdictions, as inAustria in the 1950s, Sweden in the 1950sand 1970s, Denmark and Belgium in the1970s, and Finland and the Netherlands morerecently (Box 5.2). The number of localjurisdictions was reduced sharply in Greeceand Lithuania in the 1990s, Denmark in 2007and Latvia in 2009. Amalgamations have alsotaken place at the regional level, as inDenmark.6 Initiatives in this area have beenmotivated by the need to enhance efficiencyin service delivery through the territorialenlargement and consolidation of localjurisdictions (discussed below in the contextof local government expenditure policy). Bycontrast, in some Central and EasternEuropean countries, a number of mergers oflocal jurisdictions carried out during centralplanning have been reversed since the returnto democracy in the 1990s. The averagepopulation of local jurisdictions is well below10,000 in many European countries, oftenmaking it difficult for these local authorities toreap the benefits of economies of scale in thedelivery of local services.

A final consideration when illustrating thediversity in institutional settings amonglocal governments in Europe is relatedto the legislative and executive structureof local governments. As noted in theGOLD I report, all European governmentsare administered by elected councilors,however the methods of appointment of

6. See CEMR (2009a) formore information.


157

their executives can vary: they maybe elected by the council or directlyby the people or appointed by thegovernment; the executive can beindividual or collegiate. As a generaltrend, the decentralization of revenueand expenditure functions to the localgovernments cannot be dissociated fromstronger demands for self-governmentand the devolution of political powers tothe local jurisdictions.

Support for Local Governments

There seems to be strong support for localauthorities in Europe. A number ofinitiatives have been put in place insupport of local governments, includingfirst and foremost at the supra-nationallevel the European Charter of Local Self-Government, which came into effect in1988 under the auspices of the Congressof the Council of Europe. By guaranteeinglocal governments’ political, administrativeand financial autonomy, the Chartercontributes to strengthening localauthorities politically and institutionally.Within the treaties of the European Union,and as is clear from the Treaty of Lisbon,local and regional autonomy is part of the"fundamental political and constitutionalstructures" of the Member States whichthe EU must respect (Article 4).

“Europeanization”, as the process ofaccession to the European Union and theconsequent adherence to the EU’s acquisis sometimes known, is affecting self-government at the sub-national level inmany countries. This is especially the caseof the new member States in the Balticregion and in Central and Eastern Europe,where EU integration has been an im-portant catalyst for political and economicreforms. In these countries, it can beargued that fiscal decentralization hascontributed to fostering democratic

governance through greater political re-presentation at the sub-national level –albeit often only half-heartedly at thebeginning in some countries.

Notwithstanding an overall fairly strongpolitical commitment to self-government,local governments are confronted with theemergence of middle-tier jurisdictions inmany countries (Box 5.3). Importantexpenditure mandates were devolvedto the German Länder in 2006 and tothe Italian regions since the 1990s. TheDanish reform of 2007 replaced thecounties by fewer regional jurisdictions,while maintaining the local authorities andallocating to them most of the tax bases ofthe former counties. Regional jurisdictionshave been created in the transitioneconomies that have recently joined theEuropean Union, such as the Czech andSlovak Republics, in part as a means offacilitating access by these countries tofinancial support from the EuropeanUnion’s structural and cohesion funds,which are disbursed at the regional, ratherthan local authority, level. Politicaldecentralization in the United Kingdomalso illustrates the empowerment ofregional jurisdictions, with the devolutionof important prerogatives to Scotland,Wales and Northern Ireland since 1998.

Local Government Expenditure

The Assignment of Expenditure Mandatesto Local Governments

The allocation of spending responsibilities tolocal authorities often follows two principles.On the one hand, they may interveneon any subject of local public interestwhich does not interfere with the jurisdictionof other administrative authorities andrespects the legal framework. This principle,known as the "general competence clause"is enshrined in Article 4 of the European




Box 5.3. The Regionalisation Process

Regionalisation has been in place in a number European countries for over twenty yearsto strengthen the competencies of the existing regions, to create or to reshape theregional level of government and to launch new initiatives.

Recent initiatives are noteworthy. For example, Lithuania will have two levels of localgovernment instead of one from 1 July 2010. Five self-governing regions werecreated and the 10 State regional administrations were suppressed. In Greece, asnoted in Box 5.2, the Kallikratis reform replaces the existing 54 counties by 13regional governments elected by direct universal suffrage. In Norway, the regionalreform in place since 1 January 2010 maintains the current 19 counties (instead ofreplacing them by 5 or 6 new regions, as originally envisaged), while reinforcing theirresponsibilities in the fields of road maintenance, land settlement, environmentprotection and innovation.

Several countries are considering regional reforms. In Poland, the government hasrecently announced its intention to strengthen regional competencies and to reshapetheir organisation. In Romania, the government announced in January 2009 that itintended to replace the counties by regions before 2013 (the most recent version of thedraft law refers to 16 to 20 regions). In Slovenia, despite failure of the 2008 region-alisation project, the government is preparing a new plan based on the creation of 6regions from 2011 (instead of 14, as announced in 2008). A debate is under way on thecreation of regions in continental Portugal, and a new project refers to 3 to 5 regions,instead of 8, as planned in 1998, the year when the original regional reform law wasrejected by referendum.

Regionalisation is ongoing in Sweden, where the government announced its intention tomerge from 2011 the 24 counties into 6 to 9 regions with special competencies in the fieldsof health care and regional planning. In Finland, an independent assessment of the pilotproject implemented in the region of Kainuu was to be released in the spring of 2010. Onthe basis of this evaluation, the government will present additional measures.

Regionalisation is neither linear nor general in Europe. In Latvia, for example, the 2009territorial reform originally aimed to create self-governing regions to replace the 26districts. In the end, the responsibilities of the districts were reassigned to themunicipalities and to the five State regions created in 2002 for regional development. InFrance, the 2003-2004 reforms were favorable to the regions, however the succesivelocal finance reforms and in particular the elimination in 2010 of the tax on economicactivities –taxe professionelle–, has progressively limited their fiscal power.

Source: Dexia


159

Charter of Local Autonomy. On the otherhand, they exercise powers assigned themby law, which generally represent the bulkof local government spending. They mayalso assume other duties delegated to themby the State which they execute in its name.In most countries, local governments areresponsib le for pr imary educat ion,including the construction, operations andmaintenance of schools. In some newEuropean Union Member States, includingEstonia, Latvia, Lithuania, Slovak Republicand Slovenia, education alone accounts forat least one-third of local governmentbudgets (Figure 5.1). Other responsibilitiesthat are in general under the purview oflocal governments include the supply ofurban amenities and planning, wastecollection and treatment, and distribution ofdrinking water and treatment of wastewater. In some cases, payment of socialbenefits adds to local government spending,including in selected Western Europeancountries, such as Denmark, Germany,Iceland, and the remaining Scandinaviancountries. The provision of social services,

including child care and income support forthe elderly and the disabled, is another areawhere local authorities are engaged heavilyin service delivery.

In the case of health care, which accountsfor a large share of expenditure, the role oflocal government is less uniform acrosscountries. In some cases, local authoritiesare responsible for primary and preventivecare, in particular in the Nordic and someEastern and Central European countries.The case of Finland is interesting; healthcare accounts for a large share of localgovernment budgets because themunicipalities (through municipal boards)are responsible for managing health carecentres. By contrast, in Norway, health carewas recentralised in 2002 as a means ofboosting cost-efficiency in service deliverythrough greater reliance in activity-basedfunding for the allocation of health careresources.7

The part local government spending makesup of total public spending reflects the

7. An evaluation of thereform carried out bythe NorwegianResearch Council forthe Ministry of Healthand Care Servicesnotes an improvementin access, an increasein productivity and areduction in waitingtimes after the reform.The evaluationnevertheless did notcover important areas,such as equity ofaccess and quality ofhealth care services.See also Magnussen etal. (2007) for moreinformation on theNorwegian experience,and WHO (2008) for adiscussion of countryexperiences andcommon policychallenges in thedecentralization ofhealth care systems inEurope.


0%

20%

40%

60%

80%

100%

General public service Economic affairs Health Education Social protection Other

Austri

a

Belgium

Bulga

ria

Cypru

s

Czech

Rep

ublic

Denm

ark

Eston

ia

Finlan

d

Fran

ce

Ger

man

y

Gre

ece

Hunga

ry

Icelan

d

Ireland

Ita

ly

Latvia

Lith

uania

Luxe

mbo

urg

Malta

Nethe

rland

s

Norway

Polan

d

Portu

gal

Roman

ia

Slova

k Rep

.

Slove

nia

Spain

Sweden

Switzer

land

Unite

d Kin

gdom

Sources: Eurostat (February 2010), national sources and DEXIA calculations. (cf. Annex 5.3)

Figure 5.1: Local Government Expenditure: Functional classification, 2007



difference in assignment of expenditurefunctions to local governments (Figure5.2 and Table 5.2). Local governmentexpenditure is comparatively high inp r o p o r t i o n t o G D P a n d g e n e r a lgovernment out lays in the Nord iccountries. In other countries, wheremiddle-tier jurisdictions are importantproviders of public goods and services,the relative size of local governmentsis typically lower, as in Austria,Germany and Spain. Most importantly,cross-country variations in the relativesize of local governments reflect differencesin the assignment of expenditurefunctions across the various levels ofgovernment.

Rather than being immutable, local ex-penditure patterns evolve over time. Localgovernment spending has risen in many

countries in part because expendituremandates have changed over the years. Insome cases, increase in local spendingreflects changes in the composition ofoverall expenditure and are matched bylower spending at higher levels ofadministration. This is the case ofcommunity services, including socialprotection, for example, which were fullydecentralised to the local governments inDenmark in 2007. Non-tertiary educationwas decentralised to the localgovernments in Italy during the 1990s, atrend that almost doubled the relativesize of local governments.8 In the caseof Spain, several functions weredecentralised to the middle-tierjurisdictions following the country’scomprehensive decentralization programimplemented in the 1990s, although thereis much less clarity in the assignment of

0

10

20

30

40

50

60

70

Austri

a

Belgium

Bulgar

ia

Cypru

s

Czech

Rep

.

Denm

ark

Estonia

Finlan

d

Franc

e

Germ

any

Greec

e

Hunga

ry

Irelan

d Ita

ly

Latvi

a

Lithu

ania

Luxe

mbo

urg

Malt

a

Nethe

rland

s

Poland

Por

tuga

l

Roman

ia

Slovak

Rep

.

Sloven

ia

Spain

Sweden

United

King

dom

EU27

Icelan

d

Norway

Switzer

land

% of GDP % of General Government Expenditure

Source: Eurostat (February 2010), national sources and DEXIA calculations (cf. table 5.2).

Figure 5.2: Local Government Expenditure, 2008



Expenditure Composition of expenditure (% of local government expenditure)

% of GDP % of general Compensation Use of goods Gross fixed Social Othergovernment expenditure of employees and services capital formation benefits

Austria 7.6 15.6 23.8 18.0 6.6 18.0 33.6

Belgium 6.8 13.6 53.5 14.1 11.6 11.6 9.2

Bulgaria 7.6 20.5 35.3 36.8 26.8 0.2 0.8

Cyprus 1.9 4.4 38.1 23.0 30.6 0.0 8.2

Czech Rep. 11.4 26.6 31.8 28.2 20.4 5.6 14.1

Denmark 33.6 64.8 36.6 17.6 3.7 36.1 6.0

Estonia 11.0 27.7 41.5 26.9 19.3 3.2 9.1

Finland 20.2 41.2 48.9 30.0 8.7 8.0 4.4

France 11.3 21.5 29.0 19.7 20.6 8.3 22.4

Germany 7.2 16.6 22.9 21.9 11.9 26.6 16.7

Greece 2.7 5.5 42.4 27.5 22.7 0.1 7.4

Hungary 11.4 23.2 49.1 24.8 10.7 5.0 10.4

Ireland 7.9 18.8 21.2 17.6 45.2 10.1 5.9

Italy 15.5 31.8 30.3 25.2 11.0 18.4 15.1

Latvia 11.9 30.8 46.0 23.8 25.5 2.9 1.8

Lithuania 9.4 25.0 51.5 19.0 21.3 4.8 3.4

Luxembourg 5.1 13.6 32.4 22.6 31.2 1.2 12.7

Malta 0.6 1.3 12.8 63.9 22.9 0.0 0.3

Netherlands 15.8 34.3 38.5 28.3 15.1 10.0 8.0

Poland 14.1 32.7 40.3 23.5 18.6 8.2 9.4

Portugal 6.2 13.4 30.3 25.6 24.0 5.8 14.2

Romania 9.6 25.6 35.7 22.8 25.5 10.0 6.0

Slovak Rep. 5.5 15.7 42.3 27.9 20.4 1.1 8.3

Slovenia 9.0 20.4 41.3 20.7 23.9 4.3 9.8

Spain 6.5 15.8 28.4 28.1 16.4 1.6 25.5

Sweden 24.7 47.0 46.3 24.5 7.3 14.4 7.4

United Kingdom 13.3 28.2 37.9 35.7 8.9 11.2 6.2

EEUU2277 1111..55 2244..55 3333..66 2255..33 1133..22 1144..55 1133..44

Iceland 14.0 24.2 40.3 28.3 17.0 3.4 11.0

Norway 13.3 33.3 50.9 19.3 11.8 7.4 10.6

Switzerland 8.6 24.3

Source: Eurostat (February 2010), national sources and DEXIA calculations.

Table 5.2: Local Government Expenditure: Economic classification, 2008



expenditure functions to the localgovernments. Policing was devolved to themunicipalities in Belgium in 2002.

Rising expenditure can be attributed tosome extent to an increase in delivery costs.It is difficult to compare unit costs acrosscountries, given the enormous diversity innorms and standards applicable to localgovernment provision, and information isnot readily available for most countries. Theextent to which the local authorities canexploit economies of scale in servicedelivery contributes to reducing the cost oflocally provided goods and services(discussed below). But, in general, there islikely to have been considerable costchanges over time, reflecting the increasedimportance of old age-related outlays,especially in countries where localgovernments are important providers ofhealth care and social protection services.

Because the services delivered by localgovernments tend to be intensive in labour,payroll accounts for the lion’s share of localgovernment outlays in most countries,followed by purchases of goods and servicesneeded for government operations. Relatedto the cost of payroll at the local level ofgovernment is the labour market status oflocal employees. In most countries, localofficials are civil servants and thereforeenjoy a different employment status tothose working in the private sector. This isnevertheless not the case in Cyprus, CzechRepublic, Ireland, Italy, Latvia, Malta, theNetherlands, Poland, Slovakia, Sweden, andthe United Kingdom (CEMR, 2009c). Theshare of local government employees with aseparate employment status also variesamong the different layers of sub-nationalgovernment. In Germany, for example, 61percent of those employed by middle-tierjurisdictions, such as teachers and policeofficers, have a specific status, against only14 percent of those employed at the municipal

level. There is, nevertheless, a clear downwardtrend in the number of employees with thisspecific status in many countries.

Local governments also play an importantrole in public investment, which accounts for13 percent of local government spending onaverage in the EU countries. Gross fixedcapital formation carried out at the locallevel accounts for the lion’s share of totalpublic investment, or about two-thirds onaverage among the EU countries andas high as 70 percent in countries,such as Italy and France. The Europeanlocal governments are key players innon-residential construction, includingdevelopment, improvement, operations,and maintenance of infrastructures in areasas diverse as public transport, roads,urban amenities, and in network industries,most notably water and sanitation. Localgovernments also contribute to residentialinvestment in the area of social housing.

Main Issues in Local Governmentexpenditure

Dealing with benefit spillovers across localgovernment borders

Certain services that are provided andfinanced by a given local authority maybenefit residents from adjacent jurisdictions.There are numerous practical examples ofsuch benefit spillovers. Access to urbanfacilities, such as parks and leisureareas, is open to non-residents, eventhough they do not share the associatedprovision costs by paying taxes to thesupplying authority. By the same token,individuals are often allowed to seek healthcare in jurisdictions other than the one wherethey live, even when inter-jurisdictionalrefunding mechanisms are absent. Theproblem associated with these benefitspillovers is that they discourage the localauthorities from supplying the service in

8. See Bibbee (2008) formore information.



Box 5.4. The Metropolisation Process in Europe

Several countries have aimed to strengthen the institutional framework forinter-municipal cooperation. Recent initiatives include:

• Following the mid-term evaluation of Finland’s PARAS project (report released inNovember 2009), the government announced legislation to promote cooperation inthe 7 largest urban regions, including the Helsinki metropolitan area, in the fields ofland use, transportation and housing.

• France’s territorial reform law, passed 16 December 2010, creates two new types ofinter-municipal cooperation for metropolitan areas: the “Metropolis” a supra-metropolitan area for agglomerations of more than 450,000 inhabitants, which is thecase of 8 French cities, which will benefit from the transfer of powers from thedepartments and the region, and for those agglomerations of more than 300,000inhabitants, including a member city of more than 150,000 inhabitants. With the Lawof Greater Paris, passed on June 3 2010, the State takes back the management ofdevelopment planning for the Paris region with the creation of two public entities, thestrengthening of the role of the regional prefecture and through developmentcontracts with the municipalities affected by the new regional express metro.

• In Italy, Article 23 of Law 42 (May 5, 2009) relaunched the creation of "metropolitancities” (Città metropolitana) provided for by law since 1990 and included in the 2001Constitution, which gives an additional impulse, and facilitates the creation of thesein nine major cities. These metropolitan cities will immediately assume the functionsof territorial planning and infrastructure development, the organization of systemcoordination, the management of public services, and support to economic and socialdevelopment. Article 24 creates a more detailed interim status for immediateapplication in Rome. The implementation of this reform would lead to thedisappearance of the corresponding provinces.

• Legislation was adopted in Luxembourg in November 2008 to create urbancommunities. It defines the basic rules for the creation and functioning of thesecommunities, which are public entities for inter-municipal cooperation among cities ofmore than 20,000 inhabitants.

• Among the reform proposals made by Poland’s Joint Commission of the Central andLocal Governments established in May 2008, there is a proposal for adoption of aspecial status for the 12 metropolitan cities.

Source: Dexia



question when they have to bear thecorresponding delivery costs in full.

Likewise, the costs of local services, inaddition to their benefit, can be exported toneighbouring jurisdictions. Social protectionis a case in point. What happens in practiceis that local governments may be allowed toset benefit rates while relying on fundingfrom other levels of government. In thiscase, they face the incentive to set a rateabove those of neighbouring jurisdictionswhile exporting the cost of provision

elsewhere. There are numerous examplesin Europe of countries where localgovernments have prerogatives overwelfare programs, such as Denmark,Netherlands, Sweden, and Switzerland.

Several European countries have devisedingenious ways to mitigate the perverseincentives associated with such benefit andcost spillovers. For example, the localgovernments concerned may agree to shareat least part of the provision costs inrecognition of mutual benefits. In the case of

9. Large urban areas canalso straddleinternational borders.This is the case of theEurometropole Lille-Kortrijk-Tournai, whichincludes cities on bothsides of the France-Belgium border.International initiativesto improveaccessibility, qualityand efficiency of cross-border health careinclude the EUREGIORhein-Waal created byGerman and Dutchlocal authorities.

Box 5.5. Inter-municipal Initiatives: A few examples

Europe has well developed structures in support of inter-jurisdictional ventures at thelocal level.*

In most cases, inter-municipal initiatives are voluntary. In other cases, arrangements areinstitutionalised, and selected responsibilities are legally binding, as in the case of healthcare in Finland. Inter-municipal structures are sometimes imposed upon smallermunicipalities, as in Italy and Germany, whereas in France they may be set up by amajority vote among the participating jurisdictions and have significant tax powers. Theyare the only metropolitan administrations with fiscal powers in Europe. Inter-municipalarrangements are well developed in Austria, Belgium, France, Gemany, and SouthernEurope, but they are at their infancy in the new member countries of European Union inCentral and Eastern Europe. The case of inter-municipal arrangements for water supply inCzech Republic and hospital management in Estonia are nevertheless instructive.

Inter-municipal ventures can be institutionalised as a legal entity that has managerialand financial responsibility for the shared provision of a service. They can be of a publicnature, as in the case of syndicates of different local governments, or semi-private.Semi-public arrangements are common in nearly all European countries and moreprevalent in the areas of water/sewerage, waste collection, energy and public transport.Such arrangements are often multi-purpose and cover areas related to network servicesthat cut across municipal borders, such as the management of water and transportnetworks, which are subject to inter-jurisdictional externality effects.

Public-private partnerships, which are not equivalent to the inter-municipal cooperationarrangements referred to above, are gaining ground, following the pioneeringexperience of the United Kingdom, especially in Germany, France and Southern Europe.

* See Hulst and Montfort (2007) for more information on the experiences of Belgium, Finland, France, Italy,Germany, Netherlands, United Kingdom and Spain.


165

health care and urban facilities in general,arrangements have been put in place in largemetropolitan areas, where there is plenty ofroom for inter-jurisdictional spillovers, as inFrance, Spain, and Portugal (Box 5.4).9 Cross-border compensation for the costs of healthcare is available in Denmark and Swedenwhen patients chose to seek care in adifferent jurisdiction, a practice that can beconsidered exemplary in this area. In the caseof welfare programs, benefit rates are set andservice delivery is carried at different levels ofadministration.

Lowering delivery costs through economiesof scale

Another important policy challenge for localgovernments in Europe is to make the mostof economies of scale in service delivery.Local governments are often too small toreap the benefits of economies of scale inthe provision of local services, which drivesup unit delivery costs. Operating costs arelikely to rise among smaller jurisdictions inthe case of education, becauseteacher/student ratios are higher than incommunities with larger schools. Excesscapacity, on the other hand, may also pushup operating costs, as in the case of healthcare facilities in some of the countries ofCentral and Eastern Europe, whose ratios ofhospital beds to population tend to behigher than in Western and SouthernEurope.

There are options for lowering delivery coststhrough gains in economies of scale. Localjurisdictions that are considered too smallcan be amalgamated (forcibly or not), as inthe case of Belgium, Finland, Netherlands,and Norway, for instance.10 Politicalresistance to such amalgamations by localcivil servants and residents alike shouldnevertheless not be underestimated.Central government grants can beintroduced to encourage amalgamations,

although the experience of Estonia suggeststhat these financial incentives may not besufficiently strong.

Another option is to encourage localauthorities to deliver services jointly,as in the experience of inter-municipalcooperative ventures in many Europeancountries (Box 5.5), where localgovernments join forces to provide commonservices.11 The cases of transport, urbanwaste management, water supply, firefighting and hospital administration areillustrative of efforts in this area. Jointownership of power plants in Norway alsoallows neighbouring jurisdictions to cutcosts in providing energy services. Insome cases, however, regulations ongovernment procurement pose obstacles tocost-effective inter-municipal cooperation,as argued by local governments in Swedenand other Nordic countries. An alternativestrategy to tackle excess supply is throughthe closure of underutilized facilities. This isthe case of schools in small communities inPortugal, for example. Of course, measuresshould be put in place to ensure access toservices in the case of residents that may beaffected adversely by the closure of localfacilities.

Solving coordination problems acrossdifferent levels of government

The assignment of expenditure functionsacross levels of government is not alwayswel l de l ineated. Arrangements areparticularly complex in federations, giventhe institutional autonomy enjoyed by thedifferent layers of government. In Germany,for example, the federal government setsnorms and regulations for the provision ofhealth care and finances the operating costsof hospitals. The regional governments(länder) finance hospital investment andmanage capacity, and services are deliveredby the local governments.12 Yet even in

10. For example, inIreland, according tothe Local GovernmentEfficiency Group, anumber of localauthorities should bemerged in order totrim the number ofcivil servants andstreamlineexpenditure.

11. Empirical evidence, atleast for schoolcleaning services inDenmark, shows thatthe ability to exploiteconomies of scale isthe main factorexplaining costdifferentials betweenpublic (centralized anddecentralized) andprivate provision. SeeChristoffersen et al.(2007). The evidencefor other sectors andother Scandinaviancountries reported byBlom-Hansen (2003)is nevertheless lessconclusive, althoughthe author finds a costadvantage for privateproviders in the caseof municipal roadmaintenance.

12. See Wurzel (1999) formore information.




unitary countries, assignments oftenoverlap in the areas of regulation,management and service delivery, becauseit is often increasingly difficult to unbundlethe different functional areas ofexpenditure. In France, for example, wherecurrent arrangements work rather well,the central government sets schoolcurricula and administers personnel, theregional governments are responsible forthe construction and maintenance ofupper-secondary schools and for vocationaltraining, the départements are incharge of construction and maintenanceo f lower-secondary schools, and themunicipalities deal with the constructionand maintenance of primary educationschools. A lack of clarity in the assignmentof expenditure in increasingly complexfunctional areas may also undermine theaccountability of local governments to theircitizenry.

Institutional arrangements are oftencomplex and create room for cost shiftingamong the different layers of government.A case in point is when services aredelivered and norms and standards are setby different levels of government. Forexample, education is provided essentiallyat the local level in most countries, at leastup to the secondary level, subject tocurricula set by middle-tier jurisdictionsand the central government. Of course,some degree of uniformity is required inservice delivery both within localjurisdictions and across the nationalterritory, but adherence to norms andstandards set by higher levels ofgovernment and the European Union mayplace an undue financial burden on localbudgets. In that respect, by exercising theregulatory powers attributed to them andunderstating the attendant financialburden, higher levels of government mayend up off-loading expenditureentitlements or shifting provision costs to

the local authorities, which need to befinanced locally.

Coordination failures between differentlevels of government often result inunfunded expenditure mandates at thelocal level. In Austria, child careentitlements introduced by higher levels ofgovernment have been deemed unfundedby the local authorities. In some countries,however, the local authorities are fullycompensated for initiatives taken at higherlevels of government that would createclaims on local budgets. This is the case ofDenmark, where a commitment to fullcompensation has mitigated perverseincentives for cost-shifting and theoffloading of expenditure commitments byhigher levels of administration on localgovernments. A best practice in this areais therefore to ban the creation ofunfunded mandates across the differentlevels of administration.

Local Revenue Assignments

The Mix of Revenue Sources AcrossCountries

Because the assignment of expenditurefunctions varies across countries, so doesthe mix of instruments available to localgovernments for financing provision. Ingeneral, this mix of financing instrumentsreflects decisions on the breadth and depthof the taxes to be assigned to the localauthorities. It also depends on thecorresponding transfers needed to bridgethe gap between revenue from own andshared sources, on the one hand, and localexpenditure commitments, on the other,which are based on the allocation offunctions across levels of government.

Local government revenue – including localtaxes, non-tax instruments (such as usercharges and fees for services), and


167

intergovernmental grants and transfers –accounts for close to 10 percent of GDP onaverage in the countries under examination.Nevertheless, there are differences acrosscountries (Figure 5.3). Local governmentrevenue, is much higher on average as ashare of general government receipts – aconventional gauge of the relative size oflocal governments – in some unitary States,such as the Nordic countries (except forNorway), than in the federal countries(Germany, for example). The relative sharesof local government revenue are particularlylow in some Southern European countries,such as Cyprus, Greece, and Malta,reflecting, as discussed below, differences inthe assignment of revenue sources acrossthe different layers of government amongthe European countries. Of course, thesequantitative indicators need to beinterpreted with caution, because they donot reflect the degree of autonomy enjoyedby local governments in tax policy. While

taxes and grants account for the bulk oflocal government revenue, other sources,such as user charges and fees, areadditional important sources of localgovernment revenue (Figure 5.4).

Local Taxes and Shared Revenue

The assignment of tax bases to localgovernments

A case can be made for relying on localtaxes, rather than on revenue mobilisedelsewhere, to finance the provision of localservices. It can be argued that such reliancetightens the link between the benefits andcosts of local services, which is likely tomake local officials more accountable totheir taxpayers. Local governments are bestequipped to tax immobile bases, becausetheir revenue yield tends to be fairlyinsensitive to cyclical fluctuations ineconomic activity. This facilitates fiscal


% of GDP % of General Government Revenue

Austri

a

Belgium

Bulgar

ia

Cypru

s

Czech

Rep

.

Denm

ark

Estonia

Finlan

d

Franc

e

Germ

any

Greec

e

Hunga

ry

Irelan

d Ita

ly

Latvi

a

Lithu

ania

Luxe

mbo

urg

Malt

a

Nethe

rland

s

Poland

Portu

gal

Roman

ia

Slovak

Rep

.

Sloven

ia

Spain

Sweden

United

King

dom

EU27

Icelan

d

Norway

Switzer

land

0

10

20

30

40

50

60

70


Figure 5.3: Local Government Revenue, 2008



0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Austria

Belgium

Bulga

ria

Cypru

s

Czech

Rep

.

Denmar

k

Eston

ia

Finlan

d

Fran

ce

Ger

man

y

Gre

ece

Hunga

ry

Ireland

Italy

Latvia

Lithua

nia

Luxe

mbo

urg

Malta

Nethe

rland

s

Polan

d

Portuga

l

Roman

ia

Slova

k Rep

.

Slove

nia

Spain

Sweden

United Kin

gdom

EU27

Icelan

d

Norway

Taxes Social contributions Grants Other


Figure 5.4: Local Government Revenue Composition, 2008

0

10

20

30

40

50

60

70

80

Income, Profits and Capital Gains Tax Property Tax Goods and Services Tax Other Taxes

Austria

Belgium

Bulgaria

Cyp

rus

Cze

ch R

ep.

Den

mark

Estonia

Finlan

d

Fran

ce

German

y

Greec

e

Hun

gary

Ireland

Italy

Latvia

Lithua

nia

Luxe

mbo

urg

Malta

Nethe

rland

s

Poland

Portu

gal

Rom

ania

Slov

ak R

ep.

Slov

enia

Spain

Swed

en

United-King

dom

EU 27

Icelan

d

Norway

Switz

erland

(200

7)

Sources: Eurostat and DEXIA calculations. (cf. Annex 5.5)

Figure 5.5: Local Government Tax Revenue as a Percentage of Local Government Revenue, 2008


169

13. In Finland, a proposalhas been maderecently to reduce thetax burden on labourincome, a measure tobe compensated bygreater reliance onvalue added taxation.Likewise, in Romania,the law on local publicfinances was amendedin June 2010 to,among other things,reduce the local taxburden on personalincome.

14. An example is thephasing-out of thetaxe professionelle inFrance from 2010. Tocompensate at least inpart for the associatedrevenue loss, the realestate tax will betransferred from thedépartements to the“communities of cities”and othermunicipalities. The realestate tax may wellhave to be raised,even if the centralgovernment iscommitted tomaintaining municipalrevenue throughadditional transfers. InGermany, the localbusiness tax is themain source ofrevenue for localgovernments and theonly tax rate the localauthorities have theautonomy to set. Areduction of localrevenue autonomy istherefore posingchallenges for Germanlocal governments dueto increasing pressurefor spending on socialbenefits, especiallywithin the context ofthe economic crisis,which are partlyfinanced at the locallevel.

15. Empirical evidence ontax competition amongEuropean localgovernments includesHeyndels andVuchelen (1998) andRichard, Tolkens andVerdonck (2002) forthe property tax inBelgium, Buettner(2001) for thebusiness tax inGermany, and Solé-Ollé (2003) for a fewmunicipal taxes inSpain, among others.


management over the business cycle, giventhat local governments are oftenconstrained from borrowing to financebudget deficits of a recurrent nature. As aresult, property tax revenue accounts for alarge share of local government revenue inseveral European countries, includingBelgium, France and Iceland (Figure 5.5).

Other tax bases are also important for localgovernment taxation in some countries.This is the case of mobile taxes, such as thepersonal and corporate income taxes (PITand CIT, respectively), in the Nordiccountries,13 Iceland, and most newEuropean Union Member States, such asEstonia, Latvia, Lithuania, and SlovakRepublic. Whereas PIT is a local tax in theNordic countries and Belgium, localgovernments can levy a surcharge on thecentral government PIT in Italy. In mostother countries, PIT and CIT revenue isshared among different layers ofadministration, rather than being assignedexclusively to local governments. Thefederal and quasi-federal countries (Austria,Belgium, Germany, Italy and Spain) tend torely on formula-based, fixed-coefficientarrangements for sharing PIT revenueamong the different levels of government(Table 5.3). A large proportion of PITrevenue is also shared with the localauthorities in the Central and EasternEuropean countries.

PIT and CIT revenue-sharing parametersare cast in law in most countries, but theyare not immutable. The criteriaconventionally used for sharing theserevenues include derivation (when revenueis allocated to the jurisdiction where the taxis collected), the size of the recipientjurisdiction (based on population, forexample) and estimates of expenditureneeds, when equalisation efforts arepursued (discussed below). The overallpercentage of revenue to be shared across

layers of administration may also changeover time. For example, the Portuguesereform of 2007 allows the municipalities toreceive a small share (2-5 percent) of PITreceipts collected by the centralgovernment. Revenue from other taxes,such as the corporate income tax (CIT), isshared with the local authorities in Denmarkand Finland.

In some cases, taxes on goods and servicesare additional important sources of localgovernment revenue. This is the case inAustria, Italy and Spain. As with PIT and CITreceipts, revenue from the value-added tax(VAT) is shared among different levels ofgovernment in several countries, includingGermany and Spain. With regards to levieson business transactions, several countrieshave phased them out, or are in the processof doing so, including Belgium, Germany,Spain and, more recently, France, andHungary.14

Main issues in local taxation and revenue sharing

An important consideration in a decentralisedpolicy setting is the extent of autonomy thatlocal governments should have in taxmatters. It can be argued that autonomy isbeneficial, because it may enhance taxcompetition among the local governments,which may help to constrain increasesin public spending and taxation. Whensub-national expenditure is financedpredominantly with resources mobilisedlocally, policymakers face an incentive toevaluate the benefits of an increase inspending against the costs of incrementaltaxation. It should nevertheless be noted thatautonomy in income and property taxation,where applicable, does not often lead tosignificant variation in tax rates among localgovernments in most European countries.15

In some cases this may be due to the fact thatlocal tax autonomy is curtailed by higher



PIT/wage taxes CIT VAT

Central Federated Intermediary Municipalities Central Federated Intermediary Municipalities Central Federated Intermediary Municipalities state states/regions governments government states/regions governments government states/regions governments

Austria 73.2 15.2 - 11.6 73.2 15.2 - 11.6 73.2 15.2 - 11.6

Belgium Variable Complex 0 0 100 0 0 0 Variable Complex 0 0

formula2 formula3

Czech Rep. 70.0 9.0 - 21.0 70.0 9.0 - 21.0 70.0 9.0 - 21.0

Denmark 4 86.6 0 - 13.4 100 0 - 0

Estonia 88.6 - - 11.4 100 - - - 100 - - 0

Finland 4 78.0 - - 22.0 100 - - 0

Germany 42.5 42.5 0 15.0 50.0 50.0 0 0 53.1 44.8 0 2.1

Hungary 60.0 1.0 - 39.0 100 0 - 0 100 0 - 0

Italy 92.5 05 1.0 6.5 100 05 0 0 100 0 0 0

Latvia 17 0 - 83 100 0 - 0 100 0 - 0

Lithuania 42.5 - - 57.56 100 - - 0 100 - - 0

Poland 48.8 1.6 10.3 39.3 76.0 15.9 1.4 6.7 100 0 0 0

Romania 187 13.0 - 47.0 100 0 - 0 72.5 1.5 - 26.0

Slovak Rep. 6.2 23.5 - 70.3 100 0 - 0 100 0 - 0

Slovenia 65.0 - - 35.0 100 - - 0 100 - - 0

Spain8 64.3 33.0 1.0 1.7 100 0 0 0 62.1 35.0 1.1 1.2

PIT= Personal Income Tax; CIT = Corporate Income Tax, VAT = Value Added Tax

1. Surtaxes are not considered as shared taxes but as own-source taxes if local governments have some fiscal discretion, as the municipal PIT surtax inBelgium and Italy and the municipal CIT surtax in Portugal.

2. For the Belgium regions, the fraction of PIT receipts is based on the consumer price index and real GDP growth. The regions can also apply a surtax.For the communities, the fraction of PIT receipts is based on several economic and demographic criteria.

3. VAT is shared only with the Belgium communities. The sharing coefficient depends on fluctuations in the consumer price index and the under-18population headcount.

4. In Denmark, Finland and Sweden, there is a local PIT in addition to the national PIT.5. In Italy, PIT and CIT receipts are shared between the central government and the five regions with special status (according to variable percentages).6. In Lithuania, this percentage is applied to a lump-sum representing 70% of total PIT receipts. The precise percentage of PIT transferred to the

municipal budgets varies from 40% for Vilnius to 100% for small municipalities after deducting payments to the Compulsory Health Insurance Fundand the State budget.

7. In Romania, the central government percentage is 18%, because 22% of PIT receipts are redistributed for equalisation.8. In Spain, for the regions, the PIT and VAT shares differ according to the autonomous communities. The percentages reported apply to most regions. For

some other regions, the percentages are higher (Catalonia for example). The two foral regions (Basque Country and Navarre) have a special tax regimeand raise their own PIT and CIT locally for which they set rates and bases. In return, they reimburse the central government for services provided. Forthe municipalities, only provincial capitals and cities with over 75 000 residents are affected by the PIT and VAT sharing mechanisms.

Source: DEXIA.

Table 5.3: Revenue Sharing Coefficients, 2009 (% of tax receipts)1


171

levels of government through a reduction ingrants and transfers when sub-nationalgovernments cut their own tax rates. Anotherconsideration is that tax competition can bepredatory, as when local governments arefree to set tax bases, rather than rates. In thiscase, they may try to attract businesses andhouseholds by granting tax breaks for en-terprise relocation, which results in an erosionof some tax bases.

Tensions often arise between the localauthorities and higher levels of governmentin local tax policy. The central governmentmay wish to prevent an increase in the localtax take or to alleviate the tax burden onbusinesses as a means of fosteringentrepreneurship. This is the case of thescrapping of business taxes in manycountries, as discussed above. In the UnitedKingdom, there is a cap on annual increasesin the council tax levied by the localgovernments.16 While initiatives in thisarea are in line with central governmentobjectives, they impinge on local budgets byrestricting the scope for the local authoritiesto mobilise revenue from their own sources.

The local property tax base is sometimesunderexploited. This is because ratesmay be set too low and/or the localauthorities may be unable or unwilling tokeep up-todate registers of property values.Of course, as with other taxes, it is difficultto estimate empirically the extent ofunderutilisation of the local property taxbase. To do so, actual collections wouldneed to be compared to estimates ofthe potential tax base, a task that ismethodologically complex.

Formula-based revenue-sharing arrange-ments have the disadvantage of linking localgovernment receipts to central governmentrevenue. This is the case of VAT revenuesharing in Germany and Spain, for example,which makes local budgets sensitive to

cyclical fluctuations in central or middle-tiergovernment revenue. If local governmentshave limited financial room for smoothingtransitory budgetary slippages associatedwith the business cycle, due to a ban onborrowing, for example, revenue sharingmay also make local finances pro-cyclical:spending rises with economic upswings andcontracts in downturns. Another consi-deration is that revenue sharing oftencreates incentives for higher levels ofgovernment to focus their revenue raisingefforts on non-sharable receipts so as not toexport part of the benefits of their own taxeffort to other jurisdictions.

Bridging the Gap Between LocalRevenue and Expenditure Mandates

The Use of Intergovernmental Grants andTransfers

The allocation of revenue sources tolocal governments and the financialarrangements needed to bridge the gapbetween locally raised revenue and ex-penditure commitments varies considerablyacross countries. It is often difficult to drawa precise line between the various types ofrevenue accruing to local governments,especially in so far as own, shared ortransferred revenue are concerned (Annex5.1). For example, VAT revenue isaccounted for as own revenue in Germany,because it refers to an intergovernmentalrevenue-sharing arrangement, whereas it istreated as an intergovernmental transfer inItaly. The distinction between grants andtransfers is also blurred in many cases, andrevenue sharing is treated as grants incountries such as Greece, Portugal, andSpain. Typologies of revenue structures aretherefore often fraught with problems,because quantitative indicators invariablyfail to capture important qualitativedistinctions in the assignment of revenuesources across the different layers of

16. At the same time,efforts to preventexcessive tax rises atthe local level haveincluded the need tocall a referendum inthose local authoritiesthat raise their Counciltax above a ceilingapproved byParliament each year.




government.

Notwithstanding these methodologicaldifficulties, grants and transfers account forthe bulk of local government receipts inmany countries. This is the case in mostWestern European countries, as well as afew Eastern European countries, such asRomania and Bulgaria. These countries arecharacterised by large vertical imbalances inintergovernmental fiscal relations, as localprovision is financed by revenue collected atleast in part in jurisdictions other than thatwhere the services are provided. The UnitedKingdom stands out among the morepopulous European countries on the basis ofits local governments’ reliance on grantsand transfers from the centre. By contrast,local governments tend to rely more heavilyon locally raised revenue in the Nordiccountries (except Denmark), Iceland and a

few transition economies of Central andEastern Europe, such as Estonia and SlovakRepublic.

Local governments in member states alsobenefit from transfers from the EuropeanUnion. Structural and cohesion grants focuspredominantly on financing for investmentprojects. They are disbursed in general tomiddle-tier jurisdictions and regions, andare often shared with the local authoritiesresponsible for project implementation.

Equalisation Grants and Transfers

Schemes are in place in many countries toequalise revenue capacity among the localgovernments. This is to ensure thatexpenditure needs are met across thecountry in spite of differences in revenuecapacity among the local jurisdictions

Sources: Eurostat (February 2010) and DEXIA calculations. (cf. Annex 5.6)

Figure 5.6: Local Government Grants as a Percentage of Local Government Revenue, 2008


173

arising from the fact that economic activity,and therefore tax bases, is distributedunevenly. In addition to using the taxsystem, equalisation is carried outpredominantly through intergovernmentalgrants and transfers, which may be vertical,when based on receipts from higher levelsof government, or horizontal, whenresources are shared among same-leveljurisdictions. For example, Germany has acomplex vertical transfer system combiningarrangements between the centralgovernment and the states, and betweenthe states and the municipalities located intheir jurisdictions, as well as horizontaltransfers among the states. Horizontaltransfer arrangements are also in place inAustria, Denmark, Germany, Latvia,Lithuania, Poland, Sweden andSwitzerland. Cost equalisation was intro-duced in Sweden in addition to revenueequalisation in 2004.

A problem that arises in the design ofequalisation grants and transfers is thatthey are often calculated on the basis ofactual revenue and spending, rather thane s t i m a t e d r e ve n u e c a p a c i t y a n dexpenditure needs. As such, they rely onhistorical budgeting, rather than on anassessment of potential collections andnormative costs per unit of servicedelivered. If service delivery is inefficient,a grant system based on actual expen-diture, rather than on estimated expendi-ture needs, would perpetuate deficiencies,rather than equalising spending capacityon the basis of common standards andnorms. If local tax bases are under-exploited, the use of actual, instead ofpotential, collections would overestimatethe need for equalisation of revenuecapacity without creating incentives forlocal effort in raising revenue. Toovercome this problem, equalisation isbased on potential average, rather thanactual, revenue in the Nordic countries.

Another option for dealing with perverseincentives for local tax effort is tocentralise collections. This is the case issome countries, even where localgovernments are allowed to piggyback oncentral government taxes.

An interesting policy question is how muchequalisation of revenue capacity should bepursued. It can be argued that fullequalisation would discourage recipientsfrom raising local revenue, if donorsadjusted the level of transfers in tandemwith the increase in local revenue from ownsources (the effective withdrawal rate).Partial equalisation could therefore bepursued to mitigate the tax effortdisincentive arising from full equalisation. Infact, the survey conducted by Bergvall et al.(2006) shows that full equalisation is rare inEurope.

Earmarking of Grants and Transfers

Grants to local governments are earmarkedin many countries. Based on a recent surveyof intergovernmental transfer arrangementsin the OECD area, it appears that all ornearly all grants are earmarked in CzechRepublic, Greece, Ireland and theNetherlands, whereas local governmentsin Finland, France and Portugal tend torely predominantly on general-purpose(non-earmarked) grants.17 Earmarkedgrants are also often of the matching type,when they are designed to complementlocal financing by requiring the recipientjurisdiction to match at least a share of thegrant using local sources, as in the case ofthe Netherlands and Switzerland. The useof grants also varies across countries.Although most earmarked transfers tolocal governments are used to finance theprovision of general public services, theseresources are also often allocated to morespecific functional areas, such asenvironment protection (Italy), recreation

17. See Bergvall et al.(2006) for moreinformation.




and culture (France and Italy), economicaffairs (Spain), housing (France), socialprotection (Poland), education (Belgium,Czech Republic, Finland and Sweden), andhealth care (Sweden), for example.

Earmarking is particularly common in thecase of investment programs. This is despitethe fact that most transfers to localgovernments tend to be of a recurrentnature (Figure 5.6). Only in a few Westernand Southern European countries, such asCyprus, Greece, Ireland and Portugal, andnew European Union Member States,including Hungary and Lithuania, do capitalgrants account for a larger share of receipts.The cases of Ireland and Portugal arenoteworthy, given the role played by localgovernments in infrastructure development,an area where they act on behalf of higherlevels of government.

Main Issues in the Design of Grant Systems

Dealing with benefit and cost spillovers

The design of intergovernmental grant andtransfer systems is not without practicalimplications for local government finances.The main consideration is that thesetransfers drive a wedge between the costsand benefits of local provision. While thebenefits can be internalised by localresidents, the corresponding provisioncosts can be “exported” to taxpayersresiding elsewhere. Because of theseexternalities, local governments may facea number of perverse incentives, includingto spend beyond their means and to under-utilise their own tax bases. Alternatively,as noted above, the benefits, rather thancosts, of service delivery can be exportedto other jurisdictions, an externality thatwould in general discourage local pro-vision. Of course, the experience ofdifferent countries shows that there areoptions for mitigating these perverse

incentives through practical solutionstailored to different country circumstancesand institutional settings.

Conditionality can be introduced in inter-governmental transfer systems to deal withexternalities in local government provision.Earmarked or matching grants can be usedto ensure that at least part of the costs andbenefits of provision can be fullyinternalised by local residents. As a result,the service provider is compensated byhaving the share of delivery costs thatexceeds the benefits of provision that canbe internalised by local residents covered bythe donor. Of course, in practice, the designof such grants is complicated by the factthat externalities are not directlyobservable. Matching grants may thereforeexceed the level required for mitigating thedisincentives for provision arising fromcross-border spillovers. These grants mayalso be complex to administer.

There are a number of successful countryexperiences in the use of grants andtransfer schemes to internalise inter-jurisdictional benefit and cost spillovers. TheDutch reform of 2004 is instructive in thisregard. The pre-reform grant system hadencouraged the municipalities to overspendin social protection, because they could setthe level of benefits and the centralgovernment would reimburse them forthe attendant expenditure needs in theform of grants. The current systemattenuates this perverse incentive byintroducing a block grant for socialprotection based on risk factors relatedto income and unemployment levels indifferent jurisdictions. Another example isthe introduction of a block grant in Denmarkin 2002-04 to finance municipal spending onold-age care, which aimed at encouragingcost-effectiveness in local governmentprovision.


175

Ensuring transparency and administrativesimplicity

Transfer systems tend to be formula-based,especially in so far as general-purpose grantsare concerned. This makes arrangementstransparent and predictable, which facilitatesbudget planning at the local level. However,formula-based systems are not withoutpitfalls. In some cases, formulas are basedon parameters such as the size of theresident population. An example of thispractice is the calculation of grants to financelocal government administrative costs in theCzech Republic, in place since 2005. Thismechanism assumes that service deliverycosts rise monotonically with population; assuch, it may discourage the local authoritiesfrom cutting costs on the basis of gains ofeconomies of scale, unless they are free toallocate the associated cost savings toalternative uses.

There is a trend towards simplification ofgrant and transfer systems. This isespecially so in countries where currentarrangements are admin is t rat ive lycumbersome and where local governmentautonomy is curtailed by restrictivec o n d i t i o n a l i t y i n t h e d e s i g n o fintergovernmental grants and transfers.Simpler systems based on block, asopposed to conditional, grants andtransfers go in the direction of enhancingloca l se l f-government by a l lowingrecipients greater autonomy in settinglocal policy and the level of provision.18

Nevertheless, the use of a single transferinstrument to meet different objectives(equalisation and subsidisation of servicedelivery to correct for spillover effects, forexample) may pose problems.19 The Swissreform of 2004 addressed this issue byreplacing earmarked grants by general-purpose transfers for equalisation andusing earmarking for the attainment ofpolicy objectives other than equalisation.

Another consideration is a possible loss ofaccountability to donors, which would runcounter to efforts to use grant/transfersystems to align policy objectives acrossthe different layers of administration.

Local Government Indebtedness and Financial Management

The level of local government indebtedness inrelation to GDP does not elicit concern aboutthe longer-term sustainability of local publicfinances in Europe. The implications oflocal government indebtedness formacroeconomic fiscal management aretherefore limited. Only in a few countries,such as Denmark, Iceland, Italy andNetherlands, does the stock of localgovernment liabilities amount to close to orabove 10 percent of GDP (Figure 5.7 andTable 5.4). Local governments in thesecountries, with the exception of Latvia, alsohold sizeable assets, such that their netdebt position is low in relation to GDP. Inother countries, net local asset-to-GDP ratiosare large, as in France and Italy.

Low debt positions are due to a large extent torestrictions on local government borrowing. Insome countries, fiscal rules may ban localgovernments from borrowing altogether orissuing bonds, whereas in others there areconstraints on the size of local governmentbudget deficits (Annex 5.2).20 In most cases,local governments are only allowed to borrowto finance investment (golden rule). In somecountries, golden rule provisions apply tolong-term borrowing, whereas short-termloans (often in the form of bridge financingacross budget years) may be used to financeoperating expenditures. This is the case ofEstonia and Lithuania, for example.

Local government borrowing is also subject toprudential regulations. These regulationsare often based on debt service and loanrepayment capacity, as well as the level of

18. For example, Norway’ssystem of revenuesharing has changed,resulting in an ongoingreduction in themunicipal spendingshare and less relianceon earmarking oftransfers from thecentral government.Furthermore, followingthe changes in the costcalculation system forthe municipalities,additional funds havebeen set aside by thecentral government tocompensate individualmunicipalitiespenalised by the newsystem.

19. See Carlsen (1995) fora discussion of theNorwegian reform of1986.

20. See Sutherland et al.(2005) for a typologyof sub-national fiscalrules in OECDcountries.




indebtedness in relation to local revenue.Guarantee of local government liabilitiesby higher-level jurisdictions is oftenbanned as a means of safeguarding overallfiscal sustainability and imposing hard budgetconstraints at the sub-national levels ofgovernment. Restrictions are also in place inmost countries on the type of collateralallowed in local government debt issuance.Guidelines and recommendations on localgovernment indebtedness are also availablefrom the Council of Europe.

A d m i n i s t r a t i v e c o n t r o l s o n l o c a lgovernment financial management arebeing replaced by prudential regulations. Acase in point is that of the United Kingdom’s2004 Local Government Act, whichreplaced pre-approval requirements for localgovernment borrowing in England, Scotland

and Wales by prudential regulations ondebt service capacity. The legal frameworkfor local government borrowing has alsobeen strengthened in many countries, asillustrated by Bulgaria’s Municipal DebtAct of 2005 and a government resolutionissued in the Czech Republic in 2004.Administrative restrictions have also beenloosened in the Greece with the issuance ofthe 2006 Municipal Code.

Res t r i c t i ons on loca l gove rnmentborrowing and access to financial marketshave implications for financial management,including over the business cycle. Suchrestrictions often make local budgetsoverly pro-cyclical, because local govern-ments may find it difficult to smooth cyclicalfluctuations in revenue associated with thebusiness cycle.

0

20

40

60

80

100

120

Austri

a

Belgium

Bulga

ria

Cypru

s

Czech

Rep

ublic

Denm

ark

Eston

ia

Finland

Franc

e

Ger

man

y

Gre

ece

Hunga

ry

Ireland

Ita

ly

Latvia

Lith

uania

Luxe

mbo

urg

Malta

Nethe

rland

s

Polan

d

Portu

gal

Roman

ia

Slova

k Rep

ublic

Slove

nia

Spain

Sweden

Unite

d Kin

gdom

EU27

Icelan

d

Norway

Switzer

land

Central Government Debt Intermediate Government Debt Local Debt

Sources: Eurostat (February 2010), national sources and DEXIA calculations

Figure 5.7: Central, Regional and Local Debt as a Percentage of GDP



Budget balance (Net borrowing/lending) Consolidated gross debt (Maastricht definition) Local government net debt1

General Central State Local General Central State Local Assets Liabilities Net debt

Austria -0.4 -0.6 0.1 0.1 62.6 59.5 3.4 1.8 6.0 1.8 -4.1

Belgium -1.2 -1.6 -0.1 -0.1 89.8 84.3 4.1 4.8 2.1 4.9 2.8

Bulgaria 1.8 2.0 - -0.4 14.1 13.7 - 0.7 2.0 1.3 -0.7

Cyprus 0.9 -2.6 - 0.0 48.4 83.9 - 1.9 0.3 2.1 1.7

Czech Republic -2.1 -2.4 - 0.0 27.8 25.6 - 2.3 11.0 4.1 -6.8

Denmark 3.4 3.8 - -0.4 33.4 27.6 - 5.9 10.1 10.3 0.1

Estonia -2.7 -2.4 - -0.6 4.6 1.8 - 3.2 2.9 4.0 1.1

Finland 4.5 0.9 - -0.2 34.2 31.3 - 5.4 9.9 8.5 -1.4

France -3.4 -2.9 - -0.4 67.4 60.5 - 7.5 3.5 8.3 4.7

Germany 0.0 -0.6 0.0 0.2 65.9 40.3 21.3 4.8 5.1 4.7 -0.4

Greece -7.7 -8.3 - 0.0 99.2 109.7 - 0.7 0.8 0.7 0.0

Hungary -3.8 -3.6 - 0.2 68.7 65.7 - 3.7 5.7 4.7 -1.0

Ireland -7.2 -6.8 - -0.2 44.1 43.6 - 3.0 3.6 3.8 0.2

Italy -2.7 -2.7 - -0.2 105.8 99.3 - 8.1 3.3 9.9 6.6

Latvia -4.1 -4.4 - -1.3 19.4 23.6 - 4.1 7.2 5.7 -1.5

Lithuania -3.2 -2.0 - -0.2 15.6 14.4 - 1.2 3.0 1.2 -1.8

Luxembourg 2.5 -0.2 - 0.0 13.5 12.1 - 2.2 3.4 2.6 -0.9

Malta -4.7 -4.6 - 0.0 63.6 63.5 - 0.0 0.3 0.2 -0.1

Netherlands 0.7 0.5 - -0.4 58.2 53.6 - 7.2 7.4 8.9 1.5

Poland -3.6 -3.9 - -0.2 39.9 38.4 - 2.0 8.9 3.7 -5.2

Portugal -2.7 -3.3 - -0.1 66.3 67.2 - 3.8 1.6 4.5 2.8

Romania -5.4 -4.4 - -0.9 12.2 12.4 - 1.7 1.5 2.5 1.0

Slovak Republic -2.3 -2.7 - -0.1 28.7 27.7 - 1.9 4.1 2.5 -1.6

Slovenia -1.8 -1.2 - -0.6 22.5 22.3 - 0.9 2.9 2.0 -0.9

Spain -4.1 -2.8 -1.6 -0.5 39.7 33.7 6.3 2.9 2.3 4.7 2.4

Sweden 2.5 1.4 - 0.1 33.6 29.6 - 4.8 11.0 9.4 -1.6

United Kingdom -5.0 -4.8 - -0.2 43.5 42.6 - 4.0 2.7 5.5 2.8

EEUU2277 --22..33 --22..33 --00..11 --00..22 6611..55 5533..33 55..00 55..11 -- -- --

Iceland -13.6 -13.0 - -0.9 34.7 31.4 - 3.3 9.2 10.6 1.4

Norway 18.8 20.0 - -1.2 49.9 40.4 - 9.5 8.5 12.8 4.2

Switzerland 2.2 1.1 0.8 1.1 41.1 22.5 10.3 8.3 - - -

1. Refers to 2005 for Luxembourg and 2007 for Cyprus, Iceland and United Kingdom.

Sources: Eurostat, national sources and DEXIA calculations

Table 5.4: Local Government Indebtedness, 2008 (% of GDP)



Local governments are often called upon insupport of nation-wide fiscal stabilisationprograms. In Portugal, for example,constraints began to be imposed on localgovernment indebtedness in the contextof the adjustment efforts that were putin place in 2001, when the budget deficitceiling under the Stability and Growth Pactwas first breached. The Austrian NationalStability Pact of 2005 is another initiativetowards sharing the burden of adjustmentacross the different levels of government,with the introduction of budget balancetargets for the three layers of adminis-tration. Italy’s Internal Stability Pact in-troduced constraints on sub-national debtand expenditure growth, including for themunicipalities with at least 5,000 inhabi-tants. Germany’s reform of 2006 alsointroduced provisions for sharing theburden of adjustment under the Stabilityand Growth Pact between all levels ofgovernment (Table 5.4).

Policy Challenges for European LocalGovernments

The Effects of the Global Crisis on LocalGovernment Finances

Short-term effects of the global crisis

The global financial and economic crisis thaterupted with the collapse of Lehman Brothersin September 2008, and the ensuing sharpslowdown in economic activity in 2009-10,put pressure on local government finances.21

The main channels through which localgovernments have been affected by theglobal crisis are i) lower revenue (from owntaxes, shared taxes and transfers fromhigher levels of government, which are onlyin part mitigated by stimulus packagesin some countries); ii) higher spendingassociated with rising unemployment andheightened demand for social protection; andiii) tighter financial conditions, including

through a shortage of credit and costlierborrowing, where allowed, in the aftermathof the crisis. Needless to say, the relativeimportance of these channels has variedfrom country to country and among localgovernments.

The slowdown in economic activity, whichwas very pronounced towards end-2008and in the first half of 2009, took its toll onlocal tax revenue. This is especially thecase in countries, such as the Nordiccountries and some transition economiesin Central and Eastern Europe, where localgovernments rely on taxes on income,profits and capital gains, which are veryelastic to fluctuations in the businessand financial cycles. A weakening ofconstruction activity and falling housingprices are also affecting local revenueperformance in countries that haveexperienced housing booms in recent years,including Iceland, Ireland and Spain, wherelocal governments are particularly reliant onproperty or property transfer taxes.

At the same time, local budgets havesuffered from poor collection performanceat higher levels of government, even incountries where own local tax bases are notparticularly sensitive to fluctuations inbusiness activity and in the price of financialassets. This is because in most countrieslocal governments rely on revenue-sharingschemes, grants, and transfers from higherlevels of administration, often in the formof a fixed share of revenue. Uncertaintyover the level of transfers from thecentral government is also expected tohave a bearing on local budget-making andplanning in countries where such transfersare negotiated, rather than based on a fixedproportion of central government revenue,and account for a large share of localbudgets, as in Lithuania and Greece, forexample. This is especially so as post-crisisfiscal consolidation is pursued in earnest in

21. See CEMR (2009b) foran analysis of theeffects of the globalcrisis on Europe’s localauthorities and for theCouncil’s views of howthe local governmentscould be involved inthe nation-wide policyresponses to the crisis.See also CDLR (2009)for more informationbased on a survey ofEuropean localadministrations.


179

those countries where the effects of thecrisis on public finances have been mostadverse.

Turning to expenditure, local governmentfinances have been affected adversely byheightened demand for selected socialassistance and labour market-relatedservices, whose provision is under theirpurview. A number of labour marketprograms, including unemploymentinsurance, in some countries,reintegration programs and trainingfacilities for the unemployed, have tra-ditionally been delivered by localgovernments, the role of higher levels ofadministration being restricted typically tosetting benefit rates and (partially) fundingprovision.22 Demand for these services hasrisen in tandem with the unemployed rate,which the OECD projects to peak among theorganisation’s member states in 2010(OECD, 2010). Expenditure on theseservices is likely to be particularly onerousto local governments in countries wherethese authorities play a leading role in theprovisions of social programs, such asBelgium, France, Germany, Netherlands,and the Nordic countries.

Local governments have participated in theimplementation of stimulus packages thathave been put in place in many countries inresponse to the global crisis. To the extentthat these packages have included publicinvestments that are delivered through thelocal authorities, they have required financialassistance from higher levels of government –often in the form of conditional grantsand liquidity support, by bringing forwardthe payment of transfers and allocationfrom higher levels of administration – toensure a smooth implementation of the supportpackages, as in most European countries.23

According to OECD estimates (OECD, 2009b),public investment hikes accounted formost of the spending programs announced

during 2008-10 in Czech Republic, Finland,Germany, Netherlands, Poland, and UnitedKingdom. Support for local public investmentwas particularly strong in Austria, Germany,Portugal, and Spain. This is important,because local governments account for abouttwo-thirds of public investment in Europe(over 70 percent in France and Italy).

Post-crisis budget sustainabilityconsiderations

Local governments in most Europeancountries were hit by the global crisis in afairly benign fiscal environment. As a resultof steady growth and comfortable financialconditions, public finances in most Westernand Southern European countries hadstrengthened considerably prior to thecrisis. Budget outcomes had improved ingeneral, and public indebtedness had fallenin relation to GDP. By contrast, manycountries in Central and Eastern Europewere undergoing comprehensive fiscalconsolidation to ensure the sustainability oftheir public finances. Fragile fiscal positionshad made these economies particularly ill-suited to weather the effects of the globalfinancial turmoil that culminated with thecrisis in September 2008 and the ensuingrapid deterioration in investor sentiment.

Public finances deteriorated sharplyfollowing the crisis. The OECD projected theheadline budget deficit of the Euro-areacountries to rise from close to 2 percent toabout 7 percent of GDP during 2008-10(OECD, 2010). Budget deficits rose sharplyin Europe as a whole in 2008-09 as aresult of the economic slack brought aboutby the crisis and to accommodate a widerange of anti-crisis measures. Given thatthere are lags in tax collection and in theimplementation of stimulus packages, theeffects of the crisis on the budget (at alllevels of administration) are likely to persistover the near term.

22. See OECD (1999) formore information.

23. See Blochliger et al.(2010) for moreinformation on thepolicy responses at thelocal level ofadministration on thebasis of a surveyconducted with localauthorities. Despiteconsiderabledifferences among thelocal governmentsacross – and oftenwithin – countries inthe mix of instrumentsused, the policyresponses to the crisisappear to be counter-cyclical in mostcountries.




The sharp deterioration in budget positionsacross the continent will call for remedialmeasures in the form of fiscal retrenchmentover the medium-to-longer term. Themagnitude and timeframe of fiscalconsolidation will vary from country tocountry, depending primarily on the initiallevel of debt, the size of the stimuluspackages put in place in response to theglobal crisis and the speed of recovery ineconomic activity. In any case, fiscalconsolidation efforts are likely to affect localgovernments. This is particularly the case oflocal authorities that are heavily reliant ondiscretionary grants and transfers fromhigher levels of government, which mayalso be unable, or unwilling, to support localauthorities financially and may have strongincentives to offload unfunded expenditureson local budgets in the course ofadjustment. More recently, severalEuropean countries have announced fiscalconsolidation programs to restore thelonger-term sustainability of their publicfinances, which was compromised by theeffects of the global crisis and the ensuingrecession, including cuts or freezes intransfers to sub-national levels ofadministration.

But local budgets may also stand to gainfrom fiscal consolidation at higher levels ofgovernment. Those local authorities thatrely on a fixed share of revenue collectedelsewhere may actually benefit fromrevenue raising measures implemented athigher levels of government in support offiscal adjustment. Local governments thatrely on market-based sources of financewould also benefit from nationwide fiscalretrenchment to the extent that it improvesfinancial conditions and contributes to arecovery in credit growth and animprovement in market sentiment towardsriskier assets, such as local governmentdebt.

Long-term Challenges Facing LocalGovernments

Most European countries are confrontedwith the challenge of coping with the effectson public finances of a rapidly ageingpopulation and the need to integrateimmigrants into the labour market andsociety at large. These are long-termchallenges that cut across the differentlayers of government and are particularlyimportant at the local level. This is because,in many countries, the provision of socialservices, including active labour marketpolicies, are under the remit of the localauthorities.

In many countries a rapidly ageingpopulation is continually shifting thedemand for goods and services provided bylocal governments. The share in populationof individuals aged at least 65 years is closeto 16 percent on average in the Europeancountries under examination in this chapter,which is among the highest in the world.The United Nations projects the old-agedependency ratio to nearly double to 46percent of the European population during2010-50. Demographic trends areparticularly trying in Southern Europe,where birth rates are well belowreplacement ratios and population growthrates (excluding immigration) are alreadynegative. This is also the case of thetransition economies of Central and EasternEurope, where demographic decline can beattributed only in part to net emigration.

Demand for health and old-age care is risingand delivery costs are putting increasingpressure on the public finances, often at thelocal level of administration. The OECDestimates that, even with cost containmentmeasures, public (all levels of government)spending on health and long-term carecould rise from 6-7 percent of GDP in 2005in the OECD area as a whole to around 10


181

percent by 2050.24 In some countries, theincrease could be dramatic. At the sametime, the provision of other services, suchas child care and education, which hastraditionally been in the remit of sub-national governments, continues to weighon local governments. To a certain extent,demographic decline will alleviate somespending pressures, especially if selectedservices, such as primary schooling, arescaled back in tandem with falling demand.

Immigration and the integration of long-termresidents are creating claims on localbudgets. In Southern Europe, where netemigration was the norm until relativelyrecently, immigrants already account fora high share of the labour force. Theproportion of foreign-born individualsin the population of Greece and Spain, forexample, is over 10 percent. Immigrationis particularly challenging for localgovernment finances in countries where thelocal authorities play a leading role in theprovision (and financing) of housing,employment and related social assistanceservices, which are in high demand amongimmigrants. This is despite the fact thatimmigration is mitigating some of theeffects of adverse demographics in manycountries, including labour shortages inselected sectors, rising age dependencyratios and a contraction in the pool ofworkers contributing to social security.

Overall Assessment

Local Government Expenditure Mandates

There is a great deal of variation in theassignment of expenditure functions to localgovernments among European countries.In general, local governments focus onexpenditures that are local in nature, so thatthe benefits of provision can be internalisedby residents. But it is rather difficult inpractice to unbundle expenditure functions

according to the geographical reach of thebenefits they generate. In some cases, notonly the benefits, but also the costs of localgovernment expenditures spill over acrossjurisdictional borders. This possibilitycal ls for intergovernmental financialarrangements to remedy the incentives thatlocal authorities would otherwise face torefrain from supplying these goods andservices or to shift the burden of provisionto other constituencies. The examplesdiscussed above illustrate the numerousarrangements put in place in severalEuropean countries to deal with theseimportant policy issues.

Improved inter-governmental coordinationis needed to deal with rising expenditurepressures at the local level of government.Demand for local services, including theprovision of selected health care, is risingdue to population ageing and technologicalchange. This trend is particularly challengingin Europe, where many countries arealready facing the difficulties arising from arapidly ageing population. This is also thecase of social services associated with theabsorption and integration of immigrants insome countries, who already account for acomparatively large share of the labourforce in many European countries. Dealingwith these challenges goes beyond therealm of policies that can effectively be putin place by local governments. But there is aneed to boost coordination among thedifferent levels of government to deviseworkable, cost-effective solutions to theseproblems that would take into account localgovernment expertise in service deliveryand their ability to extract information onresidents’ preferences and needs.

Regulation of service delivery by higherlevels of government needs to recognisethe benefits of local self-government. It isimportant to rely on regulations and normsset by higher levels of government to

24. The OECD argues thatmost of the upwardpressure on long-termhealth care will comefrom non-demographicfactors, includingeffects fromtechnology andrelative prices. SeeOECD (2006) for moreinformation.




ensure that minimum standards are metthroughout the national territory. Butoverly restrictive, top-down regulationsoften thwart the potential fordecentralised provision to enhance cost-effectiveness by exploiting differences inregional preferences, in addition toundermining the scope for decentralizationto strengthen accountability of localofficials to taxpayers. This drawback isparticularly important in the case ofintergovernmental financial arrangementsthat focus on the equalisation ofexpenditure and/or the use of conditionalgrants for the financing of local governmentexpenditure. Policymakers should thereforemake sure that top-down regulations donot curtail their ability to tailor servicedelivery to local preferences and needs,which may vary across regions, especiallyin territorially diverse countries. A relatedissue is that of cost shifting across levelsof government, which often arise when top-down regulations impose a financial burdenon local budgets. When unfunded, suchintergovernmental cost shifting should bebanned.

At the EU level, efforts to expand theinternal market need to take local choiceinto account. The provision of publicservices has come under increasingscrutiny from the European authoritiesas they seek to expand the Europeani n t e rna l ma rke t . 25 Tens i on s havetherefore often arisen with sub-nationalgovernments in matters that theyperceive as within the remit of their self-government prerogatives. A EuropeanCharter on Local and Regional Services ofGeneral Interest was proposed in 2009to deal with this matter. Against thisbackground, effort should be stepped upto reconcile initiatives towards buildingan internal market with the need topreserve local preferences and choice.

Options will need to be considered fordealing with inter-jurisdictional spilloversand economies of scale in servicedelivery. For example, there may becases where local provision createsconsiderable inter-jurisdictional externalitiesthat cannot be so lved e f fec t i ve lythrough the use of earmarked, conditionalor matching grants, and/or whereoptions to allow local governments tomake the most of economies of scale toreduce service delivery costs may belimited. In these cases, there can be arising temptation to recentralise servicedelivery. Nevertheless, an option that ismore in tune with local governmentpreferences in most countries includesan array of cooperative ventures amonglocal governments, such as inter-mu-nicipal consortia. While taking countryspecificities into account, greater cooperationis needed among local governments, andbetween local governments and higherlevels of administration, to search formutually beneficial solutions.

Local Government Revenue

As in the case of local governmentexpenditure functions, there are importantdifferences in the assignment of revenue tolocal governments in Europe. In somecases, local budgets are financed by localsources, as in the case where broad taxbases are delegated to the local authoritiesor shared among the different layers ofgovernment. An alternative arrangement isto assign a limited number of sources ofrevenue to the local authorities, whichwould then rely predominantly on transfersand grants from higher levels ofadministration. But, as far as the experiencein Europe is concerned, there does notappear to be a strong relationship betweenthe composition of local governmentrevenue and budget outcomes, includingthe size of budget imbalances, the level of

25. Public services,including those thataffect individualcitizens’ quality of life,sustainable economicdevelopment orsocial/regionalcohesion, are known inEU jargon as Servicesof General Interest(SGIs) or Services ofGeneral EconomicInterest (SGEIs).

26. See de Mello (2000)for a review of theliterature and somecross-countryempirical evidence.



indebtedness, and the level of localgovernment spending. This is not the case forother parts of the world, where the design ofintergovernmental fiscal relations has oftennot been conducive to fiscal discipline at thedifferent levels of administration.26

With regard to own revenue sources, theassignment of tax bases across the differentlayers of government is in general broadly inline with public finance principles. Localgovernments tend to collect immobile taxes,such as those on property, and fees forservices provided locally, which are the idealsources of revenue for local governments. Insome cases, nevertheless, as in the Nordiccountries and transition economies of Centraland Eastern Europe, mobile bases, such aspersonal and corporate income, are assignedto the local governments. The experience ofseveral countries suggests that, while the PITworks quite well as a local tax or when itsrevenue is shared between different levels ofgovernment, the CIT is best managed by thecentral government.

There may be scope for raising local revenuethrough user charges and fees for services.The experience of several countries isthat the provision of local services freeof charge (without making thecorresponding costs perceptible to users)encourages excessive demand. Co-payments may therefore be used to extractinformation on users’ willingness to pay,which is useful for local government to gaugethe marginal cost of services that wouldotherwise be difficult to price, and to raiserevenue that could be used to alleviatesupply constraints. But concern that co-payments may hinder access by selectedsocial groups to local services has oftendiscouraged local governments from levyingcharges and fees for services. Whilerecognising the constraints imposed byspecific country environments, localauthorities should continue to evaluate the

costs and benefits of user charges in theirefforts to raise revenue.

Fiscal Rules and IntergovernmentalFinancial Arrangements

Fiscal rules are effective in instilling fiscalrectitude at lower levels of government tothe extent that they are complemented byeffectively designed intergovernmentalfinancial arrangements. Constraints on localgovernment borrowing and indebtednessshould therefore be accompanied by efforton the part of policymakers at all levels ofgovernment to enhance the incentivestructure underpinning the assignment ofrevenue and expenditure functions to localgovernments and the complementaryarrangements for intergovernmentaltransfers and grants. As a general principle,intergovernmental grants and transfersshould not be perceived as gap filling orvehicles for higher level of government tobail out jurisdictions in financial distress.European countries fare well in this regard,as evidenced by the comparatively soundfinancial position of local governmentsbefore the crisis.

There is nevertheless scope for improving thedesign of fiscal rules for local governments.On the one hand, rules that restrict localgovernment access to credit markets mayencourage local governments to rely on banklending as a source of finance or to bypassrestrictions on debt issuance. In this case,the option of replacing such restrictions bymore comprehensive prudential regulationsbased on debt repayment capacity that do notdistort policymakers’ choice over financialinstruments should be considered. On theother hand, borrowing constraints orbalanced budget provisions make it difficultfor local governments to smooth thebudgetary impact of fluctuations in thebusiness cycle. This is especially the case oflocal authorities that rely on cyclical revenue,



through own collections or sharing arrangementswith higher levels of government. Thusbalanced budget provisions could beredefined on a cyclically adjusted basis, solong as local governments have access tobudget financing from non-governmentsources in bad times and instruments to savecyclical revenue windfalls in good times.

Conclusions

Europe has a long tradition of decentralisedfiscal management. There now appears to be abroad commitment to self-government,reinforced by the issuance of the EuropeanCharter of Local Self-Government andthe recognition by the European Unionof the principle of subsidiarity in thetreaties of Maastricht, Amsterdam, andLisbon. Nevertheless, tensions sometimesunderstandably occur between the localauthorities and higher levels of administration,including middle-tier governments, on policyareas where the assignment of functionsacross the different levels of administration isnot easy to delineate or rapidly evolve. To alarge extent, existing intergovernmental fiscalrelations are rooted in history, but they are notimmutable. The return to democracy in manycountries (such as in Central and EasternEurope), the need to deal with regional

diversity, and the ensuing pressures for policyautonomy (as in Belgium, Italy, Spain, and theUnited Kingdom), or economic imperatives,such as the recognition that public provisionmay be more cost-effective when it bestreflects local preferences and needs, havesustained the momentum of decentralization inEurope.

Notwithstanding considerable achievementsin this area, this paper argues thatimportant policy challenges remain for localgovernments in Europe. In particular,emerging expenditure pressures, such asthose related to population ageing andimmigration, are already bearing down– and will continue to do so – on localbudgets. At the same time, localgovernments are being affected by fiscalconsolidation in most countries following thesharp increase in public indebtedness as aresult of the global crisis and the ensuingrecession. Moreover, continued effort will beneeded to align policy objectives acrossthe different layers of administration in thedesign of intergovernmental financialarrangements so as to ensure thatdecentralised fiscal management andservice delivery are cost-effective,cooperative and consistent with principles oflocal self-government.



European Regional Policy Recommendations

Europe’s local and regional governments will wish to respond to the major fiscalchallenges and pressures which now confront them. In particular, they will need toconsider how to engage effectively with central governments, with a view to maintainingthe maximum scope for local and regional self-government for the coming decades.Some key points are likely to be:

• Local and sub-national governments should not take a disproportionate share of the fiscalreductions over the coming years, whether via reduced local revenue or through cuts ininter-governmental transfers;

• There should be recognition that local governments have generally less fiscal flexibility thannational governments, and for the most part have not incurred high levels of indebtedness;

• Central governments should also recognize that local authorities generally have major socialwelfare obligations, which cannot be escaped while unemployment and other povertyindicators are high, are pro-cyclical in impact, and over which they have little direct control;

• Central and sub-national governments should together recognize the need for good qualityfuture-oriented capital investment, mainly financed by new debt, e.g. on greening thefuture economy and urban infrastructure;

• In most countries, the local government tax base should be broadened to give more fiscalroom for maneuver and protection from excessive income losses from any one source;

• Local governments should have more freedom to raise fees and charges;

• Grants from central governments should be fewer (i.e. of a more general character), onlyear-marked in exceptional cases, and provided without imposing undue administrative andregulatory burdens;

• The EU needs to make its internal market and public procurement rules less centralised andintrusive, with more scope for local choice on local services, and with less detailedprescription. The cohesion policies must be preserved.

At the same time, local and regional governments will need to think very carefully abouthow to do the best for their citizens within the reduced resources available. This meansthat they may have a common interest in looking at new ways of delivering publicservices cost-effectively, and at new ways of reducing administrative costs. The range ofoptions must depend on local choices, whether by increased inter-communalcooperation, by some outsourcing or PPPs, by sharing some services between differentlevels of government, and so on. There will need to be an even greater emphasis onbeing responsive to the citizens and users of services.



European Regional Policy Recommendations (cont.)

The important point will be for local and regional governments to be proactive in all ofthese issues, and not cede the initiative to central governments. The role of nationalassociations of local and regional governments is of particular importance, since they cangive advice and training to their members, directly or through daughter companies, aswell as negotiate with central governments.

Amongst the points for discussion with central governments, local authorities will oftenwish to emphasize that:

• Where territorial reforms are being considered in order to enhance efficiency and cost-effectiveness,priority should be given to supporting inter-municipal forms of co-operation, rather thencentrally imposed amalgamations which often do not respect local needs and identities.

• It is essential that there are effective and fair arrangements, involving all levels ofgovernment, for co-ordination and negotiation on financial and fiscal issues, inparticular on the rules for and amounts of financial transfers, grants and redistributedresources (in particular equalisation processes), but also covering the policyframeworks for all major services / competences.

• It will probably be wise or even necessary to try to reach some form of “settlement” orcompromise with central governments, based on protecting key principles of self-government (see Article 9 of the European Charter of Local Self-government, 1985).

We should also not ignore the European Union, which is playing an increasingly importantrole in relation to economic policy, cohesion funds as well as the rules for public services.Here too, there is a need for effective and fair co-ordination arrangements, involving theEuropean Commission and the representative associations of local and regional government.The purpose would be to evaluate the financial, administrative and regulatory impact of EUlaws and policies on sub-national governments and seek improvements in the framework.

Recommendations developed by the representatives of European Local GovernmentsAssociations in the workshop on GOLD II in the city of Sceaux (Paris, March 2010)


0w2010 06 Latin America DEFcarta ang 15/9/11 07:12 Página 188

6. LATIN AMERICA

JJOORRGGEE MMAARRTTÍÍNNEEZZ--VVÁÁZZQQUUEEZZ **

GGEEOORRGGIIAA SSTTAATTEE UUNNIIVVEERRSSIITTYY,, UU..SS..AA..


* I am thankful to Gustavo Canavire-Bacarreza andGabriel Leonardo for very able assistance and HaroldVasquez-Ruiz and Mark Curtis for conductingadditional background research. I am also thankfulto Andrew Nickson, Paul Smoke, Edgardo Bilsky,Claire Frost, Eduardo Stranz, the FLACMA technicalteam, and many others attending the UCLG regionalworkshop in San Salvador in February 2010 forhelpful comments on previous drafts.


191

L ocal governance and the municipalityhave a long history and tradition in Latin

America going back to colonial times.1 Itwas only after the 1980s, and for manycountries only in the last decade, thatgenuine decentralization reform effortshave come to invigorate and enhance therole of local governments. However,despite some significant progress to date,many challenges still remain for munici-palities to play a vibrant and meaningfulrole in the delivery of public services andto contribute to improve the daily lives ofLatin American citizens.2

In the last two decades the Latin Americanregion has seen a general trend toward anincreased level of fiscal decentralization.Using the measure of sub-nationalexpenditures as percent of national ex-penditures, fiscal decentralization increa-sed from an average of 13 percent in 1985to 19 percent in 2005; using the measureof sub-national expenditures as percent ofGDP decentralization increased from 5.5percent in 2000 to 6.6 percent in 2007.However, there are significant variations inthese trends across countries in theregion.3 Overall, increased decentrali-zation can be detected in the devolution ofnew responsibilities which includes theenvironment, the fight against poverty,and an increase in decentralized ex-penditures for education, health, etc. Lessprogress can be detected in the devolutionof autonomous revenue sources.

Recent times have seen a variety ofinnovations in the region that haveattracted interest from all corners of theworld, such as ranking systems’ localperformance in Brazil and Colombia, perclient based transfers for health andeducation in Chile, or fighting poverty withdirect transfers to families administeredby municipalities in Brazil. A good numberof countries have embarked or are consi-

dering significant reforms that that willfurther strengthen municipal autonomy.For instance, Bolivia has recently approveda new Constitution to allow for betterrepresentation of different ethnic groupsat the sub-national level; Uruguay latelyintroduced a third tier of governmentmade up of 89 new municipalities; andCosta Rica only just approved the “LeyGeneral de Transferencia de Competenciasy Recursos a los Municipios” whichprovides the ability to transfer 10 percentof the national budget resources to themunicipalities, clearing the way forlocal governments to assume newcompetencies and improve the quality ofservices and infrastructure. In El Salvadorthe association of municipalities (COMURES)is maintaining an active dialog with centralauthorities to increase the funding andgeneral stability of the general transfersystem (FODES) which represents between70-80 percent of local budgets, and wasexpected to reach 9 percent of the nationalbudget in 2009 but because of the crisis itattained only 7.5 percent of the nationalbudget.

On the other hand, there are countriesin the region where some trends have movedtoward some forms of recentralization.For example, in Argentina the Law ofEconomic Emergency of 2002 and theBudget Law of 2006 have given centralauthorities increased discretion to assignfederal funds or unilaterally interrupt theirdisbursement. In the Dominican Republicthere have been elements of recen-tralization with the Municipal Law of2007 establishing fixed budget sharesfor different types of expenditures onpersonnel, services, public infrastructure,and so on; it is also feared that the newconstitution will lead to the generaltransfer fund of 10 percent of the statebudget established in 2003 (but neverimplemented). Similarly, in Peru recent

1. See United Cities andLocal Governments2008 GOLD I report.

2. The effectiveness ofdecentralization effortshas variedconsiderably acrosscountries of the LatinAmerica region. In thelast decade,decentralization hasmoved at a fast pacein countries such asColombia and Peru butit continues to bestagnant after severaldecades of planningand legal measures incountries such as theDominican Republicand Haiti. BesidesBrazil and all theSpanish-speakingcountries of LatinAmerica, this papercovers also Haiti,Jamaica, and Trinidadand Tobago. Asshorthand, all thecountries will beaddressed in thispaper as LatinAmerica.

3. See, for example,Daughters and Harper(2007).



LLAATTIINN AAMMEERRIICCAAUnited Cities and Local Governments192

legislation has revoked the municipalities’prerogative to issue building licenses andrezoning of land use. The regular transferfunds allocated to municipalities havebeen significantly reduced from 2009 to2011 (a decline of 22 percent in the lastfive years). In Colombia, the centralgovernment has recently decided todirectly allocate resources for water andsanitation that until then had beenassigned to municipal governments(through a fiduciary fund administrated byCentral Government). Finally, in Venezuelathe municipal authorities have beendenouncing the continuous curtailment ofcompetencies and resources and theincreasing encroachment of the centralauthorities in local matters.

This report takes an in-depth look at thecurrent state of the local public finances inthe Latin America region, identifies andanalyzes some of the main challenges forimproving efficiency, equity and effecti-veness in the delivery of public servicesand for promoting development and itcloses by offering a set of observationsconcerning policy reform.4

Structure and Performance of Local Government Finances in the Region

Countries in the region are highly diversealong a number of dimensions: federalversus unitary, size, colonial tradition, etc.This diversity is found first among the fourfederal countries in the region: Argentina,Brazil, Mexico and Venezuela. Among thecountries with a unitary system we canidentify clusters of countries with moresimilar institutions and current challenges,including the Andean countries (Colombia,Ecuador, Peru and Bolivia), the generallysmaller countries of Central America, theIsland States with non-Iberian traditions,5

and what we could call the southern cone

exceptions (Paraguay, Uruguay and espe-cially Chile) because of their approach tofiscal decentralization. The diversity is alsofound in population size (from the 196million of Brazil to the one million ofTrinidad and Tobago), in real GDP percapita (from $9,357 in Argentina in 2007in constant 2000 US dollars to $884 inNicaragua and only $411 in Haiti), and inother dimensions.

However, there are also many commonfeatures in the way municipalities arestructured, which enables us to observe allmunicipal governments in the region froma common perspective. An importantfeature is that for those countries withmore than one tier of sub-nationalgovernment, the relationship between thecentral government and the municipalitiesis for the most part directly between thesetwo levels as opposed to the centralgovernment dealing exclusively with theregional and local governments and thenthe latter dealing exclusively with themunicipalities.6 In most cases, the legalstatus of the municipalities is clearlystated in the constitution or specializedlaws, such as municipal codes. The mostimportant exception to this rule isArgentina where the constitution gives theintermediate level government, theprovinces, discretion to structure the fiscalarrangements with the municipalities.7 Toa lesser extent the same story is repeatedin Mexico.8 Thus, the key difference inexplaining the different approaches tocentral-local relations is between “federal”and “unitary” nations. But even in thefederal cases, the issues currently facingmunicipal governments are not essentiallydifferent from those being faced by therest of the municipalities in the region. Forthis reason, the report will not make apoint of identifying the different groups ofcountry experiences but instead we willuse a common framework for all countries,

4. The focus of this reportis on fiscaldecentralization.Issues of political andadministrativedecentralization for themost part are notcovered.

5. Naturally, thesecountries are notusually classified as“Latin” America.

6. In the technicalparlance the verticalrelationships betweendifferent levels ofgovernment arebifurcated (central tolocal and central toregional, separately)as opposed tohierarchical (central toregional to local, alllinked).

7. In contrast, forexample, the BrazilianConstitution defines itspolitical system as theunion of the centralgovernment, thestates, and themunicipalities, thusgiving localgovernments anautonomous standingvis-à-vis theintermediate levelgovernments.

8. Brazilian states alsohave some limited rolein managing themunicipalities.



Country Levels Govt. level names Intermediate Local Average Average Population in Population GDP per of gov. Level Level Population Population the largest in urban capita (#) (level 2) (level 3) Level 2 Level 3 city agglomerations (current US$)

(% of urban > 1 million population) (% of total

population)

Argentina 3 federal, provincial, 24 2218 1,654,436 18,804 35 39 6645municipality/department

Bolivia* 4 national, department, 9 327 1,058,277 29,126 26 32 1378municipality/canton

Brazil 3 federal, state, municipal 27 5564 7,041,481 34,169 12 39 7013

Chile 3 national, region, municipality (15) 345 1,109,075 48,220 39 34 9851

Colombia 3 national, department, 32 1102 1,386,232 40,253 23 35 4684municipality

Costa Rica 3 national, canton (7) 81 636,968 55,046 46 29 5891

Dominican Republic** 3 national, province, municipality (32) 155 306,677 63,314 32 22 4210

Ecuador 3 national, province, canton 22 221 606,446 60,370 29 32 3432

El Salvador 3 national, department, (14) 262 436,197 23,308 39 23 3336municipality

Guatemala 3 national, department, (22) 333 606,989 40,101 16 8 2548municipality

Haiti 3 national, department, commune 10 140 97,008 6,929 45 21 640

Honduras 3 national, department, (18) 298 398,563 24,074 .. .. 1671municipality

Jamaica 2 national, parish 14 191,128 .. .. 4802

Mexico 3 national, state, municipality 32 2454 3,290,016 42,902 23 34 9715

Nicaragua 3 national, department, (17) 153 35,003 3,889 .. .. 1023municipality (+ 2 special regions)

Panama 3 national, province/comarca, 14 75 238,810 44,577 53 38 5828district

Paraguay 3 national, department, canton 17 231 360,390 26,989 51 30 1995

Peru*** 3 national, region/special 26 1834 1,096,480 15,544 39 28 3771province, province/district

Trinidad and Tobago 2 national, region/borough/city 16 83,013 .. .. 16351

Uruguay**** 2 national, department 19 89 174,942 37,347 49 45 7297(municipality)

Venezuela 3 national, state, municipality 24 335 1,145,125 82,038 12 32 8299

Note: # computed using the number of jurisdictions in level. Between brackets when the authorities are not elected.

* In Bolivia, there are departments, provinces (not elected authorities: 112), municipalities and territories of traditional peoples “territorios indígenasoriginarios campesinos” (incorporated in the new constitution)

** In the Dominican Republic, recent constitution reforms recognize 229 municipal districts as local governments *** In Peru, are two kinds of municipalities: provincials and districts. **** In Uruguay, municipalities were created in 2009 by constitutional reform.

Sources: UCLG data collection, World Bank

Table 6.1: Decentralization in Latin America: Political and Territorial Organization (2007)



identifying particular country experiencesas lessons of what needs to be avoided orwhat may be replicated.

The Structure of Local Governments

As a rule the vertical structure of govern-ment in Latin America is organized in threetiers of government (Table 6.1), with theexceptions of Bolivia that has four levels,and Jamaica and Trinidad and Tobago thathave two levels9. The focus of this paperwill be almost exclusively on the lowesttier of government: the municipalities. Theintermediate levels (States, provinces,regions and departments) will be referredto only in issues relevant to the muni-cipalities.

As of 2010, there were over 16,000 mu-nicipal governments in Latin America.Their number by country obviously varieswith population size and territory, withBrazil counting 5,564 municipalities and atthe other extreme 16 municipalities forTrinidad and Tobago. Local governmentsvary considerably in size in each country(Table 6.1).

Even though a significant share of theLatin American countries’ population livein the largest cities (for example, 53percent in Panama, 49 percent inUruguay, 40 percent in Peru, and 35percent in Argentina), the majority ofmunicipalities in the region remain, forthe most part, small in size and of a ruralnature. For example, in Peru over 200municipalities have populations under1,000 inhabitants, and over 50 percent ofall municipalities have fewer than 5,000inhabitants. Thus the region faceschallenges at the two extremes: massivemetropolises with high levels ofpopulation density, congestion and ringsof urban poverty; and very smallmunicipalities in rural areas with low

density, little administrative capacity andlacking an appropriate scale for the pro-vision of many basic public services.

In many Latin American countries thestructure of local governments continuesto be work in progress. In the case ofBolivia, the new 2009 Constitutiondeclares autonomous governments at theregional, municipal, and indigenouscommunity level, with the added facet thatindigenous communities may fit in one ormore municipalities or regions. The legalnorms regulating this structure have notyet been enacted. The proliferation of newlocal governments, almost always throughthe fragmentation of existing ones,continues to be quite common in theregion. For example in the DominicanRepublic, between 1995 and 2006 thenumber of municipalities rose from 108 to155.10

Local Expenditures and the Assignment of Competencies

The scope of local government expen-ditures: The local government share intotal public expenditures and in GDPdiffers significantly by country but theyare generally lower than those observed inother regions of the world. However, asshown in Figure 6.1, the share of thepublic sector in GDP as measured by totalexpenditures of the general government israther high, and at levels above those ofother countries in the world at similarlevels of per capita income. This contrastof proportionately smaller localgovernment sectors in otherwise largeroverall public sectors may be explainedfirst, by fewer functional expenditureresponsibilities being assigned to localgovernments in comparison with otherregions of the world, and second, byrelatively lower levels of expenditure andservice provisions in those expenditure

9. In the case of Bolivia,the provincial levelmay not be interpretedas an additionalautonomous level.

10. A recent law in thatcountry has imposedstricter requirementsfor new potentialmunicipalitiesrequiring that theyhave 15,000 residentsand be able togenerate at least 10percent of the revenuethat their previousmunicipality wasraising.


195

responsibilities actually assigned to localgovernments, as discussed below.

As shown in Figure 6.1, there are largedifferences between the share of localgovernments in total public expendituresand the relative importance of localgovernment expenditures in GDP. Amongthe most decentralized countries, asmeasured by the municipal share in totalpublic expenditures are: Brazil, Ecuador,and Colombia11 at around 20 percent,followed by Peru and Bolivia at about 16percent and Chile with 12.8 percent. Twolarge federal countries, Argentina andMexico stand at 8.8 percent and 6.5 per-cent, respectively. At the low end we findunitary countries that are still highlycentralized such as many Centro-Americanand Caribbean countries (from 7 percentin Salvador, to 1.7 percent in Panamaand 0.9 percent in Jamaica and betweenthem Dominican Republic, Honduras,Guatemala, Costa Rica).

Perhaps a more meaningful measure oflocal governments’ role as providers ofpublic services is the share of local ex-penditures in overall GDP (Figure 6.1).This variable measures the percentage ofnational resources channelled throughlocal governments. From this perspective,Brazil at 8.3 percent and Bolivia at 7.3percent, Colombia at 5.6 percent (seenote 14) and Ecuador at 4.4 percentare currently the most municipallydecentralized countries in the region,while Argentina, Peru, Chile and Mexicoaccount for between 2 and 3 percent of theGDP. At the bottom stand more centralizedcountries: Paraguay, Honduras, ElSalvador, Dominican Republic, Costa Rica,Panama, Guatemala, and Jamaica, around1 percent of GDP.

The assignment of expenditure compe-tencies: Several features characterize ex-penditure assignments in Latin Americancountries. First, with the exception ofArgentina,12 all countries have explicit

11. For Colombia, ifdepartments asintermediate localgovernments areadded to themunicipalities, then in2008 localexpendituresrepresent as much as29 percent of generalgovernmentexpenditure and 9.5of GDP.

12. In Argentina, eachprovince regulatesthe expenditureresponsibilities ofmunicipalitiesdifferently. Theprovinces, in general,tend to enumerate aset of generalfunctionsaccompanied by aclause that may beused to expand localcompetencies. Muchless frequently theprovinces explicitlyenumerate thefunctions thatmunicipalities mustfulfill or functionsexclusively assignedto them.


Argen

tina(

2006

)

Bolivia

(200

8)

Brazil

(200

7)

Chile(

2007

)

Colom

bia(2

006)

Costa

Rica

(200

7)

Domini

can

Repub

lic(2

006)

Ecuad

or(2

007)

El Salv

ador

(200

7)

Guate

mala

(200

9)

Hondu

ras(

2008

)

Jam

aica(

2006

)

Mex

ico(2

007)

Panam

a(20

05)

Parag

uay(

2007

)

Peru(

2007

)

Local Expenditure as a % of GDP Local Expenditure as % of General Government

0

5

10

15

20

25

30

2,9

7,3 8,3

2,4

5,6

0,8 0,9

4,4

1,2 0,7 1,3 0,3

2 0,8 1,3

2,6

8,8

16,8

26,3

12,8

18,7

3,7 5,3

23,4

7

4,4 4,9

0,9

6,5

1,7

6,3

16,4

Sources: IMF, Ministries of Finance of Argentina, Bolivia, Guatemala and Peru, UCLG data collection (cf. Annex 6.2)

Figure 6.1: Local Expenditure as a Percentage of GDP and General Government



assignments for municipalities in theirnational laws. In most countries, as shownin Annex 6.1, the expenditure assignmentsare defined in the country’s constitution;otherwise, the assignments are specifiedin special laws, most commonly some formof municipal code.13 Often, in these formalassignments, municipalities are allowed toprovide any services not specificallyassigned to any other level ofgovernment.14

Second, there are enormous variations inthe assignment of responsibilities to muni-cipalities; the assignments represent amosaic of approaches, which as shown inAnnex 6.1 defy generalization. Most coun-tries provide for a set of obligatory func-tions, often exclusively assigned tomunicipalities. These range from basicurban services such as garbage collection,road maintenance, parks, market stallsand slaughter houses, and so on, as wellas some administrative functions such as,civil registry, land planning, and housingpermits. In addition, most countriesprovide voluntary functions, which oftenare co-shared with other levels ofgovernment. These may include somesocial services, such as basic education,primary health services, and publicutilities, such as water and sewerageservices. But as can be seen in Annex 6.1,in some countries (Colombia, Guatemala,Jamaica) basic education and primaryhealth can also be designated asobligatory and exclusive responsibilities oflocal governments.15

Third, in some countries (for example,Bolivia and Chile) the centralgovernments, while retaining the obligationof financing social welfare services (suchas social security, unemploymentcompensation, and welfare payments),have delegated the implementation andmanagement of several social programs

(e.g., family welfare services) to localgovernments in order to exploit theadvantage of proximity and better informationlocal governments have.16 Municipalgovernments in many Latin American countriesplay a large role in the public investmentof infrastructure at the sub-national leveloften as equal partners with upper levelgovernments in patterns similar to thoseobserved in European countries. Forexample, in Brazil local governments inrecent years have undertaken close to 45percent of all public sector investments.

Finally, many countries in Latin Americahave concurrent or shared expenditureresponsibilities, which generally results inless clarity and potentially more conflictthan exclusive assignments.

Revenue Assignments

Practically all countries of Latin America assigncertain taxes to local governments; someexceptions are Jamaica, and Trinidad andTobago. As shown in Table 6.2, the mostcommonly assigned type is the propertytax, although it varies in name and scopeacross countries.17 Other local taxes includebetterment levies, car registration and carpermits, real estate and land transfers, differentforms of business licenses,18 taxes on gambling,and in some form of sales tax or business tax.19

Practically all local governments are allowed tocharge fees for particular public services suchas building licenses, refuse collection, publicutilities, slaughter houses, and public markets.

Revenue assignments are formalized indifferent ways, usually in the general taxlaws or in special municipal laws. Theexceptions are Brazil, where it is establishedin its national constitution, and Argentinaand Mexico, where the constitutiondelegates to the provinces or states theauthority to determine local revenueassignments. This arrangement results in

13. There are some qualifiedexceptions to the rule.For example, in the caseof Colombia there is nospecial law assigningexpenditureresponsibilities atdifferent levels ofgovernment, but thereare several laws (60 of1993 and 715 of 2001)that specify certain normsregarding the assignmentof competencies.

14. A clear exception to thisrule is Chile, wheremunicipalities arecircumscribed to a closedlist of functions.

15. In El Salvador basiceducation and primaryhealth are assigned to thelocal level, but actualservice delivery worksthrough special localmechanismsadministrated jointly bythe State, thecommunities and theprivate sector.

16. On the whole theseexperiences appear tohave been positive(Bolivia, Brazil, Mexicoand Peru). In the case ofMexico some programshave been critiquedbecause of partisaninterference by centralauthorities in thedeployment of funds.

17. Of significant importanceis that just a handful ofcountries allow for thetaxation of both urbanand rural property. Thosecountries allowing onlythe taxation of urbanproperty leave ruralmunicipalities in adisadvantage. Note thatthe property tax is stillassigned to the ElSalvador’s centralgovernment and thatthere appears to be noproperty tax in theDominican Republic, onlya property transfer tax.

18. A good example is the“patente municipal” inChile which is paidannually at rate based onthe declared own capitalassets. This tax is furtherdiscussed in the nextsection.

19. In the cases of Brazil’sISS (tax on services) andColombia’s ICA (tax ontrade and industry),municipal collectionsexceed those from theproperty tax (IBI,impuesto sobre bienesinmuebles). The twocountries’ experienceswith these taxes are alsofurther discussed in thenext section.



Table 6.2: Assignment of Taxes and Fees to Local Governments in Latin America

Type of Tax Type of Fees

Country Property Others

Argentina

Bolivia

Brazil

Chile

Colombia

Costa Rica

Ecuador

El Salvador

Guatemala

Honduras

Jamaica

Mexico

Nicaragua

Panama

Paraguay public utilitiescar registration, games/gambling, wealth tax

(corporate), land transfers/subdivision

urban property (and its increased value

because of public investment)

fines, fees (cattle slaughter)tax on alcohol, economic activity and vehicles tax on unused land (urban/rural)

fines, public utilitiesSales tax (recently eliminated); patents and

business licenses

Urban/Rural property

Varies by state Car registration (all other taxes are

centralized)

Urban property

Parochial revenue fund

public utilities, firefighters, finestax on turnover of industry and trade, extraction

of natural resources (fishing, minerals, oil),

slaughterhouses

urban/rural property (and its increased value


tax on wages, advertising (banners), extraction of

products/economic activity, alcohol

urban property(3)

fines, public utilities, fees for services renderedSpecific taxes for each municipality based on

congress approval such as business taxes on

industrial, trade, and financial activities

fines, utilitiescar tags, permits (business, construction), urban/rural (and its increased value because

public infrastructure investment)

public utilitiesurban property

public utilities, finessurtax on gasoline, tax on industry/commerce,

mineral extraction, slaughterhouse, gambling

urban property (and its increased value


public utilities, fines, permitscar registration, alcoholurban property

fines, public utilitiestax on service sector (ISS), registered goods

tax, real estate transaction tax (2)

urban property (including increased value due

to infrastructure improvement)

car registration, car/property transfers,

slaughterhouse, construction

urban/rural property

public utilities, finescar registration, turnover taxurban/rural property (and its increased value

because of public investment)(1)


a variety of de facto assignments in those twocountries.20

The level of autonomy granted to localgovernments also varies. As summarizedin Table 6.3, most countries use a “closedlist” approach and do not allow the intro-duction of new taxes to local govern-ments; some exceptions include Ecuadorand Uruguay.21 On the other hand, abouttwo-thirds of the countries in the regionallow local governments the ability to setthe rates of some taxes; this practice iswidely accepted as the most desirableform of tax autonomy that can be grantedto local governments. It is interestingthat countries such as Bolivia and Peru,where decentralization reforms haveadvanced rapidly in recent years, stillgrant no discretion to set tax rates. A re-duced number of countries in the regiongrant local discretion to modify tax base.Most countries in the region allow localgovernment discretion in fixing the levelsof fees and user charges for localgovernment services. Nevertheless, animportant restriction on the revenueautonomy of local governments is the

practice by several countries to requirelocal government revenue budgets (“plande arbitrios”) to be previously approvedby a higher tier of government prior tothe start of the fiscal year. Table 6.3shows this is still practiced in Costa Rica,Nicaragua, Panama, and Paraguay; inMexico approval comes at the inter-mediate level from state governments. Interms of fiscal administration (Table 6.3),the general rule is local government isresponsible for the administration of localtaxes, fees and charges, although in somecases, tax administration responsibility isshared with the central authorities.22

In most countries in the region, theyield from property tax remains far belowits potential (Figure 6.2). While onaverage property taxes raise revenuesrepresenting 2.12 percent of GDP in OECDcountries, 0.68 percent in transitioncountries, and 0.60 percent in developingcountries, the average yield in LatinAmerica is only 0.37 percent of GDP.The reasons for low performance aremultiple, including low political will fromnational governments, local governments,


20. In the case of Mexico,the constitution assignsonly the real estate taxto the municipalities.

21. In Argentina, someprovinces may alsoallow their localgovernments tointroduce new taxesbut under quiterestrictive guidelines.

22. As an exception, itappears that in Boliviaall local taxes arecollected andadministered by thecentral authorities.

Table 6.2: Assignment of Taxes and Fees to Local Governments in Latin America (cont.)

Type of Tax Type of Fees

Country Property Others

Peru

Uruguay

Venezuela

Notes:1. Argentina: Not all provinces have delegated property taxes to their municipalities 2. Brazil: The ISS is assessed and collected by the municipality at rates set by the municipality but subject to a maximum fixed by federal law3. Tax collection authority is only given to local governments that prove to have the capacity to collect the tax

fines, fees for servicescar tags, gambling, economic activityurban/rural property

fines, fees for servicescar registration, gambling, showsurban property (and its increased value


public utilities, finescar registration, car transfer,

construction

urban/rural property



Table 6.3: Autonomy Granted in Revenue Assignments to Local Governments andResponsibility for the Collection and Administration of Local Taxes and Fees

Country Ability to Ability to Ability to Control or veto over Responsibility for introduce set tax rates change local govt. budgets by the collection ofnew taxes within legal limits tax base Central/Regional govt. Fees Taxes

Argentina Yes Yes Yes No L L

Bolivia No No No Central C/R C

Brazil Yes Yes Yes No L L

Chile No Yes Yes No L L

Colombia No Yes No No L C

Costa Rica No No No Central PS* PS*

Dominican Republic No No No No C C

Ecuador Yes Yes No No L L

El Salvador Yes Yes No No L L

Guatemala No Yes No No L C/L

Haiti na na na na na C

Honduras No Yes No No L L

Jamaica No Yes Yes Central L C/L

Mexico No No No Regional L L

Nicaragua No Yes Yes Central L L

Panama No Yes No Central L C/L

Paraguay No No No Central L C/L

Peru No No No No L L

Trinidad and Tobago No No No No L C/L

Uruguay Yes Yes No Central L L

Venezuela Yes Yes Yes No L L

Note *: Costa Rica collects, in some municipalities, through the private sector.

L= local, C= central, R=regional, PS= private sector



Parliament and the disincentive effects ofrevenue sharing and inter-governmentalfiscal transfers (IGFTs), not to mentionoutdated and poorly equipped taxadministrations. These factors translateinto generous exemptions and low taxrates, obsolete and infrequent propertyvalue assessments, incomplete registriesand cadastres and a lack of willingnessand means to enforce collections. Thislacklustre performance varies littlewith the different arrangements in theregion for discretion on rate setting oradministering property tax.23

Generally, the range of locally raisedrevenues from own taxes and fees repre-sent a relatively small share of total consoli-dated revenues in the public sector,although in terms of local budget shares,these revenues are relatively large.24 Ofcourse there is a large variation in executionfrom country to country. Figure 6.3 showsthat as percent of national GDP, localgovernments in Brazil raise 8.2 percent,Bolivia 7.7 percent, Colombia 5.2 percent.Ecuador and Peru stay at 3.8 - 3.7 percentfollowed by Guatemala and Chile 2.8 - 2.7percent and then Argentina and Mexico 2.5

23. For a discussion of theissues, see Sepulvedaand Martinez-Vazquez(2009) and De Cesareand Lazo Marín (2008).

24. These two effects arecompatible if we recallour discussion in theprevious section thatlocal governmentbudgets represent arelatively small shareof the generalgovernment budget.

Figure 6.3: Local Revenue as a percentage of GDP and General Government

2,5

7,7 8,2

2,7

5,2

1,8 3,8

1,9 2,8

1,2 0,3

2,1 1,8 3,7

7,3

17,6 18

9,2

22

6,9

22,4

9,3

16,5

4,9

0,9

7,4

2

6,8

17,8

0

5

10

15

20

25

Argen

tina(

2006

)

Bolivia

(200

8)

Brazil

(200

7)

Chile(

2007

)

Colom

bia(2

006)

Costa

Rica

(200

7)

Ecuad

or(2

007)

El Salv

ador

(200

7)

Guate

mala

(200

9)

Hondu

ras(

2004

)

Jam

aica(

2008

)

Mex

ico(2

007)

Panam

a(20

05)

Parag

uay(

2006

)

Peru(

2008

)

Local Government Revenues as % of GDP Local Government revenues as % of General Government

Sources: IMF, Ministries of Finances of Argentina, Bolivia, Guatemala and Peru, UCLG data collection, (cf. Annex 6.2)

Figure 6.2: Average Property Tax Revenue Raised as a Percentage of GDP

OECD Countries2.12

Transitional Countries0.68

Developing Countries0.60

0.37 Latin American Countries


201

- 2.1 respectively. At the lower end stand ElSalvador, Paraguay, Costa Rica, Honduras,and Jamaica (less than 2 percent).25

However, on average, municipalities raisea higher percentage of their budgets fromtheir own revenues similar to Africa, Asiaand a good portion of European countries.As shown in Table 6.4, the percentage oflocal budgets financed out of their owntaxes and fees is quite high, at or above 25percent for many countries.26 However, inBolivia, Brazil, El Salvador, Honduras, andMexico the share is much lower in relationto the other countries in the region.


As a result of limited fiscal autonomy,practically all local governments sufferfrom vertical imbalances, i.e. the expen-diture needs arising from their functionalcompetences exceed their ability to selffinance. Although the existence of verticalimbalance is not in dispute, their actualmeasure is generally a polemical issuebecause practically no country in the re-gion has introduced explicit methodologiesto measure the expenditure needs of localgovernments in a transparent and objec-tive manner. In order to address the exis-

25. See Annex 6.2 for thebreakdown ofrevenues collected byeach tier ofgovernment.

26. See Annex 6.3 in theAppendix for thebreakdown of sourcesfor revenues of localgovernments.


Country (Most recent year) Own taxes and fees Local own taxes as % of local revenues and fees as % of GDP

Argentina(2006) 49.8 1.2

Bolivia(2008) 11.4 2.7

Brazil(2007) 20.1 1.8

Chile(2007) 63.0 0.7

Colombia(2006) 41.2 2.1

Dominican Republic(2006) 58.4 0.7

Ecuador(2007) 34.6 1.6

El Salvador(2007) 13.2 0.3

Haiti(2004) 25.0 0.5

Honduras(2004) 11.1 0.2

Jamaica(2008) 100.0 0.2

Mexico(2007) 15.6 2.4

Nicaragua(2002) 44.0 0.6

Panama(2005) 49.0 0.3

Paraguay(2006) 34.1 1.2

Peru(2008) 43.2 2.6

Trinidad and Tobago(1995) 52.9 0.1

Table 6.4: Shares of Local Own Revenues (in percentages)

Sources: UCLG data collection.



ting vertical imbalances, practically allcountries in the region should implement arange of fiscal transfers, often consistingof different forms of revenue sharing, anarray of specific or conditional grants, andin some cases, equalization grants.

In addition to vertical imbalances, in prac-tically all countries in the region there arealso significant horizontal imbalancesbetween local governments. These imba-lances are the result of the different taxcapacities and economic development oflocal governments, and the different ex-penditure needs arising from disparities inthe service delivery costs and the differingresident populations’ needs arising fromtheir diverse characteristics. Horizontalimbalances are most pronounced betweenurban and rural municipal i t ies andbetween smaller and larger urban centers.As we see later, different approaches areused in the region to address thesehorizontal imbalances.

The emphasis throughout the region hasbeen to address the problem of verticalimbalances through different forms ofrevenue sharing via central governmenttax collections. There has been lessemphasis on the design of explicit equali-zation grants, although, quite often, reve-nue sharing formulas contain equalizationfeatures. Conditional grants are lesscommon in Latin America than in other re-gions of the world, but here again thereare important exceptions.

Most countries in the region use someform of general revenue sharing. The poolof funds to be shared is most frequentlydefined by total central governmentrevenues, although in some casesparticular taxes are excluded from thepool. This is the case for Bolivia,Co lombia , Domin ican Repub l i c , E lSalvador, Guatemala, Honduras, and

Nicaragua. In other cases, the pool isbased on specific central governmenttaxes; for example, 20 percent of oilproduction fees derived by the Mexicanstates must be passed on to their mu-nicipalities; Nicaragua’s additional taxsharing with municipalities is based onrevenues from natural resources; and inPeru, some of the tax sharing is fromportions of the sales tax, and proceedsfrom gas and oil extractions (canon,sobrecanon, and canon petrolero). In thelatter, actual shared revenues are subjectto considerable market fluctuations, forexample, international price levels fornatural resources.

In some cases, shared revenues aredistributed on a derivation (i.e. origin)b a s i s , f o r e x a m p l e , t h e c a n o n ,sobrecanon, and canon petrolero in Peru.This approach (sharing revenues fromnatural resources on a derivation basis)has become a significant factor forregional horizontal fiscal imbalances. Mostoften some sort of formula is used for thedistribution of resources that includesseveral variables, some of which, as notedabove, may have equalizing features.27 Forexample, in Bolivia revenue sharing isbased solely on population; in Ecuador itis according to population and relativepoverty levels; in El Salvador it isaccording to population, “equity” (a fixedamount for each municipality), poverty,and land surface area; in Guatemala it isdistributed according to a formula thatincludes equal shares (fixed amounts),population, number of settlements, andper capita income; in Honduras it isaccording to populat ion and equalamounts for a l l munic ipa l i t ies; inNicaragua it is according to population andseveral other criteria; and in Peru it isaccording to population and infantmortality rates. Frequently, the formulasare also employed by central governments

27. In some cases, like inParaguay, theallocation of funds isstill ad-hoc at thediscretion of thecentral authorities.



Figure 6.4: Composition of Local Government Revenue

to pursue several objectives other thanequalization. For example, in Ecuador thesharing formula includes elements forrewarding administrative effort andachieving goals in the nationaldevelopment plan, while in Nicaragua theformula provides incentives for increasingrevenues from property tax and forachieving more effective budget execution.

Some countries allow unconditional use ofshared revenue, including Bolivia, Ecuador,El Salvador, and Honduras. In other casesthe use of funds is conditional; Colombiauses revenue sharing funds earmarkedfor basic education, health, and water andsewerage; Guatemala, for education, healthand infrastructure; while in Nicaragua andParaguay, a share of the funds –80 percentin the latter case– must be spent oninfrastructure investment.

Revenue sharing practices in the federalcountries also have different features. InArgentina, tax sharing with local govern-ments is carried out exclusively by theprovincial governments, which can decidehow to distribute their share of federal VATand income taxes. The Brazilian states alsohave a tax sharing system funded with 25percent of their regional VAT revenues,which distributed 75 percent on a derivationbasis according to value added in themunicipalities, and 25 percent by a formulabased on population, land area, and othervariables. This same formula is used todistribute federal tax sharing with thestates (cooperation funds) to the munici-palities. In Mexico, the states are requiredto distribute to their municipalities at least20 percent of the income that they receivefrom revenue sharing in the federal funds(Fondo de Fiscalización and Fondo General

Source: UCLG data collection, (cf. Annex 6.6)



de participaciones). Mexico also has afederal grant, amounting to 1 percent offederal collections (Fondo de FomentoMunicipal) that is distributed on the basisof municipal revenue collections.

Conditional or specific transfers are lessextensively used in Latin American than inother regions of the world,28 neverthelesstheir use is increasing, especially in thosecountries where central governmentscount on being associated with localgovernments as partners for the deliveryof certain services and the implementationof national programs. For example, Boliviahas introduced a conditional health trans-fer for a national program in support ofinfants and mothers (seguro materno in-fantil). In Brazil, several conditional grantshave been introduced for public transport(funded by the sharing of federal fuellevies), basic education, and healthservices, including hospitals from thenational health system. In Chile severalhighly conditional grants have for manyyears funded local governments’ activitiesin education, health, and other socialprograms. Some conditional grants areearmarked for certain geographical areasthat are deemed to be lagging behind. Forexample, in Ecuador there is a conditionalcapital investment grant for the Amazonregion.29

A particular subgroup of conditional grantsis earmarked for investment in local infra-structure. For example, El Salvador offersgrants for municipal capital infrastructurebased on the presentation of projectproposals. In Guatemala one-eighth ofVAT revenues go to infrastructure in socialand basic services, while a share of vehicletaxes is earmarked for maintenance ofroads and drainage. In Mexico, at least 20percent of the investment grants (Fondode Compensación) from the federalgovernment must be assigned to the

poorest ten states in the country and usedby the municipalities of those states.

The practice of explicitly addressing hori-zontal disparities among local governmentsthrough equalization transfers is still notcommon but it is taking hold in the region.One reason for the slow introduction ofexplicit equalization grants is that oftenrevenue sharing schemes do incorporatesome equalization elements in theirallocation formulas. Several examples ofexisting equalization grants (above andbeyond revenue sharing schemes with someequalizing elements in their formulas) areworth mentioning. One is Bolivia’s HIPC(Heavily Indebted Poor Countries Initiative)transfers started in 1997 with funds frominternational organizations (the World Bankand the IMF) that is distributed by thecentral government to local governmentsusing a formula based on the poverty leveland population of municipalities. In Brazil,there is a federal equalization transfer to themunicipalities funded with a share of federalVAT and income tax revenues; the fund issplit into two parts, with 10 percent going tostate capital municipalities (distributedaccording to population and the inverse ofper capita income) and the other 90 percentto the rest of the municipalities (distributedaccording to an index that favours munici-palities with smaller populations). Aninteresting approach is that of Chile, wherethe formula driven equalization grant (theCommon Municipal Fund) is funded by themunicipalities’ own revenues from differentsources in what is known in the technicalparlance as a “fraternal” (or Robin Hood)system, in which the relatively richermunicipalities finance the transferredamounts to the poorer municipalities. Theallocation formula includes population size,poverty levels, exempted real estateproperty, and past revenue collections. Onekey positive feature in all these examples isthe recognition of the need to introduce a

28. Here we are referringin a conventional wayto specific fundsassigned to particularobjectives andadministeredseparately by centralgovernment agencies.This is interpreted asbeing different fromthe conditioning orearmarking of revenuesharing funds. As wehave seen above, anumber of countries inthe region conditionthe use of revenuesharing funds toinvestment ininfrastructure and soon. If the restrictionsin the use of revenuesharing funds wereincluded in the generalcategory of conditionaltransfers then thepractices in theregions would not bethat different fromthose in other regionsof the world.

29. This is a specificinstance of largerdevelopmental goalsthat can be addressedby conditionaltransfers, such as aridareas, poor areas,unexploited highpotential areas, etc.


205

separate instrument (equalization grants) toaddress the separate objective of horizontalinequalities arising from differentexpenditure needs and fiscal capacity.30 Akey common challenge ahead is the need toimprove the methodologies used to quantifythe expenditure needs and fiscal capacity ofthe different local governments. 31

Borrowing

Given their expenditure responsibilities,most municipalities have a long-termneed to finance capital infrastructure.

Local borrowing can be considered alegitimate, efficient, and equitable sourcefor financing this local infrastructure.However, it is also widely accepted thatthe local borrowing process must besubject to explicit rules and limitations inorder to ensure fiscally responsiblebehavior by local officials and to guaranteemacroeconomic stability in the country.32

Commonly applied rules include thoseabout expected behavior, such as the“golden rule” that long-term borrowedfunds must be used for capital infra-structure only, and not for recurrent

30. The use of a fraternalsystem to fund theequalization transfersin Chile is a promisinginnovation. This is acommon system to anumber of Europeancountries but it isuncertain how easily itwill be adopted byother countries in theregion.

31. For the availablemethodologies used inother regions of theworld see, forexample, Martinez-Vazquez and Searle(2007).

32. Historically this wasn’talways widelyaccepted in somecountries in the region,which in past decadessaw an accumulationof macroeconomicdifficulties associatedwith unfettered sub-national borrowing insome of thefederations andpoliticizedgovernment-runmunicipaldevelopment banks.


Country Access to Financial Markets Municipal Bank Limitations

Argentina Y Y Y

Bolivia Y N Y

Brazil Y N Y

Chile N N n.a

Colombia Y n.a Y

Costa Rica Y n.a n.a

Dominican Republic N Y Y

Ecuador N Y Y

El Salvador N Y Y

Guatemala Y n.a Y

Haiti n.a n.a n.a

Honduras Y Y Y

Jamaica N N Y

Mexico Y N Y

Nicaragua Y n.a Y

Panama n.a n.a n.a

Paraguay Y N n.a

Peru Y N Y

Trinidad and Tobago N N n.a

Uruguay Y N Y

Venezuela Y Y na

Table 6.5: Authority to Borrow by Local Governments in Latin America

Source: UCLG data collection.



expenditures, and different quantitativebudgetary limits on borrowing. Among thelatter, there are rules on non-negativecurrent budget balances, limits on thelevel of total debt and debt-servicepayments as a percent of budgetrevenues, as well as restrictions onborrowing abroad.

For a variety of reasons, ranging fromfiscal conservatism to negative past expe-rience, not all countries in the world allowtheir local governments to borrow. Yet inthe case of Latin America, as shown inTable 6.5, most countries do allow localgovernments to borrow.33 Practically allcountries allow such borrowing eventhough they impose rules and limitationson local borrowing similar to theabove-mentioned international “goodpractice” guidelines. In most cases,foreign borrowing is not allowed, insome cases it is allowed withpermission of the higher authorities, andin other cases even domestic borrowingrequires administrative approval by higherlevel authorities.34 Over time, nationalsystems have adapted to idiosyncrasies.For example, in Nicaragua, municipalitiesare able to contract short or medium termloans from public and private banks forpublic works, with long-term loans forlarge-scale public works approved by theNational Assembly. Loans must be repaidwithin the term of the elected officials;mayors and municipal councils may notleave debts to their successors, except forlong-term loans approved by the NationalAssembly. In Colombia, law 358 from 1997introduced a system of “semáforos”(traffic light) restricting the level of localdebt according to the ability to pay by thelocal units; if interest payments are below40 percent of the operational surplus andif the debt level is under 80 percent ofcurrent revenues, local governments arefree to borrow according to the law;

however, they require permission fromthe Ministry of Finance if any of thoselimits is exceeded. With law 819, whichcame into effect in 2003, the need to havea primary surplus sufficient to cover ongoing debt service was added to theexisting indicators. The three indicatorsmust be positive in every year of the loan,and this must be reflected in the mediumterm fiscal framework of the municipalityor department. In El Salvador, munici-palities can borrow from commercial banksonce they receive the proper qualityranking from the Ministry of Finance andthe semi-official organization charged withthe physical distribution of the generaltransfer funds. The municipalities thenestablish an intercept agreement for thosetransfers to work as collateral for the loansfrom the commercial banks. As in othercountries around the world, it is commonto impose limits on annual debt service aspercent of revenues (for example 20percent in Argentina and Bolivia or 40percent in Ecuador) and/or the total stockof debt as percent of total revenues (120percent in Brazil or under 100 percent inEcuador and Peru).

The actual amount of sub-national debt,which includes local and provincial/statedebt, is quite low in most countries, withthe exception of Brazil and Argentina,where sub-national debt represents inrecent times between 10 and 15 percentof GDP; Mexico, Colombia and Boliviacome behind with sub-national govern-ment debt representing less than 2 per-cent of GDP as of 2007. However, formunicipal governments alone in recentyears, Bolivia is first in debt service(interest and repayment of principal) ataround 9 percent of total municipalexpenditures, followed by Ecuador ataround 8 percent, and Argentinean andBrazilian municipalities, where debt servicestands at around 4 percent.35

33. The exceptions includeChile, DominicanRepublic, El Salvador,Ecuador, Jamaica andTrinidad and Tobago.

34. For example, foreignborrowing by themunicipalities inArgentina requiresadministrativeapproval at theprovincial level and bythe Ministry ofEconomy at thenational level, which ithas been argued hasbeen subject topolitical criteriabeyond technicalaspects.

35. See Porto (2009).


207

Budgeting

The budgeting process at the local level inmost Latin American countries is still carriedout along traditional lines with heavyemphasis on incremental budgeting andex-ante financial audit controls for thedisbursement of funds. Much less attentionis given to the planning of expenditureprograms and ex-post evaluation of theeffectiveness of funds disbursed onprogrammatic goals.36 One positive aspectwithout exception is, local budgets need tobe approved by democratically elected localcouncils. However, as we have seen above,in a significant number of countries (Bolivia,Costa Rica, Jamaica, Mexico, Nicaragua,Panama, Paraguay) at least somecomponents of the local budgets need to beapproved ex-ante by higher level authoritiesat the central or regional levels. It isquestionable whether or not theseapprovals are really needed; the bestpractice internationally is to rely onhorizontal accountability mechanismsex-ante, and on the ex-post audit and togrant full budgetary autonomy to localgovernments.37

“Participatory budgeting” is an area ofinnovation in the region that has attractedmuch international attention. The specificmeaning of this term varies among countriesintroducing this type of reform, but generallymeans additional mechanisms for citizens toinfluence local budgetary decisions beyondthe conventional vehicle of democraticelections for municipal councilors.38 Forexample, in Bolivia a 1994 lawestablished citizen committees (comités devigilancia) and community-based organizations(OTBs—organizaciones de base), that aresocial organizations of peasant communities,the indigenous population, and neighborhoodgroups. Citizen participation at the local levelis also important in Brazil, but variesconsiderably across states and municipalities.

One experiment involves groups of citizensempowered to address social andpolitical inequalities by influencing theallocation of budget resources throughneighbourhood meetings.39 Even someprovinces in Argentina, have formallyadopted participatory budgeting.40 On theother hand, participatory budgeting isgenerally appropriate for only certainelements of the budget; thus, even in PortoAlegre (Brazil) the share of the budgetsubject to this process is limited (see Box6.6). Nevertheless, the implementation ofparticipative budgeting often depends onthe will of the Mayor and the City Councilsince it is not a compulsory or permanenttool.

As for the composition of municipalbudgets, our discussion is based on asmall number of countries where dataavailability varies. In terms of the econo-mic classification of local expenditures,the high share of capital infrastructure ex-penditures in the municipal budgets of asignificant number of countries is notable(Figure 6.5). For example, in both Chileand Ecuador, local governments spend 55percent of their budgets on capitalexpenditures; in Guatemala this figure is64 percent and in Peru, 58 percent. Ofcourse, there are large variations for thesefigures across countries, including theassignment of expenditure responsibi-lities—what tier of government is respon-sible for capital infrastructure in thedifferent areas of responsibility: schools,roads, etc –and the legal restrictionsimposed on local governments for how tospend revenue sharing and other types offunds— as in Peru, where local govern-ments only can spend funds from thecanon and sobrecanon from naturalresources on capital investments. There isalso the possibility that capital expen-ditures are over reported.41 Whatever theexplanation, it is clear that many local


36. The strong emphasison ex-ante treasurycontrols has not, on theother hand, reducedcorruption, which, withsome exceptions, stillappears to be extensiveamong localgovernments (and therest of the publicsector) in the region.The difficulties lie morein poor execution of theex-ante controls than ina deficient design.

37. In some cases controlsfor checking the legalityof actions, as opposedto changing budgetallocation decisions,can be justified whenhorizontalaccountability and auditmechanisms aredeficient or not presentat all.

38. Somewhat related,citizen participationmechanisms such asreferendums, “popularinitiatives,” and electedrepresentative recallshave been operating inother regions of theworld.

39. See Afonso (2006).

40. For a positiveassessment in the caseof Bolivia, see Faguet(2004). See alsoGoldfrank (2006) andShah (2007) forgeneral assessments.

41. There are incentives inmany countries in theregion to report somecurrent expenditure ascapital expenditure. Forexample, in some casescentral legislationrestricts the share ofbudgets that can go torecurrent purposes. Inother cases, practicallyall kinds of currentexpenditures havebeen demonized asbeing inefficient so localauthorities actuallyreport some currentexpenditure as actuallybeing some form ofcapital expenditure.But, fortunately theredoes not appear to be aproblem in LatinAmerica with the off-budget programs andexpenditures that arecommon in otherregions of the world, forexample Africa andAsia.



governments in Latin America have beengiven an important role to play in the taskof providing much needed infrastructure.This highlights the importance of findingmore stable and potent instruments forinfrastructure finance.

As we have seen, the movement towardopen and participatory budgeting isspreading, increasing budget efficiency andaccountability in general. Although parti-cipatory budgeting is not directly aboutdecentralization itself, the movementtoward participatory budgeting has tendedto reinforce decentralized institutions. Thereare, however, exceptions; for example, inthe Dominican Republic, Law 176 of 2007goes a long way to introduce participatorybudgeting but the actual level of decentra-lization to local governments in that countryremains quite weak.

Even less data are available to obtain apanoramic view of the functionalclassification of local budgets in theregion; for countries where individualmunicipalities’ data are available,comparisons are hard because of thedifferent classification methods used ineach country. As shown in Figure 6.6,education expenditures, deriving from theassignment of expenditure competencies,are important items in the local budgets ofBolivia, Brazil, Chile and Colombia. For thesame reason, expenditures on healthservices are relatively important in thelocal budgets of Brazil, Colombia, andPeru, with budget shares here rangingbetween 16 and 22 percent. It is note-worthy that in most of these countries forwhich disaggregated data are available,not surprisingly “general administration” isthe most important expenditure item interms of budget shares.

Figure 6.5: Budget Expenditure by Economic Classification of Local Governments

0%

20%

40%

60%

80%

100%

Argen

tina(

2006

)

Bolivia

(200

8)

Brazil

(200

7)

Chile(

2007

)

Colom

bia(2

006)

Ecuad

or(2

007)

El Salv

ador

(200

7)

Guate

mala

(200

2)

Hondu

ras(

2004

)

Mex

ico(2

007)

Nicara

gua(

2002

)

Panam

a(20

05)

Parag

uay(

2006

)

Peru(

2008

)

Wages, Salaries and Pension fund Current Expenditure Capital Expenditure/Debt/Equipment

Source: UCLG data collection (cf. Annex 6.4)


209

Special Issues, Constraints andOpportunities for Local and Intergovernmental Finance in the Region

As shown in the introductory section, theLatin American region offers a vast arrayof different experiences and approaches tolocal finance. The kinds of issues facinglocal governments in large federal countriessuch as Argentina, Brazil, and Mexico areoften very different from those affectinglocal governments in small countries suchas El Salvador, Nicaragua, or Paraguay.Furthermore, there is also significant di-versity among the large federal countriesas well as among the unitary countries.Hence, the attempt to generalize analysisof local problems and their solutions isneither always possible nor desirable.Nevertheless, some themes common to asignificant number of countries in the re-gion clearly emerge from the description

of the local finance system presented inSection 1. In this section we identify someof the special issues, constraints andopportunities for the development of localfinance in Latin America. It is organizedaround four major themes: (a) Or-ganizational Structure; (b) Intergovern-mental Fiscal System Design; (c) BudgetProcess and Transparency; and (d) ShortTerm and Long Term Structural Challenges.

Issues on Organizational Structure

Fragmentation and sub-optimal scale42

In many countries there is an ongoingdebate on the number and size of localgovernments related to the issues ofeconomies of scales to deliver publicservices, which typically improve with size,and citizen representation, which generallydeteriorates with size. Citizens that feelmarginalized frequently call for the

42. The issues of localfragmentation andadministrativecapacity have asignificant politicalcomponent and fromthat perspective theywere reviewed inGOLD I. Our mainperspective here is onhow fragmentationand administrativecapacity may affectthe fiscal sustainabilityof local governments.


Figure 6.6: Shares of Local Government Expenditure by Functional Classification

0%

20%

40%

60%

80%

100%

Argen

tina(

2006

)

Bolivia

(200

8)

Brazil

(200

7)

Chile(

2007

)

Colom

bia(2

006)

Peru(

2008

)

General administration Education Health Sanitation Transport Others

Source: UCLG data collection (cf. Annex 6.5)



creation of new municipalities. As the currentlegal frameworks guarantee a minimumamount of funds to each municipalityregardless of size, this has promoted thecreation of new municipalities.43 In responseto this problem some countries haveintroduced legislation requiring minimumpopulation size in order to ensure the futurefiscal viability of any new local government.This action can be effective in slowing downthe process of further fragmentation but itdoes not help to address the inadequatescale of the already existing municipalities.

Perhaps a more attractive and potentiallyequally effective approach is the promotionand creation of associations of municipali-ties into mancomunidades for the deliveryof certain public services requiring certainminimum scale.44 This is an approach stilllargely unexploited but it is currently beingdeveloped in some countries, especially insome provinces of Argentina, southernBrazil, Ecuador, and Peru.45

The trade-off between economies of scaleand representation46

The issue of an optimal scale of localgovernments presents an inherent tradeoffbetween the (potential) better politicalconnection in terms of representation andaccountability of smaller jurisdictions withthe (potential) greater fiscal viability oflarger jurisdictions (in particular with regardto population size).

The essence of this trade-off between thegreater efficiency of smaller governmentsthat can better match the preferences andneeds of local residents in their expenditureallocation and economies of scale inproduction with lower costs associatedwith larger governments impl ies acompromise solution between the twoobjectives. In particular, it implies that lowercost effectiveness in the delivery of public

services may be offset by greater efficiencyin responding to the needs and preferencesof local residents.

Local administrative capacity

The problem with local governments’admin i s t ra t i ve capac i ty i s c lose lyassociated with their small size (in number ofinhabitants). Central governments (or stateor provincial governments in the case offederal countries) spend little time andresources in developing the capacity oflocal governments.47 Some of the slack hasbeen taken up by local governmentassociations, which, for example in some ofthe Andean countries and in Central America,provide their members with assistance andtraining. However, these associations oftenlack the resources necessary to address thisissue. Central (or provincial–state)governments can do much more.

To summarize, there are no exact answersor methodologies to address the issue ofoptimal size for municipalities. Severalgoals need to be pursued including costefficiency and representation andaccountability and several constraintsneed to be met regarding fiscal viabilityand administrative capacity. See Box 6.1.

Issues on Intergovernmental FiscalSystem Design

Lack of clarity in the assignment ofexpenditure responsibilities to localgovernments

One of the weakest points of manydecentralization programs in Latin Americahas been the scant attention given to clearassignment of expenditure responsibilitiesof sub-national governments, which is acrucial first step in the design of anysystem of intergovernmental fiscalrelations. Instead, the focus has been

43. For example, in ElSalvador, the Fund forEconomic and SocialDevelopment (FODES),which is the main sourceof local revenues, isdistributed according to aformula that distributes25 percent of the fundsevenly to allmunicipalities.

44. However, theseprograms can be difficultto implement. Forexample, in El Salvador,the National Plan forTerritorial Developmentand Organization(PNODT) was supposedto promote mancomuni-dades and generalcooperation among localgovernments whichwould allow them tolower administrativecosts by workingtogether to print jointlyneeded forms andgathering regularly toshare ideas. However,the results are verylimited.

45. Other approaches todealing with the problemof inadequate scaleinclude the contractingout to private companiesof some services, so thatprivate companies canbenefit from sufficientscale by supplyingdifferent municipalities,or the creation of ‘sectorspecific’ service govern-ments or districts. Theprivatization of services isbeing used in severalLatin American countries.See Martinez-Vazquezand Gomez (2008) for adiscussion of the issuesand solutions.

46. See Martinez-Vazquezand Gomez (2008) andImansyah and Martinez-Vazquez (2009) for amore extensive dis-cussion of this issue.

47. The Latin America regionshows a lower prevalencefor Ministries of LocalGovernment/Inte-rior/Home Affairs that aremore prevalent in otherregions of the world andhave a higher presenceof special agenciesdesigned to exclusivelyaddress the needs oflocal governments, suchas ISEDM y FISDL in ElSalvador or INIFOM inNicaragua. In general, itis easier for these typesof specialized agencies toplay a supporting andcapacity developmentrole than it is for sectoralministries; the downsideis that the specializedagencies tend to pull alower rank within thegovernmentadministrations in thebargain for additionalresources.


211

almost exclusively on putting in place someform of financing scheme.48 At present,there continues to be too much emphasisin many countries of the region on thedecentralization processes simplyunderstood “as the provision of some ofrevenue sharing and transfers” to localgovernments, ignoring the fundamental rulethat “finance should follow function.”

On closer analysis, the assignment offunctional responsibilities remains in manycases too general and vague. For examplein El Salvador, the municipal code givesauthority to municipalities to perform a listof responsibilities that clearly overlap withthose also assigned to centralgovernment; the same is true in Uruguay.In other cases, the vagueness lies in the

48. See Bahl and Martínez-Vázquez (2006) for adiscussion of the propersequencing ofdecentralizationreforms.


Box 6.1. Four Possible Criteria to Consider for the Minimum Size of Municipalities

Although the desirable minimum size of municipalities is a complex issue that must beaddressed in the historical and political contexts of each country, there are four basiccriteria that can be followed as guideline for policy decisions:

(1) The production/cost efficiency/population criterion. The international experienceshows that unit costs for some public services (such as water or public transportation)can be much affected by scale. Depending on the assignment of expenditureresponsibilities, in order to arrive at the lowest cost of production it is required to reacha minimum size in terms of population. Yet we must note that there are other ways toprovide services in a cost efficient manner, including buying the services from a largerlocal government, creating an association between several smaller local governments toproduce the service, or even buying the services from a large privatized producer. Whatthis means is that this criterion of a minimum population size should be administeredintelligently with flexibility to allow for these other service delivery possibilities.

(2) The representation/political responsiveness/accountability criterion. The generalpresumption is that smaller local governments will generally tend to be morerepresentative and accountable to the residing population. But consideration must alsobe given to the fractionalization of the population and adequate representation of theminority groups’ interest. Population density should also be considered. From anaccountability and representative focus it would seem that a simple but useful rule ofthumb will be the time required to travel to the location of the municipality building.

(3) The financial/fiscal capacity criterion. It seems reasonable to require that any newlocal government have a minimum level of economic capacity to self finance some of itsservice needs. Measuring this capacity is not always an easy matter, but it shouldinvolve some approximation to the “expenditure needs” and the “fiscal capacity” of eachpotential municipality and the setting of some threshold for the difference betweenneeds and capacity.

(4) The administrative capacity criterion. This can be measured in a number of differentways, but fundamentally qualified personnel available to run the business of the localgovernment efficiently is required.



way that legislation is implemented (or notimplemented). For example, Law 66 of1997 in the Dominican Republic addedprimary and secondary education to theresponsibilities of local governments; yetreal power and decision-making remainwith the regional Education and Culturecommittees, which may be considered adeconcentrated tier of the centralgovernment.49

Beyond the operational inefficienciesassociated with the unclear assignmentsof functions, an important consequence isthe ambiguity in political accountabilitythis situation introduces. The governanceimplications for attaining the purportedbenefits of decentralization are quiteconsiderable.

Another problem is that the assignment ofexpenditure responsibilities in practicallyall Latin American countries is uniform forall local governments, regardless of theirsize and administrative capacity. Asaforementioned, a good way to addressthese shortcomings is the creation ofassociations of local governments intomancomunidades. Alternatively, theremay be room for asymmetric assignmentsfor municipalities of different size andadministrative capacity.

Nevertheless, the highly asymmetricassignment of expenditure responsibilitiescan lead to confusion. For example, inEcuador, the Constitution (Article 226)establishes the obligation of centralgovernment to transfer functional compe-tencies at the discretion (by voluntaryrequest) of sub-national governments.This means that any sub-nationalgovernment can request a full or partialcompetence in a particular area at theirdiscretion, leading to great heterogeneityin central-local relations, thereby compro-mising the overall effectiveness of inter-

governmental coordination. A betterpractice could be to design two or at mostthree different packages of expenditureresponsibilities that can be devolved tolocal governments depending on theiradministrative capacity. However, animportant issue with asymmetricapproaches is the need for using verifiablecriteria, that is, differential assignmentsmust be grounded in something other thanpolitical connections.

Another factor contributing to confusion inthe assignment of expenditure res-ponsibilities is the common practice ofunfunded mandates. Frequently, line mi-nistries may partially decentralize certaincompetencies to local governments with-out providing the required resources toimplement them properly. There may alsobe increased reporting requirements onlocal governments without adequatecoordination among central governmentagencies or the provision of the technicaland financial means to make thatreporting possible.

A workable system of expenditure assign-ments, no matter how specific, is develo-ped in the laws and regulations, whichalways requires coordination and effectivedialog between the different levels ofgovernments. Because of the larger num-ber of municipalities, it is important thatthe voice of these local governments berepresented by associations of municipali-ties. On this front there has been consider-able progress since practically allcountries in the region have developedeffective municipal associations.50 How-ever, central authorities have not alwaysrecognized these associations as strategicpartners in improving policy design inexpenditure assignments and thestrengthening of other components of thefiscal decentralization systems.

49. Although the trend inthe region has beentoward moredevolution ofresponsibilities to localgovernments, thereare exceptions. Forexample, Jamaica hasbeen recentralizingfunctions previouslyassigned to the localgovernments (ParishCouncils) through thecreation of nationalentities under thetutelage of the Ministryof Local Government,which is in charge offire protectionservices, parks andmarkets.

50. See, for example, thediscussion in Porto(2009).


213

One final issue in the practice of expen-diture assignments is the practically uni-versal lack of effective and transparentmethodologies to translate the assignmentof functional responsibilities into expendi-ture needs. Sometimes, historical costs(or levels of expenditures prior to decen-tralization) are used as a guide. Differentmethodologies are used in other countriesaround the world, such as the use per-clientspending norms or bottom-up costestimates that could be successfullyimplemented in the region. The advantageof having some effective method tocalculate expenditure needs is that thediscussions on and ultimate design ofrevenue assignments, whether throughown revenue or fiscal transfers, becomesmore informed and rational.

Insufficient revenue autonomy

The level of tax revenue autonomy of localgovernments differs quite significantlyacross Latin American countries. Countriessuch as Brazil and Chile have relativelyhigh autonomy and countries such as ElSalvador, Mexico and Peru have signi-ficantly less. But, in general, as it occurs insome other regions of the world (Africa,Asia and many European countries), localrevenue autonomy in Latin America re-mains below what is desirable.

The lion’s share of financing for localgovernments in the region continues tocome from different forms of centralgovernment transfers including revenuesharing. Transfers have experienced anincrease as the most often used form ofnewly devolved financing responsibility forlocal governments. With this in mind,some important and correct policies havebeen adopted for various countries in theregion. For example, many countries inLatin America have taken steps to increasetheir share of own taxes in local budgets.

Most have now assigned the property taxto the local level, which is excellentbecause there are many features thatmake it an ideal tax.51 Unfortunately, theproperty tax remains highly underutilizedfor a variety of reasons.52 Several othertaxes have been assigned to localgovernments, including vehicle tax,betterment levies, and different forms ofbusiness taxes and licenses. Thesepositive measures should be imitated bycountries that still allow little local taxautonomy. Another practice to emulate isthe allowance of a certain degree ofdiscretion for municipalities to set therates of their local taxes, between somemaximum and minimum approved levels.

Nevertheless, it is difficult to make astrong case for policy design that allowsfor a greater degree of tax autonomywhen there is a perception that many localgovernments in the region do not makeeffective use of the tax autonomy lawgranted to them. This is most clear in thecase of the property tax for which actualrevenues collected are a small fraction ofthe revenue potential.53 So the realizationof more revenue autonomy for localgovernments may need to be accompaniedby a significant improvement in local taxeffort. However, it is important to notethat low tax effort (known in the region as“pereza fiscal”) is a complex problem. First,there is often confusion between low taxeffort (“pereza fiscal”) and low tax capacityor economic/fiscal poverty of jurisdictions.Establishing the presence of “pereza fiscal”requires a comparison between actualtax collections and potential tax collectionsof every particular jurisdiction; this is acomplex task in many cases. Second, oncethe presence of “pereza fiscal” can beestablished, it is important to understandits multiple roots, from simple politicaleconomy issues (local officials may simplybe happy spending funds but never raising

51. The property tax ishighly visible andbecause of the lowgeographical mobilityof its base andbecause propertyvalues tend to reflectwell in general thequality of localservices, the propertytax can approximatewell the concept of abenefit tax, whereresidents pay for theservices they receive(see Sepulveda andMartinez-Vazquez,2009). All of this islikely to increasepolitical accountabilityof local officials. Theproperty tax may alsohave relatively lowefficiency lossescompared to otherlocal taxes. In terms ofadministration, therecan be flexibility intaking advantage of amixed local and centraladministration andenforcementapproaches See Bahl,Martinez-Vazquez andYoungman (2008 and2010).

52. See Sepulveda andMartinez-Vazquez(2009) for anevaluation of theperformance ofproperty taxes in theregion.

53. See Sepulveda andMartinez-Vazquez(2009).




54. See Acosta (2005).

Box 6.2. A Tale of Two Cities: Bogotá’s Success Story in Raising Local TaxEffort and Lima’s Success with a New Approach to Tax Administration

Bogotá provides an example of a local government that has had success in raising localtax effort (the city was awarded a prize by the United Nations in 2002 for being the mostimproved local government in the world.) Starting in the late 1980s, The Municipality ofBogotá began a program of civic education during which it emphasized the importanceof paying taxes and the accompanying benefit for citizens that derives from a strongerlocal government. It significantly increased property tax collections through a series ofadministrative improvements, including enforced business taxes, privatizing certaingovernment run organizations, and successfully issuing own bonds, some times inforeign markets receiving a AAA rating,54 Particularly noteworthy is that the Municipalityof Bogotá in 2006 raised 42 percent of its tax revenues from the local business tax(Impuesto de Industria y Comercio, ICA). Updating the fiscal cadastre in 2009 has alsoproduced significant increases in revenues from the property tax. It is expected that theassessed tax base for this levy will increase by more than 50 percent by 2010, bringingan additional $51 million in revenues or a 13.3 percent increase over collections beforethe updating of the cadastre.

In 1996 Lima, along with other provincial Peruvian municipalities, introduced a semi-autonomous Tax Administration Service (SAT in Spanish), with the goal of increasingcollections for own local taxes. This followed the model of a national-level SAT. The SATof Lima is autonomous in its financial and human resource management and it isfinanced through a share of the taxes and fees commission it collects. The sharedcollections by Peruvian municipalities range from 3 percent to 10 percent. But note thatthe local authorities are still responsible for regulating and controlling the SATs’ work.There have been some clear benefits for those Peruvian municipalities that, like Lima,adopted a SAT approach. For example, between 1998 and 2007 those municipalitiesthat adopted a SAT increased their own revenue by 80.9 percent, or 9 percent of theannual average, by comparison over the same period the municipalities that did notadopt a SAT saw their revenues increase by 61.2 percent, or 6.8 percent of the annualaverage. The empirical studies show that the trust in tax administration in Lima andother municipalities where a SAT was adopted has increased. This could be attributed tolower political intervention in administrative processes, higher client focusmanagement, improved public relations, and a reduction of corrupt practices. But not allhave been highly regarded for their new local tax administration. The same empiricalsurvey studies identify some issues associated with the SATs: such as a limited linkbetween the revenue collection and public services, and the public perception of taxadministration as “insensible.” But some of this is to be expected since the SATs havegone against the conventions and took advantage of not always well defined rules,especially in the SAT of Lima. One of the key characteristics of the SAT agencies hasbeen their innovative drive, including internal processes, the use of moderntechnologies, human resource development, improved financial management, and thecollaboration across tax administrations.


215

them), to lack of economic resources(building an updated fiscal cadastre isexpensive), to inadequate methodologiesfor evaluation, to the lack of skilled humanresources, or even to the negative in-centives for local tax effort provided by thecentral government’s existing fiscaltransfer system. However, the regionoffers success stories in raising local taxeffort, as provided by the recent expe-riences of Bogotá and Lima (see Box 6.2).

A third issue is the need to explore taxeswith significant revenue potential thatcould possibly be assigned to localgovernments in order to increase theirrevenue self-sufficiency, akin to a flat-ratepiggy back personal income tax or localsurtaxes on some excise duties, such asthose on vehicle fuel. Another possibility is

a more intense use of betterment levies,which can complement annual real pro-perty taxes. Betterment levies are beingused quite successfully in Colombia (SeeBox 6.3). A different option is the adoptionof some form of final retail tax such as inthe case of Brazil’s ISS (Tax on Services,as discussed in Box 6.4.) Except that thistype of tax, although fine within theBrazilian tax system where the federal andstate VAT levies exclude many importantservices from their bases, may beproblematic because it would overlap withother countries’ existing VATs.55

An alternative to the ISS that would notpresent potential conflicts with theexisting national VATs, is the broader baselocal business tax (Impuesto de Industriay Comercio, ICA) in Colombia. This is a

55. For example, in therecent past Nicaraguaeliminated aproductive local salestax as part of a policyconditionality given bythe InternationalMonetary Fundprecisely because ofthe conflict presentedby the existingnational VAT.

56. I CongresoLatinoamericano deValorizacion;http://www.lonjadebogota.org.co/Portals/0/Docs/

57. Information about thistax may be found at:http://www.shd.gov.co/portal/page/portal/portal_internet/impuestos/impuestos_imp/Plus valia/INFO%20PLUSVALIA


Box 6.3. Betterment Levies (“Contribución de Valorización”) in Colombia

In general terms, a betterment tax recoups some of the benefits accrued by propertyowners due to adjacent public investment that increases the value of that property(Bird & Slack 2006). Since most real estate property is significantly affected bypublic facilities surrounding it, this tax has significant revenue potential. InColombia, this tax receives the name Contribución de Valorización and it has been inoperation for a long time. The constitution gives municipalities and other publicentities the right to a share on the added value produced by investments made inurban settings (e.g. infrastructure works). The tax is being looked at with interest byother countries and in the first Latin American conference of Valorization, held inBogotá (Colombia) in 2009.56 The levying of the tax implies a series of stepsincluding, the determination of the costs and benefits of the project, thegeographical area that is expected to benefit, and a method to distribute the costsand benefits of the project among the different properties. This distribution can usean array of “benefit factors” (use of property, closeness, access, etc.) or simply aland area, linear size of lot front, etc.

Bogotá also levies a tax called Participación en Plusvalia57 defined as the contributionowed to owners of real estate property as a result of modifications that increase thevalue of property. This is similar to the betterment tax except that it captures onlychanges in urban codes that affect the ways the property can be used or the intensity ofits use (how much can be constructed) that may increase its value.



58. See, among others,Deloitte (2010),Government of Brazil(2009), KPMG (2006),Banco Central doBrasil (2000) andPurohit (1997).

59. See KPMG (2006).

Box 6.4. The ISS in Brazil58

The ISS (Imposto sobre Servicos/Tax on Services) is a municipal level tax levied onthose services that are left out of Brazil’s state value added tax (ICMS). The servicesthat may be taxed under ISS are defined by federal law, but the states may decidewhether to tax or exempt some of those services. The base covers a wide range ofservices including, IT services, rental of premises, medical services, veterinary services,personal services (barber shops, etc), professional services (engineering, architecture,law, accounting, etc.), education and training, hotels and tourism, parking, leisureentertainment (movies, shows), repair services, financial services (by banks, etc.),municipal transportation, port, terminal, and airport services. The tax base is therevenue generated from the provision of services. The rate that municipalities maycharge for ISS is locally set but cannot exceed 5 percent (in the past, it could go as highas 10 percent). Although the tax rate to be applied is the one charged by themunicipality in which the business resides, there are exceptions in which the tax rate isthe one belonging to the municipality where the service is rendered (e.g. construction).Producers of services are charged with paying and recordkeeping of the ISS. Buyers ofservices do not directly see the tax they pay as it is included in the price charged tothem by vendors.

Although the ISS collects at the municipal level, its importance varies greatly acrosslocal governments; according to the Receita Federal (national tax administration) 1percent of municipalities (out of more than 5,500) account for 73 percent of the taxcollection. The ISS collected approximately 0.5 percent of the GDP in the mid-to-late1990s and more recently, the tax collections are nearer to 0.8 percent of the GDP. Asthe maximum tax rates were lowered sometime in the early 2000s, the observedincrease in collections as a share of GDP might come either from an expanded taxbase and/or better efforts in collecting the tax. There is some evidence of theexpansion of the tax base; when the rates were lowered in 2003. The ISS was alsoextended to services provided by financial institutions, banking services in particular.

The ISS is not without its problems. One issue is the increased tax on the production ofproducts/services for future use, since users of these services cannot identify the ISSbalance to be paid against the ISS they would receive; remembering that the price forthe services are ISS inclusive. Another issue has been the management of exports. Theimport of services is subject to this tax, and although it should not apply to exportedservices, it may become subjected to this tax.59

With regard to the future of the ISS, there have recently been calls in Brazil for thesimplification of the tax system where–one way or another- the elimination of the ISSwas contemplated. This viewpoint supports the integration of the federal-based IPI andstate-based ICMS (both value-added type taxes) and the locally-based ISS within ageneral VAT whose revenues could be shared by all three levels of government.However, the increasing importance of ISS in municipal budgets, with regard to thepotential loss in local autonomy and the difficulties of coordination at different levels ofgovernment, weigh-in on the other side of the argument.


217

60. For example,reportedly in the caseof El Salvador municipalfees and charges datefrom 1954.

61. Note that this does notmean that there are nosignificant regionaldisparities in the region;it simply means that themore frequent use ofthe derivation wouldhave made thingsworse. As discussedabove, regional fiscaldisparities arise fromthe differences ineconomic bases and themore tax autonomyprovided, the higher thepotential for enlargedfiscal disparities whichrequire a higher needequalization grants.

62. In the case of Peru, theevolution ofinternational prices fornatural resources had asignificant impact on thetransfers system. TheMinistry of Economyand Finance (MEF)currently shares 50percent of mining andhydrocarbon revenueswith local governments.Transfers from centralgovernment to regional/local governmentsincreased exponentiallyafter 2000 butplummeted in 2009.This experience hasgiven rise to verysignificant horizontalimbalances betweenmunicipalities becauseshares of miningrevenues are highlyconcentrated on sevendepartments,accounting for close to80 percent of the total.This has createdproblems with increasedunspent balances duringthe boom years and ithas continued to exposelocal governments tohigh volatility in revenuestreams.


local “direct” tax on all business activities(industry, trade and services) that uses asits presumptive tax base the monetaryvalue of annual transactions. It is alsolevied at different rates depending on thesector. The production of food pays a rateof 0.41 percent, the sales of alcoholicbeverages, tobacco products, and fuelspay a rate of 1.38 percent, and financialtransactions pay a rate of 1.1 percent. TheICA is one of the most important sourcesof municipal tax revenues in Colombia, onaverage representing approximately 42percent of municipalities’ annual taxrevenues, but as much as 72 percent inthe Municipality of Cali.

Another form of local business tax withrevenue potential is Chile’s “patente mu-nicipal.” This annual levy, administeredby the municipalities, is paid for any com-mercial activity (trade, professional,industrial, and sale of alcoholicbeverages) that requires a permanentoffice location; municipalities select ratesbetween 0.25 and 0.5 percent that fall onthe declared (to the national taxadministration) own capital of thebusiness. The “patente municipal” raisesapproximately the same amount ofmunicipal revenues in Chile as theproperty tax (“impuesto territorial.”)

In terms of revenue collection hierarchy inLatin America, the two sets of taxes thatare generally of equal importance are theproperty taxes (impuesto sobre los bienesinmuebles, IBI) and the different taxes onbusiness activities and services. In adistant third place we find those taxesfalling on the use of motor vehicles.Generally, there would appear to be roomto increase local revenues for taxes onmotor vehicles. This is also the case forlocal fees and charges in many countries inthe region; often the levels of fees andcharges are completely out of date.60

Unbundling revenue sharing

Revenue sharing is the most commonmechanism for arranging fiscal transfersto sub-national governments in LatinAmerica and in many countries provisionsfor revenue sharing are enshrined in theconstitution. As mentioned in Section 1,fiscal decentralization has often beenunderstood simply as the sharing ofcentral government revenues withoutrelating the additional revenue to anyparticular local and regional expenditureassignments.

One of the most negative aspects ofrevenue sharing in other parts of theworld is that it can exacerbate the subs-tantial horizontal disparities across localgovernments when carried out on aderivation (i.e. origin basis). Fortunately,this has generally been avoided in LatinAmerican countries.61 An importantexception has been the revenue sharingin natural resources, which in countriessuch as Peru and Ecuador is funda-mentally implemented on a derivationbasis. This has led to significanthorizontal disparities among municipa-lities.62 In most Latin American countriesfiscal transfers from revenue sharingare distributed according to a set ofparameters or formulas that try toachieve several objectives, one of whichis some degree of equalization. One ofthe positive aspects of revenue sharing isthat it manages to combine the uncon-ditional use of funds with rather plentifulsources of revenue.

The main problem with general revenuesharing is that there is some confusion overthe exact achievement of distributionformulas; pursuing many objectives withessentially one instrument tends to be thesource of that confusion. It isn’t always bestto give local governments the unconditional



use of all of these funds.63 The reformsbeing introduced or contemplated in someLatin American countries consistfundamentally of the unbundling of therevenue sharing system into two additionalseparate transfer mechanisms, namely: (i)an equalization transfer with unconditionaluse of funds and financed from a poolextracted from the shared revenues, whichwould exclusively pursue the goal ofequalization of horizontal fiscal disparities,or (ii) a system of specific or conditionalgrants for current expenditure andinvestment purposes, financed with some ofthe revenue sharing funds. The use of thesefunds would be earmarked in pursuit of avariety of sectored objectives. Advances inthis general direction have been made incountries like Brazil and Chile, whilecountries like Ecuador, El Salvador andHonduras are still using an unbundledrevenue sharing scheme as the mainfunding source of local governments.

The need to rationalize the transfersystem

The system of transfers plays a pivotal rolein drawing together the other elements ofthe intergovernmental fiscal system. Itmakes up for the vertical and horizontalgaps that own source revenues andrevenue sharing cannot meet, and when itis designed properly it does not underminelocal tax effort or the incentives forcreditworthy municipalities to borrow.

With the exception of a few countries, thecurrent system of transfers to localgovernments in Latin America lacks a clearorientation. Most countries have yet tointroduce unconditional equalization grantsthat incorporate some formula-basedmeasures for expenditure needs and fiscalcapacity.64 When some equalizationelements are introduced into therevenue sharing formulas, actual revenue

collections are often used instead ofmeasures of tax capacity, thereby creatinga negative incentive for tax effort. Thecurrent methods used to incorporatedifferent expenditure needs in the revenuesharing formulas are also problematic. Forexample, population, which is commonlyused as a good approximation for someservices needs, is not the right factor to beconsidered for other services. Forexample, the number of school agechildren provides a better approximationfor basic education needs than populationas a whole. The relative share of infantsand the elderly in the population providesa better approximation for health careneeds than the whole population per se. Insome cases, especially in Central Americancountries like El Salvador and Nicaragua,the existing formulas favor smallmunicipalities, which in turn create pro-blems of fairness and economic viabilityand regional development.

Even though many countries have someform of conditional grants, they lackstructure and consistency, especially inthe area of capital grants. Whenconditional grants are used, thecomplex system is often a problem.65

Compliance-administrative costs bylocal governments which is intricate, inmany cases penalizes smallerjurisdictions with low administrativecapacity and dilutes the achievement ofcentral government goals. A remedy tothese problems, following best inter-national practice, has been toconsolidate many of these specific orconditional programs into block grants.While specific conditional grants narrowthe use of funds (e.g. funds to buyfurniture for primary schools), blockgrants, while still conditional, have amuch broader scope for the use of funds(e.g., the funds must be spent onprimary education). The advantages to

63. The point is to quicklyachieve a moreappropriate mix ofunconditional andconditional grants,without losing sight ofthe fact that, in the longterm, it is desirable toincrease unconditionallocal governmentsfunding.

64. Some exceptions includeChile’s revenueequalization grant.


219

using just a few block grants are thesimplification and expansion of localgovernment choices, thus aligning thefinal allocation of public resources moreclosely to the priorities of individuallocal governments (e.g., repairingthe school building instead of buyingnew furniture), without undulycompromising the general goals ofcentral government’s policies (e.g.,promoting the quality and standards ofprimary education in the country).

Increasing fiscally responsible localborrowing

Historically, the Latin American regionhas suffered some of the worst incidencesof fiscally irresponsible sub-nationalborrowing in the world. The negativeexperiences of Brazil and Argentina,with uncontrolled sub-national borrowingand hyperinflation during the 1980s and1990s, are still examples of what can gowrong in this area of sub-national finance.One consequence of those experiences isthat the borrowing policy towards localgovernments in the region has becomeexcessively conservative and restrictive.

For example, in Chile, local governmentsare —in principle— not allowed to borrow ortake out loans of any kind, but either wayoutright prohibition is unnecessary oreffective. In the same country, indirectborrowing through leasing contracts or bydelaying the payment of current ex-penditure makes that norm difficult toenforce. In Uruguay, any domestic orforeign debt issued by a local governmentneeds to be approved by the nationalcongress. Peru also provides an example oflegislative conservatism in the matter oflocal borrowing: The central governmenthas established indebtedness rules tomaintain fiscal prudence by two laws (theLaw on Fiscal Prudence and Transparency

–LPTF– and the Law on Fiscal Responsibilityand Transparency –LRTF–). Besides limitson debt service ratios and total debt, thelaws also limited the rate of growth formunicipal expenditures to a maximum of 3percent per year. However, this frameworkhas not been fully enforced because ofinsufficient monitoring, and the lack ofeffective sanctions. Many local governmentsin Peru carry large budgetary arrears. At theother extreme, and more like an exception,Paraguay, has practically no restriction onlocal borrowing.

Thus a pending challenge for severalcountries in the region is how to set upinstitutions that effectively regulate borrowingwithout becoming overly restrictive of localgovernments. Many countries are stillstruggling to introduce a credible systemof penalties for lack of compliance. Thedevelopment of information and monitoringsystems covering all aspects of borrowing,65

including budgetary arrears with officialinstitutions and private providers, is urgentlyneeded. But the key ingredient for fiscallyresponsible behavior of sub-national unitsremains at the political will of the centralgovernment authorities to implement theexisting regulatory frameworks.

A second challenge for practically allcountries is how to make more creditavailable to local governments forresponsible borrowing.66 In practice, thelevel of borrowing by local governments inLatin America is far too low to meet thepresent large needs for public infrastruc-ture across the sub-continent.67 The ex-ceptions are large cities, which tend tohave ample access to domestic creditmarkets and in many cases tointernational markets with accompanyinginternational credit ratings. Thus, thecapital cities of La Paz (Bolivia) and Lima(Peru) display a very different picture fromthat of most other municipalities.

65. Monitoring systems forlocal indebtednesshave often proved adoubtful utility. InEcuador and Peruinformation on debt istaken directly from thefinancial statements ofsub-nationalgovernments and isnot crosschecked withother sources. InArgentina, the FederalCouncil for FiscalResponsibility createdby the FiscalResponsibility Law of2004 and in charge ofmonitoring compliancewith norms and rulesof fiscal and financialbehavior does notreceive timelyinformation from themajority municipalitiesas of 2009.

66. Part of the solution canbe direct on-lending tomunicipalities byregional-multinationalinstitutions such as the[Confederation Andinade Fomento] (CAF) or[the BancoCentroamericano deInversiones], orinternationalorganizations such asthe Inter-AmericanDevelopment Bank(IDB) or the WorldBank. However, alarge part of thesolution needs to bethe mobilization ofdomestic creditsources.

67. Brazil is an exception.




The absence of real access to borrowing bythe average municipality in the region is acomplex issue. It is explained by a multitudeof causes, ranging from the lack of taxautonomy for local governments to the lack ofnational financial market development. Onepotential remedy for the scarcity of localcredit is the creation of semi-official financialintermediaries or municipal banks.68 Asshown in Box 6.5, several Latin Americancountries have created this type of institutionto facilitate long-term credit to localgovernments. However, the experience of

these institutions has been mixed because ofthe difficulty of maintaining them at arms’length from central government officials andof operating them with strict banking criteria.

Issues on the Budget Process andTransparency

Streamlining the Budget Process at theLocal Level

Budgets and the budget process at thelocal level in Latin America have improved

68. Central governmentsalso guarantee on-lending to localgovernments frommultilateralinternationalorganizations,including the WorldBank, the IDB, andCAF.

69. See Peterson (1996).There are otherrelevant initiatives inthe region that forspace reasons are notdeveloped in this box,including the Bancodel Estado (BEDE) inEcuador and La Caixain Brazil. At theregional level the CAFand the IDB havebeen active supportersof on-lendingprograms for thedevelopment of localpublic infrastructure.

Box 6.5. Practice with Municipal Development Banks and Funds in Latin America69

Experience with municipal development banks in Latin America has been mixed, as hasbeen the case in many other countries around the world. Although quite a few countrieshave introduced some sort of specialized financial intermediaries or municipaldevelopment funds to raise capital financing for local governments, few of thoseinstitutions have been transformed into financial institutions with market-orientedpractices and controls channelling private savings to finance public infrastructure. Thefollowing is a summary of experiences with municipal development banks and funds inthe Latin America region.

Brazil

The Integrated Program of State Improvements (PIMES) was established as a municipaldevelopment fund administered by BADESUL, the development bank of the Rio Grande doSul, which owns and controls it. The program has two components: institutional and humanresource development and infrastructure investments. The first component comprises ofabout 10 percent of the total project costs and includes technical assistance, training andequipment for municipalities, the State Water Company (CORSAN), and other state sectoragencies, etc. The second component represents about 90 percent of the total programbudget and includes the financing of projects in water supply and sanitation, street pavingand lighting, drainage and erosion control, and so on. The Municipal Action Program (PRAM)was established in 1991 by the Parana state bank BANESTADO, which originally served asthe bank’s financial agent. PRAM was eventually converted into a revolving State UrbanDevelopment Fund (FDU), administered by BANESTADO with technical assistance providedby a legally autonomous organization that in practice functions as a department of the Stateof Parana Secretariat of Planning.

PARANACIDADE was created in 1996 as a non-profit corporate entity to provideinstitutional and technical services to municipalities in Parana; this institution alsocollects and invests financial resources from the state’s urban and regional development


221

significantly in recent times. A noteworthyinnovation has been the introduction ofparticipatory budgeting (See Box 6.6).Nevertheless, a variety of issues at diffe-rent stages of the budget process stillneed to be addressed in several countriesin the region. In terms of budget planningand formulation, there are still countrieswhere local governments must have theirbudgets, or certain aspects of them, approvedon an annual basis by higher levels ofgovernment.70 Ex-ante monitoring andapproval of local budgets by higher level

authorities is not needed where there arelocal elected councils and an effective ex-post audit system, and courts to addressirregularities. Local budget autonomy isoften limited in the case of investmentprojects.71 Another issue in the prepa-ration stage is the lack of a link betweenplanning and budgeting. Frequently, it isseen that many development plans at thelocal, regional, and national levels lackcoordination and do not relate to actualbudgets in terms of the cost of activitiesfor the fulfillment of strategic objectives.

70. These countriesinclude: Bolivia, CostaRica, Jamaica, Mexico,Nicaragua, Panama,Paraguay.

71. In Peru all publicinvestment projectsmust be approved bythe National PublicInvestment System(SNIP) which ismanaged by theMinistry of Economyand Finance (MEF).Through the GeneralDirectorate of Multi-Sector Programming(DGPM), MEF has thepower to cancel anyapproval made byregional and localgovernments if DGPMconsider that SNIPcriteria have not beenproperly applied.


Practice with Municipal Development Banks and Funds in Latin America (cont.)

programs, managing the State Urban Development Fund (FDU), which creates loans formunicipalities at maturity ranging from 8 years for urban infrastructure to 10 years forsocial infrastructure. One of the main explanations for PARANACIDADE’s success is itssupport of capacity building for municipalities.

Colombia

Colombia has been successful in using its Municipal Development Funds to accelerate thedevelopment of private credit markets for local government. The Territorial FinancingInstitution (FINDETER), which began in 1994 as an infrastructure financing window withinthe National Mortgage Bank, eventually evolved into a development bank for municipalities,working through the commercial banking system. In essence, FINDETER operates as asecond-level financing institution which re-discounts commercial bank loans tomunicipalities. The banks’ good credit experience through FINDETER has led them tocommit their own resources to municipal lending. Intermediate-sized cities anddepartments in Colombia now borrow primarily through commercial bank loans, while smallcities and towns continue to rely on FINDETER. The largest cities now finance their creditrequirements primarily through bonds.

Mexico

The federal public works bank, BANOBRAS, was founded in 1933 as the Banco NacionalHipotecario Urbano y de Obras Públicas, S.A (National Urban Mortgage and Public WorksBank), and has long had a loan program for municipal development. Its operations,however, became complex and bureaucratic and the allocation of financial resources soonresponded more to political than financial criteria. The bank’s heavily subsidized loanprogram used to focus on social housing, water supply systems and the construction ofmarkets and abattoirs. Since 1988, its interest rates have come close to market rates andBANOBRAS has switched its focus to improvements of municipal land registers.



In terms of budget execution, themisappropriation of funds by centralgovernment is still a problem in severalcountries. For example, in Haiti, 90percent of local government incomeevidently comes from transfers associatedto the Funds for the Operation andDevelopment for the Territorial Collectives(FGDCT), administered by the Departmentof the Interior. However, recent studiesshow that these funds are not beingdistributed as the Department of the

Interior claims, and that a significantshare is retained by the Department tofinance its own projects. The budgets ofcommunes (i.e. the local governments inHaiti) are the most directly hurt, receiving33 percent less than what they arebudgeted to receive. Honduras providesanother example, where the law is notrespected by the central authorities. Inparticular, while the Municipal Codeestablishes that the central governmentshould allocate 5 percent of its tax

72. Seehttp://www.internationalbudget.org/ forother experiences inparticipatorybudgeting and otherinnovations for moreopen and transparentbudgeting practices.

Box 6.6. Participatory Budgeting in Porto Alegre (Brazil)72

Participatory budgeting has been functioning successfully in the municipality of PortoAlegre, in the state of Rio Grande do Sul (Brazil) for the last two decades. Theparticipatory budget of Porto Alegre, called OPPA, is a process through which ordinarycitizens and a team of elected local government officials work together to define a list ofprojects to be included within the local government budget. Through this mechanism forthe shared management of budgetary resources, local residents perform the role ofidentifying and controlling the implementation of projects. Thus, through the OPPA,local residents are closely associated with the formulation of public policy at the initialstages, including diagnosis and needs assessment, the intermediate phase ofmonitoring and implementation, and the final phase of control and accountability.

Since its inception the OPPA has contributed to the improvements in the lives of localresidents. The number of participants in Porto Alegre has increased year by year, fromapproximately 1,000 in 1990 to nearly 15,000 in 2004. The process has also broughtopportunities to better integrate traditionally marginalized groups of the population inthe community’s development. In 2002, there was a predominance of women amongthe leaders of neighbourhood associations, delegates and counselors. In addition, mostof the OPPA participants belong to lower income groups. Other groups, such as the blackpopulation or, manual and unskilled workers have also seen higher participation rates inthe OPPA process (City, 2003). According to Abers (2000), who studied the profile ofOPPA’s participants, contrary to some expectations, the process has not given rise to theinfluence of an elite field of people with more education or income. In addition, Santos(2003) has shown that OPPA resulted in an increase in the provision of basic publicservices. In 1999 the volume of garbage collected and the number of additional lightsinstalled nearly doubled from the annual average for the period prior to the existence ofOPPA (1985-1988). In 1996, the sewer lines in the municipality were expanded to cover98 percent of households up from a coverage of about 50 percent in 1989. The WorldBank (1999) also attributed to OPPA the paving of half of the municipality streets andthe doubling in the number of students enrolled in primary and secondary schools.


223

revenue to the municipalities, in fact, only3 percent appears to have been allocatedin the most recent years.73 Othercountries, like the Dominican Republic, areexperimenting with similar situations; 10percent of their national budget must beallocated to local governments (accordingto Law 166-03), but actual transfers havenever reached this level and recently theyhave decreased from 8 percent to 6percent of the national budget.

Deficiencies related to ex-post audit of lo-cal budgets still exist.74 For example, inParaguay, many municipalities do notcomply with the requirement to send theirannual financial reports to the ComptrollerGeneral of the Republic.

Addressing the Scarcity of Data on LocalFinance

The lack of adequate data on local financeis a widespread problem in the regionwhich has major consequences. Only ahandful of countries currently make muni-

cipal data openly available to the public.Countries, such as Brazil, Peru andEcuador provide examples for bestpractice in this area; for further examplesin the region see Box 6.7.

Conclusions

The analysis in the above sections hasshown that the Latin America regioncontains a rich variety of experiences andlessons, good and bad, about de-centralization and municipal finance. Thisassortment of experiences and challenges,sometimes quite unique,75 has made itdifficult to draw up a cross-countryanalysis. Nevertheless, there are manycommon themes and challenges facingmunicipal governments in Latin Americaand each country has been able to addressthem with varying degrees of success. Forexample, in Chile, the central governmenthas made use of municipal governments’ability to increase the effectiveness ofsocial policies and encourage innovationand competition among them. In Colombia,

73. See Cardona (2006).

74. In El Salvador andother countries in theregion, municipalitiesare required toundergo a fullindependent auditonce a year to searchfor signs of corruptionand misuse of publicfunds. This practicedoes not eliminatecorruption but it goesa long way to keep itunder control.

75. For example, thechallenges faced bysome municipalities inColombia go beyondfiscal issues. Here,municipalities in thewar-torn areas mustface the challenge ofbeing on the frontlinesagainst armed actorssuch as drugtraffickers,paramilitaries and theRevolutionary ArmedForces of Colombia(FARC).


Box 6.7. Annual Publication of Executed Budgets in Colombia

Law 617, enacted in 2000, charges the National Planning Department (DNP) in Colombiawith the annual publication of budget results (revenues, expenditures and financialindicators) for all departments and municipalities, together with an explanation of wherethere have been problems and where there has been progress. Included in these recordsis a detailed recording of the municipalities’ fiscal performance and information on allincome and expenditures during the past fiscal year. This annual publication is of highquality. The DNP collects annual data on revenues and expenditures, as well as the debtlevels of all local governments. Each local government reports and certifies the accuracyof its executed budgets to the DNP through an automated system, the Sistema deInformación para la Captura de la Ejecución Presupuestal de Departamentos yMunicipios (SICEP). The DNP also receives information on debt levels from theControloría General de la República (CGR). These data are regularly used bygovernment institutions and nongovernmental organizations to monitor theperformance of sub-national governments. The comparisons in performance also allowsome form of benchmarking competition among local governments; which governmentsare doing relatively better and which are doing worse.



central legislation can provide admini-strative flexibility, with local governmentsexercising control over staff hiring andsalary decisions, and at the same timeprovide effective accountability mechanismsto maintain fiscally responsible decisionsby local officials. In Honduras, a municipalassociation can successfully providetechnical assistance and training for itsmembers. Changes in the attitudes ofmunicipal officials toward broadercommunity participation in budget deci-sions have taken place in countries such asBolivia, Brazil, and Peru.

In this concluding section we offer someobservations grouped according to theset of issues examined. Yet, there arechallenges remaining which needaddressing and it will be necessary tocontinue looking for new orientations infuture research about public local financesin the region.

Observations on OrganizationalStructure

Countries with problems of fragmentationand small municipalities, should introducelegislation and practical support for thecreation of associations of municipalitiesinto mancomunidades for the delivery ofcertain public services requiring a certainminimum scale. Other solutions to theproblem of insufficient scale that could bepursued include cooperative servicesagreements between larger and smallermunicipalities, and the contracting ofservices with private enterprises for thedelivery of services. In addition, carefulcritical study and consideration should begiven to the creation of new tiers ofvertical government (for example,regional governments) as a solution tosome of the weaknesses observed at theexisting local governments. A cheapermore efficient solution can be the

strengthening of technical assistanceand additional funding of existing governments.

In any case, existing potential incentivesto further fragmentation should beremoved. In particular, those countrieswith transfer formulas that ensure thesame amount of funds to all municipalitiesindependently of their size shoulddiscontinue this practice. Where they donot exist now, new legislation withminimum population and fiscal viabilityrequirements should be introduced toprevent any further undesirablefragmentation of local governments.

Most central governments in the region,and provincial or state governments in thecase of federal systems, should devotemore time and resources to developingadministrative capacity, especially in thecase of small and rural local governments.Some of this assistance can be providedby working together with and offeringsupport to municipal associations in orderto give quantitative and qualitativetechnical assistance and training to localofficials in the most cost-effectivemanner or by having regional universitiesand colleges contracted to tutor localgovernments.

Observations on IntergovernmentalFiscal System Design

Without a clear assignment of expenditureresponsibilities to local governments it isnot possible to have an informed judgmenton whether or not the level of financing ofthese governments is adequate. Mostsystems of intergovernmental fiscalrelations in the region would benefit froman explicit clarification of thecompetencies assigned to local govern-ments. First, this will require the clearidentification of the exclusive respon-sibilities of local governments. Second, in


225

the case of concurrent or sharedresponsibilities between the local andcentral (or intermediate level) govern-ments, it will be necessary to identifywhich attributes of the particular com-petence (regulation, financing, andimplementation) are the responsibility ofthe local governments and which belong tohigher levels of government. There will beno clear assignment of responsibilities,especially in the case of concurrentfunctions, until it is apparent which level ofgovernment is exclusively responsible forthe different sub-functions involved. Ofcourse, the implementation of servicesmay be done directly by the localjurisdiction or this unit can makearrangements for its provision, forexample by a private company or someother jurisdiction.

If there are significant differences in ad-ministrative capacity among local govern-ments it may be desirable to temporarilyintroduce two or at most three differentpackages of expenditure responsibilitiesthat can be devolved to local governmentsdepending on their administrative capaci-ties and over time, as capacity is acquired,graduate municipalities to the more com-plete levels of responsibility.

It would also be desirable to adopttransparent approaches to translate theassignment of local functional responsibi-lities into expenditure needs in order tohave a clear idea of the financing require-ments for local governments.

Greater local revenue autonomy is achallenge not yet adequately addressed bymost countries in the region. However,there is a need to find a better balancebetween the decentralization of expen-diture responsibilities and the authority tocollect local taxes from the residentsdirectly benefiting from local services. This

will lead to more fiscally responsible andpolitically accountable forms of decentrali-zation. Several options are open goingforward with this agenda.

• First, countries that have not assignedproperty tax to local governmentsshould do so. Property tax has severalcharacteristics that make it ideal as alocal tax.

• Second, other taxes that should beassigned to local governments arevehicle taxes, business licenses, andbetterment levies on real estate forfinancing basic infrastructure improve-ments.

• Third, for countries that have not doneso, some degree of discretion in settingtax rates should be granted to all localgovernments so that they can adjusttheir tax bases, within legislatedmaximum and minimum rates. Otherforms of autonomy beyond rate setting(e.g., adjustments to the tax base orthe freedom to introduce new taxes)are not generally desirable.

• Fourth, coordinated efforts of local andcentral governments should be made toincrease the revenue yield of propertytax and other taxes assigned to localgovernments. In the case of propertytaxes, these should include: regularlyupdated and improved property cadas-tres and property value assessmentmethodologies, increased effectivenessin the collection of tax bills, andremoval of disincentives for increasesin tax effort by local governments i.e.reductions in transfers when more localrevenues are collected.

• Fifth, the introduction of new taxes atthe local level should be considered,including wider use of betterment




levies and local business taxation, suchas the ICA (impuesto de industria ycomercio) in Colombia or Chile’spatente municipal.

Going forward, to improve the direction ofincreased local tax autonomy, would bethe introduction of a local piggy-backpersonal income tax with a flat ratecollection at the same time nationalincome tax is collected. This latter form oflocal tax is common, in northern andcentral Europe, but it is yet to be tried inthe Latin American region. Finally, there isa possibility of considering theintroduction of environmental or “green”taxes enabled by national legislation onthe regulation of the environment. Thisform of taxation has not taken root inmany Latin American countries though itprovides several important advantages.The first is the so-called “double dividend”since these taxes not only collect neededrevenues but also contribute a cleanerenvironment. These taxes can also fit wellin regional and local contexts. Potentiallevies in this area would include taxes onthe emission of solid waste and watercontamination.

In those countries where revenue sharingis a major source of local finance, it wouldbe desirable to un-bundle part of therevenue sharing system into separatetransfers, including: (i) an equalizationtransfer with unconditional use of fundsand (ii) a system of block conditionalgrants for current and capital purposes. Anexplicit unconditional equalization grant isneeded to address the important andincreasing problem of regional fiscaldisparities in many countries in theregion—based on differences in taxcapacity or economic base, anddifferences in expenditure needs due togeography or the population structure.Explicit conditional grants are necessary to

ensure national standards and objectivesin the provision of important services havebeen decentralized, such as in educationand health.

In those countries where local borrowing isnot allowed, new legislation shouldintroduce the possibility of responsiblelocal borrowing. In those countries thatalready allow municipal borrowing, itwould be desirable in many cases toreview the current status of regulations,streamlining them when necessary so thatthey are not overly restrictive. This reviewshould also focus on the monitoringcapabilities of the central government(including “floating debt” or budgetaryarrears with official institutions andprivate suppliers, and guarantees throughmunicipal enterprises) and theintroduction of a credible system ofpenalties for lack of compliance.

Beyond the regulation and monitoring oflocal borrowing, an even more importantchallenge for most countries in the regionis to facilitate a significant increase incredit availability to local governments forresponsible borrowing, especially forsmaller municipalities. The solution maysometimes be the creation of officialfinancial intermediaries or municipalbanks. A large amount of information isavailable within Latin America and otherregions of the world regarding the positivefeatures institutions should replicate (e.g.,operating with strict banking criteria) andthose features that should be avoided(e.g., operating with less than arms’length distance from political authorities).Policies to encourage the development ofprivate markets for local credit are equally,or even more, desirable. But it must berecognized that local credit from privatesources is unlikely to develop withoutmore revenue autonomy and greatertransparency of local budgets.


227

Observations on Budget Process andTransparency

Those countries still requiring ex-anteapproval of municipal budgets by higher-levelauthorities should phase out this practicea n d i n c r e a s i n g l y r e l y o n l o c a laccountability and effectiveness of ex-postaudits and the rule of law in order to keepan eye on the probity of local budgetexecution. The misappropriation of fundsin a selected number of countries is apractice that needs to be stopped and fullcompliance with ex-post audit rules shouldbe ensured. The ultimate effectiveness oflocal public expenditures will depend onthe adoption of modern budget evaluationpractices, which remains a pendingassignment for most countries in theregion.

The low reliability on municipal financesremains an important problem in the LatinAmerican region, affecting the quality of

policy design and of analytical work. Bestpractice in budget transparency and datadissemination in countries such asColombia and Peru, for example, should bereplicated by all countries in the regionwhere publicly available data on annualbudgets and other aspects of the localfinances are still missing. An effective wayto encourage and sustain good practices inbudget reporting and data generation is tomake good use of the data, by providinginformation to experts and ordinarycitizens on performance and by publicizingthe results in order to create benchmarkcompetition across jurisdictions.

There has been continued progress overthe past decade with the institutions thatmanage finances and with the practiceitself of municipal finance in the LatinAmerican region. Nevertheless, there is along road ahead for further improving theoverall efficiency, equity, and accountabi-lity of municipal finances in the region.




Latin America Regional Policy Recommendations

Prepared by the technical team of the Federación Latinoamericana de Ciudades,Municipios y Asociaciones (FLACMA), March 2010.

Signs of recentralization in Latin America

In various countries a backward trend has been observed with regard to decisions takenon the handover of responsibilities to local governments, affecting both local autonomyand financing.

National transfers to local governments must be stable and regular

Financial transfers to local government are a mechanism to effectively integratemunicipal participation into the national budget and constitute a right for citizens ofterritories. Universal services such as education and health are nationally designed andfinanced to ensure equality between territories and, when managed by localgovernments, merit regular and stable national transfers.

Strengthen collection and take into account the fiscal effort of local governments withregard to local poverty levels

It is often argued that local governments are “fiscally lazy” and invest little in taxcollection. These observations do not consider the low yield of economic activity andlevel of poverty in the majority of Latin American municipalities. As a result, levels ofcollection in poor areas are often confused with the efforts made to achieve them. Tocomplement these fiscal efforts, systems of income compensation should be favored,such as unconditional transfers from central to local governments that permit aredistribution of resources in favor of more vulnerable, lower income territories.

Increase the participation of local governments in public spending and their autonomyin the management of resources

It is a recurrent theme in Latin America to measure the level of decentralization of diversecountries with indicators such as local expenditure as a percentage of general governmentspending or local expenditure as a percentage of GDP. Such fiscal observations must bebalanced against real levels of local government autonomy in deciding how financialresources will be used, be they collected directly by the local government or transferredfrom the central level. In addition, access to information on municipal finance must beimproved as well as the methodologies for collecting and recording this information.

Promote association and cooperation between municipalities to strengthen municipalcapacities

In Latin America and the Caribbean there are more than 16 million local governments.Analyzing, by country, the total number of Latin American municipalities this is not



necessarily excessive to respond to the needs of citizens. However, there are significantdifferences in both the sizes and characteristics of these entities, depending on the countryand type of territory as well as the municipal human and financial resources. The mosteffective manner to balance municipal capacity – without reorganizing the territorial andinstitutional structure in each country – is through municipal associative movements, that is tosay, to encourage inter-municipal cooperation. “Mancomunidades” allow local governments tomutually support each other, manage services jointly, and undertake local developmentprograms and projects.

Increase the sources of own revenues for local government

In Latin America the main sources of municipal income –apart from fees and tariffs formunicipal services– are property taxes, business and commercial licenses, vehicle taxes,development charges and transfer systems for equalization purposes to strengthenincomes for less developed municipalities. Property tax is the most common, and is inuse across all of Latin America, with some exceptions such as the Dominican Republic,and El Salvador. It is necessary to improve and increase the sources of own revenue forlocal governments.

Prioritize strengthening and improved functioning of local governments

It is very important for the success of the decentralization process that local governmentsare effectively strengthened, helping them to better exercise their powers andresponsibilities and provide good levels of services.

Improve coordination between ministries and national institutions responsible forsectoral policy and local government; the transfer of responsibilities must beaccompanied by corresponding resources

One of the most common conflicts in public policy is that of aligning the generally sectoralvisions of national ministries, with the territorial optic of local governments. Municipalitiesare often assigned partial responsibilities, from national ministries, without the necessaryresources to successfully implement them.

Recognize and strengthen the role of local government associations

Local government associations are an important supporting structure for municipalmanagement. The national associations allow for the design and discussion of nationalpolicies and regulations on decentralization and local responsibilities, with centralgovernment and parliament; the departmental, regional or intermediary municipalassociations play a similar role with intermediary governments, and also provide technicalsupport to member governments.


0w2010 07 MEWA DEFcarta ang 15/9/11 07:16 Página 230

7. MIDDLE EAST AND WESTERN ASIA

MMEEHHMMEETT SSEERRKKAANN TTOOSSUUNN**

UUNNIIVVEERRSSIITTYY OOFF NNEEVVAADDAA

UU..SS..AA..


* The author would like to thank Serdar Yilmaz (theWorld Bank), Isam Akel (Association of PalestinianLocal Authorities), Muaffak Al-Hajaj (Jordan Ministryof Public Sector Development) and other participantsof the UCLG Gold II Workshop in Istanbul for theirhelp with this chapter.


233

T he Middle East and Western Asia(MEWA) region has unique

characteristics in terms of its economy,demography, politics, and governmentstructure.1 Some of the characteristicsthat distinguish MEWA generally fromother regions are economic dependencyon sizeable oil reserves, internal andexternal conflicts, poor governance, andsignificant growth in the working-agepopulation mainly due to high populationgrowth in the region. The MEWA region iscomposed of a complex geographical con-figuration. Turkey and Iran are thelargest countries with a combined po-pulation of close to 150 million, whichmakes more than half of the total region

population. On the other hand, Bahrainand Qatar are the smallest countries witha combined population of about 2 million(see table 7.1 for basic statistics).

While Turkey, Saudi Arabia, and Iran are thelargest economies in terms of total GrossDomestic Product (GDP), other oil exportingeconomies have higher GDP per capita.Among those, Qatar, Kuwait and the UnitedArab Emirates are not only the richest in theMEWA region in terms of GDP per capita butthey are also among the richest in theworld. On the other hand, the region alsoincludes some of the poorest countries inthe world such as Iraq, Yemen, andPalestine.

1. This chapter builds onthe MEWA regionchapter of UCLG’sDecentralization andLocal Democracy inthe World, First GlobalReport (GOLD I) byMustapha Adib(2008). That chaptergives a good overviewof the administrativeand political aspectsof decentralization inthe MEWA region. Thegoal of this chapter isto complementprevious work bygiving a comparisonof local governmentfinancing andhighlighting specialissues, constraintsand opportunities.


Country Land Area Political Population, Urban population Population GDP per (km2) Regime 2008 (% of total Growth Rate capita

population), 2008 (%), 2008 (current US$)

Bahrain 710 Constitutional Monarchy 766,926 88.5 2 21,421

Iran 1,745,150 Islamic Republic 71,956,322 68.5 1 5,352

Iraq 438,320 Parliamentary Republic 20,941,720 66.6

Jordan 88,780 Constitutional Monarchy 5,906,043 78.4 3 3,389

Kuwait (2007) 17,820 Absolute Monarchy (Emirate) 2,728,041 98.4 2 42,102

Lebanon 10,400 Parliamentary Republic 4,139,281 87.0 1 6,924

Oman (2006) 309,500 Absolute Monarchy (Emirate) 2,785,361 71.6 2 13,381

Qatar 11,000 Absolute Monarchy (Emirate) 1,280,862 95.6 12 52,690

Saudi Arabia 2,000,000 Absolute Monarchy (Emirate) 24,645,686 82.4 2 18,973

Syrian Arab Republic 185,180 Presidential Republic 21,226,920 54.2 3 2,601

Turkey 783,560 Parliamentary Republic 73,914,260 68.7 1 10,745

United Arab Emirates 83,600 Federation of Absolute 4,484,199 77.9 3 38,436

Monarchies (Emirate)

Palestinian National Authority* (2005) 6,020 Palestinian Authority 3,837,957 71.9 3 1,160

Yemen 527,970 Presidential Republic 23,053,462 30.6 3 1,153

* Figures for Palestinian National Authority correspond to the West Bank and Gaza.

Source: World Development Indicators, 2009.

Table 7.1: General Characteristics of MEWA Countries


MMEEWWAAUnited Cities and Local Governments234

Many studies have noted that countries inthe Middle East and North African (MENA)region have very centralized governmentstructures (Tosun and Yilmaz, 2008;Arzhagi and Henderson, 2005). It hasbeen argued that most MEWA countrieshave based their government organizationon the administration structure of theOttoman Empire.2 Aside from a fewisolated attempts at decentralization,the region inherited a highly centralizedsystem of taxation and public administrationwhich has left very little room for localgovernment autonomy.

To give an overview, we can say thatTurkey and Palestine have the greatestdegree of decentralization from theperspective of expenditure and revenueallocation. Of the two, it is Palestinethat gives the greatest fiscal autonomyto local governments. While this singlesout Palestine from the other countries ofregion, the situation must be givencareful consideration as the level ofdecentralization is the result of veryparticular political and security situation:strong internal and external conflicts, aweak central authority and discontinuity inits geographic border.

The system of financing local governments inthe Middle East and Western Asia must beimproved in its entirety. While Jordan,Lebanon, Turkey and Palestine use someform of rule or formula to calculate transfers,in other countries, the intergovernmentaltransfer process is largely ad hoc. Localaccess to credit is very limited. In countrieswhere there is a significant level of localgovernment borrowing, this is due to theweak regulatory frameworks set by centralgovernments.

There have been important efforts recently toreform local government finance in the MEWAregion. Examples of this include new laws that

were adopted in Turkey since 2002 as part ofits European Union membership process; areform project in Jordan to transfer theresponsibility of property tax collection andmanagement from the Ministry of Finance tomunicipalities; the advances accomplished inthe modernization of municipal administrationin Syria in the framework of the current FiveYear Plan (2006-2011); and the 1997 law onlocal authorities in Palestine. In Lebanon, inOctober 2009 the National Government, at itshighest level, committed to further supportdecentralization and local governmentreinforcement during an international Seminarorganized by UCLG in Tripoli.

Local democracy is also in progress. Localelections took place in 2010 in Lebanon,allowing democratic renewal of local leaders.In Jordan, a new municipalities' law wasissued in the year 2007 that allows for fullelection of municipal councils and Mayors,and dedicates a 20 percent quota for women.In Iraq and Yemen, provincial councils andgovernors have been elected for the first timein January 2009 and May 2008 respectively.On the other hand, local elections have not,for the moment, been renewed in SaudiArabia. While local elections in Iran werescheduled for 2011, they may be delayeduntil the presidential elections in 2013.

Given this brief background on the region,the following presents a formal overviewof the government structure and a cross-country comparison of local governmentfinance through expenditure and revenueassignment, intergovernmental transfers,and borrowing.

Centralized Government Structure and Local Government Finance

This section shows a comparison ofsubnational government finance in theM E WA r e g i o n . 3 M e a s u r i n g f i s c a ldecentralization is especially difficult in this

2. See Inalcik (1977) andBarkey (2008) forexcellent accounts oftransformation in theOttoman administrationtowards centralizationin the eighteenth andnineteenth centuries.Centralization gainedmomentum particularlyduring the Tanzimat(Reorganization) periodin the nineteenthcentury when theempire was decliningrapidly andconsolidation of powerat the center was seenas a solution to preventcollapse. Centralizationthat started in thesixteenth century is, byno means, unique tothe Ottoman Empire.

3. Ebel and Yilmaz(2003) provide anexcellent review ofmeasurement issues indecentralization. Seealso Hammond andTosun (2011) for asensitivity analysis ona variety ofdecentralizationmeasures at the localgovernment level.


235

region, firstly due to the difficulty ingathering basic data.4

Table 7.2 and Figure 7.1 show a comparisonof the size of central and local governments.


4. Jordan and Turkey aresomewhat exceptionsto this general trend.This is mainly due toJordan’s efforts tomeet the InternationalMonetary Fund’sSpecial DataDisseminationStandard and Turkey’sreform efforts in theEU membershipprocess.

Country Percentage of local Percentage of local government Percentage of central government expenditures in GDP spending in total public sector spending government expenditures in GDP

Bahrain 13.92 (2002)

Iran 6 (2004/05) 24.64 (2004)

Jordan 3.5 (2008) 9.6 (2008) 38.00 (2008)

Kuwait 37.20 (2007)

Lebanon 35.72 (1999)

Oman 3.44 (2001)

Qatar 26.23 (2005)

Syrian Arab Republic 5.6 (2007) 15.5 (2007) 36.00 (2007)

Turkey 4.8 (2008) 11.1 (2008) 23.88 (2008)

United Arab Emirates 11.95 (1999)

Palestinian National Authority* 2.3 (1999) 11.1 (1999) 10.68 (2005)

Yemen 6.4 (2004) 11.2 (2004) 27.43 (1999)

* Figures for Palestinian National Authority correspond to the West Bank and Gaza.Source: Author’s compilation from IMF-GFS datasets and national statistics of countries (for Syria, Ministry of Local Administration). Local government spending figures for Turkey from the IMF-GFS differ from the Turkish Ministry ofFinance figure of 4.8 percent for 2008.

Table 7.2: Comparison of Decentralization and Centralization Measures for MEWA Countries

Source: Author’s compilation from IMF-GFS datasets and national statistics of countries (for Syria, Ministry of Local Administration). Local government spending figures for Turkey from the IMF-GFS differ from the Turkish Ministry ofFinance figure of 4.8 percent for 2008.

Figure 7.1: Local Expenditure as a Percentage of GDPand General Government Expenditure



One notices immediately the lack of data onlocal government expenditures for most ofthe MEWA countries.5 This table shows a largevariation in the size of central governmentacross countries in the region. In thefew countries where local governmentexpenditure data is available, the size of localgovernments is significantly smaller whencompared to what we see in other developedor developing countries.

The Structure of SubnationalAdministration in MEWA Countries

MEWA countries have a variety of sub-national governments. Iran contains foursubnational tiers made up of provinces,

districts, and rural counties in addition toboth urban and rural municipalities (See box7.1). Both Jordan and Syria have a structureof governorates, districts, and sub-districts.In Palestine, both the West Bank and Gazaare subdivided into governorates and re-gions. In Turkey both provincial and sub-provincial district governors are appointedby the central government however munici-pal mayor and council members are elec-ted.6 Finally, in Yemen while both governorsand district governors are appointed by thepresident the councils at both thegovernorate and district levels are directlyelected by the people (see table 7.3). Headsof provincial governments are not generallyelected by the people but are appointed by

5. Database of PoliticalInstitutions (DPI) fromthe World Bank has anumber of federalismindicators that includevariables that showavailability of electionsin municipalgovernments andauthority over taxing,spending or legislatingat the local level.MEWA countries havesignificantly lowerfederalism scorescompared to most ofthe other countries inthe neighboringregions. See Beck etal. (2001) and Keefer(2007) for detaileddescriptions of thesevariables. Tosun andYilmaz (2010a) usethese variables intheir study ofdecentralization in thebroader Middle Eastand North Africa(MENA) region.

6. See Tosun and Yilmaz(2010b) for stylizedfacts on localgovernments inTurkey.

Box 7.1. Complex Deconcentrated System in Iran

The Iranian public administration system is composed of the central government and two typesof local administrative units—deconcentrated line agencies and the municipal authorities. TheConstitution of Islamic Republic of Iran defines the deconcentrated administrative units asgovernmental and municipalities as non-governmental units. The “public governmental” sectorincludes the line ministries and central government agencies with offices at sub-national levels. Itconsists of officials appointed by the central government. At the provincial level, deconcentratedservice delivery is coordinated through planning bodies under the supervision of theManagement and Planning Organization (MPO), the Ministry of Interior (MOI) and the Ministry ofHousing and Urban Development (MHUD). The “public non-governmental” sector includes urbanand rural municipalities, as well as the hierarchy of representative, directly and indirectly, electedcouncils. It consists of the representative bodies which include the directly elected urban andrural local councils (established in 1999) and the indirectly elected hierarchy of Islamic councils(established in 2002 and comprising of County Islamic Councils, District Islamic Councils,Provincial Islamic Councils, and the High Islamic Council of Provinces) as well as theadministrative bodies of urban and rural municipalities at the city and village levels.

The subnational administration in Iran is primarily organized at the provincial level. For admin-istrative purposes the country is divided into 30 provinces (ostan). The ostans have subdivisionscalled districts (shahrestan). Shahrestans also have further subdivisions called rural counties(bakhsh). Ostans, shahrestans and bakhsh are deconcentrated governmental units and coverthe whole territory of Iran.

The head of the Ostan administration is Ostandar, who is an official appointed by the centralgovernment. Expenditures at the ostan level are organized through line agencies and spendingunits. These units are responsible for provincial expenditures while national public services suchas defense and those public goods with significant externalities are assigned to the central units.



Governoratecouncil directlyelected by peopleDistrict councildirectly elected bypeople

Special ProvincialAdministration(SPA). Provincialgeneral assemblyis the legislativebody of the SPAand its membersare elected bypopular vote.

Governoratecouncil directlyelected by people

Reconstructionand DevelopmentCouncil (CDR - atthe national level)– appointed by thegovernmentcouncil.

Each governoratehas two councilschaired by theGovernor:- Executive Council

including theheads of the Lineministriesdirectorates..

- Advisory Councilmade up of 25appointedmembers, selectedfrom the followingbodies: theParliamentmembers from theconcernedgovernorate, themayors ofmunicipalities, theprivate sectororganizations, andthe civil society.

ProvincialPlanning andDevelopmentCouncil (Chair:Ostandar; LineMinistry reps) District PlanningCommittee (Chair:Farmandar; LineMinistry reps)

Council

Governor isappointed by thePresidentDistrict governor isappointed by thePrime Minister

Governor isappointed by thePresident.

Provincial and sub-provincialgovernors(vali/kaymakam)are proposed bythe Minister ofInterior andappointed by theCouncil ofMinisters and thePresident

Governor isappointed bypresidentialdecree

Provincial Governors(Muhafezs) areappointed by theMoIM with theapproval of thegovernment council.District Governors(Kaymakams) areappointed by theGovernors with theapproval of theMoIM

Governors areappointed by theCabinet based onrecommendationfrom the Ministerof Interior.

Provincial governor(Ostandar) isappointed by thePresidentDistrict Governor(Farmandar) isappointed by MoI.Rural countyadministrator(Bakhshdar) isappointed by theOstandar

AccountabilityArrangements

22 governorates16 governorates(11 in West Bankand 5 in Gaza)

81 provinces (il)892 sub-provincedistricts (ilce)

14 Governorates55 Districts210 Sub-districts

8 Provinces(Muhafaza) 25 Districts (caza)

12 governorates51 districts and 56sub-districts

30 provinces(Ostan)318 districts(Shahrestan)854 rural counties(Bakhsh)

Province District County

Ministry of LocalAdministration(MoLA)

Ministry of LocalGovernment(MoLG)

Ministry of InteriorMinistry of LocalAdministration(MLA)

Ministry of Interior& Municipalities(MoIM)

Ministry of Interior(MoI) / Ministry ofMunicipal Affairs

Ministry of Interior(MoI)

Central GovernmentMinistry

YemenPalestineTurkeySyriaLebanonJordanIran

Table 7.3: Deconcentration and Decentralization Systems in Selected MEWA Countries



the central government ministries and/orthe President. Among the countries thathave formal provincial councils, councilmembers are directly elected only in Syria,Turkey, and Yemen.

Municipalities are run by elected mayorsand council members in most countries,however, there are a few exceptions. InSyria, municipal council members are di-rectly elected by citizens. The mayor isthen elected by the municipal council andthen endorsed by the President, for cities,and the Ministry of Local Administrationsfor towns and villages.

In Palestine, mayors and councilors aredemocratically elected, while villagecouncils and joint service committees areappointed by the Ministry of LocalGovernment. In Iran, there is a complexdeconcentrated (and decentralized) sys-

tem, however representation of people ispartial and indirect. First, there is theinvolvement of the Ministry of Interior(MOI) in both urban and rural municipali-ties. Second, people elect city or villagecouncils that in turn appoint mayors jointlywith the MOI.

Expenditure Assignments

In the region, subnational deconcentratedor decentralized governments have very li-mited number of “own” responsibilities.Most of the local expenditure respon-sibilities can be classified as “delegated”expenditures as opposed to “own”expenditure responsibilities.7 Centralgovernment ministries make decisions onmost services that are traditionally provi-ded by local governments in other coun-tries (see Table 7.4).8 Among MEWAcountries, Turkey, Palestine, and Syria are

Source: Compiled by Mehmet S. Tosun and Serdar Yilmaz based on World Bank and UCLG consultancy missions and UCLG country datasheets.

N/AMunicipal mayorand councilmembers aredirectly elected bypeople.Village councilsand joint servicecommittees areappointed by theMoLG.

Municipal mayorand councilmembers aredirectly elected bypeople.

Municipal councilis directly electedby people. Mayor iselected by themunicipal council,election endorsedby the Presidentfor cities and theMinistry of LocalAdministration fortowns and villages



Mayor of an urbanmunicipality(Shahrdar) is jointlyappointed by theMoI and City Council,which is directlyelected by people.Mayor of a ruralmunicipality(Dehyar) is jointlyappointed by theMoI and VillageCouncil, which isdirectly elected bypeople.

AccountabilityArrangements

332 districtsmunicipality

133 municipalities251 villagecouncils49 joint servicecouncils

2951municipalities35402 villages

11 Governoratecities96 cities248 towns207 villages

42 municipalfederations944 municipalities

93 municipalitiesand GreaterAmmanMunicipality

More than 1000municipalities(Shahr) More than 68,000ruralmunicipalities(Deh/Roosta)

Urban MunicipalitiesRural Municipalities

YemenPalestineTurkeySyriaLebanonJordanIran

Table 7.3: Deconcentration and Decentralization Systems in Selected MEWA Countries (cont.)

7. Delegatedresponsibilities are thosetransferred to thedeconcentrated units ofthe central governmentfor delivery of serviceswhile actual budgetingand financing decisionsare carried out at thecentral level.

8. Tosun and Yilmaz(2008) assigned pointsto central, provincial andmunicipal governmentinvolvement inexpenditures andthereby created a basicindicator for expendituredecentralization.According to thatindicator, Turkey is themost decentralized inthe MEWA groupfollowed closely byPalestine.



Table

7.4

:Re

spon

sibilit

y for

Fina

ncing

and

Pro

vision

of P

ublic

Ser

vices

in S

electe

d M

EWA

Coun

tries

Finan

cing

Prov

ision

Func

tion

Iran

Jord

anLe

bano

non

Syria

Turk

eyPa

lestin

eYe

men

World

wide

Iran

Jord

anLe

bano

non

Syria

Turk

eyPa

lestin

eYe

men

World

wide

Socia

l Welf

are

CC

CC,P

CC

CC

CC

CC,P

CC

CC,P

,M

Hosp

itals

CC

CC,P

CC

CC,P

CC

CC,P

CC

C,PC,P

Publi

c Hea

lthC

CC

C,PC

CC

C,P,M

CC

CC,P

CC

C,PC,P

,M

Unive

rsiti

esC

CC

C,PC,P

CC

C,PC

CC

C,PC,P

CC

C,P

Seco

ndar

y Edu

catio

nC

CC

C,PC

CC

C,P,M

CC

CC,P

CC

CC,P

,M

Prim

ary E

duca

tion

CC

CC,P

CC

CC,P

,MC

CC

C,PC

CC

C,P,M

Hous

ingC

CC

CC,M

CC

C,PC

CC

CC,M

CC

C,P,M

Urba

n Tra

nspo

rtC,P

,MC

CM

MC,P

,MC,P

C,P,M

C,P,M

C,PC

MM

C,P,M

C,P,M

C,P,M

Railr

oads

CC

CC

CN/

AC

C,PC

CC

CC

N/A

CC,P

Airpo

rtsC

CC

CC

CC

C,PC

CC

CC

CC

C,P

Ports

and N

aviga

ble W

ater

ways

CC

CC

CC

CC

CC

CC

CC

CC

Urba

n High

ways

C,P,M

MC

C,P, M

C,MC,P

,MC,P

C,P,M

C,P,M

MC

C,P, M

C,MC,P

,MC,P

,MC,P

,M

Inte

rurb

an Hi

ghwa

ysC

CC

C,P,M

C,MC,P

,MC,P

C,PC

CC

C,P,M

C,MC,P

,MC,P

,MC,P

Elect

ricity

CC

CC

MM

CC,P

,MC

CC

CM

CC

C,P,M

Wast

e Coll

ectio

nM

MC

MM

MP

MM

MC

MM

MP

M

Wate

r and

Sewe

rage

P,MC

CC,P

,MM

MC,P

MP,M

CC

C,P,M

MM

C,PM

Fire P

rote

ction

CC

CC

C,MC,P

,MC,P

MC

CC

CC,M

C,P,M

C,PM

Heat

ingC

CC

CM

MN/

AM

CC

CC

MM

N/A

M

Irriga

tion

CC

CC

C,PC,P

MC

CC

CC,P

C,PM

Polic

eC

CC,M

P,MC,P

,MC

CC,P

,MC

CC,M

P,MC,P

,MC

CC,P

,M

Sou

rce:

Com

pile

d by

Tos

un a

nd Y

ilmaz

bas

ed o

n W

orld

Ban

k an

d U

CLG

con

sulta

ncy

mis

sion

s an

d U

CLG

cou

ntry

dat

ashe

ets.

C=

cen

tral

, P=

pro

vinc

ial,

M=

mun

icip

al



more decentralized on the expenditure si-de than Iran, Jordan, Lebanon, andYemen.9

In the majority of the countries examined,local governments assure infrastructuresmaintenance (roads, streetlights, andgreen areas), solid waste and sanitation,public markets, sport and cultural acti-vities, and often urban planning andbuilding control. Expenditure assignmentsin utility service functions are particularlysimilar to worldwide practice in Turkeyand, in theory, in Palestine. In the lattercountry, the local government law (1997)lists twenty seven areas of activity relatingto municipal powers and responsibilities,including: producing driving licenses,regulating businesses and industries.However in many cases (water and elec-tricity networks, building permits, muni-cipal waste deposits, road planning,etc.), the State of Israel retains controlover land within or on the edges ofmunicipal borders. In order to intervene,municipal authorities must obtain theirauthorization.

In Iran, Jordan, and Lebanon manylocal services have been significantlycentralized. In Iran, the distinctionbetween affairs regulated by the State andlocal authorities’ tasks and responsibilitieshas resulted in unclear legislativerequirements. In Jordan, after 2001, thecentral government took back 13 of the 39local government responsibilities listed inlaw, and many local services were given tothe private sector. In Lebanon, the role ofmunicipalities is also very limited. Thecentral authority exercises the majority ofpowers and responsibilities that aretheoretically assigned to the municipallevel, particularly in relation to planning,transport, education, social services,public hygiene services, water resourcemanagement and distribution, energy and

economic development. Even townplanning decisions are taken by the HighUrban Planning Council, which operatesunder the authority of the Ministry.

Syrian cities and towns have full or partialresponsibilities in local services such asroads, solid waste, water and sewerage,planning, and transportation.

Revenue Assignments

Countries in the region have largely ad-hoc local revenue systems that are understrict central government control (seeTable 7.5).

Palestine is the only country where localgovernments have some control over boththe tax rate and the tax base. Furthermore,the municipalities impose new taxes or feeswithout the explicit approval of the centralgovernment and they collect andadminister them locally.10 For certainrevenue items, there are differences inpractice between West Bank and Gaza. Forexample, while property taxes arecollected and administered by the centralgovernment in the West Bank, Gazamunicipalities collect and administerproperty taxes themselves. In Gaza, 90percent of revenues are kept bymunicipalities, while the remaining 10percent are transferred to the nationalgovernment. Other taxes, such as the fueltax, are centrally collected and should bepartially transferred to the municipalitiesaccording to current laws. Theenforcement of these provisions remainsunder discussion. Municipalities implementmost infrastructure projects with financingfrom donor countries which oversee theprojects and whose priorities often takeprecedence.11

However, municipal revenue autonomy inPalestine is not something granted by

9. The 2005 JordanCountry Profilepublished by theEconomic ResearchForum (ERF) and theInstitut de laMéditerranée (FEMISE)notes that Jordanianmunicipalities, with theexception of GreaterAmman Municipality,have limited financialindependence. Localgovernmentexpenditure is only 6%of total governmentexpenditures (ERF,2005: 17).

10. This is despite the 1997Law on LocalAuthorities whichestablished, at least onpaper, centralauthority’s strongcontrol of localadministrations (seeBox 7.3 for moreinformation on thislaw).

11. http://www.pogar.org/countries/theme.aspx?t=6&cid=14



Country Revenue Assignment

Iran All local levies are required to be consistent with the government’s annual budget and the National Five Year Development

Plans and to be in line with the capacity to pay as determined by the Ministry of Interior (ceiling for local tax/local income

ratio). With adoption of the Law on Tax Amalgamation (2003) revenue collection has been effectively re-centralized and

almost all taxes are collected by the central government. One of the main locally collected fees is the land use change and

density increase tax.

Jordan Municipalities have some control over the following revenue instruments: solid waste collection fees, business license fees,

building permits, property tax (on land and structures), municipal traffic fines, municipal parking fees.

Lebanon Local governments are not free to introduce new taxes. They are free to set tax rates for some taxes but they are not free to change

the tax base of some taxes. Central governments have control over revenue and other budget decisions of local governments. Some

local taxes are collected and administered by local governments and local fees are exclusively collected and administered by local

governments.

Syria Local governments are not free to introduce new taxes. They are free to set tax rates for some taxes but they are not free to change

the tax base of some taxes (inspection and slaughtering fees, motorcycle fees, fees for the sale/rental/auction of property, fees for

public property use, building license fees, fees for services). Central governments have control over revenue and other budget

decisions of local governments. Local fees are collected and administered exclusively by local governments.

Turkey Local governments are not free to introduce new taxes and they are not free to set tax rates or change tax bases. Limitations on

these come from separate laws on local governments and municipal budgets. Central governments have control over revenue and

other budget decisions of local governments. Some local taxes are collected and administered by local governments and local

duties and fees are exclusively collected and administered by local governments.

Palestinian National Authority According to a World Bank consultancy report, Palestinian regulatory framework provides the greatest autonomy to the local

governments. While local governments have a right to set taxes or create new ones, they can only do so through initiating

amendments to the tax law. Many taxes and fees are collected and administered locally by local governments.

Yemen Central government sets both tax rates and base, local governments can make proposals for taxes and fees. Apart from the religion

tax (‘zakat’) most taxes are not collected in many districts, particularly in rural areas.

Source: Compiled by Mehmet S. Tosun and Serdar Yilmaz based on World Bank and UCLG consultancy missions and UCLG countrydatasheets

Table 7.5: Revenue Assignment in Selected MEWA Countries



laws. Due to special security circumstanceand discontinuity in the geographic area ofPalestine, municipalities have inventedtheir own ways of raising revenues andthey often find ways of rationalizing suchpractices (World Bank, 2006). In generalPalestinian municipalities are very poorfinancially.

Iran, Jordan, Turkey, and Palestine, havesome form of property taxation at the locallevel. While local property tax is generallyknown as a key revenue instrument thatgrants fiscal autonomy to localgovernments in many developed anddeveloping countries, use of propertytaxation in the MEWA region is verylimited. This is again due to centralgovernment’s control over both the rateand base of the tax. In Turkey,municipalities have some representation inthe property valuation process of theproperty tax but they have no control overthe tax rates. In Jordan, there is a UNDPfunded project to move property taxmanagement to municipalities over time.Both Jordan and Turkey have property taxsystems that are based on the value ofland and structures, whereas Iran has aland-use change and density increase tax.

Municipalities in Jordan rely on thenational government for most revenues.Internally generated revenues representless than 40 percent of municipal financesand come from more than 70 differentsources. All municipal budgets must beapproved by the Ministry of MunicipalAffairs which raises questions aboutcatering to local needs using localbudgets. (Al-Hajaj, 2010; UNDP-POGAR,2010a.)

Municipalities in Lebanon and Syria havesome limited control over local tax ratesbut the central government has controlover the tax base and other local budget

decisions. In Lebanon municipalities havebeen assigned 16 direct local taxes (tax onrent value; tax on meeting spaces andclubs; advertising tax; etc), however localgovernments can not create additionaltaxes nor can they make changes to thetax base. In Syria local governments havebeen assigned a series of duties overwhich they can set the rate level.12

However these rate changes must beapproved by the governorate. In additionto this, local authorities have also beenassigned the right to 20 percent of thesale value of electricity. In 2003 local dutiescollected at the local level represented 9.6percent of total local budgets.

Approximately two thirds of Lebanesemunicipalities have revenues less than65,000 USD per year. It is estimated thatonly 30 out of 700 local authorities havean adequate tax base to provide localservices. Another difficulty in the localfiscal system is that municipal funds comefrom over 35 different sources (Bassil andKaram, 2009; UNDP-POGAR, 2010b).

Central government has strong controlover tax rates and tax bases in Yemen,where local governments can only makeproposals for taxes and fees. Throughoutthe MEWA countries local fees are largelycollected and administered by localgovernments.

A recent UNDP report on governance inSaudi Arabia notes that “there is no fiscalseparation between the central andmunicipal governments” (UNDP-POGAR,2010c). It was also noted in the samereport that “government spending onmunicipal services, infrastructuredevelopment and local subsidies in SaudiArabia amounted to 7 percent of thegovernment budget for 2003,” and that“the Ministry of Municipal and RuralAffairs and the Ministry of Finance are

12. The main duties are onbuilding and repairlicenses, fines, rentregistration, spirits,auction sales, streetpaving, sewerage,street cleaning,vehicles, slaughtering,advertising, streetsales, fuelconsumption, fuelstoring, ports, ondocuments,improvement dutiesetc. (Financial law1/1994)


243

currently crafting legislation to privatizemunicipal tax collection” (UNDP-POGAR,2010c).


Allocation rules for transfers to localgovernments are also mainly ad-hoc,except in Jordan, Lebanon, and Turkey andto some extent Palestine (see Annex 7.1).In these countries, the transfer system isbased on a formula that uses differentrevenue sources. They get transfers basedon 40 percent of the proceeds from vehiclelicensing fees and a special legislationfee of 6 percent on petroleum productsproduced or imported (except fuel oil) bythe Jordan Petroleum Refinery Company.These proceeds are first deposited in aspecial account in the Ministry of Financeand then transferred to the Cities andVillages Development Bank “on a regularmonthly basis to be distributed to themunicipalities according to a ratioestablished by the Prime Ministry based onrecommendation from the Minister ofMunicipal Affairs.” (Jordan Ministry ofMunicipal Affairs, 2009: 8) The ratio (ortransfer formula) is determined by takinginto consideration such general factors aspopulation, proportion of its contributionin overall revenue generation, particularimportance of location and non-localresponsibilities.

In Turkey, the transfer and allocationsystem is based on grants and sharedrevenues (specific percentages from thegeneral budget tax revenues such asincome and consumption taxes) that areset by the central government. Annex 7.1shows the percentages allocated tospecific local government units. Certaincriteria such as population, acreage,number of villages in the city, ruralpopulation and city development indexare used for distribution of transfers

among local governments.13 Palestineuses transport fees for general transfersand a separate account for discretio-nary/emergency transfers.

In Lebanon, 10 percent of the ratescollected by the national telephone, elec-tricity and water public services, as wellas 3 percent of taxes on propertytransfers are supposed to be transferreddirectly to local governments. However,this payment requires a decision from theMinister, and often undergoes importantdelays. In addition, the revenue of 13indirect taxes, collected by the State onbehalf of local governments, goes intothe Autonomous Municipal Fund (AMF).After deductions for the cost of collectionand staff salaries, the remaining fundsare transferred to local government (75percent among municipalities and 25percent among municipal federations).The Ministry of Finance distributes tomunicipalities according to a transferformula which is based on factors such aspopulation, proportion of its contributionin overall revenue generation, particularimportance of municipality’s location andnon-local responsibilities. Allocation rulesare decided annually by the Ministry ofInterior and Municipalities and theMinistry of Finance with the approval ofthe central government. Funds comingfrom the AMF make up approximately 80percent of local government budgets.14

But transfers are carried out irregularlyand with a lack of transparency (seebelow, “Significant Spatial Disparities inPublic Expenditures”).

In Syria 90 percent of local governmentrevenue is assigned revenue from thecentral government15, which is trans-ferred from the General Budget of theGovernment, through the Ministry ofFinance, to the Ministry of Local Admi-nistrations. These funds are redistributed

13. See Tosun and Yilmaz(2010b) for more onrecent law changesregardingintergovernmentaltransfer formula.




independent of the collection share ofeach municipality, resulting in a disin-centive to tax collection and to increasedinvestment that would result in greaterlocal tax revenue.

In Palestine, municipalities have beenhistorically less dependent on centralgovernment transfers compared to othercountries in the region due mainly tohistorical, as well as current political andsecurity reasons. In fact, this situation hasnot been affected significantly by theIntifada period, as there is no differencetoday in the amount of central governmenttransfers as a share of total localrevenues, which remains constant atabout 5 percent16 (World Bank, 2006).While there is a formula for the poolingand distribution of transport feesaccording to the 1997 Local GovernmentLaw, the formula is partially applied, anddiscretionary/emergency transfers are adhoc.

In other MEWA countries, there is eitherno formula or only a non-binding arrange-ment for transfers. For example, Iranmakes development transfers from its oilrevenues but allocation rules are notbased on a formula, and are decidedannually by the parliament. Yemen usesshared revenue from 28 different taxes,the religious tax Zakat being the mainsource, but the council of ministers usesonly non-binding guidelines based onpopulation density, financing gap, degreeof deprivation and performance in revenuecollection.

Borrowing

Local borrowing is either nonexistent orvery limited in most MEWA regioncountries with few exceptions. In Turkey,local administrations can engage indomestic and foreign borrowing in

accordance with the municipalities law(Law no. 5393) and the special provincialadministrations law (Law no. 5302).Municipal borrowing is, in principle,capped in proportion to annual revenues(OECD, 2004). Foreign borrowing issubject to approval by the Treasury and itis heavily concentrated on metropolitanmunicipalities. Besides the Treasury, theBank of Provinces, under the Ministry ofPublic Works and Settlement, also play animportant role as a lender of short andlong term loans to local administrations.The outstanding debt of localadministrations to the Bank of Provinceswas about Turkish Lira (TL) 5.51 billion orabout 0.6 percent of GDP in 2009 (TurkishMinistry of Interior, General Directorate forLocal Authorities, 2009). There was asignificant increase in municipal borrowingand indebtedness in the post-1980economic liberalization period, whichsubsided in the years following thefinancial crisis in 2001 but has continuedsince 2005 with a substantial jump in2008. Recent trends in domestic andforeign debt of local administrations inTurkey are shown in Annex 7.2. Foreignfinancing of large projects such as subwayconstruction seems to have played a rolein the rise in foreign debt stock. Also,short term debt has shown the largestincrease compared to long term debt.Overall, total debt has reached about 2.8percent of GDP in 2008. More than half ofthis debt has accrued to metropolitanmunicipalities. The rising indebtedness ofmunicipalities is a serious problem for thelocal administrations as well as the centralgovernment.

Jordan’s municipalities can also engage inborrowing. Cities and Villages Develop-ment Bank (CVDB) is a statedevelopment bank specialized inmunicipal lending, established in 1979 asan independent public institution. It

14. This is based on a 2001questionnaire on 350municipalitiesconducted by theLebanese Centre forPolicy Studies as citedin Bassil and Karam(2009).

15. As of 2003 transferredtaxes in Syria consist ofexport taxes andcustoms duties; 10percent of thecorporate profit tax; 10percent of income tax;7.5 percent of realestate tax; 5 percent ofduties and taxes oncars and theirregistration and 1percent of fuelconsumption tax(Doherty 2005).

16. This is the averagefigure for Palestinemunicipalities in theIntifada period.


245

provides low interest loans forinfrastructure and related projects. Thestanding loans by the CVDB were JD 61.5million (0.4 percent of GDP) in 2010.Municipal borrowing from CVDB is subjectto approval by the Ministry of MunicipalAffairs whereas the board of CVDBapproves borrowing from commercialbanks or any other financial institutions.Any foreign borrowing is subject toapproval by the cabinet. In other MEWAcountries, local borrowing is largely non-existent.

In Lebanon, aid obtained by municipalitiesfrom the central government through theAMF can also take the form of either creditgranted by the Lebanese government oraid given to carry out specific infra-structure projects (public waste, monu-ment restoration, etc.) in one or severalmunicipalities.

Issues, Constraints and Opportunitiesfor Local Government Finance in theMEWA Region

Prevalence of Deconcentrated GovernmentStructure

Local government systems in most ofthe MEWA region, with the exceptionof Turkey and Palestine, can becharacterized as a form ofdeconcentration rather than one ofdevolved local self-government. In general,the public administration system is highlycentralized, equipped with an elaboratesystem of deconcentrated field officesof line agencies. Decisions for themost part, especially service deliverydecisions, are made by the centralgovernment and the role of subnationalauthorities is largely confined to carryingthese out. In all countries, the de-concentrated units of the centralgovernment provide a big chunk of public

services, including health and education,under strict guidance of the centralgovernment. Whereas, decentralizedunits (generally municipalities) performa limited number of functions such asstreet paving and maintenance,construction of local roads, streetlighting, garbage collection, library andpark services, and issuing permits forconstructions.

A good example of the importance (andprevalence) of deconcentrated systems isthe case of Iran which was featured inBox 7.1. In Iran, line ministries providingservices, such as gas, electricity,transportation, education and health, areorganized by sector at the provinciallevel. The municipal sector providesurban municipal services including publichealth, recreational services includingparks, public safety including fire stationsand local transportation including busesand taxis as well as rural municipalservices. While Iran has an elaboratelocal council structure which is animportant component of the subnationaladministration system, the areas in whichthe local council can legislate and passbills is restricted. In fact, in relation tothe entire range of issues that impactlocal economic development, the counciland municipality has a secondary oralmost no role. The restricted interactionsand limited role of elected councilsconstitute a major obstacle to increasedinclusiveness and accountability(Tajbakhsh, 2000). This weak institu-tionalization, in conjunction with theenhanced role of the Ministry of Interior’sMunicipalities Organization is among thefactors that exacerbate Mayor-Counciltense relations, primarily because mayorsfeel increasingly dependent upon thecentral government and consequentlyless accountable to the municipal council(Tajbakhsh, 2000).




Significant Spatial Disparities in Public Expenditures

Another important issue is disparity in pu-blic expenditures found across regionswithin MEWA countries. While some ex-penditure functions are delegated to localgovernments, provinces, and municipali-ties, particularly poorer regions havedifficulty meeting their expenditure res-ponsibilities due to lack of proper revenueassignment and/or inadequate inter-governmental transfers and borrowing.This is an important issue as such dis-parities may lead to deterioration inwelfare in those regions.

For example, in Jordan the municipal ex-penditures per capita in 2007 variedbetween 207 Jordan Dinars (JD) in theprovince of Karak, followed by Mafraq(146 JD), and 7.30 JD in Aqaba or 14.4JD in Jarash, with the median falling at52.7 JD per capita. There are significantvariations in the public expenditures percapita both across and within the threemain regions.17 Household expendituresalso show significant spatial disparities insuch important expenditure items aseducation and medical care. Hence,despite government transfers, significantspatial disparities remain. Similar spatialdisparity is also observed in Syria, wherein 2003, local spending per resident wentfrom 55,523 Syrian Pound (SYP) in AlSweida to 1,575 SYP in Aleppo andDamascus), the average being 2,527 SYPper resident.

Many of the existing revenue systemssimply perpetuate these disparities.Syria’s transfer system for example, whichallocates funding based on tax registeredresidents, does not take into considerationregional development or the presence ofdisadvantaged areas with high populationgrowth.

Ad Hoc Intergovernmental Transfers and Borrowing Practices

As shown in Annex 7.1, MEWA countrieshave largely ad hoc intergovernmentaltransfer systems where most countries inthe region do not use transfer formulas.There is also often significant differencebetween de-jure and de-facto practices.Lebanon is an example where 75 percentof the Autonomous Municipal Fund (AMF)is invested in large scale developmentprojects by the central governmentthrough the Council of Development andReconstruction (CDR), and other develop-ment institutions, and hence nottransferred to municipalities. The majorityof local governments in Lebanon do nothave sufficient resources to meet theirneeds. This has led in some cases tounsustainable borrowing as depicted in thecase of the urban community of d'AlFayhaa in Box 7.2.

Local borrowing practices are also largelyad hoc. An important problem is lack oftransparency in local borrowing whichcould lead to serious indebtedness andfinancial crisis. Faced in many cases by animbalance between responsibilities andrevenues along with unpredictable orreduced transfers the environment is ripefor unsustainable debt. Local admi-nistrations in Jordan and Turkey haveengaged in significant borrowing to meettheir rising investment needs. In Jordan,there has been a substantial growth incapital expenditures as the share of capitalexpenditures rose from 41 percent of totalexpenditures in 2003 to about 58 percentin 2008, reaching a share as high as 65percent in 2007. Along with this increase,Jordan municipalities have run substantialfiscal deficits recently with an averagedeficit of 0.33 percent of GDP. Fiscaldeficit in 2007, when capital expendituresshowed the largest increase, was 0.62

17. Note that theexpenditure numbersfor the Amman regiondo not include theGreater AmmanMunicipality whichexplains the low percapita figures for thatregion. Ammanmunicipality has aspecial status and is notcounted among the 93municipalities.Separate data onAmman municipalitywas not available.


247

percent of GDP. Following different sources,this debt is due in part to the fact thatmost municipalities in Jordan, particularlythe smaller ones, do not have theresources to pay their employees andmust turn to borrowing to pay theircurrent expenditures (UCLG, 2007). Arecent report by the Jordan Ministry ofMunicipal Affairs shows that municipalindebtedness is an important problem inJordan for which a municipal debtreduction account has been established bythe central government (MoMA, 2009).

In Turkey, local administrations had afiscal deficit of TL 7.1 billion (0.75 percentof GDP) in 2008 while the totaloutstanding debt, excluding deferredpayments, of Turkish local administrationswas 2.8 percent of GDP in that same year

(Turkish Ministry of Interior, GeneralDirectorate for Local Authorities, 2009).An impo r t an t i s s ue he r e i s t ha tmunicipalities are not required to havebalanced budgets and thereby rely oncentral government to finance their debtin the case of insufficient resources.

The cases of Turkey and Jordan show thatthese countries are suffering from fiscaldeficits and indebtedness at the local levelwhich is at least partially driven by ad hoclocal borrowing practices and soft budgetconstraints.

Internal and External Conflicts andDecentralization

Conflicts have particular relevance to theMiddle East and Western Asia region as


Box 7.2: The Case of the Urban Community of d'Al Fayhaa

The Urban Community of d'Al Fayhaa is made up of the municipalities of Tripoli, El Minaand Beddaoui and has a total population of 340,000 inhabitants.

The revenue of municipal federations in Lebanon is, by law, made up of: 10 percent ofthe revenues of member local governments, an additional percentage assigned from thebudgets of local governments benefiting from common projects undertaken by theUrban Community and 25 percent of the taxes deposited in the AMF.

In practice, due to the financial constraints of most municipalities, they are neither ableto transfer 10 percent of their revenues to the municipal federations nor make addition-al contributions for service delivery in their area. Transfers from the AMF (which as men-tioned above are closer to 2 percent than the legislated 25 percent) are also limited inthat only 40 percent of funding can be used to pay for service provision (while 60 per-cent must be reserved for staff salaries).

Even if it is one of the most important municipal federations in Lebanon, the Urban Com-munity of d'Al Fayhaa is not able to finance the common services it provides (garbagecollection, management of slaughter houses, landfills, etc) which have been contractedout to private companies, nor to cover their entire staff salaries. The debt of the UrbanCommunity increases yearly and all new projects of common nature have been cance-lled in the last 6 years.



Box 7.3: Conflict, Government Structure and the 1997 Law on Local Authorities in Palestine

The local government system in Palestine reflects the realities of the Israeli occupation. Theoverriding concerns in the design of local government system have always been providingemergency services and security through central control. As a result, laws, the political system,administrative arrangements and development practices of local governments are gearedtowards these objectives rather than toward providing services to local communities.

Prior to the 1994 Oslo Peace Accord, in the absence of a sovereign state, Palestinian localgovernments have had to fend for themselves in providing services to local communities. TheMinistry of Local Governments (MOLG) was established in 1994 to help build an effective localgovernment system. However, this current legal framework in Palestine has assigned the cen-tral government strong formal controls over local governments (World Bank, 2006).

The Law on Local Authorities of 1997 (LLA) provides the legal basis for the current local govern-ment system. LLA draws heavily on other regional country legislative frameworks, particularlythat of Jordan. LLA grants significant powers to the central government, primarily Ministry ofLocal Government in its role as the sector regulatory agency, including provisions for approvalsof a wide range of activities of local governments and claw-back clauses where autonomyappears to be granted (World Bank, 2006).

The LLA provides the legal basis for municipal expenditure responsibilities andrevenue-raising authorities. However, there is a significant mismatch between the legalassignments to municipalities and the reality on the ground (World Bank, 2006). Theabsence of an effective public administration system compels the larger municipalities toassume responsibilities that are not necessarily assigned to them by law, such as fire figh-ting service and maintenance of school buildings (World Bank, 2006).

The LLA grants the central government extensive powers over municipal governments in termsof control over revenue sources. They have to obtain the approval of the central government insetting the tax rates and defining the revenue bases. Mostly within the confines of the centrallydefined tax and fee bases, assessment strategies, and rates, the local governments are provi-ded with revenue sources such as property taxes, building permits and utility revenues. Theyare also allowed to perform certain public functions and market services to raise additionalrevenues. Yet, in practice, the municipalities invent their own ways of raising revenues and theyoften find ways of rationalizing such practices in old laws (World Bank, 2006).



it’s one of the most conflict-riddenregions in the world.18 Palestine leads allother countries in the total number ofconflicts and, particularly, in number ofminor and internal conflicts. Iran and Iraqlead in the number of major conflict (orwar) years and follow with Turkey andPalestine in internal conflict years.Studies show evidence of strong negativespillovers from conflicts and point toexternal conflicts as a significant obstacleto a decentralized government structurein the region (Tosun and Sen, 2008 andTosun and Yilmaz, 2010a). The specialcircumstance of Palestine is described inBox 7.3 to show the unique form ofgovernment structure in Palestine in theface of a persistent conflict environment.

Capacity Building and CommunityParticipation

Local government officials across the regiongenerally express their need for more dataand information. However, they tend toemploy insufficiently trained staff, andtherefore do not have the capacity to workat a policy level and use local financial data.They simply do not have enough resources(time, money, or people). The need for localdata depends not only on the size of localgovernments, but also on the serviceresponsibilities assigned to them. Betterinformed and trained decision makers atthe local government level would be animportant element in implementing adecentralization strategy. In Lebanon forexample, local tax collection is hamperedby a slow tax evaluation process, non-computerized accounting practice, and alack of tax collectors (Atallah, 1999). Similarproblems are found, at least to some extent,across the countries surveyed.

There are also good practices in theregion, particularly regarding capacitybuilding and community participation

as depicted in the cases from Jordan,Syria, and Turkey. For example, EskisehirMetropolitan Municipality in Turkey is agood example of a local administrationthat gives much importance to capacitybui ld ing and tra in ing of munic ipa lemployees.19 Fuheis municipal ity inJordan provides a successful case of localcommunity involvement.20 Some of theinitiatives by the municipality are localcommunity involvement in the planningprocess (through community proposals),formation of volunteer action committees,capacity building through strengtheningof the skills of staff and departmentheads by holding workshops and trainingsessions, partnerships with the privatesector, access to information (such asfinancial statistics) and use of computertechnology –particularly GeographicInformation Systems (GIS)–, and finallyi n t eg ra t i on o f gende r and you thconsiderations in the local decision-making process. Box 7.4 describes twoother examples of good practices regard-ing capacity building.

Conclusions

For a variety of reasons (such as tradition,history and culture), the responsibilitiesassigned to local governments in theMEWA regim have not been as extensiveas those in many other parts of the world.In their efforts to reform the local govern-ment sector, governments should recog-nize that decentralization requires sharingof fiscal roles and responsibilities betweencentral and local governments accompa-nied by a robust capacity to deliver ser-vices both centrally and locally. Thechallenge is to determine how to sort-outthe responsibilities and financing amongdifferent types of local governments.

One possible suggestion to allow the de-centralization process to move forward

18. See Milton-Edwardsand Hinchcliffe (2004)for a chronology anddetailed discussion ofconflicts in the MiddleEast since 1945.

19. Eskisehir is a citysynonymous withhigher education and isvirtually a city ofstudents. It is home toAnadolu University,which is the largestuniversity in Turkey interms of studentenrollment. The mayorof EskisehirMetropolitanMunicipality, formerlypresident of theAnadolu University,maintained a goodworking relationshipbetween themunicipality and theuniversity. It is notedthat, all contractemployees in EskisehirMetropolitanMunicipality have atleast Bachelor degrees.http://www.eskisehir-bld.gov.tr/cityiseskisehir/index.html

20. Fuheis is a smallmunicipality about15km northwest ofAmman.



Box 7.4. Good Practice in Capacity Building: Jordan and Syria

Devolution of Property Tax Management to Municipalities in JordanIn 2004 the Jordanian Ministry of Finance (MOF) and the Ministry of Municipal Affairs (MoMA)began a project to transfer the responsibility for property tax collection from the MOF to themunicipalities.

The project, funded by the UNDP, was divided into two phases (phase 1: 2004-2010, phase 2:2010- 2012) to allow new strategies to be tested in larger municipalities before beingimplemented in smaller ones with lower capacity.

The project involves:• Conducting legal and procedural review

• Establishing integrated data base and network for property tax managementcenters, across the country

• Capacity building for local government staff (600 employees)

• Awareness campaigns for citizens and tax payers

• Development of administrative set up, and restructuring of property taxmanagement units.

To date 47 (out of 93) municipalities have been transferred the responsibility of property taxcollection (the remaining 46 are targeted in phase 2 of the project). Large municipalities collectboth for themselves and on behalf of the smaller neighboring municipalities.

An increase has been noted in property tax collection since the beginning of this project (Al-Hajaj, 2010).

Municipal Administration Modernization Program in SyriaThis joint initiative funded, by the Syrian Arab Republic and the European Union, has as its longrange purpose to improve the quality of life of people in urban centers across the country. Withan initial focus in 6 cities there are currently projects underway in Damascus, Aleppo, Lattakia,Tartous, Homs, Der Zour and Palmyra.

The goal is to seek out new frameworks and best practices in the long term management ofurban growth. The program will deliver a series of interconnected action plans focusing onlegislative, financial and management reform. Specific areas targeted by this program include:Decentralized governance; supporting local decision making; institutional development;financial resource management; property management; local development; twinningagreements and partnerships; urban planning and informal settlements; geographicinformation systems (GIS); public private partnerships; traffic and transportation; solid wastemanagement; Local Agenda 21 and gender.

Source: http://www.mam-sy.org/index.php?p_id=11&lang=en


251

across the region is the application ofasymmetric decentralization. Specificcriteria can be set to classify localgovernments into different categories thathave asymmetric taxing and spendingresponsibilities and borrowing privileges. Thiswould give impetus to decentralizationreform processes by which regionalgovernments (governorates) and localgovernments might be empowered withincreased autonomy in expenditure andrevenue decisions that remain in line withthe i r capac i ty to meet these newresponsibilities and build toward greaterones. However, there is also a need forsystematically reviewing legal and re-gulatory standards for “sorting out” rules andresponsibilities among different types andlevels of governments. In addition, existingand future revenue commitments on the partof central governments must be honoredboth in quantity and timeframe to allow localgovernments to plan for and deliver theirmandated services.

In the long run, the governments in theregion need to devolve expenditureresponsibilities further to local governmentswhile making them fully accountable beforetheir respective constituencies for policyresults, in terms of their effectiveness andefficiency in delivering quality publicservices. To this end, they should considers t r e n g t h e n i n g l o c a l g o v e r n m e n taccountability mechanisms by systemiccollection, analysis, and dissemination ofinformation about local fiscal performanceand compliance with financial and policygoals. Such information is essential both toinformed community participation throughpolitical process and to the monitoring ofmunicipal performance by the centralgovernment.

In reforming local government systemsthe most challenging task for thegovernments in the region would be

restructuring the overall revenue systemin a manne r t ha t p r ov i de s l o ca lgovernments “fiscal space” to strengthenown revenue and expenditure arrangements.The governments should first make surethat adequate steps are taken to establishaccountability mechanisms, then boostrevenue autonomy by giving local govern-ments adequate decision-making powerson tax rates and the determination ofsome tax bases in order to improvebudgetary predictability. They shouldgradually lift central governmentcontrols on local fees and taxes aftermaking sure that local revenue generationis maintained.

The governments in the region shouldcons ider estab l i sh ing a mul t i l eve lgovernment coordinating body that wouldoperate across the different tiers ofgovernment to launch fiscal decentralizationreforms. This coordinating body would bea mechanism for the central governmentto improve the design and gauge thedirection, pace, and extent of decentraliza-tion, and disseminate information, providetraining and directly engage municipalgovernments in the decentralizationprocess. This body would be instrumental indeveloping institutions for intergovern-mental cooperation and dialogue. It will beespecially central to increasing local publicexpenditure efficiency in areas of concurrentexpenditure responsibilities and creatingstrong incentives (financial and legal) topromote cooperative arrangements amonglocal governments for service delivery.

In public service delivery, the govern-ments could explore the participation ofthe private sector in both financing anddelivery of public services to improve theoverall efficiency of local government ex-penditures. Inter-municipal cooperationand collaboration with the private sectormight be a means of overcoming in-




efficiencies associated with small size ofmunicipalities. However the choice ofmanagement model must remain a localone to ensure the public appropriation andapproval of the chosen methods.

T h e g o v e r n m e n t s s h o u l d s t u d yrationalizing the transfer system so as tomake it a more effective instrument forthe implementation of policies of nationalinterest at the local level and reducespatial fiscal disparities. This wouldinclude an examination of both conditionaland unconditional transfer systems. Theyshould establish transparent rule-basedtransfer systems with explicit formulas forequalization. They should explore ideasfor a combination of unconditional andmatching grants that would promotemunicipal governments to exploit theirrevenue bases and improve the efficiencyof tax collection.

Another important element in the imple-mentation of a decentralization strategy iscapacity building through investment inboth staff capacity and information tech-nology. The region needs more capacitybuilding initiatives that are supported bydomestic and international funds andagencies.

Central governments in the region shouldalso credibly commit to the strengtheningof local government management capacityboth in terms of long term budget planningand the sustainability of debt. This willinvolve enforcing hard budget constraintsfor local governments and ensuring thatlocal governments receive the funds theyare assigned under law and that theseincome sources are sufficient to allowthem to provide their mandated services.This process will be important, particularlyfor local government borrowing as formany local governments in MEWA,unsustainable debt has become one of the

only options for continuing to provideservice. In some cases, lack of long-termcapacity planning has led some localgovernments to accumulate significantand sometimes unsustainable debt. Whilethe central government should setresponsibilities for local government, andensure matching long-term predictablefunding, it should also refrain from bailingout over-indebted local governments. Thismay mean letting some local governmentsfail first and then be subject to centrallyled financial restructuring rules.

Finally, a critical issue in the MEWA regionis the role of conflicts in centralization. Itseems external conflicts set a major ob-stacle to the decentralization process andoften is a key impetus for recentralizationmovements. Regional conflict preventionshould be seen as a regional or inter-national public good of which the collectiveprovision would ease the burden on thecentral and local governments of individualcountries.

This chapter presented an overview andcomparison of local government finance inthe Middle East and Western Asia (MEWA)countries followed by a discussion on someof the issues that are important in localgovernment performance. The chapteralso included region-wide proposals forreform. As has been shown throughout thechapter, MEWA countries have verycentralized government structures. Someof the challenges facing local governmentfinance and decentralization as discussedin the chapter are the preferences onthe part of central governments fordeconcentrated government units to pro-vide local services than the devolution ofpowers to the local level; spatial disparityin public expenditures that require abetter (formula based) intergovernmentaltransfer system; ad-hoc intergovern-mental transfer and borrowing practices;



internal and external conflicts that couldlimit further fiscal decentralization and drawresources away from local governmentfinances; and capacity building andcommunity participation. Despite thesechallenges and the overall bleak picture ondecentralization in the MEWA region, there

seems to be a trend towards reforminggovernment structures to allow moreflexibility in and less control of local govern-ments by the central authority. It is hard totell, however, if that trend will continue inthe future given the turbulent political andmacroeconomic environment in the region.


0w2010 08 Noram DEFcarta ang 16/9/11 19:13 Página 254

8. NORTH AMERICA

WWIILLLLIIAAMM FF.. FFOOXX

UUNNIIVVEERRSSIITTYY OOFF TTEENNNNEESSSSEEEE,, UU..SS..AA..

EENNIIDD SSLLAACCKK

UUNNIIVVEERRSSIITTYY OOFF TTOORROONNTTOO,, CCAANNAADDAA


257

B oth Canada and the United States areconstitutional democracies with a federal

structure of government. Both countries aregeographically large but the U.S. has amuch larger population than Canada (seeTable 8.1).

Canada is a federation with three levels ofgovernment: the federal government, tenprovincial and three territorial govern-

ments,1 and almost 4,000 local govern-ments (see Table 8.2). Canada'sConstitution lists the jurisdictions overwhich federal and provincial governmentshave lawmaking authority. Local institutionsare listed as one of the responsibilities of theprovinces. Each province has separatelegislation governing municipalities in theprovince and, as a result, there aredifferences across the country.

1. The three territoriesgovern the sparselypopulated northernpart of the country.Unlike the provinces,they have noconstitutionalstanding, are underthe jurisdiction of thefederal governmentand rely more heavilyon federal funding.

Canada United States

Population (2008) 33,311,400 304,059,724

Area (km2) 9,984,670 9,161,930

Population density (population per km2) 3.34 33.19

Urban population (%) (2005) 80.1 79.2

GDP per capita(USD) (2008) 45,127 46,914

Sources: Statistics Canada, CANSIM, Table 051-0001 Estimates of Population, Canada; CANSIM Table 3800030, GrossDomestic Product and Gross National Product at Market Prices and Net National Income at Basic Prices, annually; UnitedNations, World Urbanization Prospects: The 2007 Revision Population Database.

Table 8.1: Country Characteristics


National Federal government Federal government

Intermediate 10 provinces and 3 territories 50 states and 1 autonomous city

Local

- Upper tier 124 regions, 106 counties and municipal districts 3,033 counties

- Lower tier 3,524 cities, towns, villages, townships, rural 36,011 cities, towns, and townships

municipalities, district municipalities,

hamlets, parishes, etc.

- School boards 375 14,561

Sources: Information provided by the Federation of Canadian Municipalities and Canadian Education Association, PublicEducation in Canada: Facts, Trends, and Attitudes, Toronto: 2007.

Table 8.2: Government Characteristics


NNOORRTTHH AAMMEERRIICCAAUnited Cities and Local Governments258

The U.S. government structure is composedof one federal government, fifty states,and 89,476 local governments (in 2007).State constitutions and statutes are theprimary determinants of local governmentstructure, which means wide differencesexist across the country.2 Localgovernments were generally created toassist state governments in the delivery ofservices. The specific role of localgovernments is given through the stateconstitutions in some cases and by statutein others.

In contrast to the U.S., direct relations betweenthe federal government and local governmentsin Canada are limited. Cities remain creaturesof the provinces. The federal government cangive money to cities but it cannot change theirexpenditure responsibilities or their revenue-raising tools.

Local Government Finances in theRegion

What is true of many countries is also true ofthe United States and Canada: summarizingthe relative service delivery roles playedby the federal, state/province, and localgovernments is not simple because manyservices have been unbundled, with eacharena of government taking responsibility fordifferent components. Data on expendituresand revenues tell part of the story but fail toexplain fully the nuances of the intergovern-mental relationships (see Figure 8.1).

Assignment of Service Responsibilities

The powers and responsibilities of govern-ments in Canada were set out in the BritishNorth America Act, 1867 and have beenrevised since then through judicial

2. See Benton, J. Edwin.2009 “Trends in LocalGovernmentRevenues: The Old,The New, and TheFuture,” presented atthe 2009 LincolnInstitute of LandPolicy Conference,“The ChangingLandscape of LocalPublic Revenues,”CambridgeMassachusetts, June2009.

3. Simeon, Richard andMartin Papillon,“Canada,” in Majeed,Akhtar, Ronald L.Watts and Douglas M.Brown (eds.)Distribution of Powersand Responsibilities inFederal Countries.(Montreal andKingston: McGill-Queen’s UniversityPress for the Forum ofFederations andInternationalAssociation of Centresof Federal Studies,2006)

General Adminstration General

AdminstrationOtherOther

Expenditure

Sale of Goods and Services

Investment Income

0%

100%

State

Local

Federal

Revenue Canada USA

Transfers

Own Taxes

Transfers

Own Taxes

OtherRevenue

Other Revenue

CPP/QPP*

Local

Provincial

FederalEducation

Education

Debt charges

Social services

Protection

Transport

Sanitation

Sanitation HealthHealth

Figure 8.1: Income and Expenditure Breakdown

Note: illustration of data presented in Tables 8.3, 8.4, 8.5.* CPP and QPP are the Canada Pension Plan and the Quebec Pension Plan respectively


259

interpretation, constitutional amendment,fiscal arrangements, and intergovernmentalnegotiations.3 Provincial governments areempowered to control regional and localaffairs and have exclusive responsibility forservices such as education, health, socialservices, property rights, administration ofjustice, local public works (including roads,waterways, natural resources andenvironmental matters), and municipalgovernment.

Although Canada is a highly decentralizedcountry in terms of federal and provincial

powers, it is much more centralized withrespect to provincial and local powers. Localgovernments are often referred to as“creatures of the provinces” because theyhave no original powers in the constitutionand enjoy only those powers that aredelegated to them by the provinces.Nevertheless, municipalities are largelyresponsible for delivering such importantservices as police and fire protection, roadsand transit, water and sewers, solid waste,recreation and culture, and planning. Table8.3 shows the distribution of expendituresat the local level.



General administration expenditures

Education

Health

Sanitation (environment)

Transport

Protection (fire and police)

Social services

Resource conservation and development

Recreation and culture

Housing

Regional planning and development

Debt charges

Other expenditures

Total

Note: Local government revenues include municipalities and school boards in Canada.

Sources: Statistics Canada 2007. Table 385-0003 - Local government revenue and expenditures for fiscal year endingclosest to December 31, CANSIM (database).

U.S. Census Bureau. Table 1. State and Local Government Finances by Level of Government and by State: 2005-06.

Table 8.3: Distribution of Local Government Expenditures (%)

6.1

38.6

1.5

11.1

12.3

9.8

5.4

1.3

7.6

2.1

1.2

2.6

0.1

100.0

4.6

38.5

7.5

4.1

5.2

9.3

3.3

0.5

2.1

2.6

n.a.

3.4

18.7

100.0



The federal versus state/local role in theUnited States is generally based on historyand practice, and most service delivery is notspecifically articulated in the U.S.Constitution.4 The assignment of serviceresponsibilities at the state versus the locallevel can vary widely across states based onconstitutional and statutory provisions,making generalizations difficult. Some stateconstitutions provide specific assignments.For example, many state constitutions assignresponsibility for education to stategovernments, but every state except Hawaiieither assigns or delegates most provision ofprimary and secondary schools to localgovernments. The national, state, and localgovernments frequently share responsibility,

at least to some extent, for delivering mostservices. The federal government generallyplays a much smaller overall role in directservice delivery than do state and localgovernments, but the federal governmentoften has important influence over servicedelivery. Federal grants, loans, and costsharing that come with various restrictions, aswell as federal laws and regulations, arefrequently used to leverage federal prioritiesfar beyond the narrow area in which thefunding is provided.

State and local governments have nearlyexclusive responsibility for a number ofservices, including fire, education, libra-ries, solid-waste management, sewerage,

4. The power of theU.S. Congress todeliver nationalservices beyond anarrowly set ofenumeratedauthorities wasarticulated in anearly court decision(McCulloch v.Maryland [1819]).This opinion alsolimited the states’ability to tax thenational government.

Total government amount Federal Provincial Local CPP/QPP*($ USD millions) (%) (%) (%) (%)

Total revenue

Tax revenue

Property

General sales

Selective sales/excise taxes

Individual income

Corporate income

Motor vehicle license

Other taxes**

Charges and miscellaneous revenue

Health and drug insurance premiums

Contributions to social security plans

*CPP and QPP are the Canada Pension Plan and Quebec Pension Plan respectively.

**Other taxes include mining and logging taxes, taxes on payments to non-residents, payroll taxes, natural resource taxes and licenses, and othermiscellaneous taxes.

Source: Statistics Canada, CANSIM database, Table 385-0001, Consolidated Federal, Provincial and Local Government Revenues and Expenditures.

Table 8.4: Summary of Federal, Provincial, and Local Government Finances, Canada 2007-08

596,794

424,847

49,711

67,693

36,864

181,509

61,909

3,279

23,892

104,652

3,212

64,072

40.2

48.1

0.0

49.0

35.1

61.3

63.3

0.0

32.5

14.6

0.0

32.1

43.5

42.2

18.7

50.9

64.9

38.7

36.7

100.0

64.5

62.4

100.0

18.3

10.3

9.7

81.3

0.1

0.1

0.0

0.0

0.0

3.1

19.1

0.0

0.0

6.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

3.9

0.0

49.6


261

water supply, and transit. Much of healthcare is provided through the privatesector, but public-sector hospitals andclinics are also common, and regulatoryresponsibility is vested in all three ordersof government. Water, electricity, gassupply, and sewerage are exclusive stateand local responsibilities, although thefederal government plays some regulatory

and fiscal roles in all of these fields. Thepublic sector produces the services insome cases, and the private sector does soin others.

Tax Assignment

Under the Canadian Constitution, the fe-deral government has unrestricted powers


Total government amount State government Local government Federal government (USD billions ) (%) (%) (%)

Total revenue

Tax revenue

Property

General sales

Selective sales/excise taxes

Individual income

Corporate income

Motor vehicle license

Other taxes*

Charges and miscellaneous general revenue

Utility revenue

Insurance trust revenue

Unemployment compensation

Workers' compensation

Old-age and survivors insurance

Hospital insurance

Disability insurance

* Estate and Gift Taxes and Customs Duties and Fees

Sources: http://www.whitehouse.gov/omb/budget/fy2008/pdf/hist.pdf and

http://www.census.gov/govs/estimate/0600ussl_1.html

Table 8.5: Summary of U.S. Federal, State, and Local Government Finances, U.S. 2005-06

5,586,902 31.7 25.2 43.1

3,303,218 29.3 23.2 47.5

359,109 3.3 96.7 0.0

282,179 80.3 19.7 0.0

203,897 52.1 11.6 36.3

1,312,507 18.7 1.7 79.5

406,846 11.7 1.3 87.0

20,520 92.7 7.3 0.0

134,668 39.9 21.0 39.1

583,493 43.8 48.5 7.7

125,265 12.6 87.4 0.0

1,256,702 29.2 4.2 66.7

36,989 99.7 0.3 0.0

21,514 100.0 0.0 0.0

520,069 0.0 0.0 100.0

177,429 0.0 0.0 100.0

88,313 0.0 0.0 100.0



to levy taxes with the exception of taxeson provincial lands and property.Provincial governments are limited todirect taxation (income, sales, property,and commodity taxes) within provincialboundaries. The provinces were not giventhe power to inhibit interprovincial tradethrough the use of indirect taxes. Taxeslevied by provincial and local governmentsinclude the property tax. Provinces havetheir own taxes such as natural resourcesand capital taxes.

Municipal governments are largely restrictedto property taxes although somemunicipalities are permitted to levy selectivesales taxes. In addition to a property tax,Toronto levies a land transfer tax, a vehicleregistration fees, and a billboard tax andhas the authority to tax alcohol, tobaccoand amusements. Table 8.4 shows therevenues of the federal, provincial, and localgovernments in Canada.

The U.S. Constitution imposes relatively fewlimitations on taxation at the U.S. federal(Article 1 Section 8) and subnationalgovernment levels. A prohibition againsttaxing exports from a state is the onlynotable explicit restriction. The prohibitionagainst state and local taxes distortinginterstate commerce arises from thedormant commerce clause and is a verysignificant limitation on the ability of statesto tax. In turn, states generally determine,either statutorily or constitutionally, theauthority of local governments to levytaxes. For example, many states havelimited the annual growth rate inassessments for property tax purposes.

The federal government raises the greatestshare of revenue (see Table 8.5). A limitedform of specialization has developed by taxsource, though each level of governmentuses multiple tax sources. Localgovernments raise almost all of property tax

revenue and most utility revenues.Otherwise, local revenues are modestshares of total receipts for the variousrevenue sources.

Local Government Revenues

Table 8.6 shows the distribution of localgovernment revenues for Canada and theU.S. broken down by municipal revenuesand school board revenues. For municipalgovernments in Canada, by far the largestsource of revenue is property and relatedtaxes which includes the general propertytax as well as land transfer taxes (which arelevied by municipal governments in only twoprovinces), payments in lieu of propertytaxes on federal and provincial/territorialgovernment properties, lot levies (ordevelopment charges), and specialassessments. Property and related taxesaccount for more than half of municipalrevenues. Municipal governments levy fewother taxes and those taxes (such as taxeson hotels, restaurant meals, and liquor)result in limited revenues. The heavyreliance on property taxes has meant thatCanadian municipalities have notexperienced significant revenue losses as aresult of the recent economic crisis.

Municipal governments receive less than 20percent of their revenues from provincialand federal transfers with the bulk of thetransfers coming from the pro-vincial/territorial governments. The majo-rity of grants are conditional (specificpurpose). User fees are also an importantsource of revenue for municipal govern-ments. For school boards, the main sourceof revenue is provincial transfers whichaccount for almost three-quarters of totalrevenue. The property tax is the nextlargest revenue at 20 percent althoughthese estimates do not reflect provincialproperty taxes for education in someprovinces.


263

5. Hoene, Christopherand Michael A.Pagano, Cities andState FiscalStructure, ResearchReport on America’sCities, (Washington,DC: National Leagueof Cities, 2008).

6. State and localincome and sales taxdata are taken fromMikesell, John,”StateSales Taxes in theGreat Recession,”State Tax Notes, July19, 2010.


U.S. local governments generally raiserevenue using four sources: property taxes,sales taxes, income taxes, and user fees andcharges (Table 8.6). Different types of localgovernments may be empowered to usedifferent tax sources. For example, no stateallows cities unfettered use of all threetaxes, but at least five states allow somecities a form of access to all three taxes5.Property taxes are used almost exclusivelyby local governments and generate nearlythree-fourths of local revenues. Some states

also raise modest revenue with propertytaxes. Most local governments with taxingauthority can levy property taxes. Forexample, municipalities in all states exceptOklahoma are empowered to raise propertytaxes. Thirty-six states allow local salestaxes, and fourteen permit local incometaxes, which are often wage taxes ratherthan broad-based income taxes.6 Aboutone-half of local sales tax revenue iscollected by municipalities and somewhatmore than one-third by county


Municipal School Total Local Government Local Government

Own taxes

- Property and related taxes

- Consumption taxes

- Income

- Other taxes

Conditional transfers

- Federal

- Provincial

- Municipal

Unconditional transfers

Total transfers

Investment income

Sales of goods and services

Other revenue

Total revenue

Sources: Statistics Canada 2007. Table 385-0009 - School board revenue and expenditures, year ending December 31, CANSIM (database); Table 385-0024- Local general government revenue and expenditures, current and capital accounts, year ending December 31.

U.S. Census Bureau. Table 1. State and Local Government Finances by Level of Government and by State: 2005-06.

Table 8.6: Distribution of Local Government Revenues (%)

50.7

0.1

0.0

1.4

16.0

1.6

14.4

n.a.

2.9

18.9

5.3

22.2

1.5

100.0

20.6

0.0

0.0

0.0

74.0

0.2

73.4

0.4

n.a.

74.0

0.3

5.0

0.1

100.0

39.3

0.1

0.0

0.7

38.7

1.7

40.4

3.3

15.2

0.9

100.0

24.7

5.6

2.0

2.1

33.9

2.4

8.1

21.3

100.0



governments. Special districts and schooldistricts collect modest amounts. Localincome taxes are even more concentrated,with about 80 percent collected bymunicipalities. Still, municipalities obtainonly 26.6 percent of their tax revenues fromsales and income taxes, and countiesreceive only about 20.6 percent of theirrevenues from these sources, but thisdiffers dramatically across the country.Local governments in some states, suchas Arkansas and Louisiana, and somelarge cities, such as New York City andPhiladelphia, use sales or income taxesheavily. But, many other cities raise verylittle revenue from these sources.

The recent economic crisis had a veryuneven effect on local governments. Pro-perty tax revenues rose 6.0 percent onaverage during fiscal year 2009 andcontinued to rise during the last half ofcalendar year 2009. On the other hand,combined state and local personal incometax revenues fell 13.8 percent and salestaxes 4.5 percent during fiscal 2009. Thus,most local governments that are verydependent on property taxes have notexperienced substantial revenue losses, butthose dependent on sales and income taxeshave been more severely impacted.

U.S. cities also have additional powers toprovide tax incentives to attract investment:for example, tax increment financing, whichis used in most U.S. states but is onlybeginning to be used in a few cities inCanada.7 Development charges (“impactfees”, “lot levies”) are levied by municipalitiesin most Canadian provinces.

State and local user fees and chargesgenerate 19.3 percent of total revenues.Local governments collect 25.7 percent ofrevenues from user fees versus 14.2percent for states. Local governmentcharges are diverse, with hospital services

representing 26.3 percent and sewerageanother 17.1 percent. Local governments,and to a much lesser extent states, deliver anumber of utility services includingelectricity, water, natural gas and transit.These generated $125.3 billion in revenueof which 87.4 percent is collected by localgovernments.

Municipal Borrowing

Municipalities use debt financing to pay forat least part of the costs of major publiccapital works. Repayment of borrowedfunds comes from operating revenues suchas property taxes and user fees. U.S. citiesborrow more heavily than Canadian cities,on average.8

The amount that municipalities can borrowfor capital projects is almost alwayscontrolled by the province.9 These controlsare in place because cities are “creatures ofthe province” and provinces do not wish tobe responsible for unlimited borrowing andpossible repayment of debt. Moreover,unrestricted access to capital marketsmight, in some circumstances, crowd outprivate sector borrowing. The methodsthat provinces use to control municipallong term borrowing vary and may becategorized as: permitting borrowing onlyfor provincially-approved capital projects;requiring prior approval of provincialauthorities for borrowing; requiring priorapproval (through a referendum) by localtaxpayers for borrowing above a specifiedlimit; restricting annual debt servicingcosts to some percentage of municipalown-source revenues; restricting theamount of debt to some percentage ofassessed property values; and permitting(or requiring) borrowing from a provinciallycontrolled “municipal fund.”

U.S. municipalities are permitted to issuerevenue bonds but Canadian municipalities

7. Although legislation inone province(Manitoba) permitsthe use of TIFs, noneexist in that province.Recent legislation inAlberta permitsmunicipalities to use aform of TIF known asthe “communityrevitalization levy.”The Province ofOntario hasintroduced legislationto allow Toronto touse TIFs in twoneighborhoods.

8. U.S. cities haveUSD1.3 trillion in longterm debtoutstanding andUSD22 billion in shortterm debt comparedto only C$30 billion inCanada. See TDEconomics, Mind theGap: Finding theMoney to UpgradeCanada’s AgingInfrastructure,Toronto, May 2004:16 andhttp://www.census.gov/govs/estimate/0600ussl_1.html. Even account-ing for the differencein population size,U.S. cities borrowmore than Canadiancities, on average.

9. Bird, Richard, M. andAlmos Tassonyi."Constraints onProvincial andMunicipal Borrowingin Canada: Markets,Rules, and Norms ."Canadian PublicAdministration 44,2001: 84-109.


265

are not. Revenue bonds, unlike generalobligation bonds, are legally secured by aspecific revenue source. For example, arevenue bond issued by a water utility couldbe backed only by the specific revenues ofthat utility. The advantage is that these bondspromote full-cost pricing of services and shiftthe risk to the investors. The disadvantage isthat the market interest rates are often higherbecause the loans are not backed by the “fullfaith and credit” of the governments.

Municipalities in the U.S. are also permittedto issue tax-exempt bonds but these are notpermitted in Canada. Tax-exempt bondsenable municipalities to issue bonds at alower interest rate because the interestearnings are tax-free. Tax-exempt bondstend to work best in larger municipalitiesthat have access to capital markets andtheir application to smaller municipalitieshas been limited.

Intergovernmental Grants

For Canadian municipal governments,federal and provincial transfers accountfor less than 20 percent of theirrevenues. Transfers to school boards,on the other hand, account for three-quartersof their revenues. The bulk of transfers tomunicipalities and school boards comefrom the provincial/territorial governmentswith federal transfers only accountingfor 1.6 percent of municipal revenuesand 0.2 percent of school board revenues,respectively. Federal funding is usuallychanneled through the provinces butthere are some exceptions. The relativelygood fiscal situation of the federalgovernment in recent years, combinedwith the fiscal pressures on provincialgovernments (particularly to fund health),has resulted in a “… growing range offederal government initiatives that bypassthe provinces and deal directly withcitizens and cities.”10

The federal government responded to localdemands by introducing a number ofinitiatives directed to cities (and other muni-cipalities), including such measures as aproposal to share the equivalent of up to 5cents per liter of gasoline on a (roughly) percapita basis to municipalities,11 a rebate onthe Goods and Services Tax (GST) formunicipalities,12 more funding for publictransit and housing, and a commitment torenew existing infrastructure fundingprograms.

Provincial transfers are a combination ofspecific-purpose and general-purpose. For allof Canada, provincial/territorial unconditionalgrants accounted for less than 3 percent of allrevenues in 2007, whereas provincialconditional grants accounted for more than15 percent. In terms of provincial conditionaltransfers, as in the U.S., the largest transfersby far are for primary and secondaryeducation. The predominant transfers formunicipal purposes are for transportation andwater and sewers. The dependence on con-ditional transfers by municipalities is not thesame across all provinces and territories. InNew Brunswick and Manitoba, for example,unconditional grants are proportionately moreimportant than conditional grants.

At the moment, there are a variety of pro-vincial/territorial-municipal unconditionalgrant programs. Some provinces provide percapita grants. Some provinces/territoriesallocate grants to municipalities withinadequate or insufficient fiscal capacity. Stillothers consider expenditure needs and themunicipality’s ability to raise its ownrevenues. Some provinces pool municipalitiesinto different groups – arranged bypopulation, functions or services provided,rural versus urban and so on. Two provincesuse a weighting factor to differentiate thetreatment of municipalities. Some provinceshave more than one unconditional grantprogram. Finally, the territories take cost

10. Courchene, Thomas,J., “HourglassFederalism – Howthe Feds Got theProvinces to Run Outof Money in a Decadeof Liberal Budgets.”Policy Options, April,2004: 12.

11. In response to thefederal gas taxinitiative, some ofthe provinces haveoffered to share theirfuel tax revenueswith municipalities.

12. Although commonlyviewed in Canada asa federal transfer,this rebate is actuallya logical feature inthe federal value-added tax (the GST),as noted in Gendron,Pierre-Pascal, “ValueAdded Treatment ofPublic Sector Bodiesand Non-ProfitOrganizations: ADeveloping CountryPerspective,” Bulletinof InternationalFiscalDocumentation, 59(12): 2005: 514-26.




differences into account in their unconditionalgrant formulas.

As in nearly every country, the U.S. federalgovernment finances much more expendituresand services than it delivers. Total federalgrants in 2006 represented about 20.7 percentof total state and local revenue. Local govern-ments only received 12.1 percent of thesetransfers, but some of the grants to stategovernments are subsequently forwarded tolocal governments.

Over 600 federal grant programs exist forstate and local governments. The grants areprovided in many different forms, includingproject, categorical, and block grants. Somehave matching components and others arestructured through formulas. Still, except fora few specific areas, the overall federal-to-subnational intergovernmental grant systemis relatively small compared with many othercountries.13 Grants are increasingly focusedon a small number of functional areas, andparticularly health care. There is no form ofgeneral revenue sharing. The federalintergovernmental grant system is primarilyintended to provide a degree of equalizationacross people, not to equalize subnationalgovernment service delivery, with most of themoney intended to support low-incomepeople.

The composition of federal grants haschanged radically in recent years. Grants tostate and local governments for redis-tribution to individuals have risen, and othertypes of grants have fallen. The rapid rise inhealth care costs, and therefore in theMedicaid program, has been the drivingforce behind the growth in transfers forpeople. Transfers for health care have risenfrom about 21.9 percent of total grants in1983 to 48.2 percent in 2008.

The amount of transfers is decided annuallyby congressional decisions. However, some

programs, such as Medicaid and TANF, havebeen established as entitlements (withcarefully established eligibility requirements),and the basic structure is changedinfrequently. The other large grant catego-ries, transportation and education, are morelikely to support state and local servicedelivery, and these programs generally do nothave strong equalization components. Ratherthan being entitlement payments, the specificamounts are often determined through theannual budget process or by agencydecisions.

State and local governments have sought toleverage federal grants in a number of ways.First, some states appear to claim a widerange of expenditures as being appropriatefor the Medicaid program and, thus, eligiblefor the federal matching grant. Second, stateshave sought to provide their matchingcomponent through various creative means.For example, Tennessee created a “servicestax” on hospital health care during the early1990s and used this revenue to finance thestate’s share of the Medicaid program.Hospitals made the payments but receivedthe money back in Medicaid revenues,allowing the state to draw down the federalfunds with no state share. The federalgovernment disallowed this scheme based onthe argument that the state was not in factmatching the federal grant.

Unlike in Canada, state and local income,sales, and property taxes are deductibleexpenses in determining federal individualincome tax liabilities. Various interpre-tations are given to the linkage that thisestablishes between the federal and stateand local governments, one of which isthat deductibility is a form of grant to thestate and local governments, although itmay be better seen as a tax expenditure.All U.S. states provide grants and sharedtaxes to local governments. State grantsto local governments are nearly of the

13. For somecomparisons acrosscountries, see Bird,Richard M., andFrancois Vaillancourt,“FiscalDecentralization inDeveloping Countries:An Overview,” inFiscalDecentralization inDeveloping Countries,ed. Richard M. Birdand FrancoisVaillancourt,(Cambridge:Cambridge UniversityPress, 1998).



Box 8.1: The American Recovery and Reinvestment Act (ARRA) and beyond

The American Recovery and Reinvestment Act (ARRA) was passed in February of2009 and was worth a total of $787 billion U.S.

While the bulk of the ARRA was not specifically intended to prop up local budgets, $144billion was specifically targeted to state fiscal stabilization to prevent cuts to health andeducation programs and state and local tax increases. The state fiscal stabilizationfunding indirectly assisted local governments by mitigating potential cuts in state aidand funding for local programs that might otherwise have occurred in response to therecession. In some cases, ARRA funding for specific programs, such as infrastructure,broadband deployment, and green energy initiatives also was allocated through grantsand formulas to local governments. In total, to date, the White House Recovery.Govwebsite, created for the purpose of tracking ARRA funding and usage, reports that682,000 jobs have been saved or created nationwide.

While ARRA funding had some positive impacts on local government, other largeportions of the ARRA went toward entitlements, tax cuts and assistance to individuals.In addition, the obligation for rapid spending of the ARRA funds means that fundingplummets in 2011, after peaking in mid-2010. With U.S. cities facing the worsteconomic downturn in 50 years, the end of ARRA funding is expected to coincide withincreasing financial difficulties and budget short falls extending into 2011. Incomes fromtax collections, lagging behind the overall economy due both to payment periods andreassessment schedules, are expected to continue decreasing even if employment ratesbegin to rise. Local taxes will be especially hit by property reassessment as more housesare foreclosed upon and homes are reassessed to lower values.

The National League of Cities (NLC) annual State of the American City Survey on Jobsand the Economy found that 3 out of 4 local government officers reported that theeconomic and fiscal conditions of their city had degenerated over the last year. And 7 outof 10 city officials indicated that they are responding to the ongoing crisis throughpersonnel cuts, layoffs, hiring freezes and furloughs. Many will also be cutting publicsafety spending and healthcare benefits. In total, the NLC estimates that citygovernments will face shortfalls between $56 billion and $83 billion over 2010-2012,with the range of shortfalls depending on cuts in state aid over this period.

The NLC is currently supporting legislation that would provide additional stimulusfunding to state and local governments, including additional fiscal stabilization forstates through health care and education investments and the Local Jobs forAmerica Act (LJAA). LJAA would provide funds to states, local government units,and community-based organizations to save and create local jobs through theretention, restoration, or expansion of services needed by local communities. Withoutadditional stimulus for state and local governments, NLC predicts that city leaders willhave to lay off more employees, cut essential services and cancel projects.

Information facilitated by the National League of Cities



same magnitude as federal grants tostates, totaling USD 421.6 billion in2006.14 However, some of these grantsmay be the pass-through of federal grantfunds. Federal and state grants togetherprovide 38.3 percent of local governmentrevenue, with 33.9 percent coming fromthe states and 4.4 percent from thefederal government. Shared tax revenuesare often provided to local governments asone form of grant. The dominant statetransfer program in nearly every state isfor financing primary and secondary edu-cation. The specific grant structure differsacross states, but similarities exist in thebasic design. A number of states build thegrant around ensuring that local govern-ments have sufficient resources to deliveran adequate level of education. Somedegree of equalization is usually built intothe grants, along with incentives toachieve certain objectives (such as tomeet class-size expectations).Equalization is frequently based on boththe capacity to raise revenues locally andthe expenditure needs in the community.

Major Issues and Constraints for LocalFinance in the Region

Nine issues in local government finance arediscussed in this section including:

• The contribution of property taxes tolocal finance

• Sales tax base erosion in the UnitedStates

• The limited set of revenue sourcesavailable for Canadian local governments

• Vertical and horizontal tax relationshipsbetween governments

• Federal limitations on sub-nationaltaxation

• The role of intergovernmental transfers

• Federal mandates

• The importance of infrastructure finance

• Pricing of local government services

These nine issues generally apply togovernments across North America, butsome are more generally applicable toCanada, such as the limited set of availablerevenue sources, and others are moreapplicable to the U.S., such as erosion ofsales tax bases. These differences arehighlighted throughout the section.

Property Taxes

The property tax is the largest localrevenue source in both countries,accounting for 25 percent of local govern-ment revenues in the U.S. and 39 percentof local government revenues (municipaland school) in Canada (see Table 8.6). Atthe municipal level in Canada, propertytaxes account for more than 50 percent ofrevenues. The heavy reliance on propertytaxes has generated considerable contro-versy in both countries, particularly whenproperty values have risen rapidly. A num-ber of issues are addressed here includingshifts in the tax’s contribution to local fi-nance, limitations on imposition of proper-ty taxes, and the erosion of the base.

Although the property tax has been aroundfor a long time in North America and fundsimportant local services, it is a veryunpopular tax, at least in part, because ofits visibility. Unlike the income tax, theproperty tax is not withheld at source.Unlike the sales tax, it is not paid in smallamounts with each daily purchase.Instead, the property tax generally is paiddirectly by taxpayers in periodic lump sumpayments. This means that taxpayers tend

14. See the U.S. Bureauof the Census athttp://www.census.gov/govs/estimate/03sl00us.html.


269

often to be more aware of the property taxthan they are of other taxes, though it isalso possible in most places to have amonthly property tax charge folded intomortgage payments making each paymentsmaller and less visible. To a considerableextent, the property tax finances serviceswhich are also very visible, such as roads,garbage collection, and neighborhoodparks. Visibility is clearly desirable from adecision-making perspective because itmakes taxpayers aware of the costs of localpublic services. This awareness increasestaxpayer scrutiny and enhancesaccountability, which is obviously a goodthing from both an economic and politicalperspective. It does not, however, makethe property tax very popular. On thecontrary, it often appears to be harder toraise (or reform) property taxes than othertaxes.15 Property tax rate changes can beeasier than sales or income tax changes,however, because the former tend to besmall often annual changes and less subjectto media scrutiny while the latter areinfrequent, discrete, and normally subjectto considerable attention.

Erosion of the Property Tax Base

Changes in the economy, combined withconscious efforts to reduce property taxesthrough tax incentives and tax andexpenditure limits, have resulted in a re-duction in property taxes in the U.S. from87.2 percent of local government revenuesfifty years ago to 72.4 percent in 2005.16 InCanada, however, the use of limits onproperty taxes has been much lesswidespread and property taxes haveactually increased as a percentage ofmunicipal revenues over the last 20 years.In 1988, property taxes accounted for 48.4percent of municipal revenues whereas theyaccounted for 50.7 percent in 2007. Schoolboard property taxes, however, decreasedfrom 31 percent of total revenues in 1988 to

almost 21 percent in 2007, in large part,because provincial governments took overthe funding of education in many provincesduring that period. There has, nevertheless,been some erosion in the property tax baseeven in Canada.

The erosion of the property tax base derivesfrom a number of sources, some of whichgovernments have control over and someof which they do not. For example,state/province and local governments haveno control over the shift from a goods-basedeconomy to a service-based economy to aknowledge-based economy. These changesin the economy have resulted in lowerproperty tax collections in both countriesbecause, in the new information-basedeconomy, there are fewer plants and lessmachinery and equipment to tax.17

The property tax base is also declining in theU.S., though much less so in Canada,because of policy decisions to limit the useof property taxes by local governments.These policies include, for example,property tax incentives and tax andexpenditure limits. Governments havecontrol over these sources of tax baseerosion which are discussed below, thoughit is often in the hands of state rather thanlocal governments.

Narrowing the property tax base means thattax rates have to be higher to collect thesame amount of revenue. Higher tax ratesincrease the excess burden of the propertytax, make the tax even more unpopular, andcan result in greater tax arrears if there isreduced compliance. Narrowing the taxbase by targeting relief to particulartaxpayers can result in an inequitable taxsystem, for example, by shifting the burdenfrom existing to new businesses or byshifting the burden from those with rapidlygrowing market values to those withstagnant market values.

15. See Bird, Richard, M.and Enid Slack,InternationalHandbook on Landand PropertyTaxation,Cheltenham, UK:Edward Elgar, 2003.

16. Augustine, Nancy, Y.,Michael E. Bell, DavidBrunori, and Joan M.Youngman (eds.)Erosion of theProperty Tax Base,Trends, Causes, andConsequences.Cambridge, Mass.:Lincoln Institute ofLand Policy, 2009: vii.

17. See, for example,Augustine, Nancy, Y.,Michael E. Bell, DavidBrunori, and Joan M.Youngman (eds.)Erosion of theProperty Tax Base,Trends, Causes, andConsequences.Cambridge, Mass.:Lincoln Institute ofLand Policy, 2009 andThe Conference Boardof Canada, “How CanCanada Prosper inTomorrow’s World?”Performance andPotential 2004-05,Ottawa, 2004.




Property Tax Incentives

Property tax incentives to stimulateeconomic growth are used widely in the U.S.but, until recently, Canadian municipalitieswere not permitted to offer fiscalinducements (such as property taxreductions and exemptions) to new firms orbusinesses. A study of stand-alone propertytax abatements in the U.S. indicates thatover 40 states allowed for these abatementsin 2007.18 The goal set out for most of theseincentives is to increase employment and/orincome generated in the jurisdiction and, inmany cases, to increase the property taxbase of the jurisdiction and property taxrevenues.

Some authors believe that tax incentivesare justified because the firms that receivethem provide benefits to the communitythat exceed the costs to the municipalityboth for business services andenvironmental degradation caused by thebusinesses.19 Moreover, new investmentresults in benefits from agglomerationeconomies.20 When a large number of firmsare clustered together, the cost ofproduction will be lower because firms willhave many different competing suppliers,they can take advantage of greaterspecialization, and they will have a largermarket for their goods and services.Although there is some validity to thisargument, it is not clear the extent towhich an individual city will be able todetermine which firms should receive thetax incentive and how much that incentiveshould be. From a political perspective, taxincentives are an indication that themunicipality is pro-business and politicianscan take credit for job creation andinvestment.21

Notwithstanding the arguments in favor oftax incentives, there are several dis-advantages. Property tax incentives can

result in a zero-sum game whereby deve-lopment at one location is at the expenseof development to another location. Theunderlying assumptions are that theoverall supply of capital is fixed and that itis unresponsive to price (tax) changes.Under these assumptions, tax competitionwould not increase the national capitalstock but, rather, only move it around. Taxcompetition would simply result in a redis-tribution of resources from local taxpayersto industry.

Tax incentives are often wasted on firmsthat would have located there anyway. Ifthe economic activity would not haveoccurred “but for” the tax incentive,proponents argue that the tax incentive isa good thing. If, however, the economicactivity would have occurred even withoutthe tax incentive, the tax incentive isprobably wastefu l . 22 Moreover, taxincentives can lead to unfair competitionamong businesses and can lead to asituation where no major investmentsoccur without them.

Tax competition can result in inefficientlylow taxes and public services. Tax cuts needto be financed in some way and, if they arefinanced by cutting public services thatbusinesses want, the net effect on economicdevelopment could be negative.23 Theprovision of services that, at the same time,provide direct benefits to existing residentsand firms is preferable to tax incentives.Moreover, a number of studies argue thatlowering non-residential property taxes forall businesses in the municipality ispreferable to tax concessions to any specificbusiness.24

Tax and Expenditure Limits

Tax and expenditure limits (TELs) arestate-imposed limitations on the ability oflocal governments to raise property taxes.

18. Wassmer, Robert W.,“Increasing Use ofProperty TaxAbatement as aMeans of PromotingSub-NationalEconomic Activity inthe United States.”Sacramento, CA:California StateUniversity, 2007.

19. Glaeser, Edwin,Comment on “TaxIncentives and theCity,” Brookings-Wharton Papers onUrban Affairs , 2002,115-24.

20. Garcia-Mila, Teresa,and Therese J. McGuire “Tax Incentivesand the City,”Brookings-WhartonPapers on UrbanAffairs , 2002, 95-132.

21. Brunori, David, LocalTax Policy: AFederalist Perspective.Washington, DC: TheUrban Institute Press,2003.

22. Wassmer, Robert W.,“The Increasing Useof Property TaxAbatement as aMeans of PromotingSub-NationalEconomic Activity inthe United States,”Sacramento, CA:California StateUniversity, 2007.

23. Bartik, Timothy J., WhoBenefits from State andLocal EconomicDevelopment Policies?Kalamazoo, MI: W.E.Upjohn Institute, 1991.

24. Wasylenko, Michael, J.“Taxation andEconomicDevelopment: TheState of the EconomicLiterature,” NewEngland EconomicReview , March/April,1997: 37-52.


271

TELs include caps on assessmentincreases, tax rates, property taxesrevenues, and expenditures. Althoughlimitations have been used for manyyears, the proliferation of TELs in the U.S.really began when voters passedProposition 13 in California in 1978limiting property taxes to no more than 1percent of property values. Since thattime, most states have introduced someform of TEL. The political costs andbenefits of imposing TELs can beimbalanced because the state politicalleaders receive most of the benefits ofcutting taxes, but may not bear any of therevenue consequences of their actions.Some states have made an explicit effortto fund TELs with additional state trans-fers. But, many other states have reducedthe local revenue stream with no explicitplan for financing any problem created forthe local governments.

Tax and expenditures limits are notcommon in Canada but, in someprovinces, reassessments are onlyperformed every few years with the resultthat there is an assessment freeze for theperiod between assessments. Similarreassessment cycles exist in most U.S.states. The impact of not having annualreassessments is similar to cappingassessments at a zero increase for theyears between reassessments. Legislationin some provinces requires municipalitiesto reduce the property tax rate followinga reassessment so that the reassessmentis revenue neutral (known as “truth intaxation” in some U.S. states) butmunicipalities are not restricted fromincreasing property tax rates forbudgetary reasons.

Tax and expenditure limits have severelyconstrained the growth in property taxrevenue in U.S. local jurisdictions andparticularly where assessment limits and

tax rate limits are used in combination.25

There is some evidence in the U.S., forexample, that TELs have severely limitedspending on local public schools andlowered educational outcomes.26 Thegreater the increase in property valuesand the lower the assessment increasepermitted, the greater will be the erosionof the property tax base.

Notwithstanding the pervasive use ofassessment limits, they are probably theleast effective, equitable, and efficientstrategies for providing property taxrelief.27 Assessment limits are inequitablebecause properties with similar marketvalues may not be paying the same taxes.Assessment limits shift the property taxburden from those properties whosevalues are increasing rapidly to thoseproperties whose values are stagnant.28

And, the most relief goes to thoseproperties who appreciate the mostquickly. Assessment limits also shift theproperty tax burden from those who haveowned property for a long time to recentbuyers.29 In California, for example, it wasfound that by 1991 taxes on newlypurchased property in Los Angeles Countywere more than five times the taxes onproperty of equal market value ownedsince 1975.30

There are at least three other concernswith assessment limits. First, if imposeduntil time of sale, assessment limitationsreduce the incentive to move and result ina misallocation of resources. Second,assessment limits complicate theadministration of the property tax andcreate confusion among taxpayersbecause the taxes paid are no longercalculated simply as a tax rate multipliedby the tax base. Moreover, there is noincentive to review one’s assessment.Third, it is very difficult to remove afreeze: “once a freeze is imposed, the

25. The limits may not bebinding if, for example,assessment growth islimited but rates are not.Where a reduction inassessment is matchedby an increase in the taxrate, there would be nooverall decline inrevenue, although therewould be distributionalimplications. Thirty-three states placepotentially bindingconstraints on the abilityof cities to levy propertytaxes. See Hoene,Christopher and MichaelA. Pagano, Cities andState Fiscal Structure,Research Report onAilingCities.Washington, DC:National League ofCities, 2008.

26. Yuan, Bing, JosephCordes and DavidBrunori, “Tax andExpenditure Limitationsand Local PublicFinances,” in Augustine,Nancy, Y., Michael E.Bell, David Brunori, andJoan M. Youngman(eds.) Erosion of theProperty Tax Base,Trends, Causes, andConsequences.Cambridge, Mass.:Lincoln Institute of LandPolicy, 2009.

27. Sexton, Terri A,“Assessment Limits as aMeans of LimitingHomeowner PropertyTaxes,” In Augustine,Nancy, Y., Michael E.Bell, David Brunori, andJoan M. Youngman(eds.) Erosion of theProperty Tax Base,Trends, Causes, andConsequences.Cambridge, Mass.:Lincoln Institute of LandPolicy, 2009.

28. Ibid29. Winters, John. "An

Overview of PropertyTax Limitations." StateTax Notes, 2008.

30. O'Sullivan, Arthur, TerriA. Sexton, and StevenM. Sheffrin. PropertyTaxes and Tax Revolts:The Legacy ofProposition 13. NewYork: CambridgeUniversity Press, 1995.




process of thawing may be too painful tobear.”31

Sales Tax Base Erosion

Thirty-six U.S. states (including 35 of 45states that impose sales taxes) permit atleast some local governments to imposesales taxes. States normally includelimitations on the operation of thesetaxes, such as on the maximum rate thatcan be imposed. Erosion in the underlyingtax base has been an important implicitlimitation on these taxes. Data are notavailable for local sales tax bases but thesales tax bases defined for a state and itslocal governments are normally, thoughnot always, defined by the state and aregenerally very similar. The average statesales tax base has fallen from 53.2percent of personal income in 1979 to39.2 percent of personal income in 2007.At least in part as a result, state sales taxrates have risen from a median rate of3.25 percent to 6.0 percent today andaverage local sales tax rates can exceed4.0 percent.32

Three factors explain the base erosion.First, state governments have chosen tonarrow the base by granting a seeminglyongoing set of exemptions. In some casesthese were good policy choices, such as theexemption of certain business inputpurchases. Exemption of manufacturingequipment is a key example.33 Also, manyconsumer goods have been exempted. Foodfor consumption at home has beenexempted by many states and clothing by asmall set of states. The solution to thisproblem is for local governments, andparticularly states, to exert the political willto maintain the tax bases.

Second, although differences exist acrossstates, sales taxes are broadly imposed ongoods, but relatively narrowly on services.

For example, a review by the Federationof Tax Administrators finds that themedian state taxes 55 of 162 itemizedservices.34 Service consumption growth hasoutstripped goods consumption growth inrecent decades, causing the relative taxablebase to fall. In general, states have found itpolitically very difficult to broaden the baseto additional services because it requiresidentifying specific industries to draw underthe tax base.

Third, the tax is generally collected byvendors but a Constitutional provisionagainst states taxing interstate commercehas been interpreted by the U.S. SupremeCourt to mean that states can only requirevendors to collect the tax when they havephysical presence in the state. Buyersare expected to pay a parallel use tax ifthe vendor has not complied, but usetax compliance is nearly nonexistentfor individuals and relatively weak forbusinesses.35 Combined state and localsales tax rates can be as high as 11 percent,so this can be an important advantage forremote vendors. As a result, rapid growth ofInternet-based transactions has cost statesa significant share of sales tax receipts(effectively the base is narrowed),estimated at approximately USD12.0 billionin 2012.36 In addition, states lose tax basebecause of physical cross border shoppingand mail order sales. State and localgovernments have appealed to the federalgovernment to require remote vendors tocollect the tax on their behalf, but so far tono avail.

Higher rates combined with narrowing baseslikely have increased the tax’s excessburden, reduced equity, and raisedcompliance and administrative costs. Excessburdens rise more than proportionately tothe tax rate, and occur as the increasingwedge between taxable and non-taxablepurchases encourages consumers to

31. Youngman, Joan."The HardestChallenge for Value-Based PropertyTaxes: Part I." StateTax Notes, March1999.

32. Seehttp://www.taxch.com/STRates.stm

33. Nonetheless, theaverage tax base isstill thought to becomposed about 40percent of inputpurchases. See Ring(1999).

34. Seehttp://www.taxadmin.org/fta/pub/services/services.html

35. Seehttp://dor.wa.gov/Docs/Reports/Compliance_Study/compliance_study_2008.pdf

36. See Bruce, Donald,William F. Fox, andLeAnn Luna, “Stateand LocalGovernment SalesTax Revenue Lossesfrom E-Commerce,”State Tax Notes, May18, 2009: 537-53.


273

purchase more non-taxable goods. Forexample, one study found that one-eighthof the relative growth in the service sectorcan be attributed to the lack of taxation.37

Administrative and compliance costs risewith the narrowness of the base sincethe number of decisions on whether atransaction is taxable or not increasesdramatically as more exemptions areallowed.

Inadequate Revenue Tools to MeetExpenditure Requirements

Municipalities in Canada complain thatthey do not have adequate revenue toolsto meet their expenditure requirements.Several commentators have also questionedwhether municipalities will be able toprovide the services that people want atreasonable tax rates in the future.38 Thisproblem is more significant in Canada thanthe U.S. because Canadian municipalitieshave access to a narrower range of revenuetools.

These concerns originate from a series ofevents that have had an important impacton the municipal fiscal environment.First, the “offloading” of services by thefederal and provincial governments hasmeant increased responsibilities formunicipalities throughout the country.Second, the future of most countries isincreasingly linked to the fortunes of itslarge cities where employment, wealth,and productivity growth are generated. Inthe new global environment, cities have tocompete in the international marketplaceto attract business and skilled labor. To dothis, they not only have to provide sophis-ticated transportation and communica-tions infrastructure but they also have todeliver services that enhance the qualityof life in their communities. Third,municipalities that are facing rapid growthare also, in many cases, experiencing the

higher costs associated with urban sprawl.The literature on the costs of sprawl inboth Canada and the U.S. suggests thatinfrastructure and service costs are higherin sprawl developments than compactdevelopments.39 Fourth, at the same timethat municipalities are facing and willcontinue to face increased pressures onthe expenditure side of their budget, therehas been no parallel diversification of theirrevenue sources in Canada.

Some authors have argued thatmunicipalities in most provinces simplyneed to raise property taxes (on residentialproperties but not on commercial andindustrial properties for the reasons notedearlier) and user fees to solve theproblem.40 There is some truth to thisargument at the municipal level –residential property taxes have notincreased dramatically over the last 20years and user fees could probably beexpanded to include a few more services.Correct pricing, in many cases, would alsoresult in reduced demand for services andinfrastructure and remove some of thepressure on expenditures. Nevertheless, allof these measures are still likely to fallshort of meeting existing expenditurerequirements.41 Although municipalitieshave not run deficits on operating accountsbecause they are not permitted to by law,there is evidence that the infrastructuredeficit is rising as a result of insufficientrevenues at the local level. This issue isdiscussed further below.

Vertical and Horizontal Tax Relationships

U.S. federal and state governments haveindependent control over their tax basesand rates, given the limitations describedabove. Governments are not required tocoordinate their tax bases or rates, anddifferences exist in the tax bases used by

37. Merriman, David andMark Skidmore, “DidDistortionary SalesTaxation Contribute tothe Growth of theService Sector,”National Tax Journal,March (2000) 53(1):125-142.

38. See, for example, TDEconomics, Mind theGap: Finding theMoney to UpgradeCanada’s AgingInfrastructure,Toronto, 2004;Kitchen, Harry M. andEnid Slack, “SpecialReport: New FinanceOptions for MunicipalGovernments.”Canadian Tax Journal,Volume 51, Number 6,2003; and Van derPloeg, Casey. 2004."No Time to be Timid:AddressingInfrastructure Deficitsin the Western Big Six,"Western Cities ProjectReport #30, Calgary:Canada WestFoundation, 2004.

39. Slack, Enid, MunicipalFinance and thePattern of UrbanGrowth. Toronto: C.D.Howe Institute, 2003.

40. Mintz and Roberts notetwo importantexceptions to thisrecommendation,however: Albertawhere municipal userfees are relatively highand Ontario where thecost of social services isshared withmunicipalities. In thosetwo provinces, theauthors recommendconsideration of a newtax on earned income.See Mintz, Jack M. andTom Roberts, Runningon Empty: A Proposalto Improve CityFinances. Toronto:C.D. Howe Institute,2006.

41. Courchene, Thomas J.,“Citistates and theState of Cities:Political-Economy andFiscal-FederalismDimensions.” Montreal:Institute for Researchon Public Policy, 2005.




every state and by the nationalgovernment. Similar or identical basesare more common for local governmentswithin states, but wide differences existacross states and in some cases withinstates.

Federal, state, and local tax structures areoften intertwined, despite their legal andconstitutional independence. Most statesrequire individuals and corporations tobegin calculation of their income taxeswith some variant of the federal definitionof taxable activity. The relationshipsbetween federal, state and local personalincome taxes extend to administration aswell. Each state has its own taxadministration but relies heavily onfederal audits and databases to assistin collection. In many cases localgovernments administer their income taxbut also can benefit from administrativedata collected by other governments.

In the Canadian context, all municipalgovernments levy property taxes, pro-vincial governments levy property taxes ineight provinces, and school boards levyproperty taxes in five provinces. To mi-nimize the possibility of unintendedvariation in provincial assessment prac-tices within each province and to attemptto achieve intended variation where it isdesired, a central assessment authorityhas been established in each province.Every province maintains an assessmentmanual for the guidance of its assessorsand it is the practice that assessors followthe manual.42 In addition, all provincesexercise a certain measure of controlthrough the establishment of compulsoryeducational standards and trainingcourses for provincial assessors. Similarstandards have been laid down where thecities rather than the provinces assumeresponsibility.

In terms of property tax rates, each taxingauthority is free to set its own rates tomeet budgetary requirements. In manyprovinces, however, the provincialgovernment imposes restrictions on therelationship between tax rates for differentclasses of property. In particular, there areoften limitations placed on the extent towhich non-residential property tax ratesexceed residential tax rates.

The institutional linkages between taxbases mean that tax policy decisions madeby one level of government frequentlyhave implications for other levels. There isscant evidence that these vertical andhorizontal externalities are given seriousconsideration when policy decisions aremade. The federal government has madenumerous tax policy changes in recentyears by changing tax bases (frequentlynarrowing them) and lowering tax rateswith little discussion of how othergovernments are impacted.

Vertical competition between governmentsmay also exist, and it is an empiricalquestion as to how one level of govern-ment responds to policy decisions atanother level. The notion is that impositionof a tax by a higher level of governmentreduces the tax base available for lowergovernments, and vice versa.43 The affec-ted local governments could either raisetheir rate to offset the revenue loss orlower their rate because the tax is lessproductive. Research has yet to reach asolid conclusion regarding the direction ofthese relationships. Most of the researchsuggests that lower level governmentsraise their rates after increases by higherlevels, but most of the published researchhas been outside of North America and thelimited North American research hastended to be at the state/federal levels.Some evidence suggests that states tendto raise their gasoline and tobacco tax

42. Some U.S. states alsohave centralassessmentauthorities thatoversee localassessment andassess someproperties, such asstatewide utilities.

43. Other relationshipscould also exist, suchas leader/followerresponses ordemonstrationeffects.

44. See Besley, Timothyand Harvey Rosen,“States’ Responses toFederal Tax Setting:Evidence fromGasoline andCigarettes,” Journal ofPublic Economics 73(1998): 383-98.


275

rates in response to federal increases,44

suggesting that states raised their rates tooffset the base decline. Also, research onthe U.S. and Canadian corporate income,personal income and sales taxes havefound that states tend to increase their taxrates in response to federal income taxrate increases.45 Nonetheless, there hasbeen too little research to reach a firmconclusion on how federal and state taxchanges affect local governments.

Horizontal relationships between localgovernments can also be important, bothin terms of how revenues are distributedacross governments and how thegovernments compete for the tax base.Sales taxes are due in the state where thegoods and services are to be enjoyed orused – that is, on a destination basis. Thisis normally presumed to be the placewhere possession of the goods takesplace.46 But, the local tax is often collectedat the origin, creating an inconsistency.Origin taxes are well known to create thepotential for tax competition.

More than forty states, in an extraordinaryact of cooperation, worked together overthe past nine years to create theStreamlined Sales and Use Tax Agreement(SSUTA).47 On October 1, 2005, nineteenstates signed on as initial members byenacting the legislation that wasdeveloped through the process and threemore states have joined subsequently. TheSSUTA is intended to simplify the sales taxand structure it on a destination basis sothat state and local governments arebetter able to collect their sales taxes onremote transactions. The SSUTA is awonderful example of governmentcooperation, but cartels of this type aredifficult to develop and hold together, evenwhen the related structures representgood tax policy (which is generally true ofthe SSUTA).

Federal Limitations on State and LocalFiscal Activities

The U.S. Constitution imposes two basicconstraints on state and local governmentfiscal actions. First, state and localgovernments are prohibited from dis-criminating against interstate commerce.Canadian provinces are precluded fromimposing indirect taxes, which is a strongertool preventing them from distortinginterprovincial trade. Second, state andlocal governments are prohibited fromtaxing international trade. In addition, theU.S. Constitution supersedes the stateconstitutions when conflicts arise betweenthem.

Limitations arising from state and localgovernments’ inability to distort interstatecommerce are imposed both by federalcourt constraints and by congressionallegislation. The U.S. Constitution givesCongress control over interstate com-merce, which means that congressionallegislation can define when state or localgovernments violate interstate commerce.Many examples of congressional andjudicial constraints exist, but only a feware mentioned here. The constraints onstate/local governments almost alwaysprevent them from being advantaged atthe expense of others. The courts havegenerally ignored cases where national orstate policies may cause the home state tobe disadvantaged relative to others.

The U.S. Supreme Court ruled that stateand local governments can only requirevendors with physical presence in the stateto collect the state’s sales tax.48 Congresshas not acted to require remote vendors tocollect the state sales tax despite theadmonition by the Supreme Court toaddress the issue. This limitation allowseasy tax planning because vendors canpurposely sell into a state/city from remote

45. See Esteller-More,Alex and Albert Sole-Olle, “Vertical IncomeTax Externalities andFiscalInterdependence:Evidence from theU.S.,” RegionalScience and UrbanEconomics 31 (2001):247-72 andKarkalakos, Sotirisand ChristosKotsogiannis, “ASpatial Analysis ofProvincial CorporateIncome TaxResponses: Evidencefrom Canada,”Canadian Journal ofEconomics 40 (2007):782-811.

46. Some states havetraditionally taxedservices based on theplace of production.

47. The SSUTA wasformed primarily as astate response to theSupreme Court rulingin the Quill case andin light of the rapidlygrowing extent ofcross-bordershopping.

48. Quill v. North Dakota,112 U.S. 298 (1992).




locations and avoid the compliance res-ponsibility as well as the tax burden(which is either a legal burden of thevendor or the consumer, depending on thestate in question).

The U.S. courts have in some cases alsorequired state and local governments toprovide non-residents with equal access toservices. Thus, residents can move fromone state to another and can gain accessto services such as education, welfare, andhealth care for the poor within thirty days.This appears to have limited the extent towhich some state and local governmentsare willing to expand delivery of certainservices.

Congressional legislation can preemptstate or local governments from imposingtaxes in cases where Congress believesthat state or local taxation would distortinterstate commerce. The Federation ofTax Administrators (an association of staterevenue departments) has identifiedtwenty-eight examples of preemption.49

Congress passed the “4-R Act” to preventstate and local governments from taxingrailroads differently than other commercialand industrial firms of the same class. Anexample of the effect is that localgovernments are prevented from imposingheavier property taxes on railroads thanon other industrial property.


Both Canadian and U.S. municipalitiesdepend on transfers from provincial/stategovernments and, to a lesser extent, fromthe federal government. In Canada in2007, for example, transfers accounted forover 40 percent of local governmentrevenues; in the U.S., transfers accountedfor almost 34 percent of local governmentrevenues. In most cases, these transfersare for specific purposes (for example, to

pave roads or subsidize recreation pro-grams) but, in some cases, they are gene-ral purpose grants (for example, they canbe used for any expenditures or to reducetaxes).

A more detailed breakdown of transfers forCanada in Table 8.6 shows that schoolboards depend much more heavily onprovincial transfers than do municipalities.Moreover, for municipalities, the bulk oftransfers comes from provincialgovernments (14.4 percent of revenuescompared to 1.6 percent of revenues fromthe federal government). The bulk oftransfers is also conditional (16 percent ofrevenues are conditional grants comparedto 2.9 percent that are unconditionalgrants).50 All federal transfers tomunicipalities in Canada are conditionaltransfers and are mainly for targeted toinfrastructure.

Over the last 20 years, overall transfers tomunicipalities in Canada in constantdollars per capita have fallen at the annualaverage rate of 0.1 percent. This declinerepresents a 2.7 percent annual averagedecrease in unconditional transfers and0.6 percent annual average increase inconditional transfers. Unconditional trans-fers as a proportion of municipal revenueshave remained both low and constant overthe last 20 years and, indeed, over theentire post-war period. Money from theprovinces has come with restrictions andcontrols designed to “meet provincialwishes at the local level.”51 Transfers tolocal governments in the U.S., on the otherhand, rose 2.3 percent annually inconstant dollar per capita terms between1992 and 2006. Federal transfers haverisen most rapidly (3.9 percent versus 2.1percent), but state transfers still represent88.5 percent of total transfer to localgovernments. As with Canada, most ofthese transfers are conditional.

49. Federation of TaxAdministrators, June2003.

50. Conditional transfersto municipalitiesexceed 50 percent oftotal transfers in allprovinces except fortwo – Manitoba andNew Brunswick.

51. Bird, Richard M. andAlmos Tassonyi,“ConstrainingSubnational FiscalBehavior in Canada:Different Approaches,Similar Results?” inRodden, Jonathan A.,Gunnar S. Eskeland,and Jennie Litvack(eds.) FiscalDecentralization andthe Challenge of HardBudget Constraints.Cambridge, Mass.:MIT Press, 2003.


277

Although there are valid economic andpolitical justifications for intergovernmen-tal transfers, grant funding is not alwaysthe best way to address municipal fiscalproblems. First, conditional, matchinggrants lower the price of the aided serviceand encourage municipalities to spendmore on that service. Where there are noexternalities, or where the amount of thegrant exceeds the amount of theexternality, the resulting distortion in mu-nicipal behavior is inappropriate. Many ofthe matching grants in Canada havematching rates that exceed 50 percent butit is unlikely that the externalities are thathigh. Matching grants in the U.S., at leastin cases where states are willing to matchlocal contributions at the margin, are oftenat lower rates.

Second, transfers can get in the way ofproper pricing of municipal services. Thereis no incentive to use proper pricing,however, when grants cover a largeproportion of operating and capital costs.52

In parts of Canada, for example, largegrants for water treatment plants in thepast reduced the incentives of munici-palities to use volumetric pricing to reducethe demand for water or to engage inasset management.53

Third, transfers can reduce accountability.When two or more levels of governmentare funding the same service,accountability problems can arise. Fourth,transfers are rarely a stable andpredictable source of revenue for localgovernments. Lack of predictability makesit difficult for municipalities to planexpenditures.

Federal and State Mandates

Higher level governments can alter localbehavior through either a carrot or a stickapproach. The carrot approach often in-

volves providing grants that include con-ditionalities that require local govern-ments to spend the money in a particularway or alter the local price of deliveringservices that the higher government wouldlike provided. The stick approach oftenmeans requiring local governments todeliver specific services, to use specificapproaches to deliver the services, or tomeet certain input or output standards asthe services are produced.

Whether or not delivery of a service ismandatory, once the municipality deliversthe service, provincial/state or otherstandards often have to be met. Forexample, there are standards for fireprotection, water and sewerage services,solid waste disposal, building inspection,local education, day care, and housing forthe elderly. Municipalities in both Canadaand the U.S. may be more concerned withfinding sufficient funds to meet the servicestandards associated with all of thesefunctions than they are with the distinctionbetween mandated or non-mandatedservices. For example, water quality hasbeen a particular concern in recent years.The result is higher provincial standardsfor water treatment and operations,sewerage treatment and facilities, andsolid waste handling and disposal.Higher standards have led to highercosts but not necessarily to higherprovincial funding. Municipalities oftenfind such requirements and changes inthose standards difficult to meet both interms of qualified staff and capitaldemands.

The mandates are a troubling practicewhen federal/state governments imposeresponsibilities without providing fundingto meet the related expenditure require-ments. Unfunded mandates can be animportant source of expenditure growthfor local governments. The U.S. Congress

52. Bird, Richard M. andMichael Smart,“Intergovernmentalfiscal transfers:International lessonsfor developingcountries.” WorldDevelopment, 30(2002): 899-912

53. Federation ofCanadianMunicipalities,InnovativeMechanisms for FiscalTransfers toMunicipalities – TheCanadian Experiencein MunicipalFinancing, Ottawa:FCM, 2008.




enacted legislation in 1995 requiring thefederal government to determine the costsof such mandates, whether by con-gressional or administrative action. Thislegislation provides some information onthe costs of federal decisions.54 Galloexamines the act and applauds theincreased supply of information, but shequestions the long-term effect on federaldecisions as the legislation is narrowlyconstructed.55

Mandates come in two general forms. Insome cases, Congress, using its controlover interstate commerce, directly re-quires state or local governments to pro-vide certain services in certain ways.Restrictions on drivers’ licenses and voterregistration are examples. Alternatively,Congress may link conditions to grants,which state/local governments often viewas mandates, though these are variants onthe carrot approach. The No Child LeftBehind education links receipt of federaleducation grants to states establishingmeasurable standards of achievement andmaking progress to achieve these goals.

Infrastructure Funding

Most public infrastructure is the respon-sibility of municipal governments. InCanada, for example, the localgovernment capital stock represented 48percent of the total capital stock of allthree levels of government in 2002compared to 34 percent for the provincialgovernment and 18 percent for thefederal government.56 Local publicinfrastructure largely comprises roads andhighways followed by sanitary sewers andsewage treatment. Most of theinfrastructure was built between the1950s and 1970s and much of it is now inneed of replacement.

A number of Canadian studies have

attempted to measure the magnitude ofthe “infrastructure gap” or “infrastructuredeficit” and they have come up withestimates ranging from $60 billion to$125 billion. 57 Moreover, the trend in thedeficit has also been increasing from $12billion in 1985 to $60 billion in 2003 to$123 billion in 2007. Notwithstanding thedifferences in estimates of theinfrastructure deficit, there is a consensusthat there is a significant deficit and thatit is growing.

In the U.S., the American Society of CivilEngineers estimated that $2.2 trillionneeds to be invested over five years tobring the nation’s infrastructure up to agood condition.58 Since current spendingamounts to about half of the neededinvestment, they estimate that anadditional $1.1 trillion needs to beinvested over the next five years. At thefederal level, the investment gap is alsosignificant – long-term annual averageHighway Trust Fund revenues areestimated to be $32 billion compared torequired investments of nearly $100billion per year.

Some authors have argued that infra-structure is in a state of disrepair becausethe municipal financing of infrastructure isin a state of disrepair.59 Some of thereasons include: politicians prefer tosupport short-term projects because ofre-election rather than long-term capitalprojects; accounting practices fail toinclude replacement costs for depreciatingassets resulting in a fiscal shock when it istime to replace the asset; and inadequateuser fees and local taxes promote over-consumption of local services and anincreased demand for infrastructure.60

A number of different tools are used byNorth American cities to finance municipalcapital infrastructure: property taxes (in-

54. However, thelegislation does notregard conditions ongrants to bemandates and anargument can bemade that the lowerlevel governmentcould choose not toaccept the funds.

55. See Gallo, Theresa.“History andEvaluation of theUnfunded MandatesReform Act,” NationalTax Journal 57, 2004:559-70.

56. Harchaoui, Tarek, M.,Faouzi Tarkhani, andPaul Warren, “PublicInfrastructure inCanada, 1961-2002,”Canadian Public Policy,30 (3), 2004: 303-18.

57. See, for example,Mirza, Saeed, “DangerAhead: The ComingCollapse of Canada’sMunicipalInfrastructure,” Areport prepared for theFederation of CanadianMunicipalities,November 2007.

58. American Society ofCivil Engineers, 2009Report Card forAmerica’sInfrastructure, March25, 2009(www.asce.org/reportcard).

59. Kitchen, Harry, M., AState of Disrepair: Howto Fix the Financing ofMunicipalInfrastructure inCanada (Toronto: C.D.Howe Institute, 2006).

60. Ibid.


279

cluding special assessments), user fees,development charges (lot levies) andother exactions, federal andprovincial/state grants, borrowing, taxincrement financing,61 and public-privatepartnerships.62

Governments in both countries arelooking at ways to improve infrastructurefinancing. The National SurfaceTransportation Infrastructure FinancingCommission in the U.S., for example, iscalling for the federal government tomove away from the current fundingapproach based on declining federalmotor fuel taxes towards a new systemthat is built around more direct usercharges in the form of fees for milesdriven (known as a vehicle miles travelledfee system).63 The Commission proposesthat state and local governmentspiggyback their own mileage basedsystems on the federal system to raisetheir own revenues.

Canadian municipalities use mainly pro-perty taxes, user fees, and federal andprovincial transfers to pay for infra-structure costs. Since 2005, for example,the federal government has directed theequivalent of a portion of federal gas taxrevenues (now $2 billion CAD per year) ona per capita basis to municipalities forenvironmentally sustainable municipal in-frastructure (including roads). Canadianjurisdictions are beginning to think abouttax increment financing which is used inmost U.S. states. Perhaps the mostimportant source of financing infra-structure, which has been largely over-looked in Canada, is proper pricing of localservices – tolling roads, charging forwater according to use, etc. Proper pricingwould result in more efficient use of infra-structure and enable more efficientinvestment thus reducing the need tobuild additional capacity.

Pricing Policies

Collecting user fees is generally thepreferred mechanism for financing localgovernment services wherever possible.But, services can only be priced when non-payers can be excluded from consumingthe service and should only be used incases where a marginal cost is associatedwith each additional user. These factorssuggest that user fees apply best in caseswhere the local government services havesome characteristics of goods and servicesprovided through the private sector. U.S.local governments generate 25.7 percentof their own source revenues with userfees and charges and 35.0 percent whenlocal public utilities are included.64

Canadian local governments raise a similar25.6 percent from user fees.

Water, sewerage, electricity, solid wastedisposal, higher education and urban masstransit are services that can besubstantially or totally financed with userfees. Health care is often financed with acombination of user fees and insurance inthe U.S. User fees are also importantfinancing sources for parks and certainaspects of K-12 schools (books andmeals). U.S. local governments raise themost user fees from electricity, hospitals,water, sewerage, and education.

Financing service delivery with user feesoffers a number of financial and economicbenefits for local governments. First,properly set fees allow service delivery tobe (at least partially) self financing, andpreclude the need to impose additionaltaxes. Second, economic efficiency is en-hanced by setting user fees at themarginal cost of producing anddistributing the services. These fees allowthe local government to establish theappropriate amount of services to deliver(the amount that people want to buy at

61. Forty-seven states andthe District of Columbiahave TIF enablinglegislation. In Canada,a form of TIFs is usedin the Province ofAlberta and TIFs arebeing considered fortwo major projects inthe City of Toronto.There is extensiveliterature on TIFs in theU.S. See, for example,Wassmer, Robert, “CanLocal Incentives Alter aMetro City’s EconomicDevelopment?” UrbanStudies, 1994: 1251-1278 and Anderson,John, “Tax IncrementFinancing: MunicipalAdoption and Growth,”National Tax Journal,1990: 155-164.

62 Slack, Enid, “MunicipalFinancing of CapitalInfrastructure in NorthAmerica,” Journal ofProperty TaxAssessment andAdministration, Volume2, No.1, 2005.

63. National SurfaceTransportationInfrastructureFinancing Commission,2009.

64. Seehttp://www.census.gov/govs/estimate/0600ussl_1.html




the cost of provision) and ensure thatthose who most value the services obtainthem. Third, user fees follow the benefits-received principle: only recipients pay forthe services and non-consumers are notcharged. Properly set fees for utility services(such as for water and electricity) are anexcellent example of all three advantages.

Conclusions

This section contains five recommendationsfor enhanced operation of the local publicsector including:

• Providing a richer mix of taxes forCanadian cities

• Reducing the over-taxation of business

• Slowing or eliminating erosion of the taxbases

• Enhancing the pricing of services

• Designing an appropriate role for federaland provincial/state governments

A Mix of Taxes for Canadian Cities

Revenues from a mix of taxes would giveCanadian cities more flexibility to respondto local conditions such as changes in theeconomy, evolving demographics, andexpenditure needs. Other taxes are moreeffective than property taxes at linking thecosts and benefits of services when peoplecommute to work from one jurisdictionto another. Taxes that grow with theeconomy could provide cities with anincentive to make the kinds of investments(in infrastructure, for example) thatstimulate economic growth. Currently, thecontribution that municipal infrastructuremakes to economic growth is felt more bythe federal and provincial governmentsthat have access to income and sales

taxes than by municipal governments.Finally, any single tax like the property taxis almost certain to create local distortions(such as discouraging investment inhousing), some of which could be offset byother taxes.

In principle, an income tax ‘piggybacked’on the federal or provincial tax with locallyset rates has clear advantages in terms oflocal autonomy, accountability, and re-venue elasticity, but there are obviouslysome problems in imposing such taxes atthe local level. Business income is es-pecially difficult to tax both because of itsmobility across jurisdictions and becausebusiness properties are already generallyover-taxed by the property tax. Solutionssuch as taxing only employment income(as is done in a number of U.S. cities) arepossible, but have their own problems.Local surtaxes on the provincial tax inthose provinces with a retail sales taxwould be technically feasible but wouldagain clearly be economically a bad idea,not least because these taxes would againtax business inputs heavily. In provinceswith value-added taxes (which do not taxinputs), local surtaxes may also betechnically feasible.

Selective sales taxes, used by somemunicipalities in Canada, could be ex-tended to all municipalities. For example,hotel and motel occupancy taxes(which are widely used by U.S. localgovernments) could be an additional levyon the provincial retail sales tax rate onhotels and motels. The usual justificationfor imposing this tax at the local level is tocompensate cities for the services theyprovide to tourists or visitors (additionalpolice and fire protection, transit capacity,etc.). A municipal fuel tax piggybackedonto the provincial tax, if the funds areearmarked for local roads and transit, canbe viewed as a benefits tax. Moreover, it


281

should help curb urban sprawl by dis-couraging road use (although with theeffect of lowering available tax revenuesfor continued investments).

Reduce the Over-taxation of Business

There is no economic justification for thehigh taxation of businesses either throughthe property tax or the sales tax. Localgovernments should reduce the propertytax burden on non-residential propertiesto be more closely related to thebenefits received from local services. Ofcourse, a reduction in property taxes onnon-residential properties will necessarilymean an increase in taxes on residentialtaxpayers. To minimize the impact on resi-dential taxpayers, a phased implemen-tation is suggested along the lines of whatis currently being done in cities such asVancouver and Toronto.

Although some states in the U.S. haveexempted certain business inputpurchases from the sales tax and someprovinces have harmonized their sales taxwith the federal GST in Canada, manystates (and some provinces) still need toeliminate the sales tax on intermediatetransactions. However, a case can bemade for taxing the intermediatetransactions in situations where the finalgoods consumption is not part of thetaxable base. Many services are exemptfrom U.S. sales taxes, and taxation of theintermediate transactions is likely to beefficiency enhancing in these cases.

Eliminate or Reduce Erosion of Tax Bases

Property and sales tax base erosion haveresulted from a number of factorsincluding political decisions (either at thestate or local level) and underlyingeconomic trends that lower the taxablebases relative to economic growth. The

result is either declining revenues relativeto the economy or higher tax rates tomaintain revenues. Higher rates andnarrow bases increase the distortionsarising from the taxes and decliningrevenues could result in under productionof local public services.

The best approach is to keep the basesbroad to the maximum extent possible.State and national governments playimperative roles in maintaining the bases.The political courage to maintain the basesis the most important step to maintain therevenue bases. In many cases (though notall) higher level governments makedecisions that narrow local tax basesbecause they bear little or no revenueconsequences of their decisions and gainthe political benefits. There is also needfor political decisions to broaden the basesto cover evolving parts of the economy.For example, federal legislation in the U.S.is necessary to allow local governments tocollect sales taxes on remote transactionseffectively. The federal government hasfailed to act thus far, despite requests frommany states and their advocacy groups.

Price Services Properly

Services should be priced to the maximumextent feasible since the prices provide aneffective means of financing service pro-duction, providing service delivery tothose who most value the services, andhelping determine how much service toproduce. North American local govern-ments rely on pricing for many services,but should continue to seek cases wheretax financing can be replaced with user feefinancing. Further, prices should be set atmarginal cost to achieve the benefits ofpricing.

Cities should also expand pricing in newand innovative ways. Congestion pricing of




access to cities during peak times is anexample that has been used in someplaces around the world. Technologiesexist to charge vehicles that enter thecities during certain hours. These chargescan provide financing for substituteservices, such as urban transit, and rationautomobile usage to those who most valueaccess to cities during peak windows.

An Appropriate Role for Federal andProvincial/State Governments

Cities need to be empowered to meet theirexpenditure responsibilities with adequateand appropriate revenue sources. Only inthis way can they be responsible for theirown actions. To get to this point, cities (atleast in Canada) likely need more fiscalpower than they now have. The federalgovernment is very much a third wheelwhen it comes to municipal fiscal issues –only the provinces/states can change theexpenditure responsibilities and revenueraising tools of local governments.

Federal and provincial/state governmentsalso have an important role to play incities because they are involved in policyareas that have a direct impact on cities.In Canada, for example, the federalgovernment is responsible for immigrationsettlement, urban Aboriginals, andpayments in lieu of taxes on governmentproperties. If the federal andprovincial/state governments provideadequate funding in the areas under theirown jurisdiction, local governments wouldbe relieved of responsibilities that are notrightly theirs and local revenues would bereleased for truly local functions.

Regardless of the tax sources madeavailable to local governments, federaland provincial/state transfers will continueto provide an important source of revenueto local governments in Canada and the

U.S. In 2009, the U.S. implemented acountercyclical program (AmericanRecovery and Reinvestment Act of 2009)that is intended to distribute $787 billionUSD across 10 years, with 71 percent tobe expended in the first two years. Localgovernments are generally not directbeneficiaries of the ARRA but indirectlybenefit in many ways. For example, 14.7percent of ARRA funds financeinfrastructure, much of which may bedirected through local governments. Also,26.7 percent of ARRA funds are transfersto state governments, and these areallowing states to limit reductions intransfers to local governments. Similarly,as part of Canada’s $28 billion EconomicAction Plan, the 2009 federal Budget esta-blished a $4 billion Infrastructure StimulusFund to provide funding for the rehabilita-tion or construction of provincial, territo-rial, municipal, and communityinfrastructure projects over a two-yearperiod.65

Transfers are needed for those municipa-lities that do not have the fiscal capacity toprovide adequate local services with theirown-source revenues. Transfers are alsoneeded in cases of externalities becausemunicipalities generally do not considerthe benefits of their services to peopleoutside their jurisdiction. But, transferscan also distort local decision-making inways that create inefficiencies in servicedelivery. Transfers should be designed sothat they do not get in the way ofmunicipalities charging appropriate userfees for services, where appropriate.

65. The Federation ofCanadianMunicipalities hasexpressed theconcern, however,that federal andprovincialgovernments couldcut funding tomunicipalities andimpose unfundedmandates as part of adeficit-cuttingprogram coming outof the economic crisis.


0w2010 09 METROPOLIS DEFcarta ang 9/9/11 20:36 Página 284

9. FINANCINGMETROPOLITAN

AREAS

RROOYY BBAAHHLL**



* Regents Professor of Economics, Andrew YoungSchool of Policy Studies, Georgia State University.Tetyana Zelenska provided valuable researchassistance for this paper.



287

T he world population will approximatelydouble by 2050 and virtually all of this

growth will be absorbed by urban areas in lessdeveloped countries, (United Nations, 2008).The number of megacities (population greaterthan 10 million) is projected to increasefrom the current 19 to 27 in 2025. By2025, about 10 percent of the world’s urbanpopulation will reside in these cities.

Metropolitan area governments in manycountries will need to find a way to managepopulations of 5 to 20 million, and provideaffordable services (Figure 9.1). To arrive atthe right formula for governance andfinance they will be required to settle on theright degree of fiscal decentralization withinthe metropolitan area, coordinate the work

of many different government agencies andpublic companies, and find a viable plan forresource mobilization.

The financing of government services inlarge urban areas is more complicatedthan resolving the problem of financingthe expenditures of a single citygovernment because the common patternis for many different governments andpublic enterprises to provide serviceswithin a single metropolitan area. Itinvolves a balancing act in determiningwho governs, who manages and whopays. Furthermore, at almost every turn,there is a political dispute about the“right” balance of power among thesegovernments.

Figure 9.1: Population size of the 11 Largest Metropolitan Areas

Source: http://www.metropolis.org/publications/metropolitan_regions


MMEETTRROOPPOOLLIITTAANNUnited Cities and Local Governments288

This chapter is about the financing of govern-ment services in metropolitan areas. Thescope of this review is both governance andfinance. These two topics cannot beseparated because the arrangements forfinancing public services in metropolitanareas are largely driven by the servicedelivery responsibilities assigned to thevarious governments and enterprises. But,countries make different choices forstructuring and financing their servicedeliveries, and so the problems that arisecan differ greatly from country to country.In this review we consider a sample ofmetropolitan areas in both developed andemerging and developing economies. Thechoice of the sample is based on theavailability of information rather than onany formal attempt to have a “represen-tative” coverage.1

In the next section of this chapter, weconsider the theoretical underpinnings forchoosing among the various possiblemetropolitan governance structures. Insections following, the governance andfinance models that are used around theworld are reviewed, and their advantagesand disadvantages are discussed. Thechapter concludes with a discussion of therange of policy reform options that appearto be open.

Theory and MetropolitanGovernance

Can economic theory point the way to bestpractice metropolitan governance andfinance? Certainly not in any precise way,because the political economy dimension isso important. Still the economic model canprovide a useful framework for evaluatingthe practice.2 The problem is often definedas choosing the population size of a localgovernment that will maximize the welfareof its residents. The core argument is thenow-familiar “decentralization theorem”, the

basic rule of efficient expenditureassignment is to assign each function to thelowest level of government consistent withits efficient performance. The apt phase isthat “people get what they want” so theoverall public welfare is enhanced (Bahl andBird, 2008). If the story ended here,metropolitan governance would be in thehands of small, independent municipalgovernments.

For some functions, however, assignmentto the lowest level of government doesnot lead to an efficient performance. Onereason why, is the presence of externaleffects in the delivery of the service, andthe other is the presence of economies ofscale. This pretty much defines thequestions to be answered in structuringmetropolitan public finances: shouldindividual local governments carry thepublic financing load, is a metropolitangovernment necessary for managing andfinancing area-wide services, whatphysical area should the regionalgovernment encompass, and howimportant should state/federal verticalprograms be?3 Once these questions areanswered, expenditure responsibilitiescan be assigned and finance will followfunction.

Government Structure in MetropolitanAreas

Countries and metropolitan areas havereacted differently in deciding on agovernance arrangement for service deli-very. Some have created very fragmentedstructures with strong decentralization ofresponsibility and power, while othershave created a more regional approach.Almost all have tried to strike somebalance between capturing the efficienciesof area-wide government and maintaininglocal control.4 If there is a generalconclusion that can be drawn about the

1. Ideally, we wouldinclude a comparativeanalysis of fiscalindicators forgovernments providingservices inmetropolitan areas.Unfortunately there areno such data availableother than in individualcase studies, and eventhese are not readilycomparable. The IMF(various years)provides acompendium ofgovernment financesfor all countries in theworld, reported in aspecified format, butno data are reportedfor individual localgovernments.

2. Oates (1972) hasoffered an approachthat serves as astarting point for moststudents of thissubject. For applicationto metropolitan areas,see Bahl and Linn,1992; Bird and Slack,2004; and Slack, 2007.

3. A vertical program isone where the serviceis delivered in themetropolitan area by ahigher levelgovernment, andwhere the funds do notpass through thebudget of any localgovernment budget.

4. For discussions ofmetropolitan areagovernance, see OECD(2006), and Jouve andLefèvre (2002).



289

choices actually made, it would seem tobe that the sentiments for local controlhave largely held off the formation ofmetropolitan governments. One way tothink about the various approachesto metropolitan governance is in termsof the emphasis of the structureadopted: jurisdictional fragmentation,functional fragmentation, or metropolitangovernment.

Jurisdictional Fragmentation

Under this approach, many general purposelocal governments operate in the samemetropolitan area with some degree ofindependence in choosing their package ofpublic services and their tax, user charge,and debt financing arrangements. In somecases there also is an overlying metropolitangovernment, and in most cases there areregion-wide special districts.

The advantage of the jurisdictional frag-mentation model is that it keeps govern-ment close to the people, but, the welfaregains from this “home rule” model willcome at some cost, usually failure tocapture economies of scale and operatingwithin a set of boundaries that are oftentoo small to internalize important externaleffects or to allow coordinated servicedelivery. Jurisdictional fragmentation canlead to large fiscal disparities among localgovernments in the metropolitan area,since they almost surely wi l l havedifferent financing and service deliverycapacity.

Developed Countries

The jurisdictional fragmentation model bestcharacterizes governance in most U.S.metropolitan areas. The traditions of homerule in the U.S. are strong; there is anacceptance of competition among localgovernments and a higher tolerance for

fiscal disparities than is the case in manyEuropean countries. There have beennumerous attempts to establish metro-politan governments in the U.S. but almostno successes.5 The typical arrangement iswell illustrated by the New York City regionwhich includes over 2,000 governments(Benjamin and Nathan, 2001).

Strong traditions of home rule are alsofound in Europe. “Local Governments in theNordic countries fiercely defend their rightsto collect own-source taxes. They argue thattheir own-source taxation results inaccountability and makes the behavior ofthe local population and local councils moreresponsible” (Lotz, 2006, p236).

The Copenhagen metropolitan region is anexample of a jurisdictionally fragmentedstructure. Its population of 2.4 million isgoverned by 45 municipalities, which arethe dominant tier in terms of servicedelivery and taxation, and by a NationalCapital Region. The Capital Region is anelected area-wide government that hashealth care as its primary responsibility, butit has no taxing powers.

The population of the city of Paris is about 2million, but another 6 million people live inthe inner suburbs. Local governance in thisagglomeration is kept by 80 municipalities,3 departments, and numerouscompanies that provide public services.The Stockholm metropolitan regionincludes 65 municipalities and fivecounties (OECD, 2006a), and 50municipalities are contained in theRandstad (Holland) metropolitan region(OECD, 2007a). Metropolitan Vancouver includes21 municipalities and about 2 million people.

Emerging and Developing Economies

The core provision of many local servicesin Manila is the responsibility of 11 cities

5. For a discussion of thehistory of reform oflocal governance in theU.S., see Campbell andBirkhead (1976). For adiscussion of recentefforts, see Phares(2009).



and 6 municipalities whose boundaries arecontained within the metropolitan area.Each has a local council that is popularlyelected, and a defined set of expenditureresponsibilities and revenue entitlements.These 17 local governments are over-lapped by a supra Metropolitan ManilaDevelopment Authority which isresponsible for planning and coordinatingarea-wide functions.

Governance in the Mexico City metropolitanzone is another example of jurisdictionalfragmentat ion (OECD, 2004a). Themetropolitan area is overlapped by aFederal District and its 16 municipal-like subunits, the States of Mexico and Hidalgo withtheir 59 municipalities, and the federalgovernment. All of the lower-tier local unitsin the two states have elected governmentsbut the boroughs within the Federal Districthave no taxing powers.

The metropolitan municipality of Istanbuloverlaps with 73 municipalities. Themunicipalities, however, have no legislativepowers. The Sao Paulo metropolitanregion, with a population of about 18 mi-llion, is made up of 39 municipal govern-ments with no overlapping metropolitangovernment. Coordination is attempted byagreement or compact among these muni-cipalities, through a number of agencies andcouncils (World Bank, 2007).

Functional Fragmentation

A second approach to metropolitangovernance is to emphasize functionalfragmentation. Under this model, thedelivery of a single function or a particularset of functions is placed under the controlof either a public company, or a specialdistrict government. In fact, some degree offunctional fragmentation exists in almost allmetropolitan areas, but the way in whichthis is done varies widely.

A main advantage of functional fragmen-tation is that the autonomous agency is likelyto be more technically efficient because it isspecialized. The salary schedule may beoutside the normal civil service so that theagency can attract and retain higher qualityworkers. It also may be more efficient in itsoperations because it has a large enougharea of coverage to capture economies ofscale. Because it is usually the only entity inthe urban area responsible for the function,the problems of coordination for that functionare considerably less than under ajurisdictionally fragmented model. Finally, apublic company may have access to adedicated revenue stream (e.g., anearmarked tax, a compulsory transfer fromthe city government, or user charges), and ifwell-run, has arguably a greater potential fordebt finance than would a general purposelocal government.

There are drawbacks to the functionalfragmentation model, depending on theapproach taken. First, it will almost certainlybe less under the direct control of localvoters as would be an elected municipalcouncil, for example. In this respect, somedegree of local autonomy is lost. A secondconcern is that the autonomous agenciesmay be single purpose and therefore unableto contribute to coordination of servicedelivery across functions. Although thereare some exceptions, most special districtsare single purpose.

Developed Countries

Functional fragmentation can take anumber of forms, including the assignment ofseveral area-wide functions to a singlegovernment or agency. The GreaterVancouver regional district consolidated allfunctions provided previously by specialdistricts, most notably hospitals, water andsewer, capital expenditures, and solid wastemanagement. The governing board includes



291

elected local government representatives,but it is a voluntary organization and has noauthority to implement policies.

The financing of special districts and publiccompanies can take many forms. Since theservices delivered are often amenable topricing (e.g., public transportation, garbagecollection), user charges provide a base levelof revenues. In other cases, they are partiallyfinanced by compulsory transfers from thecity budget, or they might be profitableenough to subsidize the city budget.

In Stockholm, a holding company was or-ganized by the city to manage several cityowned companies that provide servicessuch as public housing, real estate ma-nagement, port operations and waterutilities. These public companies are in asurplus position and have been payingdividends to the city budget. The same istrue in the case of two energy companiesin which the City of Oslo holds equity.

The City of Paris participates (or is partowner) in several enterprises that provideservices ranging from transportation tosocial services. These are financed by usercharges and by compulsory transfers fromthe city budget. The City of Paris pays aboutone-third of its subsidies to the publictransport companies and covers almost halfthe budget for the Prefecture de Police.Transfers to the municipal company incharge of social programs accounts for asignificant percent of budget expendituresby the city government.

The City of Madrid makes compulsory trans-fers to the two public companies thatprovide transportation services. In theItalian metropolitan cities, the transfers tothe companies providing transportation,waste collection and disposal, and watertreatment services account for about 25percent of total metropolitan city govern-

ment expenditures. Milan, however, earnssignificant dividends from its companies.The City of Lausanne has fully incorporatedthe electric company into its budget, andthe company maintained a surplus positionduring the late 2000s.

The water boards in the Randstad region inthe Netherlands – with responsibility forflood control, water quality, and wastewatertreatment – are another example (OECD,2007a). These are local, independent publicauthorities that are democratically elected.The eleven boards in the Randstad region donot have administrative boundaries that arecoterminous with municipalities. The waterboards have taxing powers: a water boardcharge and a pollution levy.


Public companies play an important role indelivering services in the metropolitan areasin transition countries. Sometimes therelationship between the city governmentand the public companies is quite complex.For example, the City of Riga providesservices through 42 companies in which itholds ownership and through the heatingcompany where its equity stake is 49percent. Most of these companies are self-supporting, but the transport enterpriseclaims about 10 percent of the operatingbudget of the city.

In Zagreb, most capital spending (and somecurrent spending) is the responsibility of aholding company that was created followingthe merger of 22 municipal companies. TheCity of Zagreb uses more than 15 percent ofits budget for subsidy payments to theholding company. In other eastern Europeanmetropolitan cities, it is more a matter of thecity supporting the loss-making activity of asingle company, notably transportation. Themetropolitan cities of Sofia, Budapest, andOdessa are examples.



Special purpose agencies can be importantin managing and financing public servicedelivery in developing countries. Sometimesthis is because the special district statusgets the service delivery function separatedfrom the politics at the local level, sometimesit makes management easier and arguablymore professional, and sometimes it is aneasier route to a dedicated revenue streamand debt finance.

Public companies are set up by the localgovernments, as is the case of publictransportation in Bogota, Colombia. Theyalso can be multi-function, as for the water,energy and telecommunications company inMedellin, Colombia. In some cases, thespecial purpose agencies can become thedominant player in local governmentfinance. Webster (2000, p7) points out thatover 65 percent of urban infrastructureexpenditure in metropolitan Bangkok ismade by state enterprises, as comparedwith approximately 25 percent by thenational government and less than 10percent by the city government.

Metropolitan Government

The third general approach is metropolitangovernment. Under this model, generalservices are provided by an area-widemetropolitan government. Typically themetropolitan government is elected and hassignificant powers to regulate the servicedelivery and financing in the metropolitan area.While there are a number of area-widegovernments in large urban areas, few of themhave this range of powers. More often,they have a limited range of functionalresponsibilities, and govern alongside lowertiers of government. The emphasis under thisapproach is regional governance but usuallywith some degree of local autonomy protected.

There are significant advantages to thisapproach, most notably a built-in coor-

dination in the delivery of all functionsassigned to the metropolitan government.This gives a potential for better resourceallocation by comparison with the casewhere responsibility for local services isdivided among multiple municipalities andspecial purpose governments. The me-tropolitan government form also offers agreater potential for equalization becausethe quality of local services is not tied to thewealth of each local jurisdiction as it is in thecase of jurisdictional fragmentation.

Metropolitan governments often have alarge enough area of coverage to captureeconomies of scale and to internalizeexternalities. This could result in both lowercosts of service delivery and efficiencygains.6 Finally, because factors are lessmobile across than within metropolitanareas, there are more choices for efficienttaxation. There also may be significant taxadministration economies in an area-wideapproach to revenue raising.

The metropolitan form of governance alsohas significant drawbacks. The most im-portant is that it diminishes the power oflocal voters to influence the local budget. Ineffect, the election of the local council isreplaced by election of local representativesto the more distant metropolitan council. Asecond drawback is that metropolitangovernance often brings intergovernmentalconflict. If lower tier local governments existunder a metropolitan arrangement, theymay resist the leadership (and especiallythe dominance) of the metropolitangovernment.

Finally, the boundaries of the metropolitangovernment may not be drawn large enoughto fully capture the benefits of area-widegovernance. In this situation, one of themost significant advantages of metropolitangovernment may be substantiallydiminished. 7 For example, New York City

6. If it does not result inlower cost of providingservices, it may resultin the provision of ahigher level of servicesdue to a “leveling up”effect.

7. A comparison ofmetropolitan areaboundaries withgovernance boundariesin many cities is givenin OECD (2006).



293

has a population of 8 million but accountsfor less than 40 percent of the population ofthe New York metropolitan region. In thiscase, services provided outside themetropolitan government boundaries arenot coordinated with those provided insideand some of the advantages of area-widegovernance are lost. A similar situationexists in Toronto, which holds only aboutone-third of the population of themetropolitan region. In Copenhagen, where45 municipalities make up the greaterCopenhagen metropolitan area, the region-wide government – the National CapitalRegion – includes only 33 of these.

Developed Countries

About the closest the U.S. has come toarea-wide governance is the MetropolitanService District in Portland, Oregon. Thismetropolitan government includes 25cities and provides area-wide servicesfor transportation, solid waste and a numberof environmental concerns. But Portland(“metro”) is a far cry from a comprehensiveregional government. As Lefèvre (2008)notes: “By U.S. standards, metro is aninnovative metropolitan arrangement; yetby European standards, it is critiqued as aweak metropolitan governance arrange-ment with limited responsibilities and re-sources.”

Canada presents some interesting con-trasts in the structuring of metropolitangovernment. Toronto approximates a truemetropolitan government. It replaced theformer two-tier metropolitan governmentwith a single tier metropolitan city in 1998(Slack, 2000; OECD, 2009a). All localgovernment functions, including thosepreviously invested in special districts andunderlying municipalities, rest with thenew metropolitan government. Two otherstructural reforms in Canada, however,took a less centralizing approach.

Vancouver created a regional governmentwith some service delivery responsibilitybut the lower tier municipalities were leftas the dominant local government units.Montreal used an amalgamation ofmunicipalities to create two stronger corecities, but left in place a fragmented localgovernment structure.

There are other examples of area-widegovernments in OECD countries. In Madrid,the Autonomous Community of Madrid isresponsible for the metropolitan area, whichis about the same size as the functionalurban region of Madrid (OECD, 2007).Underneath the Community is 179municipalities, including the City of Madrid,which accounts for about half of thepopulation of the metropolitan area. Thefunctions of the Community are broaderthan those of the municipalities.

The Tokyo metropolitan government hassubstantial responsibility for serviceprovision to a population of about 12 millionpersons (Togo, 1995; Tokyo MetropolitanGovernment, 2010). It has a prefecture(state) status in Japan’s intergovernmentalfiscal system. Below the metropolitangovernment are 23 special wards in the corearea, in addition to 26 cities, 5 towns and 1village. All have elected assemblies. Thespecial wards carry out service delivery fordesignated functions on behalf of themetropolitan government, while themunicipalities are general purpose localgovernments.

The Greater London Authority (GLA) wascreated in 1999, as a senior level ofgovernment in metropolitan London, with aprovision to elect a mayor and, separately,an assembly. The GLA is responsible for anumber of functions, including transport,economic development, environmentalprotection, and police. About 80 percent ofexpenditures are made for transport and



police. It is financed by central governmentgrants (63 percent), user charges (20percent) and property taxes (10 percent)(Bird and Slack, 2004). In part becauseresources are so limited, it would bedifficult to classify London as a strongmetropolitan government. The underlying33 boroughs are independent of the GLAand provide basic urban services such aseducation, housing, social services, streetcleaning, and roads. There is a clear se-paration of expenditure responsibilitiesbetween the upper and lower tiers ofgovernment in the metropolitan area.


Metropolitan government has had an easierroad in many emerging and developingeconomies. Oftentimes, area-widegovernments were in place and theirboundaries simply grew with their popu-lations, while in other cases they werecreated to meet specific needs. In manycases, democratically elected local govern-ments are relatively new, and home ruletraditions are much less entrenched.Moreover, the weak level of infrastructure inplace and the strains placed on city financesby migration, make area-wide governmentan easier sell.

Before 1994, Cape Town, South Africa wascomposed of 61 local government entities.This number was reduced to 6 generalpurpose governments and a metropolitanauthority in 1996, and finally to a singlelocal authority, the “Unicity” of Cape Town in2000 (OECD, 2008). The gross inequity inservices provided and the need for localinput and coordination of area-wide ser-vices, were driving forces behind theconsolidation.

A somewhat different model was adopted inManila, where the Metropolitan ManilaDevelopment Authority (MMDA) exists to

manage area-wide functions while the localgovernment units are responsible for localfunctions. The local government units (citiesand municipalities) are governed by electedcouncils while the Chair of the MMDA isappointed by the President and itsmembership is prescribed by law. Theformation of the MMDA (and its predecessorbodies) was a result of the concern fordelivery of area-wide services and theperception of government that the well-being of Metropolitan Manila is a nationalpriority. The history of metropolitangovernance in Manila has been one of astruggle for power between the metro-politan government and the lower level localgovernments.

Taxes, Charges, and Transfers8

Culture, economic structure, and politics allplay a role in determining the particulars ofa public financing regime in metropolitanareas. But there also is a theory of taxassignment that points the way to “bestpractice” in financing metropolitanservices.9 The guidelines from this theoryare generally followed in many (most)developed countries, but are less oftenfollowed in the emerging and developingcountries.

The Theory of Tax Assignment

Is there a best way to finance public serviceprovision in metropolitan areas? Are thereguidelines for identifying those tax revenuesources most appropriate for financing localand area-wide governments in metropolitanareas? The answer to both questions is aqualified “yes”.

Four considerations might guide taxassignment decisions in metropolitan areas.The first is accountability. In order to makeelected local government officials moreaccountable to their voting constituents, it is

8. For other papers onthis topic, see Bahl andLinn (1992), Bird andSlack (2004) andChernik andReschovsky (2006).

9. One of the beststatements of thetheory is McLure(1998). For reviewsand applications, seeMusgrave, 1983; andMartinez-Vazquez,2008.



295

necessary to give these officials someindependent taxing powers. Theaccountability rule for tax assignment fitslower tier local governments in those metro-politan areas where the city council iselected. Those who hold fast to the repre-sentation rule of taxation would tend to limitnon-elected governments to cost-recoveryuser charges. If there is to be generaltaxation by a non-elected body, it should beenabled by a referendum.

The issue is more complicated when it comesto metropolitan government or area-widespecial purpose governments. If thearea-wide government is elected, taxingpowers will enhance accountability tovoters. This would apply, for example, to thecases of the Portland, Toronto, or Londonmetro areas, where the leadership of the metrogovernment is elected. Some area-widegovernments (special districts such asNew York’s Port Authority or Toronto’sServices Board) are led by appointedofficials. In this case, accountability tovoters will not be enhanced by taxingpowers. These agencies should charge forservices rendered, but their monopolypowers should be regulated. In yet anotherarrangement, the council of the metro-politan government may be made up of allelected mayors in the metropolitan area.Because of the large membership of thecouncil, this arrangement may allow a singleelected mayor to hide from beingaccountable to his home constituency. Thereare many examples here, including theGreater Vancouver Regional District.

A third principle, and the one that seemsto be followed most religiously indeveloped countries, is “correspondence”,i.e., local governments should not levytaxes whose burden can be exported tothose who do not benefit from servicesdelivered by the local government(McLure, 1998). This principle imposes a

tight restriction on local governments injurisdictionally fragmented metropolitanareas. It suggests that lower tier localgovernments should rely only on benefittaxes10 and on taxes on immobile factors.Metropolitan governments and area-widespecial districts, on the other hand, can begiven access to some broader-based taxesbecause labor is less likely to cross juris-dictional boundaries.

Finally, the theory of tax assignment alsocalls for consideration of the relative costsof tax administration in deciding on thelevel to which a tax will be assigned. Localgovernments, particularly in developingcountries, might be denied access to cer-tain taxes for this reason, while area-widegovernments in metropolitan areas couldhave some inherent advantages in taxadministration.

Tax Assignment: The Practice

Do countries follow the general “rules” laiddown by the theory discussed above? Mostdeveloped countries do make tax assign-ments that are in step with good practice(though there are exceptions). Metropolitanarea governments in developing countrieshave many fewer taxing options, andappear to be less in step with what manypolicy analysts see as best practice.

Developed Countries

Higher income countries appear to have givenmore attention to the issue of structuringgovernance in large metropolitan areas and tofinding ways to finance these structures.Examples of the “special” financial arrangementsthat have been put in place include (a) grantingmetropolitan governments both city and statelevel status (Tokyo, Shanghai, Berlin), (b) pro-viding for special taxing powers (New York City)and (c) instituting special intergovernmentaltransfer arrangements (London, Rome). A



sample of such special arrangements isdescribed in Table 9.1.

One underlying objective in many developedcountries is to increase the fiscal self-sufficiencyof metropolitan local governments. In somecountries in the sample reported in Figure 9.2,this strategy has succeeded. The Tokyometropolitan area government has both city andprefecture (state) status, hence it has access to abroader tax base than do other localgovernments in Japan. About 70 percent of allmetropolitan government revenue is from localtaxes.11 Toronto has a more traditional financingstructure for a local government. It reliesprimarily on the property tax and user charges.The Toronto metropolitan city funds about 60percent of its budget from property taxes and

user charges. The property tax alone accountsfor about 41 percent of revenues (OECD, 2009a).

Metropolitan local governments in some theNordic countries and Spain rely primarily onindividual income taxes, and New York Citymakes heavy use of a combination of retailsales tax, personal and corporate incometaxes, and business taxes. Stockholm’s localgovernments cover about 80 percent of theirexpenditures from local sources, primarilyfrom an earned income tax. In Paris, the prin-cipal local tax was applied on economicactivities –a tax on inmobile capital but limitedto added value. This was replaced in 2010 bycombination of a tax on property, a share oflocally collected VAT and a variety of lowreturn charges on network activities, which

10. A “benefit tax” in thiscase could refer to anytax where the revenuesraised are borne bythose who benefit fromthe services financed. Aresidence-basedincome or payroll taxwould qualify, but anorigin based businesstax would not.

11. The largest revenuesource is the corporateincome and registrationtax. Tokyo and NewYork are the largestmetropolitangovernments studiedhere that rely to anysignificant degree ontaxes on corporations.

Table 9.1: Selected Special Revenue Treatments for Metropolitan Local Governments

Source: http://www.metropolis.org/publications/metropolitan_regions

May impose a water board charge and a pollution levy.Netherlands: Independent, elected Water Boards

Special central Transfers cover about two-thirds ofExpenditures

London (GLA)

Has both State and Metropolitan Government Features, andState Taxing Powers

Community of Madrid (OECD, 2007)

Metropolitan City with no underlying tier of government,created by a Provincial Act.

Toronto (Slack, 2000)

Metropolitan government has both prefecture (state) andmunicipal taxing powers and expenditure responsibilities

Tokyo Metropolitan Government (Tokyo, 2010)

Financed 75 percent by Central Transfers and 25 percent byMunicipal Transfers

Copenhagen National Capital Region (OECD, 2007)

SSppeecciiaall TTrreeaattmmeennttGGoovveerrnnmmeenntt



297

globally have meant a loss of autonomy and adrop in revenue.

Metropolitan local governments in otherdeveloped countries do not have significanttaxing powers (Slack, 2007). The GreaterLondon Authority receives most of itsrevenues from central government grants.The Stuttgart Regional Authority has notaxing authority. The Greater VancouverRegional District is financed primarily byuser fees and intergovernmental transfers.


In practice, large urban governments inmost emerging and developing economiesdo not rely heavily on local taxation. Despitethe arguments that local governments inmetropolitan areas could feasibly handle agreater range of taxes, most are limited to

property taxes and user charges as themain sources of revenue. There are someexceptions to this general pattern, notablyBrazil, and these are taken up below.

Property Tax

Almost everyone’s choice for a major ins-trument of local government taxation isthe property tax. It passes many of thetheoretical tests of a good subnationalgovernment tax, but it is costly toadminister and it is politically unpopular.

Developed Countries

Among the developed countries, propertytax is a favorite among the English speakingfederal countries, but it is much lessimportant among non-English speakingcountries and unitary countries in general

0

20

40

60

80

100

City of Stockholm (2006a)

Copenhagen Municipalities

(2007)

Copenhagen Capital region

(2007)

Grand Lyon (2006c)

Madrid Community

(2007)

Tokyo Metropolitan Government

(2010)

Toronto Metropolitan Government

(2009)

Figure 9.2: Percent of Revenues Raised from Own Sources by Selected Metropolitan Local Governments

Source: All information taken from OECD Territorial Reviews with the exception of the Tokyo Metropolitan Government

taken from (http://www.metro.tokyo.jp/ENGLISH/PROFILE/index.htm)



(Lotz, 2006). Property tax revenues accountfor one-half or more of local governmentfinancing in the Toronto, Montreal, andMelbourne metropolitan areas, and itaccounts for 34 percent of the budget inNew York City.

Different patterns emerge for somemetropolitan area local governments inother OECD and transition countries.Municipalities in the Netherlands, includingthose in the Randstad region, receive lessthan 5 percent of revenues from theproperty tax. There is no local governmentproperty tax in Sweden or Norway (OECD,2006a, p176). In Copenhagen, theprimary revenue source of municipalities isthe income tax, and property taxes playonly a minor role. The same is true inStockholm, Tokyo, and in the Swiss cities.The property tax is somewhat moreimportant in Madrid at the city level, butfinancing is dominated by income taxeslevied at the regional government level. InBusan and Daegu, Korea, the property taxis an important source of local financing,but most of the revenue is derived fromproperty transfers.


Governments in most emerging and deve-loping countries do not seem to have boughtinto the idea that the property tax is a good fitfor financing services provided in metropo-litan areas. While it is true that property va-lues are growing in most metropolitan areas,valuation in most countries fails to capturethis growth. This seems to be the case even incountries with large metropolitan areas(Mathur, et. al., 2009; de Cesare, 2004).Moreover, delays in general revaluation arecommonplace, significantly lowering therevenue-income elasticity of the property tax.The property tax as practiced in developingcountries generally fails the tests for a goodsubnational government tax in terms of its

high administrative cost and its unpopularitywith voters.

There is a great deal of variation in the ex-tent to which the property tax generatesmeaningful revenues for metropolitan cities.In Cape Town, about 20 percent ofmetropolitan government revenues are de-rived from a tax levied against the capitalvalue of land and improvements. This isabout the same share of revenues that isreceived from intergovernmental transfers.

The primary source of revenue for muni-cipalities in the Mexico City metropolitanarea and in the Istanbul metropolitan area,is the property tax. However, in neither caseare the local governments empowered toset the tax rate or determine the tax base.The result is that there is relatively littleautonomy for the metropolitan localgovernment to determine its revenue level,and in both cases the property tax falls wellshort of its potential. There is some localgovernment discretion in metropolitan citiesin India but the results are much the same.The low yield is largely attributed to thepoor administration of the tax. For example,in Mumbai only about 70 percent ofproperties pay taxes, and in Kolkataproperties are assessed at about 20 percentof their value (Mathur, et. al., 2009).

Income and Payroll Taxes

The individual income tax can meet many ofthe tests for a good metropolitangovernment tax. It can generate significantrevenue from an elastic tax base. It isroughly consistent with the correspondenceprinciple in that the burden falls mostly onthose who benefit from the servicesprovided, though correspondence problemsdo arise with respect to those who crossprovincial borders to reach their place ofwork.12 Its administration can be simplifiedby a direct piggyback on the income tax of a

12. The correspondenceprinciple would call fora residence-basedincome tax, and fornon-residents to filereturns and pay anamount that wouldserve as a benefitcharge for localservices received. For adiscussion, see McLure(1998).



299

higher level government, or by informationsharing with a higher level government.

Perhaps the major drawback of the personalincome tax, as a metropolitan local govern-ment revenue, is that it is cyclically sensitiveand can leave a local government in adifficult financial position during an eco-nomic contraction. This sensitivity is ofgreatest concern where there is little diver-sification in the city revenue structure andwhere the central government does nothave the financial strength to compensatefor the revenue losses. During the recenteconomic down turn, personal income taxesin the cities of Riga, Bucharest, andBudapest all declined significantly and had amajor budget impact. By contrast,Stockholm and Lausanne also rely on inco-me taxation, but neither suffered as muchbudgetary stress during the economiccontraction because their revenue structu-res were more diversified.

Developed Countries

New York City has long levied an earningstax, and until 1999, the liability was withcommuters as well as residents. Theearnings tax now finances about 16 percentof the city budget. The major source of localgovernment revenue in Cleveland is anearnings tax.

Urban government income taxes are moreprevalent in Europe than in the U.S. andCanada, and in many cities are the do-minant sources of local government re-venue. The piggy back approach to incometaxation offers considerable advantages tosome metropolitan local governments. Itallows local rate determination whileavoiding the issue of defining the tax baseor administration of the tax. The primaryrevenue source for Swiss cities is apiggyback personal income tax. Rome leviesa residence-based income tax, on a base

defined by the central government. Theprincipal municipal government revenuesource in metropolitan Copenhagen is theindividual income tax (OECD, 2009). Thetax base is defined by the centralgovernment, and collections are made bythe central government. In theory, the 45municipalities in the metropolitan area arefree to set the tax rate, but since 2002 acentrally imposed freeze on the tax rate hasbeen in place. About 80 percent of municipalrevenue is raised from the income tax. TheCapital Region, the metropolitan area-widegovernment in Copenhagen, has no taxingpower.

Local governments in the metropolitan re-gion of Stockholm rely almost exclusively ona local tax on the earned income ofresidents. The base is defined by the na-tional government, but local governmentsare free to set the tax rate. The major con-cerns with the earnings tax in Stockholmare (a) that such complete reliance on itleaves the municipalities vulnerable tocyclical movements in the economy, and (b)the equalization formula that limits therevenues a local government can receive,provides a significant disincentive to reve-nue mobilization.

The area-wide government in Madrid (TheCommunity of Madrid) relies on the indi-vidual income tax for most of its revenues.The tax base is defined by the centralgovernment,13 but the regional government(community) may choose the tax rate(subject to some restrictions) and is entitledto one-third of revenue collections (OECD,2007).

There is less use of corporate income ta-xes at the local government level in metro-politan areas, arguably because of cyclicalsensitivity of the revenues and of a fear ofdriving away investment. However, thereis some practice. Both Geneva and Lisbon

13. The Community mayprovide certainpreferentialtreatments, therebylowering the tax base.



derive significant revenue from a sur-charge on the corporate income tax. NewYork City derives a significant amount ofits budget revenue from the corporateincome tax.


The Eastern European Cities use a differentmodel of local income taxation. In thesecases, it is a sharing of central income taxesbased on collections in the city (Zagreb andBucharest) or collections from residents(Riga). The central government sets the“local” tax rate, e.g., 26 percent in Riga in2008. The City of Zagreb may levy a surtaxof up to 30 percent on personal income taxcollections, and presently it levies a rate of18 percent.

In general however, subnational govern-ment income taxes are rarely found inemerging and developing economies. The-re are four reasons for this. The first is theadministrative problems that would beposed. The second is tax competition.Central governments in some developingcountries rely heavily on this source ofrevenue, and even have trouble collectingmuch from the personal income tax (Birdand Zolt, 2005). The third reason is thatincome taxes often carry incomedistribution goals and these are perceivedto be the exclusive responsibility of thecentral government. Finally, incomegeneration in the formal sector isconcentrated in most developing coun-tries, and it is not likely that much revenuewould be generated outside the metro-politan areas.

States and the Federal District withinmetropolitan Mexico City, finance a part oftheir budgets from a payroll (wage) tax.They are free to choose the tax rate,define the tax base and administer thetax. The tax is collected from employers.

There are three problems with the Mexicanpayroll tax that need to be reckoned with.First, it is a tax on wages, and if levied at ahigh enough rate, could drive someemployment to the informal sector.Second, since it is levied at the place ofwork, it will result in some tax exporting tonon residents.14 Third, it is cyclicallysensitive and can create significant budgetproblems during an economic contraction.

Sales Taxes

Sales taxes are an attractive option for fi-nancing the provision of local public servicesin metropolitan areas, because with asignificant share of consumption takingplace within the large urban areas, therevenue potential is considerable. Also, asthe point of tax collection for mosttransactions is identifiable, administration isfeasible in developed countries and in somedeveloping countries. However, metro-politan government sales taxes raise thepossibility of introducing unwanteddistortions in resource allocation, if they arenot properly structured.

Developed Countries

The retail sales tax is an important sourceof revenue for many local governments inU.S. metropolitan areas (e.g., about 11percent of the total revenue budget in NewYork). In Canada, local governments inmetropolitan areas are mostly limited tothe property tax, and cannot levy a generalsales tax, though many local governmentsdo impose various selective sales taxes. InItaly, regional governments levy a valueadded tax on an origin basis (Bordignon,Giannini and Panteghini, 2001). Bird andSlack (2004) have argued that this form ofbusiness taxation deserves wider atten-tion. Local government sales taxes are notwidely used in the metropolitan areas ofEastern Europe.

14. For a discussion of thepayroll tax in Mexico,see Diaz-Cayeros andMcLure (2000)



301


In theory, a destination-based sales taxcould satisfy the conditions for a good localgovernment tax in emerging and developingeconomies. In theory, it could be leviedat either the metropolitan governmentlevel or at the lower tier of metropolitangovernment. Theory and practice, however,do not always come together because ofadministrative constraints in most emergingand developing economies.

A value added tax (VAT) is usually levied bycentral governments as a destination-basedsales tax. Unfortunately, in most developingcountries, there is no realistic prospect thatthe tax administration will be able tosupport a subnational government VAT.15 Avalue-added tax levied at the metropolitancity level is almost certainly not feasible,except perhaps in the special case where ametropolitan city covers a large area and hasthe legal status of a state/province, e.g.,Shanghai or Berlin. A better route would be asurcharge on the state government levy or arevenue sharing arrangement with the stategovernment. Here there is some experience.One fourth of the state value added taxin Brazil is distributed to municipalitieson a basis of point of collection. A similararrangement exists for province-level citiesin China (Fu, 2007).

Retail sales taxes are not administrativelyfeasible in developing countries, even inlarge metropolitan areas. A large percent ofsales takes place in the informal sector, anda destination-based retail sales tax wouldlikely swell this number even more. Informalsector retailers usually do not keep accuratebooks of account, and small merchantsoften keep a “special” set of accounts for taxpurposes.

A gross receipts tax levied on all sellers inthe formal sector, on an origin basis, can

be revenue productive. But, this willcreate distortions by shifting tax burdensfrom producing to consuming regions, byintroducing a cascading effect on prices,and by discriminating against formalsector sellers. An origin-based sales taxis also subject to the “headquartersproblem” or the problem that arises whenfirms pay their tax bill at the central officelocation. While all of these reasonssuggest that a gross receipts tax is not agood choice for financing governments inmetropolitan areas, reason is sometimesoutweighed by the appeal of a significantrevenue take.

The major own source revenue ofBrazilian municipalities is a tax on ser-vices (ISS), almost all of which is collec-ted by the largest municipalities (Rezendeand Garson 2006). The ISS and the urbanproperty tax together account for about60 percent of total local tax revenue.Buenos Aires, both city and province, levya gross receipts tax. The tax is com-plicated because the tax rate varieswidely by type of product. Bogota derivesmuch of its revenue from a gross receiptstax. The business tax in the Philippinesis levied on gross receipts and accountsfor about 30 percent of local revenues(Taliercio, 2005).

Charges and Fees

Most analysts argue that benefit chargesof one form or another are the mostappropriate revenue source for localgovernments (Oates, 1972; Musgrave,1983; Bahl and Bird, 2008). It constitutes acharge for benefits received, and may leadto recovery of the cost of providing theservice in question. Central governments inmany countries cede this revenue source tourban local governments because it doesnot crowd out central revenues as might alocal income tax or a consumption tax.

15. There are efforts andeven some experiencewith implementation.Brazil has long reliedon a state level valueadded tax. India alsohas implemented astate level value addedtax, but is still workingout the details of how itwill operate,particularly withrespect to interstatetrade.



Developed Countries

User charges are particularly important forfinancing the operating expenditures ofspecial purpose districts and public com-panies in metropolitan areas. The com-parison for selected metropolitan areaspresented in Figure 9.3 gives some idea ofthe relative importance of user charges infinancing metropolitan services. Eventhese amounts are an understatement,however, because they do not include theuser charges levied by public companies.European cities often are full or partialowners in various types of public servicecompanies (e.g., public transit, water,energy, housing) that finance theiroperations with user charges. Sometimesthe charges are supplemented by asubsidy paid from the city budget, andother times the user charge revenuesgenerate a surplus that is distributed backto the city.


User charges are a particularly attractiverevenue option for metropolitan localgovernments in emerging and developingeconomies. The levy can be linked toservice benefits, hence there is morewillingness to pay, and both assessmentand collection can be easier than in thecase of general taxes. On the other hand,charges often are made for essentialservices, and sentiments can run highwhen increases are necessary to coverrising costs. Public housing rents, waterrates and bus fares are examples.

The City of Cape Town collects about 35percent of revenues from user charges,mostly electricity, water, sanitation, andrefuse collection. While the revenue take isquite large by comparison with othermetropolitan cities, there is poor complianceand low collection rates (OECD, 2008). This is

20

40

60

80

100

0

Toro

nto

Met

ropo

litan

(2

009)

Toky

o M

etro

polit

an

Gov

ernm

ent

(201

0)

City

of M

adrid

(2

007)

Mel

bour

ne

Met

ropo

litan

R

egio

n (2

003)

Cop

enha

gen

Met

ropo

litan

M

unic

ipal

ities

(2

009)

Por

tland

Met

ro

(200

6)

Gre

ater

Lon

don

Aut

horit

y (2

004)

Gre

ater

V

anco

uver

R

egio

nal D

istr

ict

(200

4)

City

of

Sto

ckho

lm

(200

6a)

Gra

nd L

yon

(200

6c)

Figure 9.3: User Charges as Share of Total Revenue for Selected Metropolitan Local Governments

Source: All information taken from OECD Territorial Reviews with the exception of Tokyo

(http://www.metro.tokyo.jp/ENGLISH/PROFILE/index.htm) and Greater London Authority and Greater

Vancouver Regional District (Bird and Slack 2004)



303

a not uncommon outcome in emerging anddeveloping economies. For example, theBangkok Metropolitan Administration collectsonly about 20 percent of charges due forgarbage pickup (Webster, 2000, p17).Mohanty, et al. (2007) reported a low rate ofcost recovery for the Indian metropolitancities of Mumbai and Kolkata. By contrast,user fees in Bogota are sufficient to coveroperating costs for the city’s urban bustransport company (RTI, 2005).


A special treatment as regards the flow ofintergovernmental transfers may be reservedfor metropolitan cities, sometimes to excludethem from certain flows to encourageself-sufficiency, and sometimes to recognizetheir special needs. In Rome, for example, aspecial central transfer equivalent to about15 percent of the current revenues of thecity is given as a recurrent grant torecognize Rome’s capital city status. Somemetropolitan areas have the same statusas states or provinces, in which case theyhave both a state and a city entitlement tointergovernmental transfers. Also differentfrom other local governments, a specialprogram of intra-metropolitan governmentrevenue sharing may be in place.

Developed countries

Metropolitan area governments in somedeveloped countries do not depend asheavily on intergovernmental transfers asdo other local governments. For example,the cities of Stockholm, Paris, Madrid, andLausanne all raise more than two-thirds oftheir own financing from local taxes andcharges.

A few illustrations may help describe thegeneral practice. Grants play a minor role infinancing municipalities in metropolitanCopenhagen. Specific grants account for

about 10 percent of revenues, and these areprimarily for reimbursement of agencyfunctions performed on behalf of the centralgovernment. On the other hand, the area-wide capital region government, which isprimarily responsible for health care, has noindependent taxing powers. About 75percent of its financing comes from centralgovernment transfers and 25 percent frommunicipalities’ transfers.

The Madrid regional government receivesabout 20 percent of its revenues in sharedtaxes and grants. The municipal govern-ments, however, depend on transfers forabout 40 percent of revenues. Large muni-cipalities receive most of their transfers inthe form of shared taxes, while smallermunicipalities receive formula grants basedon indicators of expenditure need and taxeffort. Melbourne and Toronto finance onlyabout 15 percent of their respective budgetswith intergovernmental transfers.

There are some exceptions to this generalpattern. One is Germany, where localgovernments have few independent reve-nue sources, and rely almost exclusively ontransfers. In the Netherlands, localgovernments have only limited taxingpowers. Equalization is done throughtransfers.


Metropolitan local governments in emergingand developing economies are heavilyreliant on intergovernmental transfers fromfederal and state/provincial governmentsthough usually less so than are other localgovernments. There are several reasonsfor this. Metropolitan cities have a strongertax base and are sometimes given morelegal taxing powers than other localgovernments. Often times large urbangovernments have a better ability to collecttaxes. They also may have more incentive



to levy taxes than do other localgovernments. If the higher level govern-ments use equalization grants, metropolitancities will receive less relative to theirpopulation size and therefore will be pushedtoward raising more own source revenues.

In Cape Town, only 20 percent of metropo-litan city revenues are derived from grants.The major transfer in the system – “theequitable shares grant” – is allocated on anequalizing basis. The result is that Cape Townand the other metropolitan cities in SouthAfrica receive about half the per capitaamount that goes to smaller cities.

The Metropolitan Manila DevelopmentAuthority has no taxing powers and limitedauthority to levy user charges. It reliesalmost exclusively on grants from thecentral and provincial governments and onmandated contributions from the underlyinglocal government units. In effect, the lowertier local governments pay the metropolitangovernments for services delivered.

Mexico has a highly centralized financingstructure. Sub-national government taxesaccount for less than 1 percent of GDP. Mostmetropolitan services are financed fromconditional (22 percent of spending) andunconditional transfers. However, thedependence on transfers is significantly lessin the Federal District of Mexico than in theother states in Mexico.

About 50 percent of revenues of themetropolitan municipality of Istanbul comefrom intergovernmental transfers. The mostimportant of these (50 percent of revenue)is the revenue sharing grant which isdistributed on a derivation (origin ofcollection) basis. This basis for distributionfavors Istanbul greatly because of its largetax base, and because it receives a share ofthe tax paid by all companies that areheadquartered in Istanbul.16

Brazil uses both discretionary grants andequalization grants to support localgovernments. The former, for educationand health, probably favor metropolitancities, but the latter do not. Rezende andGarson (2006, p20) report that the tenlargest metropolitan areas, which house 30percent of the Brazilian population andgenerate about half of the national GDP,receive only about 13 percent of thedivisible pool from shared income andindustrial products tax.

The core municipalities in metropolitanareas in Eastern Europe are heavily finan-ced by intergovernmental transfers, oftenin the form of a shared personal income tax(e.g., Bucharest, Budapest, and Zagreb).Revenues from these transfers werebuoyant during the economic expansion ofthe early 2000s, but contracted signi-ficantly in the later part of the decade. Inother cases, the transfers take the form ofconditional grants that are restricted toparticular uses. The City of Zagreb pro-vides decentralized services, but under astrict set of central government earmarks.The revenue structure of Budapest is do-minated by intergovernmental transfers ofvarious forms, including both revenuesharing and conditional grants. The centralgovernment provides about 70 percent ofSofia city revenues through intergovern-mental transfers.

Conclusions: Is There an Easy Way Forward?

Removing the constraints to providing anadequate level of public services inmetropolitan areas is a subject that is highon the policy reform list in many countries.But the “right” way to deliver public servicesin large urban agglomerations, and tofinance these services, is still an openquestion. This review shows that theinternational practice is quite diverse.

16. For a discussion of this“headquartersproblem” see Bahl andSolomon (2003).



305

What is the Real problem?

The underlying problem to deal with inmetropolitan governance and finance isfinding a way to marry two very differentspatial units. The functional economic regionhas boundaries that are informal and alwayschanging, as one would expect of a labormarket area. The “champion” of the region asa government entity is the planner or socialreformer who sees the great efficiency andequity gains that would come with regionalservice delivery. The other spatial unit, thelocal government, has fixed boundaries. Thechampions of the local government arevoters and their elected officials, who want tomaintain control over services provided in thelocal area. It seems unlikely that these twounits, with their differing priorities, will cometogether easily in support of a generalpurpose region wide government. The issueis even more complicated by overlappingspecial districts or public companies whoseservice boundaries may not be coterminouswith either the metropolitan area (labormarket area) or the general purpose localgovernments.

The public policy solution lies in finding away to deliver some services with a degreeof local control and financing, while de-livering others on a region wide basis andwith a broader finance base. All govern-ments will likely identify with a model thatproduces better prospects for long runeconomic growth, and better transportationservices and public amenities. Localgovernments can be moved by strategiesthat give them some voice and a promiseto hold a lid on taxes. But none of thesearguments seem to be convincing when itcomes to moving basic services away fromthe local level, or more drastic yet,abolishing local units of government. Thepractice shows that governance and financein some metropolitan areas have movedtoward this solution, but almost no one

would declare that the delivery of regionalservices is properly coordinated.

Is there a next step in improving andrationalizing the financing of large cities inmetropolitan areas? Three policy directionswould seem worth considering.

Governance

It may be time to move away from the goodbut academic advice that area-wide, generalpurpose local government are the answer tothe public financing problems inmetropolitan areas. Home rule is too firmlyentrenched to be dismissed, at least in theforeseeable future. Where this shift inemphasis would take public policy is towardemphasizing a two-tier metropolitanstructure, probably overlaid by a capitalinfrastructure district(s) for services thatlend themselves to pricing, and a regionalcoordination and planning district. Attentioncould then shift to designing a system oftaxing and charging that would best fit atwo-tier governance.

The above model might also work inemerging and developing economieswhere there is a tradition of local govern-ment, e.g., Manila, Calcutta or São Paulo.In all three metropolitan areas, a kind oftwo-tiered governance is in place but therevenue mobilization system is not welldesigned. Where something akin to ametropolitan government is in place, itcould be overlapped with an infrastructuredistrict(s) and a coordinating body. Amajor struggle in such cases would be toinsure that the government boundariesgrow with the boundaries of the economicregion. Again, the question would be howto design a financing system that wouldmatch up with the governance system.

How would this differ from the presentsystem? One difference is that the regional



districts would need to be elected and to beviewed as financing districts as well asservice delivery areas. Enabling legislationby higher level governments would berequired to make this happen. Second, thetaxation instruments used by the lower tiergovernments should be designed to fit thebasic efficiency rules of taxation, e.g., notax exporting. Some sort of monitoringmight be put in place to regulate intra-metropolitan practices that distort tradebetween the member communities. Third,higher level governments should ensure abetter match between expenditure assign-ment and revenue assignment. This wouldrequire important changes on the expendi-ture side – more clarity and fewer mandates– and would require passing more economicgrowth-responsive tax bases to themetropolitan local governments. Finally, theprovincial and national intergovernmentalsystem should be redesigned to give moreautonomy to metropolitan area localgovernments, and at the same time to limitthe flow of grants to them in order toencourage local revenue mobilization.

Metropolitan Finance

As always, part of the problem inmetropolitan areas is that resources arescarce and do not match up well with thedemand for public services. But the level oftaxation in some OECD countries is veryhigh and additional taxing space may belimited. In the United States, there may beroom for tax increases, but a combination ofthe federal deficit and political backlash maycrowd out most opportunities for increasedlocal government taxation.

Ideally, one could compare the level andstructure of taxation across metropolitanareas. Unfortunately, there is scant eviden-ce on the finances of overlapping govern-ments in metropolitan areas, as indicated bythe relatively few (and somewhat dated)

special studies reviewed above. More timelyevidence is available for core citygovernments and from the credit ratingagencies that regularly evaluate thesemunicipalities.17

The major stumbling block in developing amodel system of metropolitan finance is thelower tier local governments. Those weddedto fiscal decentralization within metropolitanareas will argue for heavy assignment ofexpenditure responsibility to these localgovernments. If these local governmentsare to be moved toward budgetaryindependence, significant tax assignment isimplied. The property tax and user chargesalone will not likely carry the expenditureload, so long as the property tax remains sopolitically contentious. But broad-basedtaxes in jurisdictionally fragmentedmetropolitan areas are likely to becharacterized by a significant amount ofexporting of burdens to residents in otherjurisdictions. The recourse is to use aresidence-based earnings tax with somesort of compensating mechanism as is thecase in Denmark, or to rely heavily onintergovernmental transfers to finance localservices.

There is room for metropolitan (area-wide)governments to contribute more to thefinancing of services in the metropolitanarea. A residence-based income tax, with anappropriate commuter charge, would be anattractive alternative. This might be justifiedfirst, on grounds that area-wide servicesprovide benefits to non-users, hence lessthan full financing from user charges wouldbe efficient. Second, this tax could supportan intra-metropolitan equalization fund.This would remove the disincentive to lowertier tax effort, as has been observed insome Nordic countries. Intra-metropolitanrevenue sharing has much to recommend itrelative to a national or provincial programof grants.

17. The credit rating data,however, is focused onindividual localgovernments. It doesnot take account of theimpact of overlappinggovernments on coremunicipalities. Forexample of a creditrating report, seeStandard & Poors.



307

Some other taxes that might be appropriatefor a metropolitan taxing district are the propertytax, or at least the commercial-industrialportion of it, and taxes on motor vehiclelicenses. User charges would continueto play a major role in financing theregional district, but grants from higherlevel governments would not.

For single purpose special districts, revenuemobilization could continue to be centeredon user charges, as is the case now. Specialbenefit taxes, raised within the metropolitanarea, could be dedicated to the specialdistrict. There is much more room here forpublic financing. Finally, general taxation tosupport a service is a possibility (and onethat has been tried in several places), butshould only be done if authorized by a voterreferendum.

In emerging and developing economies,metropolitan cities need to ratchet up thelevel of own source revenues but have lessadministrative capacity or legal authority todo so. Unfortunately, the will to increaselocal taxes is often not in place. If some ofthese constraints are removed, significantrevenues might be raised in the largestmetropolitan cities. The most viable optionsfor increasing the rate of revenue mo-bilization might include:

• Increase the rate and the collectionefficiency on user charges. Give metro-politan local governments discretion toset the level of user charges.

• Improve the administration of theproperty tax so as to raise the overalleffective rate. There are many ways todo this, and there is a literature that hasexplored this at length in developingcountries. Among the options thatmight be considered, in most countriesthey are removing preferentialtreatments, increasing valuation rates,

bringing untapped properties into thebase, installing a “payment in lieu” forgovernment properties, simplifying pay-ment options and increasing penalties fornon-payment.

• Levy taxes on wage income, either di-rectly or as a piggyback on the nationalgovernment income tax.


Central and state (provincial) governmentsmight adopt an explicit strategy of usingequalization grants to discriminate againstlarge cities in order to induce them to makea greater level of tax effort. This is alreadydone in some developed countries. In fact,integrated transfers might be limited toconditional grants that would stimulatespending for a national priority. Thisfinancing strategy will of course be limitedby the extent to which expenditureresponsibilities have been assigned to localgovernments in metropolitan areas.

In the emerging and developing economies,the metropolitan local governments tend tobe more dependent on transfers from higherlevel governments. This often dampensthe enthusiasm of local governments tomobilize more revenues from their ownsources. A strategy of replacing mosttransfer revenue with increased localtaxing power would not be met withgreat enthusiasm by local politicians,because of the political pain associatedwith imposing taxes. It has been argued inmany countries that a significant amountof untapped taxable capacity resides inthe metropolitan cities. Authorization ofa model of local government taxation,such as that discussed above, couldprovide an incentive for metropolitan areagovernments to find a way to tap thiscapacity.


0w2010 10 Conclusion DEFcarta ang 15/9/11 23:23 Página 308

10. CONCLUSION

JJOORRGGEE MMAARRTTÍÍNNEEZZ--VVÁÁZZQQUUEEZZ


PPAAUULL SSMMOOKKEE

NNEEWW YYOORRKK UUNNIIVVEERRSSIITTYY,, UU..SS..AA..


311

Overview

Local governments around the world havebecome key public sector actors during thepast two decades, and decentralizationnow ranks among the most common andconsequential global reforms (See FirstGOLD Report on Decentralization and LocalDemocracy). This trend could ultimately beas influential as other major institutionaltransformations of the past century, suchas decolonization in Africa and Asia or thetransition from planned to marketeconomies in the former Soviet Union,China, and elsewhere. Indeed, localgovernments have in many respects trulycome of age. Their role is reinforced inglobal policy circles, including throughmajor multilateral proclamations, suchas the European Charter of LocalSelf-Government (1985) and the UNHabitat Guidelines on Decentralization andReinforcement of Local Governments (2007).

In many regions of the world decentraliza-tion has enhanced the functions and auto-nomy of local entities. Local governmentsplay increasingly more critical roles in deli-vering basic infrastructure services, suchas roads, transportation and water, and so-cial services, such as education and health.These developments have contributed inminor and major ways to the progressivedeepening of local democracy, the allevia-tion of internal regional tensions in conflictprone areas, the promotion of broader anddeeper citizen participation in publicaffairs, and the overall strengthening andefficiency of the public sector.

Decentralization has also generated a dramaticupsurge in expectations. Citizens look more tolocal governments for those public services thatimprove daily living conditions. Centralgovernments depend on local governments tosupport priority development and povertyreduction goals. Private firms increasingly rely

on local governments to deliver infrastructureand other services that support production andstimulate job creation.

One of the most critical factors underlyingthe ability of local governments to meetthe growing expectations placed on them isthe quality of the architecture andoperation of the intergovernmental fiscalsystem. This Second GOLD Report focuseson local government finance worldwide.Local government finance is important notonly because the role and impact of localgovernments have dramatically increased,but also because this progress has recentlybeen confronted by daunting challenges.

The global economic and financial crisis thatemerged in 2008 —the most significant crisissince the Great Depression— has imposedmajor financial constraints on localgovernments in many countries. Equallyimportant, central authorities in somecountries have responded to the crisis bytaking recentralization measures to deal withtheir own fiscal problems and increasingcontrol over local governments. It is too earlyto say whether these actions represent adurable change in the decentralization trend,but they clearly pose immediate challenges tothe viability and effectiveness of localgovernments. Resource constraints during aperiod of greater responsibility and need poseobvious threats, but so does the damagingcurtailment of local government autonomy,which is a necessary condition for the full rea-lization of the promise of decentralization.

Beyond the impact of the global financial crisis,local governments are confronted with othersignificant structural challenges. As substantialurbanization continues unabated in some partsof the world, public service demands aregrowing faster than many local governmentscan keep up with. New needs are also arising asa result of an emerging understanding of theonerous implications of global environmental


CCOONNCCLLUUSSIIOONNUnited Cities and Local Governments312

challenges, as well as from major demographicchanges, such as the increasing number of theelderly in some countries and the explosion ofyouth as a share of the population in others.These challenges, however, also presentopportunities to strengthen and boldly reinventthe role of local governments. They are often ina unique position of strength to deal withpressing local problems, the solutions to whichhave important national consequences.

The preceding regional chapters documentstrengths and weaknesses of local governmentfiscal frameworks in different parts of the worldand examine the capacity of local governmentsto mobilize resources and manageexpenditures. The chapters also assessintergovernmental relations and developmentsin governance, such as broader and deepercitizen participation in local planning and bud-geting. This chapter summarizes keychallenges and issues discussed in thepreceding regional chapters, and points topossible broad-based policy solutions thatcould both alleviate problems and weaknessesexperienced to date and help to improve overalllocal government performance.

The next section outlines basic contextualfactors that affect fiscal decentralizationworldwide. This is followed by a summary ofrecent influential trends, experiences, andpolicy issues. Building on the review of fiscaldecentralization parameters outlined in theintroduction and discussed in the regionalchapters, common and noteworthy localgovernment finance issues and challenges areconsidered. Finally, the chapter closes withrecommendations and concluding thoughts onthe way forward in local finance reform and thenext steps for UCLG.

The Context of Reform: Diversity,Politics and Change

The potential for local governments to serve asfull partners in managing public functions and

to contribute to local governance and improvedservice delivery remains a promising, but onlypartially fulfilled process in many countries. Tosome extent this should be expected, asdecentralization occurs under differentcircumstances, is subject to powerful politicalforces, and requires some minimum capacity tobe effective. Even in the most conduciveenvironments, decentralization is a highlydynamic process that demands ongoingadaptation to evolving economic, social andpolitical conditions.

Understanding Diversity

As highlighted in the introductory chapter,countries have been subject to differenthistorical influences, so they are buildingfrom diverse institutional and governancetraditions. This includes their experienceswith and inclinations towards decentra-lization, as well as their ability to absorbdecentralization reforms. The role of localgovernments in public finance variesconsiderably across regions (Figure 10.1 &Table 10.1), and there are also largedifferences within regions. An importantimplication of these various differences isthat desired local finance reforms varyconsiderably across regions and countries.Clearly, the reforms needed to strengthenlocal finances differ between countries thathave a long tradition with decentralizationand those with a shorter history of relevantexperience.

Local government finance is prospering inmuch of Europe, North America, and parts ofEast Asia and the Pacific (Korea, Japan,Australia, New Zealand). It remains at anearly stage in some regions, such as theMiddle East & Western Asia, where most localgovernments are deconcentrated units of thecentral government with limited autonomy.South Asia has a recent tradition ofdemocracy, but local governments facecontrols by higher level governments. In Latin


313Second Global Report on Decentralization and Local Democracy.GOLD 2010

Note: while localexpenditures as aproportion of publicexpenditure may berelatively high in East Asia,Eurasia, and South Asia thisdoes not necessarily meanthat local goverments haveautonomous control overspending, which is animportant aspect ofdecentralization(devolution).

Figure 10.1: The Comparative Fiscal Role of Local Government: Expenditure and Revenue as a Percentageof General Government Expenditure and Revenue

4020105

expenditurerevenue

North AmericaEurope Eurasia

East Asia

South East Asia

South Asia

MEWA

África

Latin America

Table 10.1: The Comparative Role of Local Government

Local expenditures as percentage Local expenditure as percentageof total revenues of total expenditures

Africa 3.2 7.9(3.6) (6.8)

AsiaSouth Asia 1.5 16.0

(0.9) (0.9)East Asia 20.0 40.0

(0.3) (0.3)South –East Asia 5.3 15.5

(0.8) (0.6)Eurasia N.A. 26.5

(15.1)Europe (2008) 13.0 23.9

(0.7) (0.5)Latin America 4.0 11.1

(4.5) (7.3)Middle East & Western Asia N.A. 4.6

(1.7)North America 17.8 26.8

(0.6)

Notes: Coefficient of variation in parentheses. The means include : Africa: Benin, Burkina Faso, Egypt, Ethiopia, Gabon, Ghana, Kenya, Malawi, Mali, Morocco,Mozambique, Niger, Nigeria, Senegal, South Africa, Tanzania, Togo, Tunisia, Uganda, Zambia, Zimbabwe. South Asia: Bangladesh, India, Pakistan; East Asia: China,Japan, Korea; South -East Asia: Indonesia, Philippines, Vietnam, Thailand. Eurasia: Armenia, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia, Ukraine.Europe: Austria, Belgium, Bulgaria, Cyprus, Czech Rep., Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania,Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Romania, Slovak Rep., Slovenia, Spain, Sweden, Switzerland, United Kingdom, Iceland, Norway,Switzerland ; Latin America: Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, El Salvador, Honduras, Jamaica, Mexico, Panama, Paraguay, Peru. Middle East andWestern Asia: Bahrain, Iran, Jordan, Lebanon, Palestine, Syrian Arab Republic, Turkey, United Arab Emirates, Palestine, Yemen. North America: Canada, U.S.A.

Source: Data from Table 10.1



America and Eurasia, local finances aregenerally improving, but still face challengesassociated with past centralized traditions.China and most of Southeast Asia have madeprogress, but intergovernmental fiscalrelations are unevenly developed and stillexperience significant challenges. Africanlocal governments are rarely wellempowered, but there are hopeful advances,especially in some Anglophone countries,such as South Africa or Tanzania.

One of the critical inferences emergingfrom the diversity of local governmentsystems is that there is a need for diverseapproaches to deal with fiscal challenges,even those that are relatively common.There is no magic formula to ensure thatlocal government systems will functioneffectively. The road to success requiresconsistent and appropriately sequencedapplication of basic local public financeprinciples outlined in the introductorychapter. These are relatively few, and theyleave adequate flexibility for each countryto structure its intergovernmental financesystem to fit its history and national goals.

Responding to Political Reality

Decentralization is an intensely political processsince it involves the central governmentassigning powers and granting autonomy tolocal governments. While political forces canoften open the door to decentralization, asdiscussed throughout this report, they can alsopose challenges. These include reluctance ofcentral politicians to devolve powers to localgovernments for fear of losing control, theopposition of central bureaucrats whoseinstitutional and personal goals conflict withdecentralization, or resistance to legallymandated decentralization reforms duringimplementation from elites and pre-existingdeconcentrated agencies. At the local level, localpoliticians can undermine decentralization ifthey are not sufficiently accountable to their

constituents. These political realities must beunderstood and responded to if decentralizationis to be effective and prosper.

Developing Capacity

Effective local governments require admin-istrative capacity. Local government capacitycan be an important constraint, particularly indeveloping countries. At times, perhapssomewhat paradoxically, decentralizationunderperforms because of weak centralinstitutions, either due to political instability or alack of control of basic functions of government,such as unified tax administration or treasuryand budget implementation controls. If decen-tralization is to meet its promise, capacityconstraints and their consequences must berecognized and efforts to develop appropriatecapacities need to be adopted.

Adapting to Change

Conditions and motivations for decentralizationchange, sometimes rapidly and dramatically.These changes can be relatively routine, suchas the adoption of new legislation or theturnover of a government power after anelection. They can also be momentous, such asa major political shift or a sudden economic andfinancial crisis. Local government policies andsystems need to respond effectively to thesechanges, and adapt as necessary to shiftingcircumstances. At the same time, the 2008global financial crisis showed that adapting toshifting circumstances can also damage localgovernments. Local governments and theiradvocates must be vigilant and be prepared todefend their legitimate interests when theycome under threat.

Broad Policy Trends and Issues

Before reviewing major findings on localgovernment finance, it is important tocontextualize those findings by noting somebroader trends and common decentralization


315

issues the report shows can affect local fiscalperformance. Some of these are experiencedglobally, while others are particularly relevantin certain regions or some subset of countriesacross regions.

Global Crisis

The financial and economic crisis noted above isaffecting local governments globally. Emergingcountries of Asia and some in Latin America havebeen less impacted, but others have suffereddrastic effects. In March 2010, for example, theGreek government reduced by decree thenumber of local governments from 1,034 to 355in order to save an estimated 1.2 billion eurosannually.

The pains of fiscal adjustment due to the crisisare being strongly felt by local governments in allthe continents. In a number of countries inAfrica, Eurasia, and Latin America, centralgovernments have cut transfers or introducedrecentralization measures. In some regions, theeffects of the crisis simply compound the effectsof existing challenges. In Africa, for example,trade liberalization and fiscal transition, and inless developed countries more generally, povertyand informality have long presented challengesfor public finance in general and localgovernments in particular.

Even in the most advanced countries,stabilization policies to reduce publicindebtedness, such as those being adopted inEurope and North America, are deeply impactinglocal finances. Local governments in many ofthose countries fear that a disproportionateshare of the costs of further fiscal consolidationwill fall on them in the form of cuts inintergovernmental transfers, restrictions on localcredit, and other austerity measures.

The financial and economic crisis is not the onlyglobal crisis with relevance for localgovernments. Financing climate changemitigation policies and the investments required

for the associated risk management wouldconsiderably increase the resource needs of localgovernments worldwide. The financialimplications for local governments of theresponse to environmental challenges are onlybeginning to be understood.

Partial or Interrupted DecentralizationReform

The global crisis provides one example of howdecentralization can be stalled or reversed, butthis is a more general problem taking differentforms as evidenced in the regional chapters.Fiscal decentralization frameworks involvecomplex systems with many interrelatedcomponents, and some are easier to implementor politically more feasible than others. Thus,some local finance systems are only partiallydesigned (relative to best practice principles)and some are only partially implemented even ifthey are mandated in the legal frameworkdesign.

If only certain elements of the system areimplemented or partially implemented,problems can arise because of theinterdependencies involved. Deficiencies withone component often undermine the ability ofthe overall system to function effectively. Forexample, lack of clarity with functionalassignment can lead to uncertainty regardingthe financial needs of local governments.Similarly, problems with the design andimplementation of intergovernmental transfersystems can compromise incentives andcapacities for local service delivery, localrevenue generation, and local borrowing.

Among the most pervasive and damaginginstances of incomplete decentralization is theassignment of too few revenues to financeassigned functions. At a global scale, very fewcountries escape dealing with major gapsbetween local expenditure and local revenues.This can result from a flaw in system design,but revenue inadequacy tends to occur for




political or capacity reasons even in countrieswhere constitutional or legal provisions prohibitunfunded local government mandates. Theproblem tends to be more significant in someregions. African countries, for example, generallyhave much less decentralization of revenuesthan of expenditures, leading to particularlysevere revenue-expenditure gaps.

Demographic Shifts

Many European countries are confronted withthe challenge of coping with the effects onpublic finances of a rapidly ageing populationand the need to integrate immigrants into thelabor market and society at large. The ageingpopulation challenge is also relevant in severalcountries in Asia, such as China and Japan, andin Eurasia, such as Russia and Ukraine. In somedeveloping countries, the growth of youth as ashare of population poses different types ofservice challenges that also have seriousfinancial implications.

Rapidly increasing urbanization, particularlyin many of the developing countries ofAfrica, Asia, and Latin America, continues tocreate demand for public services andinfrastructure that requires a huge volumeof resources. The needs are even larger ifinvestments for climate change adaptationare included. Given available information onmaintenance and development costs of urbaninvestments, it would seem reasonable toexpect a need for about 200 billion USD peryear over the next 25 years for thedeveloping countries alone1 ; this representsonly one-third of the total spending need fortotal public infrastructure spending estimatedby the World Bank.2

Jurisdictional Fragmentation

Fragmentation is a major issue for manycountries in most regions of the world. In manycountries the appropriate structure and size oflocal governments is an ongoing debate. Small

local governments cannot independently takeadvantage of economies of scale in the deliveryof some services, resulting in higher costs.Smaller local governments, however, generallyprovide a stronger political connection tocitizens and may be better able to respond tolocal demands. Getting the right balance, e.g.by maintaining smaller local governmentsbut providing mechanisms for cooperativearrangements among them and links to higherlevels for large scale services, is a criticalchallenge in many counties.

Thailand has more than 7,500 bottom tier localgovernments with an average population ofless than 10 000, and there are concerns thatthese are too small for service delivery. Insome cases, such as Uganda or DominicanRepublic, new local governments are beingconstantly created, diluting the capacity oflocal governments that were only recentlyempowered. In a number of countries perverseincentives, such as offering equal lump sumtransfers to all local governments regardless ofsize, create incentives for further jurisdictionalfragmentation.

On the other hand, in countries such as France(with 36,600 local governments), citizensstrongly identify with smaller local governments(communes). These are said to bringgreater representation and accountability, thuspotentially balancing the additional costsrepresented by the inability to realize economiesof scale, particularly if the latter can be realizedby creating cooperative arrangements amongthe smaller units. When local governments inSouth Africa were substantially consolidated in2002, some analysts expressed concernthat the new larger, more fiscally viable localgovernments had damaged political connectivityto citizens in some areas.

Deconcentration and Devolution

Devolution of spending and taxing powers toautonomous local governments is generally

1. UCLG. 2007. UCLGPolicy Paper on LocalFinance. UCLG.

2. World Bank (2005)estimated theinvestment needs inpublic infrastructurein developingcountries, amountingto 600 billion USDper year over thenext 25 years.However, thesefigures include allpublicinfrastructures,whether national(energy,communications andinformationtechnology,transport, water andsanitation, etc.) orurban (local roads,local water supply,and sanitation, wastedisposals, schools,street lighting...).(World Bank. 2005.Infrastructure andthe World Bank: Aprogress report. TheWorld Bank). TheUCLG Committee onLocal Financeestimated one thirdof this amount, i.e.0.4 percent of WorldGDP, needs to bechanneled to urbaninfrastructure (UCLG. 2007. UCLGPolicy Paper on LocalFinance. UCLG).


317

held up as the standard for decentralization,but even some countries with elected localcouncils maintain deconcentrated administration.In Kenya, for example, district administrationsexist in the same territory as elected countycouncils. There is little clarity with respect totheir distinct functions, sometimes resulting inservice redundancy or gaps (although thissituation should be corrected by forthcomingreforms based on the 2010 constitution).

In other cases, empowered local governmentshave not been created. In the MEWA region, forexample, deconcentrated local administrationprevails except in Turkey and Palestine. Similarsituations can be found in countries in otherregions, such as Egypt, Pakistan, Bangladesh,Thailand, and Kazakhstan. The use of localgovernments as deconcentrated units of thecentral administration leaves unexploitedefficiency gains in the delivery of public servicesthat are achievable with devolved systems bybetter matching the needs and preferences oflocal residents and making local officials moreaccountable to citizens.

Intermediate Governments in Federal and Hierarchical Systems

While a federal country is often associated withhigh fiscal decentralization, many federalconstitutions do not recognize directly the rightof local entities to self government. Instead,they empower states or other intermediategovernments to establish fiscal relationshipswith local governments. This approach has ledto considerable fiscal powers for local govern-ments in Brazil, Canada, South Africa, and theUnited States.

In other cases, such as Argentina, Australia,India, Mexico, Nigeria, Pakistan, andRussia, local governments enjoy (oftenconsiderably) less fiscal autonomy eventhan those in some unitary countries withmore centralized traditions. Depriving largepopulous countries like India or Pakistan (in

the latter local authorities were suspendedin 2009 by agreement between federal andprovincial authorities) of accountable localgovernments diminishes their chances forattaining the potential benefits ofdecentralization. Limited authority for localgovernments is also present in unitarysystems with strong hierarchical linksbetween intermediate and local tiers ofgovernments, such as China or Vietnam.

The Role of International DevelopmentAgencies

International development agencies oftencreate challenges for the very developingcountries in Africa, Asia, and Latin Americathat they are supposed to be assistingthrough support for decentralization and localgovernment reforms. There are various issuesin this regard, but two are particularlyimportant. First, these agencies have oftenpushed particular types of reforms, sometimesbased on particular objectives of the agenciesor simply what has worked in other countries.As a result, in some cases, the reforms beingpromoted have been inappropriate for recipientcountries or have proven unsustainablebecause there is no strong national ownership.

Second, the donors have commonly createdparallel mechanisms to implementprograms that support the financing anddelivery of local services because of con-cerns about low local administrative capa-city, corruption, and other institutionalweaknesses in the host country. Thesemechanisms can be based at higher levelsor at local levels, but in either case theybypass the regular decision-making andresource management procedures of localgovernments. They can improve servicedelivery and may be appropriate in someform at early stages of decentralizationwhen local governments are weak, but theyultimately undermine the legitimacy andeffectiveness of local governments unless




the systems and procedures they use areadapted and institutionalized into regulargovernment operations.

Local Government Finance: Main Issues and Challenges

As outlined in the introduction of the report,several key aspects of fiscal systems needto be in place and meet certain basicprinciples for local governments to performeffectively. These include expenditureassignments and management; local ownsource (autonomous) revenues; properlystructured intergovernmental transfers; and,where appropriate, access to borrowing andother alternatives to mobilize resources fordevelopment expenditures. This sectionoutlines key issues and challenges identified inthe regional chapters with respect to each ofthese issues.

The emphasis in this section is on identifyingproblems and challenges that require attention,but it is important to acknowledge, as noted inthe Introduction and some regional chapters,that there have been very significantimprovements in local finance over thepast decades in many developed and alsodeveloping countries. These improvementsrange from increased efficiency in publicexpenditures to greater revenue mobilization,and to innovations in public management, suchas the more general adoption of the type ofparticipatory budgeting that began in LatinAmerica.

Expenditure Assignment and Management

A clear assignment of responsibilities andexplicit methodologies to translate expenditureresponsibilities into financial needs arefundamental for local finance. Deficiencies onthis front weaken local governments andundermine the rest of the local fiscalframework. The challenges commonly fall undera number of categories outlined below.

Clarity in expenditure assignment: Insufficientclarity occurs in many regions, particularly indeveloping countries in Africa, Asia, Eurasia,and Latin America. This results from poorlydrafted laws and conflicts betweendecentralization laws and sectoral lawsregarding specific services. Sectoralresponsibilities may continue to beimplemented by line ministries withoutcoordination or in competition with localgovernments, duplicating efforts by keepingdeconcentrated offices and staff at pre-decentralization levels; this is a commonoccurrence, for example, in African and LatinAmerican countries. The ambiguity ofexpenditure assignments can be more severewhere there are more levels of government, asin China, and in federations where intermediatelevels have significant but inadequately definedcontrol over local governments under theirjurisdiction, such as in the case of India. Arelated institutional issue in some countries, suchas Australia and Argentina, is whether localgovernments should obtain separate legalstatus from their intermediate level governments,provinces, or states.

Suitability of and compliance with expendi-ture assignment: In some cases centralauthorities still play an unwarranted role inthe delivery of basic local services,sometimes contrary to decentralization law.This can result in levels and types of servicesthat differ from those desired by local people.In other cases, services with benefit spillovers(affecting people of jurisdictions beyonddirect beneficiaries) or a heavy focus onredistribution lack coordination of tasks withhigher levels of government; this can result ininsufficient or uneven provision of services.This happens, for example, in China, whichassigns responsibility for social security andpublic pensions to local governments.

Funding expenditure mandates: Lack of clarityin functional assignment creates room for costshifting among levels of government, often


319

through unfunded local expendituremandates that can be can be extremelyburdensome. These can involve requiringlocal governments to deliver specific services,use particular delivery approaches, or meetcertain input or output standards in servicedelivery. This is a common occurrenceamong developing and developed countries.Sometimes such mandates may involveservices that local governments are notrequired to provide under local governmentlegislation.

Budget approval and control by higher levelauthorities: The central or regional autho-rities assist with and closely oversee —andultimately may even approve— the budgetsvoted on by local elected councils in manycountries, particularly in Africa, Asia, Eurasia,Latin America, and MEWA regions. Thispractice of ex-ante control weakens thebudgetary autonomy of local authorities.

Incentives for local expenditure efficiency:Particularly in developing countries, localgovernment spending quality is often low interms of the outcomes produced relative tothe costs incurred. This is partly attributableto resource constraints and the often-excessive administrative shares of the localgovernment budget. But other factors notedabove (lack of clarity in functionalassignments, unfunded mandates, etc.)and below (conditional transfers and lowrevenue autonomy) also undermine localaccountability and incentives to use resourcesefficiently.

Local Revenue Generation/Autonomy

Local revenue generation and autonomy arecritical for local governments to be able tomeet their expenditure responsibilities in anaccountable and efficient way. Yet there arevery few countries in the world that so far haveprovided local governments with the meansand autonomy needed to raise adequate

revenues. This problem is manifested in variousways related to the design and use of localrevenue systems.

Vertical fiscal imbalances: The transfer of ex-penditure responsibilities to local govern-ments has often not been accompanied bydevolution of corresponding revenue sources(including intergovernmental transfers, whichare discussed below). As noted above, localgovernment revenues in many regions play aminor role in national public budgets. This hasresulted in increasing financial pressures onthe local government expenditure, and evenwhere resources are more adequate, greaterlocal government dependence on centraltransfers.

Revenue autonomy: Autonomy is highlyconstrained in most of Africa, Asia, and theMiddle East and West Asia. The situation isbetter in Eurasia and Latin America, but notuniformly. Local governments have limited orno authority to introduce new taxes, and todecide on some or all tax rates, fees, anduser charges. Even some decentralizedcountries, such as Australia, limit local revenueautonomy. A number of prominent attempts toenhance tax autonomy and reduce transferdependence, such as recent “Trinity Reforms”in Japan, have only partially succeeded.Revenue autonomy is stronger, but not withoutchallenges, in advanced economies of WesternEurope and North America.

Property taxation challenges: The property taxis the most commonly recommended andglobally used local government tax, but itssignificant revenue potential often remainsunrealized. On average developing countriesraise 0.5 percent of GDP from property taxcompared to two percent in developedcountries. This is partly because the tax isunpopular— even in some developed countrieswhere it plays a significant role (U.S., Canada,U.K.), citizen opposition has been strong.In addition, it is difficult and expensive to




administer, all the more so in many developingcountries without well defined propertyregisters, with sizable informal areas, and withweaker local capacity for value assessments,enforcement, and collection.

Diversification of the local tax base: Local taxbases are often narrow, especially given theproblems with heavy reliance on the propertytax. A number of countries in Europe andNorth America use local personal incometaxes. A local piggyback, flat-rate personalincome tax is collected with the nationalincome tax in Nordic countries and sometransition economies of Central and EasternEurope. In Latin America, several countries,such as Brazil, Chile, and Colombia usevarious types of local business taxation. Localsales taxes are used in a few countries,notably in Canada, with the presence of anational VAT, and in the United States, wherethere is no VAT. Poor diversification of thelocal tax base is often aggravated by the lackof flexibility to adapt to evolving circumstan-ces (for example, growth in the servicesector). Inelasticity (lack of revenue responseto changes in the economic base) of manylocal taxes over time is problematic asprogressively increasing demand for servicesand costs outstrip revenue growth. In anumber of African countries (Tanzania,Uganda, and Zambia) some viable local taxeshave been recently eliminated and partiallyreplaced with transfers, and many countries,prominently Korea, suffer from a proliferationof “nuisance taxes” that yield low revenuesrelative to collection costs.

Fees and user charges: Local governmentsneed to establish fees for services, ideally on acost-recovery basis where this is feasible. InCanada and the U.S., local governmentsgenerate one-quarter of their own revenueswith fees and charges, which is all the moresignificant given their broad high levels of localown tax revenue. The situation is very differentin many developing countries. In some African

countries, such as Algeria, Benin, Egypt andTunisia, local governments have no authority toset local fees and charges.

Politics of local revenues: Political barriers tolocal revenue generation can be seen in boththe reluctance of local government to raisetaxes (for instance, in some EU countries) aswell as in the limitations imposed on localrevenue generation legislated by higher levelsof government or citizen referendums (in manystates in the U.S.). To some extent these phe-nomena result from poor taxpayer educationand general expectations by citizens for moreand better quality services with the same orlower taxes.

Local and central roles in revenue collection:International practice varies as countriesseek to maximize revenues while minimizingadministration and compliance costs(which favor a role for higher levels inadministration and enforcement) andmaximizing local accountability and localinformation advantages (which favor localgovernments’ direct involvement inadministration and enforcement). Althoughcentralized mechanisms are in principledesirable for certain taxes, central agenciesdo not in some regions, including MEWA andWest Africa, transfer the resources theycollect to local governments in a timelymanner. The lack of incentives for central taxauthorities to collect local revenues can alsobe a problem. The experience of a variety ofcountries (Costa Rica, Jordan, and somecountries in Eurasia) shows significant increasesin revenue collections when tax administrationresponsibilities are transferred from central tolocal authorities.


A properly structured system of intergovern-mental transfers is a critical component of alocal finance system. The use of transfers,however, faces a number of challenges that are


321

dealt with in different ways and to varyingdegrees across regions and countries.

Appropriate and stable revenue sharing: Mostcountries share some central taxes with localgovernments, an arrangement that is simpleand has high revenue potential. This can be apartial solution to vertical imbalances, butshared revenues suffer from variousconstraints. Revenue sharing on a derivationbasis can be seen as a stimulus for localeconomic activity, but it can also reinforcehorizontal disparities and leads to highervolatility of local revenues. Particularly indeveloping countries, the amounts shared maybe uncertain or lack transparency, making longterm planning difficult for local governments.This is the case in a number of West Africancountries, where central governments withholdfor their own purposes or delay resources towhich local governments are entitled. Perhapsmost importantly, substantial revenue sharingcan create perverse incentives for local revenuegeneration, undermining both local autonomyand the accountability of local governments totheir constituents.

Horizontal fiscal imbalances: Despite the oftensignificant differences across local governmentsin expenditure needs and the ability to financethem, many countries lack effective equalizationgrants. In Africa, just a few countries (Moroccoand South Africa) have introduced them, and inMEWA there are none. The situation is a littlebetter in Latin America, where a few countries(e.g. Brazil and Chile) use explicit equalizationschemes, although more countries in the regionemploy only limited redistribution elements inrevenue sharing schemes. Some Asian countriesuse equalization transfers (e.g. Australia,Indonesia, Japan), while others virtually ignorefiscal disparities (e.g. China, India, Philippines,New Zealand). Equalization grants are commonin Eurasia, Europe, and North America (except atthe federal level in the United States), but withvarying effectiveness. Some Eurasian countrieshave not used transparent methodologies for

equalization transfers, although the situation isimproving.

Equalization transfer design: Where equalizationschemes exist, they often present problems; forexample, (1) only differences in fiscal capacity orexpenditure needs, instead of both, areconsidered; (2) actual revenues, instead of fiscalcapacity, are measured, creating disincentivesfor local revenue mobilization; (3) the pool offunds are not stable or well defined, or (4) theuse of funds is subject to rigid conditions that ineffect make the equalization grants, which arenormally general purpose grants without userestrictions, into conditional transfers. Infederal countries, such as Australia, there areissues regarding how second tier governments(the states) interpret federal policies regardingequalization.

Conditional transfer design: Conditional grantsfrom upper level governments are a keyelement of local fiscal frameworks. Suchgrants can (1) promote national standardsand goals in the provision of importantservices that have been decentralized, forexample, some aspects of education andhealth; (2) address inter-jurisdictional ex-ternalities with respect to, for example,environmental concerns; or (3) support localgovernment infrastructure investments.Conditional or earmarked grants exist in manycountries, especially for capital infrastructurepurposes. However, in certain regions, such asEurasia, conditional grants are not welldeveloped. In other countries, such as Egypt,Nigeria, Tanzania, and Uganda, conditionaltransfers excessively dominate total transfers.Several problems are often associated with thistype of grant, including their number and com-plexity, which impose high compliance costs onlocal governments; lack of transparency,stability or timeliness; and sometimes politicalmanipulation. In addition, excessive reliance onconditional grants can overly constrain localgovernment autonomy and move their focusfrom local to national priorities, potentially




reducing their own comparative advantage infocusing on local needs.

Performance based grants. A relatively recentinnovation in the field of transfers has been theintroduction of performance based transfers insome African and Asian countries with supportfrom international organizations. Performancebased grants condition the transfer of funds tomeeting certain standards and objectives,generally leaving local governments to decidehow best to use the funds. This type of transfercombines the flexibility of unconditional grantswith an unconventional form of conditionality.On the downside, these transfers may privilegejurisdictions with greater administrativecapacity, and they may suffer from theproblems associated with voucher programs.Thus far they have been used more to promotecompliance with financial and administrativemanagement procedures than to improveservice delivery outcomes. It is too early todefinitively judge the effectiveness ofperformance based transfers but they are apromising mechanism and further trials arecertainly desirable.

Local Government Borrowing and Accessto Financial Markets

Perhaps the most neglected aspect of localgovernment finance in many regions of the worldis borrowing. In the context of the rapidurbanization discussed earlier, especiallydeveloping countries in Africa and Asia, the needfor infrastructure investment is paramount. Inthis context, borrowing, with theintergenerational equity that it entails, ispotentially an important means to finance longerterm investments. At the same time, there aremultiple factors that need to be considered.

Local government borrowing and fiscal respon-sibility frameworks. These frameworks areoften weakly developed and poorlyimplemented. Some frameworks are highlyrestrictive, effectively precluding local

government borrowing (e.g. Denmark,Chile, Kenya and Tunisia). Otherframeworks are too lax, potentially allowingfor the type of risky behavior that occurredin the 1990s in Brazil and Argentina. In afew case such as South Africa, robustframeworks that promote responsibleborrowing have been developed.

Access to credit. In many cases, especially inpoor developing countries, local governmentsoften have poor and unreliable access to credit.Financial markets are not well developed, andmany local governments do not have credithistories or do not meet technical standardsrequired by lenders. The situation is brighter inthe short and medium term in emergingeconomies where financial markets tend to bemore developed with the introduction ofsystems for disclosure, credit ratings, pricingbenchmarks, and so forth.

Special institutions. Special credit institutionsthat have been set up to lend to localgovernments (as is the case in more than 60countries, often with support from internationalorganizations in developing countries) haverarely performed well. Their often disappointingresults have been associated with thepoliticization of lending decisions andproblematic design issues. Many of theintermediary institutions are not sufficientlyindependent from the government, and theyare not allowed or have not attempted to linkwith domestic credit markets. In this regard,local governments are not supported inlearning how to become familiar with anddevelop capacity to comply with marketexpectations regarding financial capacities,disclosures, provisioning, and so on.

Central government practices. A number ofcentral government practices, such as weakappraisal mechanisms for loans fromgovernment affiliated credit institutions, localgovernment bailouts and automatic intercepts,have disrupted the normal development of local


323

credit markets. There has been a pervasiveproblem with approval by governmentassociated lending mechanisms of inadequatelyvetted loans for non-viable projects. Thepractice of bailouts when local governmentscannot or will not repay their loans underminestheir fiscal discipline and distorts the creditmarket. Although reliance on automaticintercepts from transfers are generallyassociated with better repayment to specialcredit institutions and can help to developaccess to credit, maintaining them for longperiods without encouraging local governmentgraduation to more market oriented sourcescan create poor incentives for local govern-ments to properly consider and lenders toproperly appraise local government projects.

Links to the broader intergovernmental fiscalsystem. Other aspects of local governmentfinance covered above are sometimes notconducive to borrowing. Borrowing can becurtailed if local governments have insufficientaccess to discretionary sources of revenue tomake loan payments or if intergovernmentaltransfers undermine incentives for evenrelatively wealthy local governments to borrow,for self-financing development projects.Lack of appropriate financial managementpractices also undermines the ability of localgovernments to properly prepare developmentprojects, qualify for credit, and manage theirdebt portfolios.

Recommendations

The findings of GOLD II clearly indicate thatlocal governments around the world–from the most industrialized to the leastdeveloped countries– suffer from problemsand challenges in their local governmentfinance systems, and in some respects thesituation has stagnated or worsened in recentyears. In Africa local governments representwell under 10 percent of public expendituresand less of pubic revenue. MEWA countries alsohave limited resources and even more limited

autonomy. In many countries in Latin America,Asia, Eurasia, and even in Europe, local govern-ments lack legitimacy because they cannotmeet important responsibilities with availableresources. Although some needed actions willbe difficult to quickly implement, there is muchthat can be done.

Expenditure Assignment and Management

A clear assignment of expenditure respon-sibilities should be at the top of nationalreform agendas for local governmentfinance. There are some important politicaleconomy issues noted throughout thereport that often make this step difficult.Several basic measures need to be followedfor this foundational reform that will insome cases require a revision of thelegal framework and harmonization ofdecentralization and sector laws.

Identify the exclusive responsibilities oflocal governments is needed to increase theclarity required for accountability. In caseswhere there is legal clarity and theassignments have not been implemented,action is needed to enforce the provisions ofthe legal framework. In cases where it isdeemed necessary to have concurrentresponsibilities for particular services, it isimportant to identify which level has specificresponsibilities for various aspects — i. e.regulation, financing, and implementation.

Limit higher controls on local expenditures. Inthe EU, for example, the Commission shouldnot excessively control or interfere with localservice delivery. In multi-tier systems the roleplayed by intermediate tiers (states andprovinces) in controlling local expendituresshould be appropriate and restrained. Thereshould be limited infringement on localautonomy, and with specifically local services,intermediate levels should not be interfering. Indeveloping environments where there aresignificant differences in administrative




capacity across local governments, asymmetricassignment of responsibilities may bejustified, at least temporarily. Over time localgovernments can graduate —with appropriateincentives and support— to more complete levelsof responsibility as their capacity is developed.

Determine financial requirements. For a clearassignment of expenditure responsibilitiesto become useful for other aspects of thelocal fiscal framework, they must betranslated into expenditure needs/financingrequirements through application of anappropriate standardized methodology. Asystematic evaluation of the cost of transferredresponsibilities should precede the transfers oftask and resources.

Fund all mandates. It is important to be explicitthat the level of government with the power toregulate a function also has the obligation to payfor it. Increasing coordination and dialogueamong levels of government regarding functionalassignment is also needed, and there should beex-ante review of all government legislationregarding local governments to detect anyunfunded mandates.

Ensure that human resources follow functions.Funding/staffing of deconcentrated offices ofline ministries should be downscaled oreliminated. This will reflect the functionstransferred to local governments and ensurethat they have the staff to execute them, whileat the same time reducing the existence of staffat other levels who might interfere with localgovernment functions.

Reduce and progressively eliminate ex antecontrol of local government budgets. In somedeveloping environments this may not bepossible to do quickly, but as the local financesystem matures it is important to shift froman emphasis on ex ante control to an emphasison ex post control, such as audits, and to agreater focus on developing accountability tothe citizens.

Implement expenditure decentralization stra-tegically. It may be appropriate to use the typeof asymmetry noted above, and bothperformance incentives and capacity buildingmay be needed. Capacity building and technicalassistance should support local governments toestablish a foundation in the first stages ofdecentralization and then help them adapt toperformance incentives in later stages.

Local Revenue Generation and Autonomy

Autonomous local revenue generation is themost serious fiscal challenge faced in a majorityof countries globally. Althought the exact set ofrevenue reforms will vary across countries, thistype of reform is to some extent needed inmost countries.

Increase reliance on own revenues withmeaningful discretion. This strengthens thelink between benefits and costs of localservices, making local officials moreaccountable to taxpayers and more fiscallyresponsible. This can be done throughreforms to existing sources of revenue and/orthe addition of new sources, and appropiatesystems and capacity must be developed inconjunction with expanded revenue authority.

Reform and modernize property taxadministration. Clearly the poor revenueperformance of the property tax has a heavyadministrative component. But there arepolitical limits to using this source, so thenature and extent of reforms must be decidedon a case-by-case basis.

Diversify the local tax base. This is needed inmany countries but reforms should targetviable and productive local revenues (i.e. notnuisance taxes) as well as ensure that localeconomic activity will not be impeded.

Only a limited number of local taxes meet thesecriteria, including vehicle taxes, businesslicense taxes, piggyback income taxes and


325

betterment levies on real estate. Vehicle taxescan be based on registration, licensing, parkingand similar bases. Business taxes can takedifferent forms, but typically use sales turnoveras a proxy for the tax base; care must be takennot to convert them into sales taxes thatconflict with other consumption taxes,particularly national VAT. Going further in thedirection of increasing local tax autonomywould be the introduction of a local piggy-backpersonal income tax with a flat rate collected atthe same time as the national income tax iscollected. Betterment levies are an importantmeans for financing infrastructure investmentsin countries where they are used. All of thesesources can be extremely productive, but theyare most relevant for urban and intermediate-level (states, provinces) governments.

In addition to these more traditional resources,a potentially valuable but relatively unexploitedsource in most regions is environmental or“green” taxation related to waste management,water and air polluting activities, and theproduction of energy. Green taxes wouldprovide a so-called “double dividend” since theypromote both revenue generation and a cleanerenvironment. There are also opportunities todevelop sources of revenue based on theincreasingly important knowledge economy.

There is often an opportunity to adapt the fiscalsystem to include some taxation on activitiesfrom the informal sector, particularly in de-veloping countries.

Increase freedom to raise fees and user charges.There are economic, technical, and politicalchallenges and limitations associated with suchrevenues, but they could be more extensivelyused in most countries. Better and more explicitpricing for public services may help to improveefficiency if political obstacles to charges can beovercome. The principle of cost recovery onpublic services should be promoted wherefeasible, but in a way that does not undermineaccess to basic services by the poor.

Carefully organize local tax collectionresponsibilities. The challenges of gettingthe right arrangements between central andlocal governments, as noted above, areconsiderable. With local collection, robustsystems and incentives are needed for thepotential benefits to be realized. Whencentralized administration of local taxes isappropriate, it is important to establishthe right incentives for central taxadministrations. Extensive dialogue andcooperation between different levels of taxadministration is always desirable and shouldbe institutionalized. This includes informationsharing on collections with local governmentsand allowing their participation in someaspects of management.

Engage local government officials more fullyin mobilizing local resources, linking them toservice delivery, and using them moretransparently. Local officials must assumetheir responsibility to mobilize the localresources required to improve local serviceprovision. The tax morale of local residentsand their willingness to contribute to the localfunds can be improved through campaigns offiscal awareness that inform citizens abouthow resources are used and how decisionsare made. Local officials should also ensurethe transparent management of funds andencourage citizen participation in order toincrease their confidence on the budgetprocess.


Considerable challenges and weaknesses inintergovernmental transfer systems wereoutlined above. Multiple steps could commonlybe taken to improve the structure and operationof intergovernmental transfer systems.

Assure predicable, regular, and transparenttransfer mechanisms. A legal framework shouldestablish a minimum level of public resourcesthat the central government must transfer every




year to local authorities and offer sufficientassurance that they will be allocated in a clearand fair manner.

Secure an appropriate balance among thevarious types of transfers. There is no hard andfast rule about derivation based versus formulaallocated tax sharing, although the former mayworsen fiscal disparities, reinforcing the needfor equalization (see below). Similarly, thereis no normatively ideal balance betweenunconditional and conditional transfers. Asignificant share of unconditional funds hoveverreinforces local government autonomy andaccountability and it is the better option tosupport local autonomy and locally drivendevelopment when local governments haveacquired minimum capacities.

Expand and improve the use of equalizationtransfers. Countries that do not use them shouldconsider doing so to offset the differentialabilities of local governments to meet basicservice needs. Countries that do use themshould take stock of their approach and movetowards a system that uses an explicit and stablerule to determine the pool of funds; takesexpenditure needs and revenue capacity (asopposed to actual expenditures or revenues)into account when allocating funds; and allowsunconditional use of transferred funds. Incountries where elements of equalization areimbedded in revenue sharing, as is common inLatin America, it would be desirable, followingthe rule of using a single instrument for eachobjective, to unbundle those schemes andseparately introduce an explicit equalizationtransfer with the properties listed above.

Review and improve mechanisms used forallocating resources under conditional grants.Beyond the basic guidelines on equalizationgrants noted above, best practice forconditional grant systems calls forsimplification, moving toward using fewerseparate block grants with clear sectoralobjectives and providing governments with

sufficient flexibility for deciding on the bestuse of the funds while meeting the broadersectoral objectives defined by the upper levelauthorities.

Consolidating grants where large numbers ofpoorly coordinated programs exist. In somecountries in Europe and Asia, for example,there are too many grants that are not clearlydistinguished and the resources could bemore productively used in a more consolidatedsystem.

Local Government Borrowing andInvestment Finance

In many countries, there are considerableopportunities for increasing the use of borrowingand other investment finance mechanisms aswell as expanding and improving sources offunding for this purpose. A number of specificpolicies and reforms can often support this goal.

Promote local government borrowing.Borrowing is one of the necessary pillars of localfinance. Responsible local borrowing, guided byprudent rules and regulations (a fiscalresponsibility framework) should be allowedwhere feasible, in parallel to the strengtheningof local capacities.

Develop and strengthen legal and regulatoryframeworks for local government borrowing.Rules regarding debt level and debt serviceratios need not be overly restrictive, butcentral authorities need to enforce hardbudget constraints and avoid bailouts. Centralmonitoring of local borrowing is critical,especially where private market disciplineis not operational. Such monitoring shouldcover not only regular debt but also “floatingdebt” or budgetary arrears with officialinstitutions and private suppliers, and localgovernment guarantees for municipalenterprises. Monitoring should be complementedwith a credible system of penalties for lack ofcompliance.


327

Expand and improve options and support me-chanisms for local government borrowing,including support where appropriate tointermediate financial institutions or municipaldevelopment funds. Beyond the regulation andmonitoring, an even more important challengefor most developing countries is to facilitate asignificant increase in credit availability to localgovernments for responsible borrowing, es-pecially for smaller municipalities. The solutionmay be the creation of official financialintermediaries or municipal lending institutions,such as Municipal Development Banks orFunds. International experience, however,suggests that they must focus on lendingoperations rather than get involved in othermatters (such as technical assistance to localgovernments), should be operated followingstrict banking criteria (including projectappraisal), and should increase the share ofprivate capital in their pool of resources overtime. Policies to encourage the development ofprivate markets for local credit are equallyimportant. The exact mix of these activities willdepend on the context of a particular countryfollowing the general rule to use the market tothe extent feasible and to use public or mixedlending mechanisms in a way that prepareslocal governments for eventual commercialborrowing.

Reform other aspects of the local finance systemas necessary to enhance the prospects for localgovernment borrowing. Local governmentsmust have access to and effectively use existing(and as needed additional) local taxes, usercharges, and central government grantsearmarked to local infrastructure. In addition, itis necessary to have good financial managementpractices in place.

Consider other investment financing mechanismswhere feasible. Tax increment financing,betterment levies (valorization), and publicprivate partnerships can also provide necessaryinvestment finance for local governments. Thesemechanisms, however, also require certain

capacities and conditions and should not be seenas an easy alternative to borrowing.

Determine an appropriate role in infrastructurefinance for International Financial Institutionsand other development agencies. Theseinstitutions have long played an important rolein developing and some transition countries,and in many cases they will continue to do sofor the foreseeable future. Such resources havetraditionally flowed to central governments withonlending to local governments. Suchonlending should comply with the basicprinciples outlined above, and there should bean increasing role for direct sub-sovereignlending, especially to larger cities in countrieswhere this is feasible.

Framing Institutional Reform

The finance system reforms outlined above willneed to be reinforced by other measures of amore institutional nature, most of which werediscussed earlier in this chapter to set the stagefor the discussion of fiscal decentralization. Anumber of key institutional issues often impactlocal finances and merit consideration.

Assess and respond as necessary to localgovernment jurisdictional fragmentation.Fragmentation is neither inherently desirablenor undesirable, but as discussed above it cancreate problems. There are two types of issues.

The first is ensuring that any creation of newjurisdictions is done according to clear criteria toprevent the proliferation of non-viable entities.In some cases there are perverse incentives(e.g. in the transfer system) to create newjurisdictions. These should be avoided.

The second is coping with existing frag-mentation that is deemed to be problematic.Where politically feasible, consolidation ofsmall, non-viable units may be considered, butthis can undermine political connection andlocal accountability. An alternative policy is to




enable the creation of voluntary municipalpartnerships to deliver public services requiringa minimum scale. Such associations andagreements can also help to address benefitspillovers across local government or theexporting of the costs of local services toneighboring jurisdictions by, for example,through agreements that provide for sharingservice provision costs in accordance withbenefits. Other solutions include voluntaryjurisdictional consolidation, the creation ofspecial districts to take advantage of economiesof scale in selected services, or jointlycontracting with private firms.

Identify the right roles for and interactionsbetween deconcentrated and devolvedgovernment entities. In cases where bothdeconcentrated and devolved entities coexistside by side, it should be made clearwhat functions each is responsible for, andthey should respectively be provided withappropriate staff, funding, and capacity to meettheir obligations. In countries where there hasbeen heavy reliance on deconcentration alone,consideration could be given to introducingdemocratically elected local governmentswith devolved autonomy to prioritize theirbudgets in accordance with the expressedneeds of local residents. It is important tonote that there can be room for bothdeconcentrated and devolved levels in somecases, but the system must be set up to tapthe advantages of each and prevent one type—usually deconcentrated administration—from undermining the other.

Assess the appropriate role for and operations ofexternal development assistance agencies andfinancial institutions in developing countries. Asdiscussed earlier, there are two broad types ofproblems —the heavy handedness of externalagents in promoting certain types ofdecentralization reforms, and their tendency tocreate parallel institutions and mechanisms forimplementing their programs that at leastpartially bypass normal decision making and

resource allocation procedures of local govern-ments. The latter measure is generally intendedto compensate for real and perceived problems,such as weak local government capacity,corruption, and ineffective and bureaucraticcentral government agencies. Parallelmechanisms can help to deliver services and maybe appropriate in some form at early stages ofdecentralization when local governments arevery weak, but ultimately they undermine thelegitimacy and effectiveness of local govern-ments. Neither of these donor approaches isconsistent with current thinking on aideffectiveness, as reflected in the Paris Declarationon Aid Effectiveness (2005), Accra Agendafor Action (2008) and the upcoming SeoulHigh Level Forum on Aid Effectiveness (2011).The underlying philosophy highlights theharmonization of development assistance withnational policy and stresses the importance ofusing national systems to deliver services,thereby reinforcing both national andsubnational governments’ capacity developmentand their accountability to citizens.

Ensuring that external development partnersfollow national policies is ideally the role of thenational government. In countries with weakcapacity and significant need for assistance,however, this may be difficult. Under suchcircumstances, the development partnersthemselves need to take steps to ensure thatthey align with national priorities.

Ideally parallel institutions should not be used.If it is necessary to use them for reasons notedabove, they must be framed as temporaryarrangements with a clear plan for phasingthem out in favor of greater reliance on localmechanisms as they become institutionalized.When local mechanisms are sufficientlycredible, external agencies should foreseebudget support that empowers local decisionmaking.

International agencies need to ensure thatbudget support programs contribute to the


329

strengthening of local governments and thedevelopment of their autonomy. Likewise,sector-wide approaches are often a centralizingforce in practice, but they can be instrumentalin strengthening and implementing the specificlocal powers and responsibilities as defined inthe legal frameworks for decentralization.

Create a regular and systematic dialoguebetween local governments and the centralgovernment on intergovernmental and localfinancial policy. Although this has not beenpreviously discussed in an extensive way,this report clearly leads to the conclusionthat local governments in many countries arenot sufficiently consulted on nationalpolicies of great consequence for them.Local governments could be consulted annuallyduring the national budget process on allquestions that directly or indirectly affect theirfinancing. This would require a mechanismcreated to bring together the national actors(legislature and executive) and local govern-ments. For such an approach to be effective, itwould be important to ensure access toappropriate information on public finances,both in general and specifically regarding localgovernment matters.

The Way Forward

Local governments have become moreimportant and more autonomous in manycountries around the world and higherexpectations have been placed on them.Because this has happened in a challengingglobal environment of substantial urbanization,demographic shifts, climate change, and in-creasing risk, more attention needs to be givento developing the basic fiscal architecturethat serves as a foundation for goodlocal government performance. As highlightedthroughout the report, there has been goodprogress on many fronts in many countries, butthere are still major deficiencies and challengesin most cases, both in terms of the elements ofthe fiscal system that need to be in place and

the capacity of local governments to functioneffectively. Unless these are confronted headon, there are great dangers of social andeconomic decline in the more advancedeconomies and a failure to meet key in-creasingly urgent needs in developing coun-tries, including poverty reduction targets andthe Millennium Development Goals.

Although diversity is great across countries,there are some shared challenges common tomany places. Clarity of functional assignmentin law and practice is a challenge in manydeveloping countries, and unfunded mandatesare a more general problem. In many countriesthere is a pressing need to reassess thestructure of local taxes, and the degree ofautonomy that local governments have indefining and using them. In many cases it will bedesirable to move beyond traditional localrevenue bases, and to search for a moreappropriate distribution of transferred and own-source resources between national and localgovernments, as well as among subnationalgovernments in the context of the emergence ofnew tiers and new units at particular levels.Growing investment requirements necessitatean expansion of local government access tocapital, increasingly through market-oriented annon-traditional mechanisms. There is also aneed for developing more innovative approachesto raising resources and delivering services,including through new and expanded forms ofpartnership with different actors (private sectorand civil society).

As countries around the world strive to improvetheir local government systems, they will haveto keep in mind some daunting short-term andlonger term challenges. The most immediatechallenge is the global financial and economiccrisis that started near the end of 2008, whichhas resulted in revenue shortfalls for manylocal governments and even attempts torecentralize in some cases. Countries also facelonger term challenges that cut across all levelsof government, some of which can have




particularly important implications for localgovernments because of the increasing role thisgovernment level plays in the provision of socialservices, environmental control, and so on.Some of these challenges are common (globalwarming, energy crisis, etc.) but others differ byregion of the world. Rapidly increasingurbanization, for example, particularly in manyof the developing countries of Africa, Asia, andLatin America, is creating complex demands onpublic services and infrastructure, yet localgovernments in many countries in these regionsdo not have the necessary authority andautonomy to meet these demands. In additionthey too frequently cannot even cover theoperating costs of existing services muchless the costs of the substantial additionalinvestments needed.

Although many suggestions to improve localgovernment finance systems have been madein the regional chapters and in this concludingchapter, in closing this volume it is important toreiterate again a few fundamental pointsregarding the approach to reform.

First, each country is unique and the basicprinciples for reform need to be tailored to theeconomic, political, fiscal, and social realities ofindividual countries. In Europe, for example,substantial capacity exists, but there is a needfor system reforms and increased access toinvestment finance. At the other end of thespectrum, less developed countries in severalregions need to build basic institutionsgradually if reforms are to take root and besustained, although more capacity may exist inlarger cities for more immediate assumption offunctions and resources.

Second, consultation and collaboration amonglevels of government and other actors will becritical as efforts to strengthen local financesystems advance —each actor has animportant role, but no actor alone can do whatneeds to be done. In particular, centralgovernments need to treat local governments

as partners, with full consultation in all issuesof shared responsibilities. Local governmentsalso need to continue the efforts they arealready pursuing in many countries to reachout to citizens, to develop partnerships withnon-governmental organizations and privatefirms, and to seek innovative means to dealwith the challenges they face.

Third, while political factors are critical and thereis no point in pursuing reforms that arepolitically infeasible, it is also important tomake decisions about reform based ongood information and evidence, the lack ofwhich created considerable challenges for thepreparation of this report. Better informationand analysis and broader and more transparentdissemination of such inputs can create andnurture a better environment for pursuing theright reforms over time. In addition, the successof initial modest reforms can create political mo-mentum for the adoption of more advancedreforms with greater impact over time.

Finally, there is considerable value added fromregional and global cooperation, sharingexperiences, and learning by doing in pursuinglocal finance reform. The role of UCLG, itsregional member organizations, and theirindividual country members, provides a strongfoundation for collaborative learning at variouslevels, and these actors need to continue tostrengthen those links going forward.Global and regional events, online access toinformation, country specific, regional andglobal networking activities, diagnostics to helpcountries and local governments to planconcrete productive action, and forums andmechanisms for sharing experiences andexpertise would all be productive ways tosupport better local government finance.Some opportunities in these areas alreadyexist, but much remains to be done toconsolidate and improve knowledge aboutthem, enhance access to them, and deepenan understanding among all stakeholders ofhow to effectively use them.


ANNEX

0w2010 11 ANNEXOS DEFcarta ang 15/9/11 23:04 Página 331

Second Global Report on Decentralization and Local Democracy. GOLD 2010

333

Legality/reality indicators Budgetary autonomy indicator Personnel autonomy indicators

Written into Difference Restrictions Responsibility in hiring Responsibility in fixing salariesthe Constitution between imposed by local managers

Constitution higher-level and Reality (1) authorities (2)

Algeria Yes No Lines State/Region/LG State/Region/LG

Benin Yes Yes Structure Local authorities Condition

Burkina Faso Yes Yes Structure State/ LG State/ LG

Côte d’Ivoire Yes No Line/structure State/ LG State/ LG

Egypt Yes No Line/structure State State

Ethiopia Yes No No Local authorities Regions

Kenya Yes No Structure State Local authorities

Ghana Yes Yes Line/structure State/ LG State/ LG

Malawi Yes Yes Lines State/LG State

Mali Yes Yes Line/structure State/ LG State/ LG

Mauritania Yes Yes Line/structure State/ LG State/ LG

Morocco Yes Yes Line/structure Local authorities State

Mozambique Yes Yes No State State

Niger No Yes Line/structure State/ LG State/ LG

Rwanda Yes No Line/structure LG State

Senegal Yes Yes Line/structure State/ LG State/ LG

South Africa Yes No Lines LG LG

Tanzania Yes Yes Lines LG LG

Togo Yes Yes Line/structure State/ LG State/ LG

Tunisia No Yes Line/structure State State

Uganda Yes Structure Local authorities State

Source: Compilation by authors, from data collected at national level by UCLG and from other sources (see bibliography).

(1) “yes” means that the reality differs from that of the Constitution

(2) The supervisory authority exercises control on the budget "Lines" or on the overall budget "structure" or both

Annex 2.1: Three indicators of expenditure jurisdiction for 21 African countries, 2009

2. AFRICA


AANNNNEEXXUnited Cities and Local Governments334

Annex 2.2: Some characteristics of transfers to municipalities for 20 African countries, 2008

Country Number of Number of Number of Brief observations about transfer systems transfer conditional grants/ transfers with systems Number of ad hoc financing /

unconditional number of grants transfers financed

by % of taxes

Algeria 2 2/0 2/0 One of the funds does not have a formula, the other is linked to fiscal balance (revenues)

Benin 5 4/1 5/0 The Fund of support to the development of local governments (FADeC) was created in 2009. Two funds use the population criteria. One is a fiscal balance fund and the other replacement a former tax

Burkina Faso 8 7/1 2/0 General endowment fund and general capital expenditure grant (GDF/DGE) allocated according to population size; payments for salaries and payments for capital expenditure and operating expenses. Only one (investment) fund takes poverty into account.

Côte d’Ivoire 4 4/0 3/1 A system for sharing local taxes based on the origin of the tax and three other funds. One of these,the DGF uses a formula based, in practice, on the population.

Egypt 1 1/0 1 A system for sharing some surtaxes based on population size, surface area, needs and fiscalbalance transfers. The surtax associated with the Suez canal is shared between the governoratesthat lie along the canal

Kenya 2 2/1 2/2 A system that takes into account the share of the population and the proportion of the urbanpopulation and a system based on road construction projects

Malawi 3 2/1 2/1 An unconditional transfer based on population and poverty, a transfer that finances devolvedagriculture expenditure and a transfer for education based on the number of children of school-going age

Mali 3 3/0 3/0 No formula would be used

Mauritania 1 1/0 1/0 The FRD is divided into an operating grant and a capital expenditure grant. It is distributedaccording to population size, poverty and infrastructure

Morocco 3 0/3 0/3 30% of VAT goes to municipalities and provinces/prefectures. Municipalities receive a lump sum, atax equalisation and a tax effort, while P/P receive funding to cover payroll expenses.1% of income tax (CIT and PIT) goes to regions

Mozambique 3 2/1 2/1 The FCA (Municipal Compensation Fund) transfer distributes 1.5% of state revenues based onpopulation (75%) and surface area (25%). The other two do not use formulas.

Niger 2 1/1 2/0 No formula would be used

Senegal 2 2/0 0/2 There is a Decentralisation grant (FDD) distributed according to population size and the FECL thatfinances the domestic counterparty of internationally funded projects

South Africa 15 14/1 13/0/Other Most funds have precise formulas that factor in population and needs; some are ad hoc single formulas (FIFA 2010)



335

Annex 2.2: Some characteristics of transfers to municipalities for 20 African countries, 2008

Country Number of Number of Number of Brief observations about transfer systems transfer conditional grants/ transfers with systems Number of ad hoc financing /

unconditional number of grants transfers financed

by % of taxes

Tanzania 3 2/1 3/0 The three transfers are each distributed using a formula, which include such elements aspopulation, poverty, rurality

Togo 1 1/0 1/0 The FACT is not operational.

Tunisia 3 3/0 3/0 A general transfer (FCCL) is used, based on population size and revenues, a transfer via grants andreduced interest rates for investments (CPSCL) and transferred ministerial credits

Uganda 3 2/1 3/0 88% of grants are conditioned for sectoral spending in reduction of poverty, 11% of transfers areunconditioned, and 0.5% of the amounts are used for equalisation

Zambia 3 3/0 3/0 Since 2007 there are two grants for restructuring and functioning. This was subdivided in 2010into three components: institutional support, service provision and abolished tax replacement. Theportion on services considers 6 variables including population and poverty index. The 3rd grant isfor investment but to date it has had a weak impact.

Zimbabwe 2 2/0 2/0 Payment of salaries of health staff and granting of financing depending on rural council budgets

Source: National information collected by UCLG; Hugounenq, Rocaboy and Vaillancourt 2010 (Kenya); Dafflon and Madies 2009 (Burkina Faso); Gilbert,Hugounenq Taugourdeau 2009 (Senegal); Vaillancourt 2008 (Côte d’Ivoire); Martinez-Vazquez and Timofeev 2008 (Egypt)



(1) General Government

Expenditure Revenue

General Government Local Government General Government Local Government

Africa Million USD % of GDP Million USD % of GDP % of GG (1) Million USD % of GDP Million USD % of GDP % of GG

Benin (2007) 1301 23.5 53 1 4.1 1149 20.7 49 0.9 4.3

Burkina Faso (2007) 1820 26.9 45 0.7 2.5 1407 20.8 31 0.5 2.2

Cote D'Ivoire (2007) 3826 19.3 237 1.2 6.2 3775 19.1 252 1.3 6.7

Egypt (2007) 55600 42.1 6067 4.6 10.9 36496 27.6 na na na

Kenya (2007) 5541 20.4 252 0.9 4.6 na na na na na

Malawi (2007) 1312 36.6 111 3.1 8.5 na na 99 2.8 na

Mali (2007) 1858 27.1 64 0.9 3.4 1599 23.3 57 0.8 3.6

Mauritania (2008) 957 33.4 5 0.2 0.6 691 24.1 5 0.2 0.8

Morocco (2007) 21929 29.2 1680 2.2 7.6 27691 36.8 2912 3.9 10.5

Mozambique (2009) 3568 34.1 86 0.8 2.4 na na 28 0.3 na

Niger (2007) 877 20.7 21 0.5 2.4 640 15.1 23 0.5 3.6

Rwanda (2008) 1188 26.7 274 6.1 23.1 1246 28 264 5.9 21.2

Senegal (2007) 3017 26.6 95 0.8 3.2 2369 20.9 97 0.9 4.1

South Africa (2007) 94625 33.3 16513 5.8 17.4 104826 36.9 18633 6.6 17.7

Tanzania (2007) 4008 23.8 647 3.8 16.2 na na 707 4.2 na

Togo (2006) 522 20.9 9 0.4 1.8 482 19.3 8 0.3 1.6

Tunisia (2008) 7906 19.6 762 1.9 9.6 12394 30.7 1020 2.5 8.2

Uganda (2007) 2011 16.4 684 5.6 34 na na 901 7.3 na

Annex 2.3: Public Finance Indicators for 18 countries



337

Annex 4.1: The Structure of Local Government Outlays (2008)

Annex 4.2: The Structure of Local Tax Revenues (2008)


General public services 26.3% 19.9% 8.1% 9.1% 13.4% 9.5% 9.5% 12.5%

Defense 0.1% 0.1% 0.3% 0.7% 0.5% 0.1% 0.1% 0.0%

Public order and safety 0.0% 1.6% 2.3% 3.0% 1.4% 3.2% 1.3% 1.6%

Economic affairs 8.0% 10.5% 16.4% 19.8% 1.7% 7.4% 6.6% 10.6%

Environmental protection incl. in “Economic affairs” 0.3% 5.5% 0.7% 0.0% 0.0% 0.2% 0.5%

Housing and community amenities 26.4% 13.8% 40.5% 13.5% 10.6% 8.8% 21.0% 6.0%

Health 0.1% 18.9% 2.4% 18.6% 5.2% 2.3% 11.5% 19.2%

Recreation, culture, and religion 6.7% 4.5% 8.3% 6.3% 4.3% 5.2% 5.7% 3.8%

Education 18.0% 24.6% 11.1% 27.6% 58.8% 53.4% 36.6% 28.7%

Social protection 2.5% 5.9% 5.3% 3.6% 4.7% 10.2% 7.5% 17.1%

Source: Authors’ calculations based on IFS database, for Armenia – on "Local Self-Government Reforms in Armenia (2007 and 2008) Book 3", for Russia –on RF Ministry of Finance reports. Data for Kyrgyzstan is for 2006.


Taxes on income, profits, and capital gains: 0.0% 47.8% 36.9% 36.2% 46.9% 85.9% 75.0% 87.2%

payable by individuals 0.0% 26.2% 36.9% 36.2% 28.0% 69.6% 60.3% 84.8%

payable by enterprises 0.0% 18.0% 0.0% 0.0% 15.7% 5.7% 14.7% 2.2%

Taxes on payroll and workforce 0.0% 0.0% 0.0% 34.0% 0.0% 0.0% 0.0% 0.3%

Taxes on property (incl. land tax) 100.0% 12.5% 62.5% 14.0% 18.3% 13.0% 10.5% 4.3%

Taxes on goods and services 0.0% 34.6% 0.0% 6.6% 34.8% 1.1% 14.5% 8.2%

General taxes on goods and services 0.0% 30.0% 0.0% 0.0% 31.0% 0.0% 1.6% 0.0%

Excises 0.0% 1.1% 0.0% 4.8% 1.8% 0.1% 0.2% 0.2%

Other taxes 0.0% 5.1% 0.6% 9.1% 0.0% 0.0% 0.0% 0.0%

Source: Authors’ calculations based on IFS database.

4. EURASIA



Annex 4.3: Structure of Local Transfers

Country Armenia Belarus Georgia Kazakhstan Kyrgyzstan Moldova Russia UkraineRevenue item (2008) (2005) (2008) (2008) (2008) (2009) (2008) (2008)

Transfers 47.6% 20.9% 35.6% 55.8% 39.6% 56.9% 45.5% 46.4%

General purpose transfers 40.6% 13.8% 23.5% 24.5% 5.6% 55.1% 10.2% 23.0%

incl. equalization grants 40.6% 13.8% 23.5% 24.5% 5.6% 55.1% 10.1% 22.4%

Targeted transfers 7.0% 6.4% 12.1% 26.4% 34.0% 1.8% 35.2% 23.4%

Incl. financing of delegated responsibilities 1.8% n.a. 11.3% n.a. n.a. n.a. 18.8% n.a.

Source: Authors’ calculations based for Armenia – on "Local Self-Government Reforms in Armenia (2007 and 2008) Book 3", for Georgia – on “Local Self-Government in Georgia” for other countries – on the countries’ Ministry’s of Finance report.



339

Main own-source taxes (by order of importance in revenue) Main shared taxes (by order of importance in revenue) Main grants/transfers

Austria Municipal business tax, real estate tax. VAT, wage tax, property acquisition tax, CIT, General purpose, support based on financial

petrol tax, PIT need, earmarked grants and equalisation grants

(from the middle-tier jurisdictions and other

municipalities).

Belgium Surtaxes on PIT, vehicle traffic circulation tax, General purpose (municipal and provincial

regional withholding tax on real estate, funds) and earmarked grants (operating and

wide range of local taxes. capital expenditure subsidies).

Bulgaria Tax on the acquisition of property by way of PIT (until 2008). Since 2008, general supplemental subsidy,

donation, real estate property tax, transport vehicle general equalisation grant, capital

tax, inheritance tax, patent tax (since 2008) investment grant.

Cyprus Waste collection tax, municipal corporate tax, General purpose, compensation for the abolition

property tax. of taxes and earmarked grants

(for capital spending).

Czech Rep. CIT paid by municipal companies, property tax. VAT, PIT and CIT. Most are earmarked, designed for current

spending (State delegated functions)

and capital expenditure.

Denmark Municipal PIT, land tax, business property tax. CIT General purpose, equalisation and earmarked

grants (including VAT compensation,

reimbursement of social welfare expenditure

and municipalities with specific needs).

Estonia Land tax. PIT Block grant , specific purpose (education and

social services), equalisation fund, earmarked

grants (for capital spending)

and amalgamation grant.

Finland Municipal PIT, real estate tax. CIT General purpose, specific purpose (education,

health care, social welfare) and equalisation.

France Local business tax (until 2009), property tax on Domestic tax on petroleum products, General purpose, equalisation, general

buildings, residence tax, property tax on land, special tax on insurance contracts. decentralisation grant, vocational training grant,

transfer taxes on property transactions, and capital expenditure grants (VAT

transport contribution. compensation and specific purpose such

as rural development, school equipment).

Germany Local business tax, property tax on real estate. PIT, VAT and tax on interest. General purpose, compensation for functions

delegated by the middle-tier jurisdictions,

investment transfers (from middle-tier

jurisdictions).

Annex 5.1: Local Government Revenue, 2009

5. EUROPE



Annex 5.1: Local Government Revenue, 2009 (cont.)


Greece Waste management and street cleaning fee, General purpose (Central Autonomous Funds

street-lighting fee, property tax, charge on hotel composed of the shares of national taxes such

and restaurant. as PIT, vehicle tax, property sales tax, VAT, etc.)

and earmarked grants (for capital spending).

Hungary Municipalities : local business tax, property tax on PIT and motor vehicle tax. Normative grants (both general and earmarked),

buildings (from 2010, a new property tax will other earmarked grants (“targeted” and

replace other taxes related to property), tourism tax. “addressed” for capital spending) and grants for

municipalities in financial trouble.

Iceland Local PIT, real estate tax. Municipal equalisation fund, housing fund,

general purpose fund.

Ireland Commercial rates tax (business property tax). General purpose (Local Government Fund –

General Purpose Grant financed partly by the full

proceed from the motor tax) and earmarked grants

(roads, housing, water supply and sewerage).

Italy Municipal property tax, municipal tax on building PIT (municipalities and provinces). Municipalities and provinces : operating grants

licenses, household waste tax, surtax on PIT. PIT and CIT (regions with special status). (ordinary fund, consolidated fund and

Provincial vehicle insurance tax and registration equalisation fund).

tax, surtax on electricity consumption. Regions : operating grant coming partly from the

Regional tax on productive output, surtax on PIT, VAT transfers, national health fund,

regional automobile tax, fuel duty administrative federalism fund.

Capital expenditure grants.

Latvia PIT and real estate tax. General purpose, earmarked grants (teachers' compensation,

road maintenance, public investment programme),

Local government Finance Equalisation Fund.

Lithuania Tax on the immovable property of enterprises PIT and tax on pollution. General purpose (based on equalisation),

and organisations, land tax. specific purpose (for most delegated

responsibilities, including schools and social

benefits) and health care.

Luxembourg Municipal business tax, property tax. Municipal grant fund (composed of the share of

the PIT, VAT, vehicle tax and a flat-rate amount),

operating and capital earmarked grants.

Malta – – General grant, funds for special needs.

Netherlands Municipal property taxes, refuse collection rate, General purpose (Municipal and Provincial Funds

sewer tax. based on a equalisation system), specific

Provincial surtax on the national motor vehicle tax. purpose (numerous grants but mainly targeted

at education and social services).



341


Norway Local PIT, property tax (municipalities), CIT, wealth tax. General purpose grant, earmarked grants to

natural resource tax. kindergartens (to be included in the general

purpose grant system form 2011)

Poland Municipalities : property tax, farming tax, PIT and CIT. General purpose comprising 4 shares (balancing,

tax on vehicles. regional, education, equalisation); Earmarked

grants, notably to carry out State delegated

responsibilities and own specific responsibilities.

Portugal Municipal property tax, property transfer tax, Since 2007, possibility to retain 2% of the PIT General block grant made of shares of state

vehicle registration tax, surtax on CIT. of local residents. taxes (PIT, VAT, CIT) and divided into the General

Municipal Fund, the Municipal Cohesion Fund

and the Social Municipal Fund.

Romania Building and land taxes held by legal persons and VAT and PIT. Earmarked grants and special funds (e.g.

by individuals, patent tax support for the disabled).

Slovak Republic Real estate tax, vehicle tax. PIT Earmarked grants to cover delegated responsibilities

(mainly education and social services).

Slovenia Property tax, gift and inheritance tax, tax on profits PIT Equalisation grant, investment grant and specific

from gambling, sales tax on immovable property grant (support for ethnic minorities).

Spain Municipalities : tax on property, tax on Municipalities with more than 75 000 Municipalities and provinces : general grant

constructions, installations and works, tax on motor inhabitants and provinces : PIT, VAT and excises. (composed of shares of central State taxes)

vehicles, tax on economic activities, tax on capital based on a equalisation system and specific

gains in urban areas. purpose grants (transport infrastructure).

Provinces: surtax on the municipal tax

on economic activities.

Sweden Local PIT, property tax (since 2008). Equalisation grants for both municipalities and

county councils, a grant for pharmaceutical

benefits for county councils , several targeted

grants to municipalities and a VAT refund.

United Kingdom England, Scotland and Wales : Council tax England, Scotland and Wales : business rates England, Scotland and Wales: “Aggregate External

(tax on property paid by residents). (business property tax). finance” comprising a general purpose grant

Northern Ireland: district rates (property tax) (Revenue Support Grant), redistributed business

rates (Formula Grant) and specific purpose grants

(unfenced and ring-fenced).

Northern Ireland: general grants and various specific grants.

Sources: National sources and DEXIA (see bibliography).

Annex 5.1: Local Government Revenue, 2009 (cont.)



Golden rule1 Approval required from higher-level jurisdiction2 Debt ceiling in place

Austria Yes Depending on the Länder regulation Depending on the Länder regulation

Belgium Yes No No

Bulgaria Yes No Cap on debt service

Cyprus Yes Yes No

Czech Rep. Yes No No

Denmark Yes No No

Estonia Yes No (Yes, on an interim basis since 2009 and until 2012) Caps on debt service and on outstanding debt

Finland No No No

France Yes No No

Germany Yes Depending on the Länder regulation Depending on the Länder regulation

Greece No No No

Hungary No No Cap on new annual borrowing

Iceland No No No

Ireland No Yes Cap on new annual borrowing

Italy Yes No Cap on debt service

Latvia Yes Yes Cap on new annual borrowing

Lithuania Yes Yes Cap on outstanding debt and on new annual borrowing

Malta Yes Yes No

Netherlands Yes No No

Norway Yes No No

Poland No No Caps on debt service and on outstanding debt

Portugal Yes Yes Cap on outstanding debt

Romania Yes Yes Cap on debt service

Slovak Republic Yes No Caps on debt service and on outstanding debt

Slovenia Yes Yes Caps on debt service and on outstanding debt

Spain Yes Depending on the Autonomous Communities regulation Cap on outstanding debt and on new annual borrowing

Sweden Yes No No

Switzerland Depending on the cantons Depending on the cantons regulation Depending on the cantons regulation

regulation

United Kingdom Yes No No

1. According to the golden rule, borrowing is allowed only to finance capital expenditure.2. In general and for ordinary loans. For other types of loans (e.g., in foreign currency, in foreign markets, bonds, for a non-eligible spending, exceeding

debt ceilings, etc.) or for local governments in financial difficulty prior approval of the supervisory authority may be necessary.

Sources: National sources and DEXIA.

Annex 5.2: Local Government Long-term Borrowing Rules, 2009



343

Composition of expenditure (% of local government expenditure)

General public services Economic affairs Health Education Social protection Other

Austria 18.1 13.2 17.7 16.2 18.7 16.1

Belgium 23.5 10.0 2.9 19.5 16.6 27.5

Bulgaria 4.6 14.1 10.0 30.4 7.2 33.6

Cyprus 42.9 0.0 0.0 0.0 0.0 57.1

Czech Republic 12.2 21.5 2.2 29.3 11.7 23.1

Denmark 4.2 3.6 22.8 10.8 54.1 4.5

Estonia 6.1 13.0 14.9 38.9 6.3 20.9

Finland 13.7 6.5 28.5 20.2 23.8 7.3

France 18.4 12.5 1.0 16.4 16.2 35.4

Germany 16.3 13.1 2.3 14.6 33.0 20.7

Greece 40.2 17.8 0.0 2.9 11.0 28.1

Hungary 18.0 7.2 14.4 29.2 12.5 18.7

Iceland 11.4 6.5 14.6 27.3 26.0 14.2

Ireland 10.8 28.0 0.0 15.4 3.5 42.3

Italy 14.7 14.3 44.5 8.3 4.5 13.7

Latvia 10.1 13.3 9.5 37.0 6.6 23.5

Lithuania 6.5 10.8 21.1 37.8 6.3 17.5

Luxembourg 20.8 15.8 0.5 24.1 3.8 35.0

Malta 57.5 8.9 0.0 0.0 0.0 33.6

Netherlands 15.0 17.1 1.7 28.1 14.2 23.8

Norway 11.9 6.3 14.7 27.3 25.7 14.2

Poland 9.2 15.3 15.2 28.5 12.9 18.8

Portugal 35.0 19.2 5.5 8.3 2.6 29.3

Romania 11.5 25.5 0.5 27.5 13.2 21.9

Slovak Rep. 14.9 14.6 0.5 39.3 7.2 23.5

Slovenia 10.5 11.5 11.1 38.8 8.9 19.2

Spain 34.3 11.9 1.3 3.9 9.1 39.6

Sweden 11.3 5.9 27.0 21.4 26.8 7.7

Switzerland 23.5 8.7 20.2 21.0 16.5 10.1

United Kingdom 6.5 8.6 0.0 31.8 27.9 25.2

Sources: Eurostat, World Bank (2009), Dexia and CEMR (2009).

Annex 5.3: Local Government Expenditure: Functional classification, 2007



Revenue Composition of revenue (% of local government revenue)

% of GDP % of general government revenue Taxes1 Social contributions Grants Other2

Austria 7.7 15.9 61.4 4.7 20.0 13.9

Belgium 6.7 13.8 29.6 5.4 53.7 11.3

Bulgaria 7.2 18.5 13.6 0.0 74.3 12.1

Cyprus 1.8 4.2 28.4 0.0 45.8 25.8

Czech Rep. 11.4 27.9 46.2 0.0 36.9 16.9

Denmark 33.2 60.0 36.0 1.5 55.5 7.0

Estonia 10.4 28.0 47.5 0.0 42.8 9.7

Finland 20.0 37.4 47.4 0.1 28.8 23.7

France 10.9 22.1 45.5 0.3 37.0 17.1

Germany 7.5 17.1 42.9 1.3 38.9 17.0

Greece 2.7 6.7 9.6 0.0 64.2 26.2

Hungary 11.6 25.5 22.4 0.2 65.8 11.6

Ireland 7.7 22.1 9.8 3.7 70.4 16.1

Italy 15.3 33.2 42.8 0.6 48.5 8.1

Latvia 10.7 30.8 51.9 0.0 37.2 10.9

Lithuania 9.1 26.7 37.5 0.2 58.1 4.2

Luxembourg 5.1 12.8 31.3 0.2 48.0 20.5

Malta 0.6 1.4 0.0 0.0 92.6 7.4

Netherlands 15.3 32.9 8.4 2.3 69.6 19.8

Poland 14.0 35.2 33.2 0.0 53.3 13.4

Portugal 6.1 14.0 39.4 1.4 40.3 18.9

Romania 8.7 27.1 10.4 0.0 84.7 4.9

Slovak Rep. 5.4 16.6 59.6 0.5 27.9 11.9

Slovenia 8.5 19.9 39.2 1.0 46.1 13.8

Spain 6.0 16.3 48.9 0.6 40.5 10.0

Sweden 25.3 45.5 64.8 2.1 19.6 13.5

United Kingdom 13.1 31.0 13.0 1.9 70.0 15.1

EU27 11.3 25.4 36.1 1.2 48.7 14.0

Iceland 13.1 29.6 72.7 0.0 10.3 17.0

Norway 12.2 20.7 41.8 0.0 39.5 18.7

Switzerland 8.7 24.2

1. Includes own-source and shared revenue, even where a sub-national government has virtually no power to set rates or bases.2. Includes primarily fees and user charges, but also revenue associated with physical and financial assets (such as sales, dividends and interest).

Sources: Eurostat (February 2010), national sources and DEXIA calculations.

Annex 5.4: Local Government Revenue, 2008



345

Total Taxes on:1

Income, profits and capital gains2 Property3 Goods and services4 Other5

Austria 61.4 22.1 3.9 24.4 11.0

Belgium 29.6 9.3 17.6 2.7 0.0

Bulgaria 13.6 0.4 12.0 1.2 0.0

Cyprus 28.4 0.0 15.6 6.4 6.4

Czech Rep. 46.2 25.2 1.5 19.4 0.0

Denmark 36.0 32.2 3.7 0.0 0.0

Estonia 47.5 43.9 2.9 0.7 0.0

Finland 47.4 44.9 2.5 0.0 0.0

France 45.5 0.0 27.9 14.3 3.3

Germany 42.9 16.9 5.3 20.6 0.0

Greece 9.6 0.0 5.1 4.5 0.0

Hungary 22.4 0.0 3.0 19.3 0.0

Ireland 9.8 0.0 9.8 0.0 0.0

Italy 42.8 11.1 4.3 27.5 0.0

Latvia 51.9 47.2 4.1 0.6 0.0

Lithuania 37.5 33.4 3.0 1.0 0.0

Luxembourg 31.3 28.5 1.6 1.1 0.0

Malta 0.0 0.0 0.0 0.0 0.0

Netherlands 8.4 0.0 6.6 1.8 0.0

Poland 33.2 20.6 10.3 2.4 0.0

Portugal 39.4 11.2 11.1 17.1 0.0

Romania 10.4 0.1 6.9 3.4 0.0

Slovak Rep. 59.6 48.3 6.9 4.4 0.0

Slovenia 39.2 29.7 5.3 4.3 0.0

Spain 48.9 10.9 15.3 22.7 0.0

Sweden 64.8 63.2 1.6 0.0 0.0

United Kingdom 13.0 0.0 13.0 0.0 0.0

EU 27 36.1 13.3 10.4 11.8 0.7

Iceland 72.7 54.8 17.9 0.0 0.0

Norway 41.8 36.6 4.5 0.7 0.0

Switzerland (2007) 56.1 45.9 9.1 1.1 0.0

1. Includes own-source and shared revenue, even where a sub-national government has virtually no power to set rates or bases.2. Includes all taxes on income (d51 in ESA95 classification)3. Includes taxes on land, buildings and other structures (d29a), current taxes on capital (d59a), other current taxes n.e.c. (d59f) and capital taxes

(d91).4. Includes taxes on products such as VAT, import taxes and other consumption taxes (d21), taxes on the use of fixed assets (d29b), business and

professional licences (d29e), taxes on pollution (d29f) and other taxes on production (d29h).5. Includes taxes on payroll and workforce (d29c).

Sources: Eurostat (Feb. 2010) and DEXIA calculations.

Annex 5.5: Local Government Tax Revenue, 2008 (% of local government revenue)



Total Current Capital

Austria 20.1 16.1 4.0

Belgium 53.7 48.4 5.3

Bulgaria 81.8 74.3 7.6

Cyprus 45.8 27.6 18.3

Czech Rep. 36.9 30.5 6.3

Denmark 56.5 56.0 0.6

Estonia 42.8 36.6 6.1

Finland 28.8 28.5 0.3

France 38.3 33.6 4.7

Germany 39.1 32.3 6.8

Greece 65.8 44.1 21.6

Hungary 66.0 56.8 9.2

Ireland 70.4 27.6 42.8

Italy 48.5 42.2 6.3

Latvia 37.2 30.8 6.3

Lithuania 58.2 43.5 14.7

Luxembourg 48.0 41.2 6.9

Malta 92.6 89.2 3.4

Netherlands 70.0 65.7 4.4

Poland 53.5 49.7 3.8

Portugal 40.3 20.0 20.3

Romania 84.7 81.6 3.1

Slovak Rep. 27.9 19.9 8.1

Slovenia 46.4 41.7 4.7

Spain 43.3 32.6 10.6

Sweden 20.0 19.7 0.3

United Kingdom 70.0 62.2 7.8

EU27 49.2 43.1 6.0

Iceland 10.3 8.7 1.6

Norway 39.5 39.5 0.0

Sources: Eurostat and DEXIA calculations.

Annex 5.6: Local Government Grants, 2008 (% of local government revenue)



347

Anne

x 6.1

:As

signm

ents

of e

xpen

ditur

es fu

nctio

ns

Functi

ons

Argen

tina

Bolivi

aBra

zilCh

ileCo

lombia

Costa

Rica

Domi

nican

Repu

blicEcu

ador

El Salv

ador

Guate

mala

Hond

uras

Jamaic

aMe

xico

Nicara

gua

Peru

Urugu

ayVen

ezuela

Cons

titut

ion de

fines

assig

nmen

tsNo

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

NoNo

Yes

Yes

Yes

Yes

Gene

ral a

dmin

istra

tion

Secu

rity,

polic

eC/

RC

RC

CC

CC/

LNA

LL

CC/

R/L

CC/

LC

C/R/

L

Fire p

rote

ction

C/R

CR

CL

CL

NAL

C/L

LL

LC

LC

C/R/

L

Civil p

rote

ction

C/R

CR

CL

CC/

LC/

LL

CL

CC/

R/L

C/R/

LC

C/L

Just

ice

Civil s

tatu

s reg

ister

CC

CC

CC

CC

LL

LC

R/L

LC/

LC

L

Statis

tical

offic

eC/

RC

C/L

CC

CC

CL

CL

CC/

R/L

CC

C

Elect

oral

regis

ter

C/R

CC

CC

CC

CL

CL

CC

CC

C

Educ

ation

Pre-

scho

ol ed

ucat

ionR

C/L

LC

LC

CC

LL

C/L

LC/

R/L

C/R

C/L

C/R/

L

Prim

ary e

duca

tion

RC/

LL

CL

CC

CL

LC/

LL

C/R/

LC/

RC/

LC/

R/L

Seco

ndar

y edu

catio

nR

C/L

RC

LC

CC

LNA

CL

C/R/

LC/

RC

C/R/

L

Voca

tiona

l and

tech

nica

lR

C/L

RC

LC

CC

LNA

CC

C/R

C/R

C/R/

L

High

er ed

ucat

ionC

C/L

CNA

CC

CC

CNA

CC

C/R

C/R

CC

Adult

educ

ation

RC/

LR/

LNA

CC

CC

CL

CC

RC/

RC

Publi

c Hea

lth

Healt

h pro

tect

ionR

C/L

C/R/

LNA

LC

CC

LNA

C/L

LC/

RC/

RC/

LC/

R/L

Prim

ary c

are

RC/

LC/

R/L

C/L

LC

CC

LNA

C/L

LC/

R/L

C/L

C/R

CL

Hosp

itals

RC/

LC/

R/L

C/L

RC

CC

NANA

C/L

C/L

C/R

C/R

CC/

R

Socia

l Welf

are

Kind

erga

rten a

nd nu

rser

yR

C/L

LL

CC

NAC

LNA

NAC

C/R

CC/

R/L

C=

cent

ral,

R=

regi

onal

, L=

loca

l

Sou

rce:

IM

F;M

inis

trie

s of

Fin

ance

of A

rgen

tina,

Bol

ivia

, Gua

tem

ala

and

Peru

, UCLG

dat

a co

llect

ion.

6. LATIN AMERICA


AANNNNEEXXUnited Cities and Local Governments348An

nex 6

.1:

Assig

nmen

ts of

exp

endit

ures

func

tions

(con

t.)

Functi

ons

Argen

tina

Bolivi

aBra

zilCh

ileCo

lombia

Costa

Rica

Domi

nican

Repu

blicEcu

ador

El Salv

ador

Guate

mala

Hond

uras

Jamaic

aMe

xico

Nicara

gua

Peru

Urugu

ayVen

ezuela

Fam

ily w

elfar

e ser

vices

C/R

C/L

C/R/

LL

CC

NAC

LNA

NAC/

LC

C/L

CC/

R/L

Welfa

re ho

mes

C/R

C/R

C/R/

LNA

NAC

NAC

LNA

NAC

C/R

LC

C/R/

L

Socia

l sec

urity

C/R

CC

C/L

CNA

CC

NANA

NAC

CC

C

Hous

ing an

d tow

n plan

ning

Hous

ingL

C/L

C/L

LC/

LC

CL

LL

LC

LL

CL

C

Town

plan

ning

LL

C/L

LL

CL

LL

LL

C/L

LL

LL

L

Regio

nal/s

patia

l plan

ning

RC/

LC/

LR

C/R

NANA

LL

NAL

C/L

RR

LR/

L

Envir

onm

ent, s

anita

tion

Wate

r & se

wage

RC/

R/L

LL

R/L

C/R

CL

LL

LC/

LR/

LL

C/R/

LL

L

Refu

se co

llect

ion &

disp

osal

RC/

R/L

LL

LL

LL

LL

LC/

LL

LL

LC/

L

Cem

eter

ies &

crem

ator

iaR

C/R/

LL

LL

LL

LL

LL

NAL

LL

LL

Slaug

hter

hous

esR

C/R/

LL

LL

LL

LL

LL

NAL

LL

LL

Envir

onm

enta

l pro

tect

ionR

C/L

C/L

LR/

LC/

RC/

LL

LL

LC

C/R/

LL

C/R/

LL

C/L

Cons

umer

prot

ectio

nC

CL

NAC

CNA

NAL

NAL

CC/

R/L

CC

Cultu

re, le

isure

& sp

orts

Thea

tres &

conc

erts

NAC/

R/L

R/L

NAL

LC

NAL

LL

NAR/

LC/

LL

C/R/

L

Muse

ums &

libra

ries

NAC/

R/L

C/R/

LNA

LC

CNA

LL

LNA

R/L

LC/

LL

C/R/

L

Park

s & op

en sp

aces

NAC/

R/L

C/R/

LL

LC\

LL

NAL

LL

C/L

LL

LL

R/L

Spor

ts &

leisu

reNA

C/R/

LC/

R/L

LL

NAC/

LNA

LL

LC/

LC/

R/L

LL

C/L

C/R/

L

Relig

ious f

acilit

iesNA

C/R/

LC/

R/L

NANA

NANA

NAL

NAL

NAC

LL

Othe

r cult

ural

facil

ities

NAC/

R/L

C/R/

LL

NANA

NANA

LNA

LC/

LC/

R/L

LL

L

C=

cent

ral,

R=

regi

onal

, L=

loca

l

Sou

rce:

IM

F;M

inis

trie

s of

Fin

ance

of A

rgen

tina,

Bol

ivia

, Gua

tem

ala

and

Peru

, UCLG

dat

a co

llect

ion.



349

Anne

x 6.1

:As

signm

ents

of e

xpen

ditur

es fu

nctio

ns (c

ont.)

Functi

ons

Argen

tina

Bolivi

aBra

zilCh

ileCo

lombia

Costa

Rica

Domi

nican

Repu

blicEcu

ador

El Salv

ador

Guate

mala

Hond

uras

Jamaic

aMe

xico

Nicara

gua

Peru

Urugu

ayVen

ezuela

Traffi

c, tra

nspo

rt

Road

sC/

R/L

CC

LC/

R/L

C/L

CR/

LNA

C/L

LC/

LL

C/L

C/R/

LL

C/R/

L

Trans

port

C/R

CC/

LL

C/L

NAC/

LR/

LNA

LL

C/L

C/R

C/L

C/R/

LL

C/R/

L

Urba

n roa

d tra

nspo

rtC/

R/L

CL

LL

C/L

C/L

R/L

NAL

LC/

LR/

LC/

LL

LL

Urba

n rail

tran

spor

tC

CC

NAL

NANA

NANA

NAC

C/L

LL

Ports

CC

CNA

C/L

CC

NANA

NAC

CC

LR

C/R

Airpo

rtsC

CC

NAC/

LC

CNA

NANA

CC

CC

C/R

Utilit

ies

Gas

RC

CNA

C/L

NANA

NAL

NAL

CC

LC/

L

Dist

rict h

eatin

gR

CC

NANA

NANA

NAL

NAL

CL

Elect

ricity

RR

C/R/

LNA

C/L

CC

NAL

NAL

CL

LC/

R/L

C/L

C/L

Wate

r sup

plyR

R/L

NAR/

LC

CL

LL

LC

LL

C/R/

LL

C/L

Econ

omic

Agric

ultur

e, fo

rest

s, fis

hing

C/R

C/R/

LC/

LNA

C/L

CC

LL

NAL

CC

C/R

LC

Econ

omic

prom

otion

C/R

C/R/

LC/

LNA

C/R/

LC

C/L

LL

NAL

CC/

R/L

C/R

LC/

R/L

Trade

& in

dust

ryC/

RC/

R/L

C/L

NAC/

LC

CL

LNA

LC

C/R

C/R

LC/

R/L

Tour

ismC/

RC/

R/L

C/L

NAC/

R/L

CC

NAL

NAL

CC/

RL

C/R

LL

Othe

r eco

nom

ic se

rvice

sNA

C/R/

LC/

LNA

LC

NANA

NANA

LC

C/R/

LC/

R/L

Emplo

ymen

tNA

C/R/

LC/

LL

C/L

CC

NANA

NAL

CC/

R/L

C

C=

cent

ral,

R=

regi

onal

, L=

loca

l

Sou

rce:

IM

F;M

inis

trie

s of

Fin

ance

of A

rgen

tina,

Bol

ivia

, Gua

tem

ala

and

Peru

, UCLG

dat

a co

llect

ion.



Country Total Total Local Local Local Local Local LocalExpenditure in Expenditure in Expenditure Expenditure Expenditure Revenues government government General Government General Government (Mill USD) (% of GDP) % of General (Mill USD) revenues revenues as % of (Mill USD) (% of GDP) Government as % of GDP General Government

Argentina(2006) 70,468 32.9 6,204 2.9 8.8 5,277 2.5 7.3

Bolivia(2008) 7,262 43.6 1,223 7.3 16.8 1,278 7.7 17.6

Brazil(2007) 420,253 31.5 110,693 8.3 26.3 108,748 8.2 18

Chile(2007) 31,094 19.0 3,982 2.4 12.8 4,417 2.7 9.2

Colombia(2006) 48,405 29.8 9,046 5.6 18.7 8,451 5.2 22

Costa Rica(2007) 5664 21.5 208 0.8 3.7 468 1.8 6.9

Dominican Republic(2006) 7,612 17.9 403 0.9 5.3 na

Ecuador(2007) 10,357 18.9 2,423 4.4 23.4 2,087 3.8 22.4

El Salvador(2007) 3,533 17.3 249 1.2 7.0 385 1.9 9.3

Guatemala(2009) 5,620 15.5 245 0.7 4.4 1,016 2.8 16.5

Haití(2004) 802 15.0 na na na na

Honduras(2008) 3,770 26.6 184 1.3 4.9 177 1.2 4.9

Jamaica(2006) 3,912 32.6 34 0.3 0.9 34 0.3 0.9

Mexico(2007) 339,502 31.2 21,969 2.0 6.5 23,007 2.1 7.4

Nicaragua(2006) 1,461 28.0 na na na na na na

Panama(2005) 6,855 44.9 137 0.8 1.7 na na 2.0

Paraguay(2007) 2,109 20.3 133 1.3 6.3 182 1.8 6.8

Peru(2007) 17,137 15.9 2,812 2.6 16.4 3,977 3.7 17.8

Uruguay(2007) 8,611 30.8 na na na na na na

Venezuela(2007) 58,888 25.8 na na na na na na

Annex 6.2: Public Finance Indicators

Source: IMF; Ministries of Finance of Argentina, Bolivia, Guatemala and Peru, UCLG data collection.



351

Country (Most recent year) Revenues (% total) Expenditures (% total)

Federal / Central Regional Local Federal / Central Regional Local

Argentina(2006) 55 38 7 58 33 9

Bolivia(2008) 71 12 17 72 11 17

Brazil(2007) 54 28 18 45 29 26

Chile(2007) 91 9 87 13

Colombia(2006) 64 14 22 67 14 19

Costa Rica 93 7 96 4

Dominican Republic(2006) 95 5

Ecuador(2007) 78 22 77 23

El Salvador(2007) 91 9 93 7

Guatemala(2002) 84 16 96 4

Haiti(2004) 100 100

Honduras(2004) 95 5 95 5

Jamaica(2008) 99 1 99 1

Mexico(2007) 68 25 7 69 25 6

Panama(2005) 98 2 98 2

Paraguay(2006) 93 7 94 6

Peru(2008) 66 16 18 66 18 16

Annex 6.3: Revenues and Expenditures by Government Level (%)



Country Wages and Salaries/ Pension funds Current expenditure other Capital Expenditure/debt/(Most recent year) (% of total) than wages and salaries (% of total) equipment (% of total)

Argentina(2006) 47.40 29.27 23.33

Bolivia(2008) 65.80 26.90 7.30

Brazil(2007) 46.74 53.05 0.21

Chile(2007) 29.11 15.66 55.23

Colombia(2006) 74.05 1.96 23.98

Ecuador(2007) 23.60 20.80 55.60

El Salvador(2007) 45.61 39.78 14.60

Guatemala(2002) 0.00 36.00 64.00

Honduras(2004) 30.42 19.03 50.55

Mexico(2007) 82.83 12.29 4.88

Nicaragua(2002) 23.08 38.46 38.46

Panama(2005) 41.68 55.24 3.08

Paraguay(2006) 71.99 10.94 17.07

Peru(2008) 11.73 30.31 57.95

Annex 6.4: Budget Expenditure by Economic Classification of Local Governments

Country (Most recent year) General Education Health Sanitation Transport Othersadministration (% of total) (% of total) (% of total) (% of total) (% of total)

(% of total)

Argentina(2006) 28.69 4.31 9.93 na 7.35 49.72

Bolivia(2008) 4.77 22.15 7.20 1.41 0.4 64.47

Brazil(2007) 13.6 25.99 21.94 2.97 3.05 33.6

Chile(2007) 42.84 36.86 11.89 na na 8.42

Colombia(2006) 18.56 26.51 19.58 3.76 na 31.58

Peru(2008) 30.72 9.79 16.06 na 20.08 23.35

Annex 6.5: Share of Local Government Expenditure by Functional Classification





353

Country (Most recent year) Own taxes Shared Conditional Unconditional Own taxes Shared Conditional Unconditional and fees revenues Transfers transfers/Aid and fees revenues Transfers transfers/Aid ( % of the total) ( % of the total) ( % of the total) ( % of the total) ( % of GDP) ( % of GDP) ( % of GDP) ( % of GDP)

Argentina(2006) 49.80 42.40 0.00 7.80 1.22 1.04 0.00 0.19

Bolivia(2008) 11.40 17.20 0.00 71.40 2.65 4.01 0.00 16.63

Brazil(2007) 20.10 76.50 0.00 3.40 1.75 6.67 0.00 0.29

Chile(2007) 63.00 0.00 0.00 37.00 0.66 0.00 0.00 0.39

Colombia(2006) 41.20 0.00 58.80 0.00 2.11 0.00 3.02 0.00

Dominican Republic(2006) 58.40 10.40 31.20 0.00 0.69 0.12 0.37 0.00

Ecuador(2007) 34.60 0.00 0.00 65.40 1.62 0.00 0.00 3.07

El Salvador (2007) 69.90 0.00 0.00 30.10 0.00 2.07 0.00 3.51

Guatemala(2002) 25.00 5.00 60.00 10.00 0.53 0.11 1.28 0.21

Honduras(2004) 58.10 11.00 6.00 24.90 0.95 0.18 0.10 0.41

Jamaica(2008) 100.00 0.00 0.00 0.00 0.16 0.00 0.00 0.00

Mexico(2007) 15.60 45.30 0.00 39.10 2.38 6.91 0.00 5.98

Nicaragua(2002) 44.00 5.00 11.00 40.00 0.56 0.06 0.14 0.51

Panama(2005) 49.00 46.10 4.90 0.00 0.33 0.31 0.03 0.00

Paraguay(2006) 34.10 1.80 10.80 53.30 1.23 0.07 0.39 1.91

Peru(2008) 43.20 48.30 0.00 8.50 2.62 2.92 0.00 0.52

Annex 6.6: Origin of Revenues of Local Governments (millions US$)




Source: Compiled by Mehmet S. Tosun and Serdar Yilmaz based on World Bank and UCLG consultancy missions and UCLG country datasheets.

Iran

Jordan

Lebanon

Palestine

Syria

Turkey

Yemen No formula, but Council of Ministers uses non-binding guidelines – population density,financing gap, degree of deprivation, performance in revenue collection.

Shared revenues from 28taxes, mainly – Zakat.

30% of grants from extra-budgetaryfunds that are earmarked, others are not.

According to a new law of July 2, 2008:from the overall budget tax revenue2.85% for Municipalities (other than Metropolitan)2.50% for District Municipalities of the Metropolitan, 30% of which is to be spared for Metropolitan1.15% for Special Provincial Administrationplus:5% for Metropolitan Municipalities from the tax revenue collected within the borders of theMetropolitan Cityextras:0.1% of the finally decided overall budget tax revenue is handled to the Ministry of Finance asMunicipal Balancing Payment (to be distributed through 2 payments in March and July) , of which:60% for Municipalities with a population not exceeding 5,00040% for Municipalities with a population between 5,001-9,999No conditions attached, but according to certain decided criteria such as population, acreage,number of villages belonging to the city, rural population, city development index.

National taxes such asincome and consumptiontaxes.

System based on unconditionaltransfers. Specific percentages fromthe Overall Budget Tax Revenues thatis set by the central government aredistributed among different localunits.

National amount of the transfer is ad hoc in both types of transfers.Oil revenuesCustoms taxesIncome taxesReal estate taxes

Two types of transfers. First isunconditional and second isconditional based on expropriationreimbursement

A formula for the pool and distribution of transport fees exists on paper (1997 LocalGovernment Law). Yet, the formula is only partially applied. Discretionary/emergencytransfers are ad hoc.

Transport fees and a separateaccount assigned fordiscretionary/emergencytransfers.

Current transfers and grants for transportfees are not earmarked.Discretionary/emergency transfers arechanneled to specific projects.

The Ministry of Finance (MoF) distributes to the municipalities according to a ratio (or transferformula) based on such general factors as population, proportion of its contribution in overallrevenue generation, particular importance of location and non-local responsibilities.Allocation rules decided annually by the MoIM + MoF with the approval of the centralgovernment.

Indirect taxes accumulatedinto the AutonomousMunicipal Fund. 10 % ofnational public servicerevenue (water and energy)

Transfers to all municipalities from thecentral government (autonomousMunicipal fund)

Cities and Villages Development Bank distributes to the municipalities according to a ratio(or transfer formula) based on such general factors as population, proportion of itscontribution in overall revenue generation, particular importance of location and non-localresponsibilities.

Special legislation fee of 6% onpetroleum products produced orimported (except fuel oil) by theJordan Petroleum RefineryCompany40% of the proceed of VehiclesLicensing fees.

Transfers to all municipalities fromcentral government’s fuel taxcollections

No formula. Allocation rules decided annually by the Parliament.Oil revenues.Transfers to urban LGs appear to benegligible, particularly, in the larger cities.Development transfers: 60% earmarked30-40% discretionary.

Allocation RulesSourceTransfer Characteristics

Annex 7.1: Transfer Characteristics and Allocation Rules in Selected MEWA Countries

7. MEWA



355

Million TL 2006 2007 2008

Domestic Debt: 16,520 23,577 21,666

Short Term 5,140 7,384 8,375

Long Term 11,379 16,193 13,290

Foreign Debt: 2,298 3,602 4,797

Short Term 167 984 386

Long Term 2,131 2,617 4,410

Total Debt 18,819 27,179 26,463

Domestic Debt (% of GDP) 2.18 2.80 2.28

Foreign Debt (% of GDP) 0.30 0.43 0.50

Total Debt (% of GDP) 2.48 3.22 2.78

Source: Turkish Ministry of Interior, General Directorate for Local Authorities. http://www.mahalliidareler.gov.tr/Home/Dokumanlar/faaliyet_raporu.pdf

Annex 7.2: Domestic and Foreign Debt Stock of Turkish Local Administrations


0w2010 12 BIBLIOGRAFIA DEFcarta ang 16/9/11 19:39 Página 356

BIBLIOGRAPHY



359

AfricaAfrican Center for Gender and Social Development.2009. African Perspectives on the Global Economicand Financial crisis, including the impact on health.

Agence Française de Développement, AFD. 2005.Financer les investissements des villes des pays endéveloppement. Synthèse des travaux du groupe detravail, Financement des investissement descollectivités locales. Paris, AFD.

Akindele S. T. and Olaopa. 2002. Fiscal federalismand local government finance in Nigeria. Republic ofNigeria.

Banque Africaine de Développement (BAD), leCentre de développement de l'Organisation decoopération et de développement économique(OCDE) et la Commission économique des NationsUnies pour l'Afrique. 2008. Perspectives Économiquesen Afrique. Tunis.

Bird, Richard, and François Vaillancourt. 1998. FiscalDecentralization in Developing Countries. CambridgeUniversity Press, Cambridge, 318 pages.

Boex, Jameson, and Jorge Martínez-Vázquez. 2005.Local Government Finance Reform in DevelopingCountries: The Case of Tanzania. Palgrave MacMillan.

Boex, Jamie, Randson Mwadiwa, R. Kampanje.2001. Malawi Intergovernmental Fiscal TransfersStudy. Final Report. Lilongwe and Atlanta, GeorgiaState University, UNDP.

Boschmann, Niña. 2009. Fiscal Decentralization andOptions for Donor Harmonisation. DevelopmentPartners Working Group on Local Governance andDecentralization. Berlin, Germany.

Boutayeb N. 2007. Fiscalité: principal levier demobilisation des ressources locales. Note présentéeau 7° Global Forum on Reinventing Government.Vienne, Autriche.

Brun, Jean-François, Gérard Chambas, Martial Laurent.2010. Économie politique de la réforme de transitionfiscale: le cas du Maroc. Document de travail de lasérie Études et Documents.

Busse, Mathias, Axel Bormann, Simon Mevel. 2004.The Impact of ACP/EU Economic PartnershipAgreements on ECOWAS Countries: An EmpiricalAnalysis of Trade and Budget Effects. HamburgInstitute of International Economics, Hamburg.

Chambas, Gérard. 2005. Afrique au Sud du Sahara :Mobiliser des ressources fiscales pour ledéveloppement. Ministère délégué à la Coopération,au Développement et à la Francophonie. Economica.

Champagne, Eric and Ben Mamadou Ouedraogo.2008. “Decentralization in Burkina Faso: A Policy ReformProcess in Slow Motion, Obstacles to Decentralization:Lessons from selected countries, International StudiesProgram”. Andrew Young School of Policy Studies,

International Studies Program, Georgia StateUniversity, Atlanta, September 23, 2008.

Cités et Gouvernements Locaux Unis. 2008. PremierRapport Mondial sur la Décentralisation. Paris, L’Har-mattan.

Commission Économique pour l’Afrique. 2004.Renforcer la gestion financière locale et la décen-tralisation fiscale en Afrique : Etat des lieux etproposition d’un plan d’action. Addis-Abeba.

Dafflon, Bernard, et Thierry Madies. 2009. Lesfinances publiques locales des collectivitésterritoriales au Burkina Faso. miméo AFD.

Fonds d’Équipement Communal. 2003. Noted’information, Émission obligataire réservée auxinvestisseurs institutionnels. Royaume du Maroc.

Gagnon, Gérard. 2002. Le financement dudéveloppement local ; Un état des lieux en Afrique del’Ouest, Éléments de comparaison Ghana-UEMOA.Contributions issues des ateliers de validation duchapitre Club du Sahel – OCDE, Paris.

Gilbert, Guy, Réjane Hugounenq et EmmanuelleTaugourdeau. 2009. Le système de financement descollectivités locales au Sénégal. miméo AFD.

Hinkle, Laurence, Mombert Hoppe and RichardNewfarmer. 2006. “Beyond Cotonou : EconomicPartnership Agreements in Africa”, in Trade, Doha, andDevelopment - A Window into the Issues. Chapter 22,The World Bank, Trade Department, 2006.

Hugounenq, Réjane, Yvon Rocaboy et FrançoisVaillancourt. 2010. Les finances publiques localesdes collectivités territoriales au Kenya. miméo AFD.

JICA. 2006. Local Level Service Delivery,Decentralization and Good Governance: Acomparative study of Uganda, Kenya and Tanzania :education, health and agriculture sectors. SynthesisReport, Research Group, Institute for InternationalCooperation, Tokyo, accessed 08 May 2009.

Martínez-Vázquez, Jorge and Andrey Timofeev.2008. Decentralizing Egypt: Not Just AnotherEconomic Reform. International Studies Program,Andrew Young School of Policy Studies, GeorgiaState University. Working Paper Series, at AYSPS,GSU, number paper 0833.

Mahjoub, Azzam. 2010. L’implantation de la zone delibre-échange. Situation en 2010 et perspectivespour 2010-2015. Université de Tunis El-Manar.

Meddeb, Radhi. 1996. La zone de libre-échange Tu-nisie-UE, un modèle et un défi. Chambre de com-merce franco-arabe, Paris.

Minister of Finance of the Republic of South Africa.2007. Municipal fiscal powers and functions bill.Republic of South Africa.

Odd-Helge Fjeldstad. 2001. Fiscal decentralisation inTanzania: For better or for worse? Bergen, Chr.


BBIIBBLLIIOOGGRRAAPPHHYYUnited Cities and Local Governments360

Michelsen Institute (CMI Working Paper WP2001:10) 15 p.

Organisation for Economic Co-operation andDevelopment (OECD), African Development Bankand United Nations Economic Commission for Africa.2010. African Economic Outlook 2010 - PublicResource Mobilisation and Aid.

Oriakhi, D.E. 2006. “Fiscal Decentralization, PublicExpenditure and Efficient Service Delivery: Anassessment of the performance of sub national levelsof governments in Nigeria”, in Workshop on Publicexpenditures and service delivery in Africa: Managingpublic expenditures to improve service quality andaccess. Lusaka, Zambia, 9-11 october 2006. UnitedNations Economic Commission for Africa.

Partenariat pour le Développement Municipal (PDM).2002. La Decentralisation Financière en Côted’Ivoire.

Presentations made during the Seminars on LocalFinance in Africa, organized by UCLG in Harare, 3May 2010, and Rabat, 7 May 2010 :• Akoete Massan, Emilie, Les Finances locales du

Togo.• Awal, Mouhamadou, Le Financement local au

Gabon.• Ben Akezouh, Chabane, Projet de loi portant code

communal déposé à l’A.P.N. : innovations(Algérie).

• Ben Letaieff, Mustapha, La Décentralisation fiscaleen Tunisie.

• Djacoto, Victorin, Grands traits des finances descollectivités locales en république du Bénin.

• El Madani, M., Organisation des finances locales auMaroc.

• Kebd, Yahya Ould, Atelier sur le Rapport Mondial(GOLD) : Mauritanie

• Mactar Gaye, Amadou, Le Financement local auSénégal (Ministère de l’Intérieur, Direction descollectivités locales).

• Itsana, Pacôme, Description du système definancement local au Gabon.

• Kadewere, Felix, Fiscal Decentralization in Malawi.• Kouassi, Charles Yao, La Décentralisation fiscale en

Côte d’Ivoire.• Maganlal, Mayur, State of Fiscal Decentralization in

South Africa.• Mboga, Hamisi, Comments on the African Chapter

GOLD II on Kenya.

San, Olisa Agbakoba, and Hilary Ogbonna. 2004.Local Government Administration and Developmentin Nigeria – A Capacity Building Manual for theHuman Rights Law Service. Hurilaws, Lagos, Nigeria.

Schoeman, Niek J. 2005. Rethinking FiscalDecentralization in South Africa. Department ofEconomics, University of Pretoria, South Africa. ESSAconference held in Durban, 7-9 december 2005.

Steffensen, Jesper, et Svend Trollegaard. 1999.Sénégal - Décentralisation Fiscale et Finances des

Collectivités Locales en relation avec la fournitured’infrastructures et de services.

United Nations Population Fund (UNFPA). 2007.State of World Population 2007, Unleashing thePotential of Urban Growth.

Vaillancourt, François. 2008. La décentralisation enCôte d’ivoire; état des lieux en 2007 et observations.Banque Mondiale. miméo.

Yatta, François. 2009. La décentralisation fiscale enAfrique ; enjeux et perspectives. Karthala 324pp.

Yatta, François Paul. 2006. Villes et Développementéconomique en Afrique : une approche par lescomptes économiques locaux. Editions Economica,315 pages. Paris, France.

Asia Pacific2011 to 2013 Implementation Plan of the NationalProgram for Sub-National Democratic Development,National Committee for Democratic Development,Royal Government of Cambodia, May 2010.

Amornvivat, Sutapa. 2004. Fiscal Decentralization:The Case of Thailand. Ministry of Finance. Thailand.Miméo.

An, Jongseok. 2003. Local Public Finance andIntergovernmental Fiscal Relations. Korea Instituteof Public Finance. Miméo.

Aoki, Ichiro. 2008. Decentralization andIntergovernmental Finance in Japan. PRI DiscussionPaper Series No. 08A-04. Research DepartmentPolicy Research Institute, Ministry of Finance.

Asian Development Bank. 2008. Philippines: Sourcesof Local Government Unit Financing. Sector StudyReport, Manila.

Asian Development Bank. 2009. Decentralizationand Devolution in the Philippines: Status, Triumphs,Tests, and Directions. Manila.

Association of District Development Committeesof Nepal. August 2007. State Restructuring andIssues of Local Self Governance in Nepal.

Australian Government. Assessing Local GovernmentRevenue Raising Capacity. Productivity CommissionResearch Report, April 2008.

Australian Government. Local Government NationalReports, 2005-06 to 2007-08, Department ofInfrastructure, Transport, Regional Development andLocal Government.

Australian Government. Report on GovernmentServices 2008, Steering Committee for the Review ofGovernment Services.

Azfar, O., S. Kähkönen, and P. Meagher. 2001.Conditions for effective decentralized governance: Asynthesis of research findings. Working Paper No.256. IRIS Center, University of Maryland.



361

Bahl, Roy and Jorge Martínez-Vázquez. 2009. FiscalFederalism and Economic Reform in China. Miméo.

Bahl, Roy and Musharraf Cyan. 2008. LocalGovernment Taxation in Pakistan. Miméo.

Bahl, Roy, Eunice Heredia-Ortiz, Jorge Martínez-Vázquez, and Mark Rider. 2005. India: FiscalCondition of the States, International Experience,and Options for Reform. Georgia State University.Andrew Young School of Public Policy.International Studies Program. Working Paper 05-13 (1) May.

Bahl, Roy, Sally Wallace, and Musharraf Cyan. 2008.Pakistan: Provincial Government Taxation. GeorgiaState University. Andrew Young School of PolicyStudies. World Bank. Miméo.

Bjornestad, Liv. 2009. Fiscal Decentralization, FiscalIncentives, and Pro-Poor Outcomes: Evidence fromViet Nam. ADB Economics Working Paper Series. No168. Asian Development Bank, Manila.

Boex, Jamie, Peter Grudgeon, and Roger Shotton.2002. Role of UNDP in Promoting Local Governance& Decentralization in Bangladesh. Mimeo.

Canel, Patrick, Sakon Varanyuwatana, and FrancoiseBrunet, 2007. Étude Sur Les Collectivités Locales enThaïlande. Agence Française de Développement.Miméo.

Doi, Takero. 2004. A Missing Link in DecentralizationReform in Japan: Trinity Reform Package. PRI DiscussionPaper Series No. 04A-08. Research Department PolicyResearch Institute, Ministry of Finance.

Fox, William F. and Balakrishna Menon. 2008.Decentralization in Bangladesh: Change Has BeenElusive. Mimeo.

Garcia, Beatriz Carrillo. 2008. China’s FiscalDecentralization: Consequences for Promotion ofLocal Development. Analisis, vol. 11, no. 31.

Hanai, Kiyohito and Bach Thi Minh Huyen. 2008.Revenue Assignments between Central and LocalBudgets in Vietnam. Miméo.

Hossain Md. Awal. 2006. AdministrativeDecentralization: A Framework for Discussion andIts Practices in Bangladesh. Miméo.

Interim Constitution of Nepal. 2007.

International Monetary Fund. 2002. Thailand:Selected Issues and Statistical Appendix. IMF.Washington.

JICA. 2008. Trends and Developments inDecentralization in Thailand. Miméo.

Kardar, Shahid. 2006. Local Government Finance inPakistan Post 2001. Miméo.

Keefer, Phillip E., Ambar Narayan, and TaraVishwanath. 2003. The Political Economy ofDecentralization in Pakistan. Miméo.

Law on Administrative Management of the Capital,Provinces, Municipalities, Districts and Khans, (TheOrganic Law on Decentallization and DemocraticDevelopment), 2009.

Lee, Yongmo. 2003. Fiscal Decentralization,Institutional Arrangements, and EconomicPerformance in Korea. Ph.D Dissertation. Robert F.Wagner School of Public Service. New YorkUniversity.

Lee, Yongmo. 2004. “Adequacy of Korea’s TaxAssignment: A Fiscal Federalism Approach”,International Journal of Public Administration, Vol. 9,No. 2.

Lee, Yongmo. 2006. “Decentralization and FiscalDisparities in Korea: With Special Reference to SeoulMetropolitan City”, Journal of Local GovernmentStudies, Vol. 8, No. 2.

Lewis, Blane. 2002. “Indonesia”, in P. Smoke and Y.H. Kim eds., Intergovernmental Transfers in Asia:Current Practice and Challenges for the Future.Manila, Asian Development Bank.

Lewis, Blane. 2003a. “Local Government Borrowingand Repayment in Indonesia: Does Fiscal CapacityMatter?” World Development, vol. 31, no. 6.

Lewis, Blane 2003b. “Property Taxation inIndonesia: Measuring and Explaining Administrative(Under-) Performance”, Public Administration andDevelopment, vol. 23, no. 3.

Lewis, Blane. 2003c. “Tax and Charge Creation byRegional Governments under FiscalDecentralization: Estimates and Explanations”,Bulletin of Indonesian Economic Studies, vol. 39,no. 2.

Lewis, Blane. 2005. “Indonesian Local GovernmentSpending, Taxing, and Saving: An Explanation ofPre- and Post-Decentralization Fiscal Outcomes”,Asian Economic Journal, vol. 19, no. 3.

Lewis, Blane. 2006. “Local Government Taxation: AnAnalysis of Administrative Cost Inefficiency”, Bulletinof Indonesian Economic Studies, vol. 42, no. 2.

Lewis, Blane. 2007a. “On-Lending in Indonesia:Past Performance and Future Prospects”, Bulletin ofIndonesia Economic Studies, vol. 43, no. 1.

Lewis, Blane. 2007b. “Revisiting the Price ofResidential Land in Jakarta”, Urban Studies, vol. 44,no. 11.

Lewis, Blane. 2009a. “Fiscal Decentralization inIndonesia: Accountability Deferred”. Miméo.

Lewis, Blane. 2009b. “Philippine Local GovernmentUnit Surpluses and Reserves: Facts, ComparativePerspective, Preliminary Questions, and a ResearchAgenda”. Manila. Miméo.

Lewis, Blane and Andre Oosterman. 2009. “TheImpact of Decentralization on Sub-National



Government Fiscal Slack in Indonesia”, PublicBudgeting & Finance, vol. 29, no. 2.

Lewis, Blane and Daan Pattinasarany. 2009a.“Determining Citizen Satisfaction with Local PublicEducation in Indonesia: The Significance of ActualService Quality and Governance Conditions”, (withDaan Pattinasarany) Growth and Change, vol. 40,no. 1.

Lewis, Blane and Daan Pattinasarany. 2009b. “AStochastic Cost Frontier Analysis of PrimaryEducation in Indonesia: Do Schools Need MoreMoney?”, (December) Education Economics.

Lewis, Blane and Paul Smoke. 2009. “IncorporatingSub-National Performance Incentives in theIndonesian Intergovernmental Framework”, NationalTax Association Proceedings, 101st Conference onTaxation, November.

Lewis, Blane and David Woodward. 2010. “DebtRestructuring in Indonesia: Reform or Reversion?”,Bulletin of Indonesian Economic Studies, vol. 46,no. 1.

L im, Sung I l . 2010. Recent Trends inIntergovernmental Fiscal Relations in Korea and theFuture Tasks. (publication pending)

Local Government and Shires Associations of NewSouth Wales. Independent Inquiry into FinancialSustainability of NSW Local Government: AreCouncils Sustainable? Final Report, May 2006.

Local Government New Zealand. 2006. Devolution– Fact or Fiction: A Report on the Extent of NewFunctions and Costs Acquired by Local GovernmentSince Reform in 1989.

Local Government New Zealand. July 2005.Local Authority Funding Issues: Report of theJoint Central Government/Local AuthorityFunding Project Team.

Local Government Rates Inquiry. Trends in the Levelof Local Government Funding and Expenditure,Thomas Consulting, July 2007.

Localcouncils.govt.nz, Statistical reports.

Ma Jun and John Norregaard. 1998. China’s FiscalDecentralization. Miméo.

Manasan, Rosario G. 2004. Local Public Finance inthe Philippines: In Search of Autonomy withAccountability. Philippine Institute for DevelopmentStudies. Discussion Paper Series No. 2004-42.

Manning, Nick, Doug Porter, Jackie Charlton,Musharraf Cyan, and Zahid Hasnain. 2003.Devolution in Pakistan-Preparing for Service DeliveryImprovements. Working Paper. Mimeo.

Martínez-Vázquez, Jorge. 2005. Making FiscalDecentralization Work in Vietnam. Georgia StateUniversity. Andrew Young School of Policy Studies.Working Paper Series 05-13.

Martínez-Vázquez, Jorge and Bob Searle. 2007.Fiscal Equalization: Challenges in the Design ofIntergovernmental Transfers. Springer.

Mathur, Om Prakash and George Peterson. 2006.State Finance Commissions and Urban FiscalDecentralization in India. Miméo.

Meemakshisundaram. 2005. Delivering BasicServices through Local Governments in Rural India.Miméo.

Mochida, Nobuki. 2008. Fiscal Decentralization andLocal Public Finance in Japan. Routledge.

New Zealand Government. Draft copy ofObservations and Trends from the 2009-2019 Long-term Consolidated Council Plans, Department ofInternal Affairs, Received November 2009.

New Zealand Government. Policy Statement on LandTransport Funding 2009-10 to 2018-19, Minister ofTransport, May 2009.

Oommen, M.A. 2008. Fiscal Decentralization to LocalGovernments in India. Cambridge ScholarsPublishing.

Peterson, George. 2002. Pakistan’s FiscalDecentralization: Issues and Opportunities. Miméo.

Ping Xin-Qiao and Jie Bai. 2005. FiscalDecentralization and Local Public Good Provision inChina. China Center for Economic Research WorkingPaper Series. No.E2005004.

Pradhan, Hemanta K. 2004. India: LocalGovernment Finance and Bond Markets. Miméo.

Rao, M. Govinda. 2000a. “Fiscal Decentralization inVietnam: Emerging Issues”, Hitotsubashi Journal ofEconomics, Vol. 41, no. 2.

Rao, M. Govinda. 2000b. Fiscal Decentralization inIndian Federalism. Miméo.

Rao, M. Govinda. 2003. “Fiscal Decentralization inChina and India: A Comparative Perspective”, AsiaPacific Development Journal, vol. 10, no. 1.

Rao, M. Govinda and Nirvkiar Singh. 2006. ThePolitical Economy of India’s Fiscal Federal Systemand its Reform. Miméo.

Reports on and Analysis of the Operations of theCommune/Sangkat Fund. Provided by theSecretariat of the National Committee for DemocraticDevelopment, Royal Government of Cambodia forauthor research, May 2010. Reports on and Analysisof the Operations of the Commune/Sangkar fund.

Saigal, Suraj. 2002. Literature Review on ServiceDelivery in India. Background Reference for WorldDevelopment Report 2004. Mimeo.

Shen, Chunli and Heng-fu Zou. 2007. FiscalDecentralization and Public Service Provision inChina. Miméo.



363

Shen, Chunli, Jing Jin, and Heng-fu Zou. 2006. FiscalDecentralization in China: History, Impact,Challenges, and Next Steps. World Bank. Miméo.

Shun’ichi Furukawa. 2003. Decentralization in Japan.Japan Center for International Exchange. Miméo.

Takero Doi and Shunichiro Bessho. 2003. “Japan’sReform for Decentralization: Trinity ReformPackage”. International Symposium on theLegislation of Fiscal Transfers, Ministry of Finance, P.R. China.

Weerasak, Krueathep. 2006. Fiscal Strategies of ThaiLocal Governments during the DecentralizationMovement: A Preliminary Exploration ofCharacteristics and Causes. Institute for LocalGovernment Initiatives Working Paper 01-2549.

World Bank. 2004. India State Fiscal Reforms. PREM.Report No. 28849-IN. South Asia Region. New Delhi,India.

World Bank. 2006. Spending for Development:Making the Most of Indonesia’s New Opportunities.World Bank Public Expenditure Review, Jakarta,Indonesia.

World Bank. 2007. West Bengal: Fiscal Decentralizationto Rural Governments: Analysis and Reform Options.Miméo.

World Bank. 2009. Background Notes on FiscalDecentralization. Miméo, Manila.

World Bank and Asian Development Bank. 2005.Decentralization in the Philippines: StrengtheningLocal Government Financing and ResourceManagement in the Short-Term. Manila.

Yagi, Kinnosuke. 2004. Decentralization in Japan.Policy and Governance Working Paper Series No. 30.Graduate School of Media and Governance, KeioUniversity, Japan.

Yao, Yi and Xiaobo Zhang. 2007. “Race to the Topand Race to the Bottom: Tax Competition in RuralChina”. International Food and Policy ResearchInstitute. Discussion Paper 00799.

Zaidi, S. Akbar. 2005. The Political Economy ofDecentralization in Pakistan. NCCR North SouthWorking Paper No. 1. Miméo.

Zhang, Xiaobo. 2006. Fiscal Decentralization andPolitical Centralization in China. United NationsUniversity. World Institute for DevelopmentEconomics Research. Research Paper No. 2006/93.

Zhang, Xiaobo and Jorge Martínez-Vásquez. 2003.The System of Equalization Transfers in China.Miméo.

EurasiaAlymkulov, Emil and Marat Kulatov. 2001. “LocalGovernment in the Kyrgyz Republic”, in Victor Popaand Igor Munteanu ed., Developing New Rules in

the Old Environment: Local Governments inEastern Europe, Caucasus, and Central Asia.Budapest: Local Government Initiative of the OpenSociety Institute.

Asanakunov, Bolotbek and Gulnara Bekbasarova.2009. The budgetary process and management ofmunicipal resources. Toolkit for local governments.Bishkek: The Association of Municipalities of theKyrgyz Republic (in Russian).

Assignments of Responsibilities to Different Levelsof Government in Countries of Central Asia:Prospects for Further Development. 2003.Budapest: Fiscal Decentralization Initiative forCentral and Eastern Europe.

Bakanova, Marina and Jorge Martínez-Vázquez.2007. Taking Stock of Intergovernmental FiscalRelations: Issues and Challenges. Belarus PolicyNote. World Bank, unpublished manuscript.

Caucasian Institute for Economic and SocialResearch - CIESR. 2009. Expenditure policy ofLocal self-governments in Georgia. Analyticalresearch.

Chiriac, Liubomir, Igor Munteanu, Victor Popa andVictor Mocanu. 2001. “Local Government inMoldova”, in Emilia Kandeva, ed., Stabilization ofLocal Governments: Local Governments in Centraland Eastern Europe. Budapest: Local GovernmentInitiative of the Open Society Institute.

De Silva, Migara, Galina Kurlyandskaya, YelenaAndreeva and Natalia Golovanova. 2009.Intergovernmental Reforms in Russian Federation:One Step Forward, Two Steps Back? Washington,DC: World Bank.

Decentralization and local democracy in the world.First UCLG Global Report. 2008. World Bank andUnited Cities and Local Governments.

Deryugin, Aleksandr. 2007. “The Kyrgyz Republic —Government Structure and Reform of the BudgetarySystem” in Toft, David, ed. The challenges ofdecentralization in the Kyrgyz Republic. Budapest:Fiscal Decentralization Initiative for Central andEastern Europe.

Expert Analysis of the Impact of the LocalGovernance Reform and the IntergovernmentalFiscal Relations Reform on the Financial Standing ofMunicipalities in the Russian Federation. 2009.Moscow: The Institute for Urban Economics (inRussian).

Expert-Group. 2009. Impact of the global financialcrisis on local communities in Moldova. UNDP-Moldova.

Government Finance Statistics Yearbook 2009, IMFStatistics Department.

Khachikyan, Sos. 2009. “A Simple Look on Local Self-government System in Armenia” in V. Klimanov, ed.,



Foreign Experience of Intergovernmental FiscalInnovations. Public Finance Series: Vol. 19, Moscow:IPFR.

Kirn, Tanja and Elguja Khokrishvili. 2008. Will anAsymmetrical System of Fiscal DecentralisationResolve the Conflicts in the Republic of Georgia?Finanzwissenschaftliche Diskussionsbeiträge.

Kurlandskaya, Galina, Yelena Nikolayenko, andNatalia Golovanova. 2001. “Local Governments inthe Russian Federation”, in Victor Popa and IgorMunteanu ed., Developing New Rules in the OldEnvironment: Local Governments in EasternEurope, Caucasus, and Central Asia. Budapest:Local Government Initiative of the Open SocietyInstitute.

Kurlandskaya, Galina. 2001. “Budgetary Pluralism ofRussian Authorities”. Local Government and PublicService Reform Discussion Paper Series 17.Budapest: Local Government Initiative of the OpenSociety Institute.

Kurlyandskaya, Galina, and Natalia Golovanova.2006. “Decentralization in the Russian Federation”,in Economic Change and Restructuring 39 (3):213–33.

Losaberidze, David, Konstantine Kandelaki and NikoOrvelashvili. 2001. “Local government in Georgia”, inVictor Popa and Igor Munteanu ed., Developing NewRules in the Old Environment: Local Governments inEastern Europe, Caucasus, and Central Asia.Budapest: Local Government Initiative of the OpenSociety Institute.

Losaberidze, David. 2007. “Kyrgyzstan and Georgia— Perspectives Based on Georgia’s DecentralizationExperience” in Toft, David, ed. The challenges ofdecentralization in the Kyrgyz Republic. Budapest:Fiscal Decentralization Initiative for Central andEastern Europe.

Losaberidze, David. 2009 [2]. AdministrativeStructure of Local Self-Government: Civil SectorPerspective. Caucasus Institute for Peace,Democracy and Development, Policy Brief, March.

Losaberidze, David. 2009. “Local Self-Government inGeorgia”, in Tumanyan, David, ed., Local Self-Government in South Caucasus: local finance andservices, central-local relations (in Russian).

Martínez-Vázquez, Jorge, Jamie Boex andAndrey Timofeev. 2005. “Country Report: AnAssessment of Fiscal Decentralization Reform inArmenia”, in Fiscal Decentralization in TransitionEconomies: Case Studies from the Balkans andCaucasus. UNDP.

Martínez-Vázquez, Jorge, Jamie Boex and MichaelSchaeffer. 2005. “Country Report: An Assessment ofFiscal Decentralization Reform in Georgia”, in FiscalDecentralization in Transition Economies: CaseStudies from the Balkans and Caucasus. UNDP.

Martínez-Vázquez, Jorge and Wayne R.Thirsk. 2008.Fiscal Decentralization in Ukraine: Accomplishmentsand Challenges in the Transition. unpublishedmanuscript.

Martínez-Vázquez, Jorge, Andrey Timofeev, andJameson Boex. 2006. Reforming Regional-LocalFinance in Russia. Washington, DC: World Bank.

Meruert Mahmutova. 2001. “System of LocalGovernment In Kazakhstan”, in Victor Popa andIgor Munteanu ed., Developing New Rules in theOld Environment: Local Governments in EasternEurope, Caucasus, and Central Asia. Budapest:Local Government Initiative of the Open SocietyInstitute.

Miroslav Kobasa, Alexander Karamyshev andValentin Dritz. 2001. “Local Government in Belarus”,in Victor Popa and Igor Munteanu ed., DevelopingNew Rules in the Old Environment: LocalGovernments in Eastern Europe, Caucasus, andCentral Asia. Budapest: Local Government Initiativeof the Open Society Institute.

Morozov, Boris. 2008. Property Tax in Republic ofMoldova – Basis for Local Autonomy? Unpublishedmanuscript.

Movsisyan, Vahan. 2007. The Development ofFinancial Equalization Mechanism in Armenia.Communities Finance Officers Association.

Narmania, David. 2007. Rights of Local Jurisdictionsand Tax Revenue Distribution in Georgia.Finanzwissenschaftliche Diskussionsbeiträge.

Navruzov, Yury. 2001. “Local Government inUkraine”, in Victor Popa and Igor Munteanu ed.,Developing New Rules in the Old Environment:Local Governments in Eastern Europe, Caucasus,and Central Asia. Budapest: Local GovernmentInitiative of the Open Society Institute.

Popa, Victor. 2007. “The Endless Reform of LocalSelf-government in the Republic of Moldova” in Toft,David ed., The challenges of decentralization in theKyrgyz Republic. Budapest: Fiscal DecentralizationInitiative for Central and Eastern Europe.

Problems of Reforming Local Self-governance:Structural and Financial Aspects. 2005. Consortiumfor Economic Policy Research and Advice. Moscow:IET.

Regional Peculiarities of the Municipal Reform in thePeriod of Transition. 2007. Consortium for EconomicPolicy Research and Advice. Moscow: IET.

Shergelashvili, Tengiz and David Narmania. 2006.“Fiscal Equalization in Georgia”, in Natela,Mshvidobadze ed., Fiscal Equalization in Caucasus.Budapest: Fiscal Decentralization Initiative forCentral and Eastern Europe.

Slavgorodskaya, Margarita, Tatiana Letunova, AlexeyKhrustalev. 2008. Analysis of Financial Aspects of



365

Local Self governance Reform Implementation.Moscow: IET. (in Russian)

Slukhai, Sergii and Ihor Tsekhanovskiy. 2011.“Ukrainian local finance under pressure: central andlocal governments respond” in Lucie Sedmihradska,Nadezhda Babcheva, Mihaly Lagos ed. LocalGovernment Finance in Times of Crisis: An EarlyAssessment. NISPAcee publication.

Tumanyan, David. 2010. “Local Self-government inthe Republic of Armenia: Current Developments”,inD+. Diputació Barcelona.

Tumanyan, David, David Losaberidze and MaiseGyullaliev. 2007. Local Self-Government in SouthCaucasus: current situation and Prospects. GTZ (inRussian).

Tumanyan, David ed. 2007. Local Self-governmentReforms in Armenia (2004-2006). Book 2.Communities Finance Officers Association.

Tumanyan, David. 2006. “Fiscal Equalization inArmenia” in Natela, Mshvidobadze ed., FiscalEqualization in Caucasus. Budapest: FiscalDecentralization Initiative for Central and EasternEurope.

Tumanyan, David, ed. 2009. Local Self-governmentReforms in Armenia (2007 and 2008). Book 3.Communities Finance Officers Association.

Tumanyan, David. 2009 [2]. “Local Self-Governmentin Armenia”, in Tumanyan, David ed., Local Self-Governmment in South Caucasus: local finance andservices, central-local relations (in Russian).

Veverita, Eugenia. 2006. “Fiscal Equalization: TheCase of the Republic of Moldova”, in Gabor Peteri ed.,Fiscal Equalization in South Eastern Europe.Budapest: Fiscal Decentralization Initiative forCentral and Eastern Europe.

World Development Indicators, World Bank,Washington, DC.

EuropeBergvall, Daniel, Claire Charbit, Dirk-Jan Kraan andOlaf Merk. 2006. Intergovernmental Transfers andDecentralised Public Spending. Working Paper, No. 3,OECD Network on Fiscal Relations Across Levels ofGovernment, OECD, Paris.

Bibbee, Alexandra. 2008. Making Federalism Work inItaly. Economics Department Working Paper, No.590, OECD, Paris.

Blöchliger, Hansjörg, Claire Charbit, José MariaPinero Campos and Camila Vammalle. 2010.Sub-Central Governments and the EconomicCrisis: Impact and Policy Responses. EconomicsDepartment Working Paper, No. 752, OECD,Paris.

Blom-Hansen, Jens. 2003. “Is Private Delivery ofPublic Services Really Cheaper? Evidence from Public

Road Maintenance in Denmark”, Public Choice, Vol.115, pp. 419-38.

Buettner, Thiess. 2001. “Local Business Taxation andCompetition for Capital: The Choice of the Tax Rate”,Regional Science and Urban Economics, Vol. 31, pp.215-45.

Carlsen, Frederik. 1995. “Why Is Central Regulationof Local Spending Decisions so Pervasive? Evidencefrom a Case Study”, Public Budgeting and Finance,Vol. 15, pp. 43-57.

CDLR. 2009. “The Impact of the Economic Downturnon Local Government: What is Happening and WhatCan Be Done about It?” Report to the Council ofEurope, European Committee on Local and RegionalDemocracy, Strasbourg.

CEMR. 2009a. “Balancing Democracy, Identity andEfficiency: Changes in Local and Regional Structuresin Europe”, Council of European Municipalities andRegions.

CEMR. 2009b. “The Economic and Financial Crisis:Impact on Local and Regional Authorities”, Council ofEuropean Municipalities and Regions.

CEMR. 2009c. “Social Dialogue in the Local andRegional Government Sector: An Overview”, Councilof European Municipalities and Regions.

Christoffersen, Henrik, Martin Paldam and Allan H.Wurtz. 2007. “Public versus Private Production andEconomies of Scale”, Public Choice, Vol. 130, pp.311-28.

de Mello, Luiz. 2000. “Fiscal Decentralization andIntergovernmental Fiscal Relations: A Cross-Country Analysis”, World Development, Vol. 28,pp. 365-80.

de Mello, Luiz. 2008. “Local Government Finances:The link between Intergovernmental Transfers andNet Worth”, in G.K. Ingram and Y.H. Hong eds., FiscalDecentralization and Land Policies (Cambridge, MA:Lincoln Institute).

DEXIA 2008. Sub-National Governments in theEuropean Union: Organisation, Responsibilities andFinance, DEXIA Editions, Paris.

EUROSTAT. 2009. February 2009. EuropeanCommission.

Heyndels Bruno and Jeff Vuchelen. 1998. “TaxMimicking among Belgian Municipalities”, NationalTax Journal, Vol. 51, pp. 89-101.

Hulst, Rudy and André van Montfort. 2007. Inter-Municipal Cooperation in Europe. Springer.

Magnussen, J., T.P. Hagen and O.M. Kaarboe . 2007.“Centralized or Decentralized? A Case Study ofNorwegian Hospital Reform”, Social Science andMedicine, Vol. 64, 2129-37.

OECD. 1999. Decentralising Employment Policy:Trends and Challenges. OECD, Paris.



OECD. 2003. OECD Economic Outlook, Chapter 3,OECD, Paris.

OECD. 2006. Projecting OECD Health and Long-TermCare Expenditures: What Are the Main Drivers?.OECD Economics Department Working Papers, No.477, OECD, Paris.

OECD. 2009a. OECD Employment Outlook, OECD,Paris.

OECD. 2009b. OECD Interim Economic Outlook,March, OECD, Paris.

OECD. 2009c. OECD Economic Outlook, No. 86,December, OECD, Paris.

OECD. 2010. OECD Economic Outlook, No. 87, June,OECD, Paris.

Richard, Jean-François, Henry Tolkens and MagaliVerdonck. 2002. “Dynamiques des InteractionsFiscales entre les Communes Belges 1984-97”,Économie et Prévisions, Vol. 156, pp. 1-14.

Solé-Ollé, Albert. 2003. “Electoral Accountability andTax Mimicking: The Effects of Electoral Margins,Coalition Governments and Ideology”, EuropeanJournal of Political Economy, Vol. 19, pp. 685-713.

Sutherland, Douglas, Robert Price and IsabelleJoumard. 2005. “Sub-Central Government FiscalRules”, OECD Economic Studies, Vol. 41, pp.141-81.

WHO. 2008. Decentralized Health Systems inTransition, World Health Organization. Geneva.

World Bank. 2009. World Development Indicators2009. The World Bank.

Wurzel, Eckhard. 1999. Towards More EfficientGovernment: Reforming Federal Fiscal Relations inGermany. OECD Economics Department WorkingPaper, No. 209, OECD, Paris.

Latin AmericaAbers, R. N. 2000. “Inventing Local Democracy.Grassroots Politics in Brazil”. Boulder: Lynne RiennerPublishers.

Acosta, Olga Lucía and Richard Bird. 2005. “TheDilemma of Decentralization in Colombia” in RichardBird, James Porteba and Joel Slemrod ed., FiscalReform in Colombia: Problems and Prospects, 247-286. Cambridge. MA: The MIT Press.

Afonso, José Roberto. 2006. “Decentralization andBudget Management of local Government in Brazil”.http://info.worldbank.org/etools/docs/library/233842/Afonso.pdf

Afonso, José Roberto, Érika Amorim Araújo, andGeraldo Biasoto Jr. 2005. Fiscal Space and PublicInvestments in Infrastructure: A Brazil Case Study.Texto para Discussão n. 1141. Brasília: Institute ofApplied Economic Research. http://www.ipea.gov.br/pub/td /2005/td_1141.pdf

Arze, F. y J. Martínez-Vázquez. 2004. “Descentralizaciónen América Latina desde una perspectiva depaíses pequeños: Bolivia, El Salvador, Ecuador yNicaragua”.Gestión y Política Pública, 13, 3: 619-661.

Autonomia Local, Descentralización y Desarrollo Mu-nicipal en Latinoamérica. Capitulo Latinoamericanode la Union Internacional de Municipios y Autorida-des Locales (IULA). Cuaderno del Centro Latinoame-ricano de Capacitación y Desarrollo de los GobiernosLocales, IULA/CELCADEL.

Bahl, Roy, and Johannes Linn. 1992. Urban PublicFinance in Developing Countries. Washington, D.C.World Bank and Oxford University Press.

Bahl, Roy, and Jorge Martínez-Vázquez. 2006.“Sequencing Fiscal Decentralization Reform” in R.Ebel and G. Peteri eds., Decentralization BriefingNotes. Open Society Institute. Budapest.

Bahl, Roy and Jorge Martínez-Vázquez. 2008. “TheDeterminants of Revenue Performance”, in Bahl, Roy,Martínez-Vázquez and Youngman eds., Making theProperty Tax Work: Experiences in Developing andTransitional Countries. Lincoln Institute of LandPolicies, Cambridge, Massachusetts.

Bahl, Roy, Jorge Martínez-Vázquez and JoanYoungman. 2008. Making the Property Tax Work.Lincoln Institute of Land Policy, Cambridge,Massachusetts.

Bahl, Roy, Jorge Martínez-Vázquez and JoanYoungman. 2010. Challenging the ConventionalWisdom on the Property Tax. Lincoln Institute ofLand Policy, Cambridge, Massachusetts.

Banco Central do Brasil.2000 – Focus (February 6),http://www4.bcb.gov.br/Pec/Gci/Ingl/Focus/F20000605-Tax%20Re form.pdf

Banco Mundial. Abril de 2007. Paraguay - ImpuestoInmobil iario: Herramienta Clave para laDescentralización Fiscal y el Mejor Uso de la TierraVolumen I: Informe Principal. Informe No. 37456-PA. Unidad de Reducción de la Pobreza y GestiónEconómica Región de América Latina y el Caribe.

Banco Mundial. 2004. Perú: Estudio de la Situaciónde la Deuda Municipal.

Batista Domínguez, Abel A. Panamá. RégimenJurídico del Municipio. In Régimen JurídicoMunicipal en Iberoamérica. Cienfuegos Salgado,David (coordinador). Instituto de investigacionesJurídicas, Universidad nacional Autónoma deMéxico, 2008. http://www.bibliojuridica.org/libros/6/2545/20.pdf

BID, Federacion de Municipios del IstmoCentroamericano (FEMICA), Fondo Suizo deCooperacion Tecnica/BID. Departamento Regionalde Operaciones para el Istmo Centroamericano,México, República Dominicana y Haití. Lamodernización de las finanzas municipales un pasoesencial para la consolidación institucional. Editores



367

Huáscar Eguino y Fabrice Henry. Anales delseminario realizado en la quinta reunión de la RedCentroamericana para la Descentralización y elFortalecimiento Municipal. Antigua, Guatemala 18-20 de septiembre de 1998.

BID. Agosto, 2007. Aportes para el Desarrollo delSistema Financiero Municipal: El papel de losingresos propios municipales, estudio para lospaíses del cono sur. País: Uruguay. ProyectoRegional RG-P1186. Estudio preparado por:Daniel Sureda, Consultor Coordinador delProyecto: Huáscar Eguino Especialista enDesarrollo Urbano y Municipal.

Bird, Richard M. 2003. Taxation in LatinAmerica: Reflections on Sustainability and theBalance between Equity and Efficiency.International Tax Program Papers 0306,International Tax Program, Institute forInternational Business, Joseph L. Rotman Schoolof Management, University of Toronto.

Bird, Richard M. and Enid Slack. The Role of theProperty Tax in Financing Rural Local Governments inDeveloping Countries. ITP Paper 0608, http://www.rotman.utoronto.ca /iib/ITP0608.pdf

Cardona Recinos, Rakael. 2006. “Honduras: Debili-dad estructural de la autonomía municipal” in LosGobiernos Locales y la Decentralización del Estado enCentroamérica. Diputació Barcelona.

Caribbean Update, June 1, 2002.

Castaneda, Tarsicio. 1986. Innovations in theFinancing of Education: The Case of Chile. A WorldBank Discussion Paper, Education and TrainingSeries. Washington, DC.

CEPAL. David Glejberman. Diciembre del 2005. Elsistema municipal y la superación de la pobreza yprecariedad urbana en Uruguay. División DesarrolloSostenible y Asentamientos Humanos. Santiago deChile.

Constitución de la Provincia de Córdoba (Argentina),http://www.mininterior.gov.ar /municipales/regimen/archivos/cordoba.doc

Constitución de la nación Argentina, http://www.argentina.gov.ar/argentina/portal/documentos/constitucion_nacional.pdf

Constitución de la República (Uruguay), http://www.parlamento.gub.uy/constituciones/const004.htm

Constitución de la República Bolivariana de Ve-nezuela, http://www.venezuela-oas.org/Constitucion%20de%20Venezuela.htm

Constitución Política de Colombia, http://www.banrep.gov.co/regimen/resoluciones/cp91.pdf

Constitución Política de Nicaragua y sus reformas,http://www.bcn.gob.ni/banco/legislacion/constitucion.pdf

CSIS. Katherine Bliss Peter DeShazo. Enero de 2009.El control de la corrupción en los gobiernos locales deAmérica Latina. Center For Strategic & InternationalStudies. Documentos de Política sobre las Américas.Volumen XX Estudio 1.

Daughters, R. and L. Harper. 2007. “Fiscal andPolitical Decentralization Reforms”, in E. Lora ed., TheState of State Reform in Latin America. Inter-American Development Bank, Washington, D.C.

De Cesare, Claudia, and José Francisco Lazo Marín.2008. Impuestos a los patrimonios en AméricaLatina. Serie Macroeconomía del Desarrollo No.66,Comisión Económica para América Latina y el Caribe(CEPAL), Naciones Unidas.

Deloitte Touche Tohmatsu. 2010. International TaxBusiness Guide Brazil 2010.

http://www.deloitte.com/assets/DcomGlobal/Local%20Assets/Documents/Tax/Intl%20Tax%20and%20Business%20Guides/2010/dtt_tax_guide_2010_Brazil.pdf

Dunin, Lubomir Ficinski. 2000. Decentralizing CityManagement: A Successful Experiment. World Bank,Washington, D.C.

Eaton, Kent. 2004. “Designing SubnationalInstitutions: Regional and Municipal Reforms inPostauthoritarian Chile”. Comparative PoliticalStudies, Vol. 37, No. 2, March 2004, pp. 218-44.

Faguet, Jean-Paul and Fabio Sanchez. 2006."Decentralization’s Effects on Educational Outcomesin Bolivia and Colombia”, STICERD - DevelopmentEconomics Papers 47, Suntory and ToyotaInternational Centres for Economics and RelatedDisciplines, LSE.

Faguet, Jean-Paul. 2004. "Does decentralizationincrease government responsiveness to localneeds?: Evidence from Bolivia”. Journal of PublicEconomics, Elsevier, vol. 88(3-4), 867-893,March.

French Ministry of Foreign Affairs. 2005.Decentralization Policies in Latin America.

Gamba Ladino, Julio C. Colombia. 2008. “Régi-men municipal”, in Régimen Jurídico Municipal enIberoamérica. Cienfuegos Salgado, David (coordinador).Instituto de investigaciones Jurídicas, Universidadnacional Autónoma de México. http://www.bibliojuridica.org /libros/6/2545/9.pdf

Goldfrank, Benjamin. 2006. “Los procesos depresupuesto participativo en América Latina: Exito,fracaso y cambio”. Revista de Ciencia Política PoliticalScience Journal, Political Science Institute, CatholicUniversity of Chile, 26:2 (December).

Government of Brazil. 2009. Taxes on Consumptionof Goods and Services in Brazil. https://www.receita.gov.br/publico/estudotributarios/Eventos/EspecialistasPoliticaTributaria/Taxes%20on%20Consumption



%20of%20Goods%20and%20Services%20in%20Brazil%20-%20Raimundo %20Carvalho.pdf

GPC, Grupo Propuesta Ciudadana. Abril de 2007.Vigilancia del proceso de descentralización. ReporteNacional Nro. 11, Balance 2003-2006. Participa Perú,Vigila Perú.

Gustafson, Lowell S.. 1990. "Factionalism, Centralismand Federalism in Argentina". Publius: The Journal ofFederalism, 20 (Summer): 163-176.

Imansyah, Muhammad Handry and Jorge Martínez-Vázquez. 2010. Understanding Sub-NationalGovernment Fragmentation in Indonesia. AsianDevelopment Bank and University of Indonesia.

International Budget Partnership. 2005. The LatinAmerican Index of Budget Transparency. http://www.internationalbud get.org/themes/BudTrans/LA05.htm

International Monetary Fund. January 2006.Paraguay: Report on observance of standards andcodes—Fiscal Transparency Module. IMF CountryReport No. 06/14.

KPMG. 2006. International Corporate Tax Brazil #16,(November). http://www.kpmg.com.br/publicacoes/tax /ITB/KPMG_ITB_16_Recent_Changes.pdf

Letelier, Leonardo. 2003. “La descentralizaciónfiscal en el contexto de los gobiernos regionaleschilenos”, in A diez años de la creación de losgobiernos regionales: evaluaciones y proyecciones.Santiago: Subsecretaría de Desarrollo Regional yAdministrativo.

Ley 136 de 1994 (ley municipal). Colombia.http://www.medellin.gov.co/ManualContratacion/Normatividad/LEY 20136 20de 20199 4.doc

Ley No. 40 Ley de los Municipios. Nicaragua. http://www.femica.org/archivos/codigonicaragua.pdf

Ley Orgánica de las Municipalidades. Prov. De Bue-nos Aires-Argentina. http://www.tribctas.gba.gov.ar/secciones/inf_especifica/informacionlegal/leyOrganicaMunicipales/LEY%20ORGANICA%20DE%20LAS%20MUNICIPALIDADES.doc

Ley Orgánica del Poder Público Municipal. Venezuela.http://www.fides.gob.ve/DOCS /LOPPM.doc

Ley Orgánica Municipal. Prov. De Córdoba-Argentina.http://www.villasantarosa.gov.ar/bajando_legislacion.php?id=ley_8102.pdf

Ley Orgánica Municipal 9515. Uruguay. http://www.parlamento.gub.uy/leyes/AccesoTextoLey.asp?Ley=09515&Anchor

López Ghio, María C. Argentina. “Algunos aspectosdel régimen municipal”, in Régimen Jurídico Muni-cipal en Iberoamérica. Cienfuegos Salgado, David(coordinador). Instituto de Investigaciones Jurídicas,Universidad Nacional Autónoma de México, 2008.http://www .bibliojuridica.org/libros/6 /2545/5.pdf

Martínez-Vázquez, Jorge and Juan Luis GómezReino. 2008. “El tamaño importa: la estructuravertical de gobierno y la gestión del gasto pú-blico local”. Presentado en la conferencia delCAF: “Las Promesas y las Realidades de la Des-centralización Fiscal en América Latina” Lima 4de Diciembre 2008.

Martínez-Vázquez, Jorge and Bob Searle eds., 2007.Fiscal Equalization: Challenges in the Design ofIntergovernmental Transfers. Springer.

Martínez-Vázquez, Jorge and Andrey Timofeev.2005. “Choosing between Centralized andDecentralized Models of Tax Administration”, in N.Bosch and J.M. Duran eds., Financiación, solidaridadinterterritorial y políticas tributarias de lascomunidades autónomas. Edicions i Publicacions dela Universitat de Barcelona : Barcelona (in Spanish)and Internacional Studies Working Paper #05-2,Andrew Young School of Policy Studies, GeorgiaState University, Atlanta (in English).

Ministerio de Economía y Finanzas – MEF. Glosariodel SIAF – MEF. Peru.

Ministerio de Economia y Finanzas - MEF. Boletín detransparencia fiscal - Informe especial procesos dedescentralización en Latinoamérica: Colombia,México, Chile y Perú. Perú.

Ministerio de la Presidencia. Julio 2004. Ecuador: Lospresupuestos de los gobiernos seccionales y laAsamblea Constituyente de 1997-8 Silke Lechner.Unidad de Coordinación de la AsambleaConstituyente - Componente Descentralización yAdministración Pública. La Paz.

Ministry of Finance. n.d. The Brazilian Tax System.Ministry of Finance, Brasília. http://www.receita.fazenda.gov.br/principal/Ingles/SistemaTributarioBR/BrazilianTaxSystem/default.htm

Nassuno, M. 2006. Burocracia e Participao: AExperienca Do Orcamento Participativo Em PortoAlegre. Tese de doutorado. Universidade de Brasilia.

Nickson, R. Andrew. 1995. Local Government in LatinAmerica, Lynne Rienner Publishers, Boulder.

Oxford Analytica. December 2006. Peru FiscalTransparency. Country Report 2006.

Oxford Analytica. December 2006. Venezuela FiscalTransparency. Country Report 2006.

Parry, Taryn Rounds. 1997. “Achieving Balance inDecentralization: A Case Study of EducationDecentralization in Chile”. World Development, Vol.25, No. 2, pp. 211-25.

Peterson, George. 1996. Building Local CreditSystems. UNDP/UNCHS (Habitat) /World.

Porto, Alberto. 2009. Perfiles del sector públicomunicipal en países del cono sur y de la regiónandina. Banco Interamericano de Desarrollo,Washington, D.C.



369

Purohit, Mahesh C. 1997. “Value Added Tax in aFederal Structure: A Case Study of Brazil”. Economicand Political Weekly, Vol. 32, No. 7 (Feb. 15-21,1997), pp. 357-362, The World Bank, Washington,D.C.

Rezende, Fernando, and José Roberto Afonso. 2006.“The Brazilian Federation: Facts, Challenges, andProspects”, in Jessica Wallack and T. N. Srinivasaned., Federalism and Economic Reform: InternationalPerspectives, 143–88. New York: CambridgeUniversity Press.

Rotondo Tornaría, Felipe. 2008. Uruguay. El munici-pio. En Régimen jurídico municipal en Iberoamérica.Cienfuegos Salgado, David (coordinador). Institutode Investigaciones Jurídicas, Universidad NacionalAutónoma de México. http://www.bibliojuridica.org/libros/6/2545/23.pdf

Santos, B. (org.). 2004. Democratizar a Democracia.Os Caminhos da Democracia Participativa. Rio deJaneiro: Civilização Brasileira.

Saravia Aldana, Mario Danilo. 2008. “Nicaragua.Régimen del municipio”. In Régimen Jurídico Mu-nicipal en Iberoamérica, Cienfuegos Salgado,David (coordinador). Instituto de investigacionesJurídicas, Universidad nacional Autónoma deMéxico. http://www.bibliojuridica.org/libros/6/2545/19.pdf

Sepulveda, Cristian and Jorge Martínez-Vázquez.2009. Explaining Property Tax Collections inDeveloping Countries: The Case of Latin America.Paper prepared for CEPAL.

Shah, Anwar (editor). 2006. Public expenditureanalysis for citizen-centered governance. The WorldBank Washington D.C.

Shah, Anwar (editor). 2006. Local public financialmanagement. The World Bank Washington D.C.

Shah, Anwar (editor). 2006. Local governance indeveloping countries. The World Bank WashingtonD.C.

Shah, Anwar (editor). 2007. Participatory Budgeting.Washington, DC: World Bank Institute, Public SectorGovernance and Accountability Series.

Sirtaine, S. 2005. Financing Sources forInfrastructure Projects in LAC. World Bank,Washington, D.C.

Smoke, Paul, Eduardo J. Gomez, and George E.Peterson (Editors). 2006. Decentralization in AsiaAnd Latin America: A Comparative InterdisciplinaryPerspective. Studies in Fiscal Federalism and State-Local Finance, Edgard Elgar.

Subsecretaria de Desarrollo Regional y Administrativo.2007. Memoria de la Descentralización. Gobierno deChile.

Tulchin, Joseph S. and Andrew Selee. 2004.Decentralization and Democratic Governance in Latin

America. Woodrow Wilson Center Report on theAmericas #12. Washington, D.C.

United Cities and Local Governments. 2008.Decentralization and Local Democracy in the World:First Global Report. World Bank and CGLU.

Uribe, Maria C. Bogota’s Betterment Levy.Conference on Unlocking Urban Land Values forInfrastructure Finance, Bangalore (India) September14th, 2009. http://indiausp.org /brookings/Bogota-Betterment-Levy.pdf

Villegas Moreno, José L. Venezuela. 2008. “Elmunicipio. Aproximación a su configuración en elnuevo ordenamiento constitucional y legal”, inRégimen Jurídico Municipal en Iberoamérica.Cienfuegos Salgado, David (coordinador). Institutode Investigaciones Jurídicas, Universidad NacionalAutónoma de México. http://www.bibliojuri dica.org/libros/6/2545/24.pdf

Von Haldenwang, Christian, Elke Büsing,Katharina Földi, Tabea Goldboom, FerdinandJenrich and Jens Pulkowski. 2009. Administracióntributaria municipal en el contexto del proceso dedescentralización en el Perú. Deutsches Institutfür Entwicklungspolitik.

World Bank. 1998. Impact Evaluation Report,Building Institutions and Financing LocalDevelopment: Lessons from Brazil and thePhilippines.

World Bank. 1999. World Development Report1998/99. Knowledge for Development. Summary.Washington: Oxford University Press.

Middle East West Asia Adib, Mustapha. 2008. “Middle East/ Western Asia”,in Decentralization and Local Democracy in theWorld, First Global Report. United Cities and LocalGovernments: 200-229.

Al-Hajaj, Muaffak. 2010. Director of Follow up andEvaluation, Ministry of Public Sector Development“Local Government Finance in Jordan” Presentationduring the UCLG organized workshop in Istanbul 15march 2010.

Al-Hajjeh, Tamer. 2009. “Development of theLocal Administration Structure in Syria”, in EuroSyrian Cities Congress. Minister of LocalAdministration.

Arzhagi, Mohammed and J. Vernon Henderson.2005. “Why Countries are FiscallyDecentralizing”, in Journal of Public Economics,89, 1157-1189.

Atallah, Sami. 1999. The legal and administrativedimensions of municipalities in Lebanon,Interventions during CERMOC symposium in Beirut17 April 1999, Cermoc 1999 document nº4,Municipal finance in Lebanon (as referenced byBassil, Najwa & Karam Karam. 2009)



Barkey, Karen. 2008. Empire of Difference: TheOttomans in Comparative Perspective. New York, NY:Cambridge University Press.

Bassil, Najwa & Karam Karam. 2009. Local Finance inLebanon. Study realized for UCLG/BTVL.

Beck, Thorston, George Clarke, Alberto Groff,Phillip Keefer and Patrick Walsh. 2001. "NewTools in Comparative Political Economy: TheDatabase of Political Institutions”, The WorldBank Economic Review, 15 (1): 165-176.

Doherty, Pat. 2005. Municipal AdministrationModernization Project in Syria: Report onFinancial Resources Management, Local Taxes /Duties, Evaluation of Current Position, Proposalsfor interventions. Municipal AdministrationModernization Programme (MAM).

Ebel, Robert D. and Serdar Yilmaz. 2003. “Onthe Measurement and Impact ofDecentralization”, in Jorge Martínez-Vázquezand James Alm, Public Finance in Developingand Transitional Countries: Essays in Honor ofRichard Bird. pp 101-119. Cheltenham, UK:Edward Elgar.

Ecevit, Bulent. 1990. Mithat Pasa ve TurkEkonomisinin Tarihsel Sureci – Mithat Pasha andHistorical Continuum of the Turkish Economy,Demokratik Sol Parti. Istanbul: Tekin Yayinevi.

Economic Research Forum (ERF). 2005. JordanCountry Profile: The Road Ahead for Jordan(August). Cairo, Egypt: the Economic ResearchForum and Institut de la Mediterranee.

Hammond, George and Mehmet S. Tosun. 2011.“The Impact of Local Decentralization onEconomic Growth: Evidence from U.S.Counties”, Journal of Regional Science. 51 (1):47-64.

Hanna, Nelly. 1995. The State and its Servants:Administration in Egypt from Ottoman Times tothe Present. Cairo, Egypt: The AmericanUniversity in Cairo Press.

Inalcik, Halil. 1977. “Centralization andDecentralization in Ottoman Administration”, inThomas Naff and Roger Owen ed., Studies inEighteenth-Century Islamic History. Carbondale:Southern Illinois University Press, 27-52.

Jordan Ministry of Municipal Affairs. 2009.Jordan Resource Transfer Pool and Allocation toMunicipalities Bulletin, No. 01. RegionalPlanning/Development and InternationalPrograms Department, September 15.

Keefer, Philip. 2007. DPI2006, Database of PoliticalInstitutions: Changes and Variable Definitions.http://siteresources.worldbank .org/INTRES/Resources/469232-1107449512766/dpi2006_vote_share_variable_definitions.pdf

Milton-Edwards, Beverly and Peter Hinchcliffe. 2004.Conflicts in the Middle East Since 1945. New York,NY: Routledge.

Municipal Administration Modernization Syria (n.d.).Retrieved from http://www.mam-sy. org/index.php?p_id=11&lang=en.

OECD. 2004. OECD Economic Surveys Turkey.Volume 2004/15-December. OECD.

Saliba, Najib E. 1978. “The Achievements of MidhatPasha as Governor of the Province of Syria, 1878-1880”, International Journal of Middle East Studies 9(3): 307-323.

Tajbakhsh, Kian. 2000. “Political Decentralizationand the Creation of Local Governments in Iran:Consolidation or Transformation of the TheocraticState?”, Social Research 67 (2): 210-222.

Tosun, Mehmet S. and Arzu Sen. 2008. “PreventingRegional Conflicts as a Global Public Good: the Caseof the MENA Region”. manuscript.

Tosun, Mehmet S. and Serdar Yilmaz. 2010a.“Centralization, Decentralization and Conflict in theMiddle East and North Africa”, Middle EastDevelopment Journal 2 (1):1-14.

Tosun, Mehmet S. and Serdar Yilmaz. 2010b.“Decentralization, Economic Development, andGrowth in Turkish Provinces”, Emerging MarketsFinance and Trade 46 (4): 23-44.

Tosun, Mehmet S. and Serdar Yilmaz. 2008.Centralization, Decentralization and Conflict in theMiddle East and North Africa. World Bank PolicyResearch Working Paper No.4774. Washington, D.C.:The World Bank.

Turkish Ministry of Interior, General Directorate forLocal Authorities. 2009. Mahalli Idareler GenelFaaliyet Raporu, 2008. http://www.mahalli-idareler.gov.tr/Home /Dokumanlar/faaliyet_raporu.pdf.

UCLG. 2007. UCLG Country Profiles: The HashemiteKingdom of Jordan. Retrieved from http://www.cities-localgovernments.org /gold/Upload/country_profile/Jordan.pdf

UNDP Programme on Governance in the ArabRegion (POGAR). 2010a. Country Theme: LocalGovernment: Jordan. Retrieved from http://www.pogar.org/countries/theme.aspx?t=6&cid=7

UNDP Programme on Governance in the ArabRegion (POGAR). 2010b. Country Theme: LocalGovernment: Lebanon. Retrieved from http://www.pogar.org/countries/theme.aspx?t=6&cid=9

UNDP Programme on Governance in the ArabRegion (POGAR). 2010c. Country Theme: LocalGovernment: Saudi Arabia. Retrieved from http://www.pogar.org/countries/theme .aspx?t=6&cid=16

World Bank. 2006. West Bank and Gaza:Intergovernmental Fiscal Relations and Municipal



371

Finance Policy Note. Report Number. 36519-WBG,World Bank, Washington, DC.

North AmericaAmerican Society of Civil Engineers. 2009. 2009Report Card for America’s Infrastructure, March 25.

Anderson, John, 1990. “Tax Increment Financing:Municipal Adoption and Growth”, National TaxJournal, 155-164.

Augustine, Nancy Y., Michael E. Bell, David Brunori,and Joan M. Youngman eds., 2009. Erosion of theProperty Tax Base, Trends, Causes, and Consequences.Cambridge, MA: Lincoln Institute of Land Policy.

Bartik, Timothy J. 1991. Who Benefits from Stateand Local Economic Development Policies?Kalamazoo, MI: W.E. Upjohn Institute.

Benton, J. Edwin. 2009. “Trends in LocalGovernment Revenues: The Old, The New, andThe Future”, presented at the 2009 LincolnInstitute of Land Policy Conference, The ChangingLandscape of Local Public Revenues. Cambridge,MA: June 2009.

Besley, Timothy and Harvey Rosen. 1998. “States’Responses to Federal Tax Setting: Evidence fromGasoline and Cigarettes”, Journal of Public Economics73.

Bird, Richard, M. and Enid Slack. 2003. InternationalHandbook on Land and Property Taxation.Cheltenham, UK: Edward Elgar.

Bird, Richard M. and Michael Smart, 2002.Intergovernmental Fiscal Transfers: Internationallessons for developing countries. WorldDevelopment, 30.

Bird, Richard, M. and Almos Tassonyi. 2001."Constraints on Provincial and Municipal Borrowing inCanada: Markets, Rules, and Norms”. CanadianPublic Administration, 44.

Bird, Richard M. and Almos Tassonyi. 2003.“Constraining Subnational Fiscal Behavior inCanada: Different Approaches, Similar Results?” inRodden, Jonathan A., Gunnar S. Eskeland, andJennie Litvack eds., Fiscal Decentralization and theChallenge of Hard Budget Constraints. Cambridge,MA: MIT Press.

Bird, Richard M., and Francois Vaillancourt 1998.“Fiscal Decentralization in Developing Countries: AnOverview”, in Richard M. Bird and FrancoisVaillancourt ed., Fiscal Decentralization in DevelopingCountries. Cambridge, UK: Cambridge UniversityPress.

Bruce, Donald, William F. Fox, and LeAnn Luna.2009. “State and Local Government Sales TaxRevenue Losses from E-Commerce”, State TaxNotes, May 18, 2009.

Brunori, David. 2003. “Local Tax Policy: A FederalistPerspective”. Washington, DC: The Urban InstitutePress.

Courchene, Thomas J. 2005. Citistates and the Stateof Cities: Political-Economy and Fiscal-FederalismDimensions. Montreal: Institute for Research onPublic Policy.

Courchene, Thomas, J. 2004. “Hourglass Federalism– How the Feds Got the Provinces to Run Out ofMoney in a Decade of Liberal Budgets”. PolicyOptions, April, 2004.

Esteller-More, Alex and Albert Sole-Olle. 2001.“Vertical Income Tax Externalities and FiscalInterdependence: Evidence from the U.S”., RegionalScience and Urban Economics 31: 247-72.

Federation of Canadian Municipalities. 2008.Innovative Mechanisms for Fiscal Transfers toMunicipalities – The Canadian Experience inMunicipal Financing. Ottawa: FCM.

Gallo, Theresa. 2004. “History and Evaluation of theUnfunded Mandates Reform Act”, National TaxJournal, 57.

Garcia-Mila, Teresa, and Therese J. Mc Guire. 2002.Tax Incentives and the City. Brookings-WhartonPapers on Urban Affairs.

Gendron, Pierre-Pascal. 2005. “Value Added Treatmentof Public Sector Bodies and Non-Profit Organizations:A Developing Country Perspective”, Bulletin ofInternational Fiscal Documentation, 59 (12).

Glaeser, Edwin. 2002. Comment on Tax Incentivesand the City, Brookings-Wharton Papers on UrbanAffairs.

Harchaoui, Tarek, M., Faouzi Tarkhani, and PaulWarren. 2004. “Public Infrastructure in Canada,1961-2002”, Canadian Public Policy, 30 (3).

Hoene, Christopher and Michael A. Pagano 2008.Cities and State Fiscal Structure, Research Report onAmerica’s Cities. Washington, DC: National Leagueof Cities.

Karkalakos, Sotiris and Christos Kotsogiannis. 2007.“A Spatial Analysis of Provincial Corporate IncomeTax Responses: Evidence from Canada”, CanadianJournal of Economics 40.

Kitchen, Harry, M. 2006. A State of Disrepair: How toFix the Financing of Municipal Infrastructure inCanada. Toronto: C.D. Howe Institute.

Kitchen, Harry M. and Enid Slack, “Special Report:New Finance Options for Municipal Governments”.Canadian Tax Journal, Volume 51, Number 6, 2003.

Merriman, David and Mark Skidmore. 2000. “DidDistortionary Sales Taxation Contribute to the Growthof the Service Sector”, National Tax Journal, March.

Mikesell, John. 2010. “State Sales Taxes in the GreatRecession”, State Tax Notes, July 19, 2010.



Mintz, Jack M. and Tom Roberts, 2006. Running onEmpty: A Proposal to Improve City Finances,Toronto: C.D. Howe Institute.

Mirza, Saeed. 2007. Danger Ahead: The ComingCollapse of Canada’s Municipal Infrastructure. Areport prepared for the Federation of CanadianMunicipalities, November 2007.

O'Sullivan, Arthur, Terri A. Sexton, and Steven M.Sheffrin. 1995. Property Taxes and Tax Revolts: TheLegacy of Proposition 13. New York: CambridgeUniversity Press.

Ring, Raymond. 1999. “Consumer’s share andproducer’s share of the general sales tax” in NationalTax Journal, 52, 79-90.

Sexton, Terri A. 2009. “Assessment Limits as aMeans of Limiting Homeowner Property Taxes”, InAugustine, Nancy, Y., Michael E. Bell, DavidBrunori, and Joan M. Youngman eds., Erosion ofthe Property Tax Base, Trends Causes, andConsequences. Cambridge, Mass.: LincolnInstitute of Land Policy.

Simeon, Richard and Martin Papillon. 2006“Canada”, in Akhtar Majeed, Ronald L. Watts, andDouglas M. Brown eds., Distribution of Powers andResponsabilities in Federal Countries. GlobalDialogue on Federalism, Vol. 2. Montreal McGill-Queen’s University Press. 91-122.

Slack, Enid. 2003. Municipal Finance and the Patternof Urban Growth. Toronto: C.D. Howe Institute.

Slack, Enid. 2005. “Municipal Financing of CapitalInfrastructure in North America”, Journal of PropertyTax Assessment and Administration, 2(1).

TD Economics. 2004. “Mind the Gap: Finding theMoney to Upgrade Canada’s Aging Infrastructure”,Toronto, May.

TD Economics. 2004. “Mind the Gap: Finding theMoney to Upgrade Canada’s Aging Infrastructure,Toronto”, 2004.

Vander Ploeg, Casey. 2004. "No Time to be Timid:Addressing Infrastructure Deficits in the Western BigSix”, Western Cities Project Report #30, Calgary:Canada West Foundation, 2004.

Wassmer, Robert W. 2007. Increasing Use ofProperty Tax Abatement as a Means of PromotingSub-National Economic Activity in the United States.Sacramento, CA: California State University.

Wassmer, Robert. 1990. “Can Local Incentives Alter aMetro City’s Economic Development?” UrbanStudies, 1994: 1251-1278.

Wasylenko, Michael, J. 1997. “Taxation and EconomicDevelopment: The State of the Economic Literature”,New England Economic Review, March/April.

Winters, John. 2008. "An Overview of Property TaxLimitations”. State Tax Notes.

Youngman, Joan. 1999. "The Hardest Challenge forValue-Based Property Taxes: Part I”. State Tax Notes,March 1999.

Yuan, Bing, Joseph Cordes and David Brunori. 2009.“Tax and Expenditure Limitations and Local PublicFinances”, in Augustine, Nancy Y., Michael E. Bell,David Brunori, and Joan M. Youngman eds., Erosionof the Property Tax Base, Trends, Causes, andConsequences. Cambridge, MA: Lincoln Institute ofLand Policy.

Financing Metropolitan Areas Acosta, Olga Lucia and Richard Bird. 2005. “TheDilemma of Decentralization in Colombia” inRichard Bird, James Porterba and Joel Slemroded., Fiscal Reform in Colombia: Problems andProspects, 247-286. Cambridge, MA: The MITPress.

Bahl, Roy. 2010. Financing Service Delivery inMetropolitan Areas. International Studies Program,Andrew Young School of Policy Studies, GeorgiaState University, working paper.

Bahl, Roy and Johannes Linn. 1992. Urban PublicFinance in Developing Countries. New York: OxfordUniversity Press.

Bahl, Roy and David Solomon. 2003. “TheRegional Service Council Levy”, in Roy Bahl andPaul Smoke ed., Restructuring Local GovernmentFinance in Developing Countries: Lessons fromSouth Africa, 127-172. Cheltenham UK: EdwardElgar.

Bahl, Roy and Richard Bird. 2008. “SubnationalTaxes in Developing Countries: The Way Forward”Public Budgeting & Finance, 28(4):1-25.

Benjamin, Gerald and Richard P. Nathan. 2001.Regionalism and Realism: A Study of Governmentsin the New York Metropolitan Area. Washington:Brookings Institution Press.

Bird, Richard and Enid Slack. 2004. Fiscal Aspects ofMetropolitan Governance. Joseph T. Rotman Schoolof Management, ITP Paper 0401, Toronto: Universityof Toronto.

Bird, Richard and Eric Zolt. 2005. “Redistribution viaTaxation: The limited Role of the Personal IncomeTax in Developing Countries”, UCLA Law Review,August, 52 (6): 1627-1695.

Bordignon, Massimo, Silvia Giannini, and PaoloPanteghini. 2001. “Reforming business Taxation:Lessons from Italy?”, International Tax and PublicFinance, March, 8(2): 191-210.

Campbell, Alan K. and Guthrie Birkhead. 1976.“Municipal Reform Revisited: The 1970sCompared with the 1920s”, in Alan Campbell andRoy Bahl ed., State and Local Government: ThePolitical Economy of Reform, 1-15. New York: TheFree Press.



373

Chernik, Howard and Andrew Reschovsky. 2006.“Local Public Finance: Issues for MetropolitanRegions” in Competitive Cities in the GlobalEconomy. Paris: OECD Territorial Reviews, 415-432.

De Cesare, Claudia M. 2004. “General Characteristicsof Property Tax Systems in Latin America”. Paperpresented at 7th International Conference onOptimizing Property Tax Systems in Latin America,Guadalajara, Jalisco, Mexico.

Diaz-Cayeros, Alberto and Charles McLure. 2000.“Tax Assignment”, in Marcelo Giugale and StevenWebb ed., Achievements and Challenges of FiscalDecentralization, 177-196. Washington: The WorldBank.

Fu, Gao. 2007. “Urban InfrastructureInvestment and Financing in Shanghai” inGeorge Peterson and Patricia Annez ed.,Financing Cities: Fiscal Responsibility and UrbanInfrastructure in Brazil, China, India, Polandand South Africa, chapter 6. Los Angeles andWashington DC: Sage Publications and TheWorld Bank.

International Monetary Fund Various Issues.Government Finance Statistics, Washington DC:International Monetary Fund.

Jouve, Bernard and Christian Lefevre, eds. 2002.Local Power, Territory and Institutions in EuropeanMetropolitan Regions London: Frank Cass.

Lotz, Jorgen 2006. “Local Government Organizationand Finance: Nordic Countries” in Anwar Shah ed.,Local Governance in Industrial Countries, 223-264.Washington DC: World Bank.

Martínez-Vázquez, Jorge. 2008. “RevenueAssignments in the practice of FiscalDecentralization”, in Nuria Bosch and JoseDuran, eds. Fiscal Federalism and PoliticalDecentralization: Lessons from Spain, Germanyand Canada. Edward Elgar Cheltenham, UK andNorthampton, M.A.

Mathur, Om, Debdulal Thakur, Nilesh Rajadhyakshaand Roy Bahl. 2009. Urban Property Tax Potential inIndia. New Delhi: National Institute of Public Financeand Policy.

McLure, C.E. 1998. “The Revenue AssignmentProblem: Ends, means and constraints”. PublicBudgeting, Accounting and Financial Management.Winter.

Mohanty, P.K., B.M. Misra, Rajan Goyal, and P.D.Jeromi. 2007. “Municipal Finance in India: AnAssessment”, Study No. 28. Development ResearchGroup, Department of Economic Analysis and Policy,Mumbai: Reserve Bank of India.

Musgrave, Richard A. 1983. "Who Should Tax, Whereand What?" in Charles E. McLure, Jr., ed., TaxAssignment in Federal Countries. Canberra: Centre

for Research on Federal Financial Relations,Australian National University.

Oates, Wallace. 1972. Fiscal Federalism New York:Harcourt Brace Jovanovich.

OECD. 2003. “The Metropolitan Region ofMelbourne, Australia”, OECD Territorial Reviews.

OECD. 2004. “Busan, Korea”, OECD TerritorialReviews.

OECD. 2004a “Mexico City”, OECD Territorial Reviews.

OECD. 2006. “Competitive Cities in the GlobalEconomy”, OECD Territorial Reviews.

OECD. 2006a. “Stockholm, Sweden”, OECDTerritorial Reviews.

OECD. 2006b. “Milan, Italy”, OECD TerritorialReviews.

OECD. 2007. “Madrid, Spain”, OECD TerritorialReviews.

OECD. 2007a. “Randstadt Holland, Netherlands”,OECD Territorial Reviews.

OECD. 2008. “Cape Town, South Africa”, OECDTerritorial Reviews.

OECD. 2008a. “Istanbul, Turkey”, OECD TerritorialReviews.

OECD. 2009. “Copenhagen, Denmark”, OECDTerritorial Reviews.

OECD. 2009a. “Toronto, Canada”, OECD TerritorialReviews.

Phares, Don. 2009a. “Prologue on MetropolitanGovernment and Governance” in Don Phares ed.,Governing Metropolitan Regions in the 21st Century.London: Sharpe, xi-xvii.

Rezende, F. and S. Garson. 2006. "FinancingMetropolitan Areas in Brazil”, Revista de economiacontemporanea, 10(1): 5-34.

RTI International. 2005. “Case Study on Mobilizationof Private Capital in Bogota, Colombia”, ResearchTriangle Park, North Carolina: RTI.

Slack, Enid. 2000. “A Preliminary Assessment of theNew City of Toronto”, Canadian Journal of RegionalScience Spring, 2000 XXIII: 1, 13-29.

Slack, Enid. 2007. Managing the Coordination ofService Delivery in Metropolitan Cities: The Roleof Metropolitan Governance. Policy ResearchWorking paper 4317 Washington: The WorldBank.

Taliercio, Robert. 2005. “Subnational Own-Source Revenue: Getting Policy andAdministration Right” in Roland White and PaulSmoke ed., East Asia Decentralizes: MakingLocal Government Work, 107-128. Washington:The World Bank.



Togo, Hisatake. 1995. “The MetropolitanStrategies of Tokyo: Toward the Restoration ofBalanced Growth” in L.J. Sharpe ed., TheGovernment of World Cities: The Future of theMetro Model, 177-202. Chichester, UK: Wiley.

Tokyo Metropolitan Government: 2006. 2010.http://www.metro.tokyo.jp/ENGLISH/PROFILE/index.htm

United Nations. 2008. World Urbanization Prospects:The 2007 Revision, New York: United Nations.

Webster, D. 2000. “Financing City Building: TheBangkok Case Discussion Paper” Asia PacificResearch Center, Institute for International Studies,Stanford University.

World Bank. 2007. Brazil Sao Paulo: Inputs for aSustainable Competitive City Strategy. Report No.37324-BR Washington: The World Bank.

World Bank. 2009. Reshaping Economic Geography,World Development Report Washington: The WorldBank.


Local Government Finance: The Challenges of the 21st Century

The 2nd Global Report of United Cities and Local Governments (UCLG) on Decentralization and Local Democracy describes and

analyzes the architecture of fiscal decentralization in more than one hundred countries as well as in major metropolitan areas. In

the majority of these countries, local authorities have been taking on more and more responsibilities for public investment and

the provision of services that are essential for both economic development and the well being of their citizens.

If increasing fiscal decentralization has been a global trend in recent decades, there are significant variations across and within

regions and countries. Local budgets make up on average 25% of public expenditure in the countries of the European Union but

less than 5% in many developing countries. Although revenue and expenditure autonomy have both generally increased, this has

occurred unevenly across countries and in many cases is stronger on the expenditure side. Revenue sources too often remain

limited and uncertain, especially for small and middle size cities.

Accelerating urbanization and important shifts in the global context (climate change, increasing risk of natural disaster,

migration and demographic changes, among others) present both threats and opportunities for local governments in the years

ahead. Local governments need to be able to respond effectively to the urbanization of poverty, growing investment

requirements and other pressing needs. The economic and financial crisis that began in 2008 only made the situation more

difficult and urgent.

This report, offered as part of UCLG’s ongoing commitment to deepen understanding of the level of government closest to

citizens, identifies universal, as well as more regional and country specific challenges facing local government finance. Some of

the most common and consequential, beyond inadequate sources of revenue include, unpredictable and problematically

structured transfers and grants, excessive budget controls imposed by higher-level governments, and unfunded mandates. In

addition to analyzing challenges and outlining opportunities that local governments could take advantage of, the Report

proposes recommendations for national and local policy reforms and innovations intended to strengthen the fiscal role and

performance of local governments around the world.

United Cities and Local GovernmentsCites et Gouvernements Locaux UnisCiudades y Gobiernos Locales Unidos

Generalitat de Catalunya


II_ENG.pdf · DISCLAIMER "The terms used concerning the legal status of any country, territory,...

Documents

Transcript of II_ENG.pdf · DISCLAIMER "The terms used concerning the legal status of any country, territory,...