Fiscal Decentralization and Local Finance in Developing Countries: Development from Below [1 ed.] 1786435292, 9781786435293

Citation preview

Fiscal Decentralization and Local Finance in Developing Countries

M4439-BAHL_9781786435293_t.indd 1

01/03/2018 11:08

STUDIES IN FISCAL FEDERALISM AND STATE–LOCAL FINANCE Series Editor: Jorge Martinez-Vazquez, Regents Professor of Economics and Director, International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University, USA This important series is designed to make a significant contribution to the development of the principles and practices of state–local finance. It includes both theoretical and empirical work. International in scope, it addresses issues of current and future concern in both East and West and in developed and developing countries.   The main purpose of the series is to create a forum for the publication of high-quality work and to show how economic analysis can make a contribution to understanding the role of local finance in fiscal federalism in the twenty-first century.   Titles in the series include: The Political Economy of Financing Scottish Government Considering a New Constitutional Settlement for Scotland C. Paul Hallwood and Ronald MacDonald Does Decentralization Enhance Service Delivery and Poverty Reduction? Edited by Ehtisham Ahmad and Giorgio Brosio State and Local Fiscal Policy Thinking Outside the Box? Edited by Sally Wallace The Political Economy of Inter-Regional Fiscal Flows Measurement, Determinants and Effects on Country Stability Edited by Núria Bosch, Marta Espasa and Albert Solé-Ollé Decentralization in Developing Countries Global Perspectives on the Obstacles to Fiscal Devolution Edited by Jorge Martinez-Vazquez and François Vaillancourt The Challenge of Local Government Sizes Theoretical Perspectives, International Experience and Policy Reform Edited by Santiago Lago-Peñas and Jorge Martinez-Vazquez State and Local Financial Instruments Policy Changes and Management Craig L. Johnson, Martin J. Luby and Tima T. Moldogaziev Taxation and Development: The Weakest Link? Essays in Honor of Roy Bahl Edited by Richard M. Bird and Jorge Martinez-Vazquez Multi-level Finance and the Euro Crisis Causes and Effects Edited by Ehtisham Ahmad, Massimo Bordignon and Giorgio Brosio Fiscal Decentralization and Budget Control Laura von Daniels The Future of Federalism Intergovernmental Financial Relations in an Age of Austerity Edited by Richard Eccleston and Richard Krever Fiscal Decentralization and Local Finance in Developing Countries Development from Below Roy Bahl and Richard M. Bird

M4439-BAHL_9781786435293_t.indd 2

01/03/2018 11:08

Fiscal Decentralization and Local Finance in Developing Countries Development from Below

Roy Bahl Professor Emeritus, Andrew Young School of Policy Studies, Georgia State University, Atlanta, Georgia, USA and Professor Extraordinarius, University of Pretoria, South Africa

Richard M. Bird Professor Emeritus, Rotman School of Management and Senior Fellow, Institute on Municipal Finance and Governance, Munk School of Global Affairs, University of Toronto, Canada

STUDIES IN FISCAL FEDERALISM AND STATE–LOCAL FINANCE

Cheltenham, UK • Northampton, MA, USA

M4439-BAHL_9781786435293_t.indd 3

01/03/2018 11:08

© Roy Bahl and Richard M. Bird 2018 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2017959423 This book is available electronically in the Economics subject collection DOI 10.4337/9781786435309

ISBN 978 1 78643 529 3 (cased) ISBN 978 1 78643 530 9 (eBook)

02

Typeset by Servis Filmsetting Ltd, Stockport, Cheshire

M4439-BAHL_9781786435293_t.indd 4

01/03/2018 11:08

Contents Prefacevii PART I  WHY DECENTRALIZATION MATTERS 1 Fiscal decentralization 101

3

2 Has decentralization worked?

36

PART II  DECENTRALIZING EXPENDITURE 3 Expenditure assignment and management

73

4 Decentralizing and financing infrastructure

121

PART III FINANCING LOCAL GOVERNMENT: THE KEY TO THE PUZZLE 5 Financing local and regional government

167

6 Taxing land and property

227

7 Intergovernmental transfers

279

PART IV  SUMMING UP 8 Financing metropolitan areas

341

9 Giving decentralization a chance

394

References422 Index489

v

M4439-BAHL_9781786435293_t.indd 5

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 6

01/03/2018 11:08

Preface We came up with the idea for this book a few years ago. We thought there was room for a volume on fiscal decentralization in developing countries that focused on design, implementation and sustainability, and tried to balance the economic, public management, and political economy factors that determined success. An old saw is that ‘we know some things because we have seen some things.’ We thought we had seen quite a lot during the several decades we have worked, occasionally together, but mainly independently with international agencies, officials and political leaders in many countries around the world. We have been fortunate enough to know and work with many smart people and have (we hope) learned much from them over the years, and we thought we were familiar with much of the literature. Moreover, each of us has published several books before – indeed, a colleague once said several forests must have fallen to produce all the paper we have generated over the years. Yet another book, especially one done jointly, did not seem to be such a big deal. However, we did not quite understand what we were taking on with this book. There is a mountain of literature on this subject, much of it good and most of it relevant. Some may of course be found in scholarly books and articles, but most is buried in reports to or by governments or agencies. No one – certainly not us – can find everything relevant, let alone read and understand it. Moreover, some of the most critical points emerge only when one is involved in the actual experience of designing and especially implementing fiscal decentralization. Getting as far into this material as we could, attempting to digest it and put it into a proper framework and perspective turned out to be a far more difficult task than we had anticipated. What we have finally come up with in this book is thus inevitably far from being the complete or the definitive work that we may initially have had in mind. Many caveats must be noted. We try to cover all subnational governments, but focus more on local than on regional governments. We try to cover both unitary and federal countries, but we do not attempt to treat federalism thoroughly. We try to cover the developing (and transitional) world, but miss many important places. We try to be as accurate and up to date as we can, but things change frequently and our discussion of practices in specific countries may be a bit dated at times. We do not cover some important topics – borrowing, user charges, rural vii

M4439-BAHL_9781786435293_t.indd 7

01/03/2018 11:08

viii

Fiscal decentralization and local finance in developing countries

local governments and so on – in the depth they deserve. We would have liked to do more to offset these lapses, but time waits for no man, and our long-patient wives were about to banish us to our respective doghouses, so we had to call a halt. Finally, we have tried to make this book as readable as we could, although there is only so much that can be done with a topic like ours, at least by authors whose lives have been spent in spinning out academic and official prose. Despite these (and no doubt other) lapses, we hope this book may find its way to the desks (or screens) of those who deal with these issues in practice, as well as to those whose research may, in time, solve some of the problems that we never could. As we already mentioned, this book owes much to the many colleagues, officials, and politicians around the world from whom we have learned so much over the last half century, especially those with whom we have had the pleasure of working on some of the topics discussed here – not least the editor of this series, our long-time friend and collaborator, Jorge Martinez-Vazquez. Though few match Jorge in terms of the number of works we cite here, many other scholars with whom we have worked and from whose research we have learned are included in the extensive list of references included in this book. We have also learned as much or more from the many officials and policy-makers with whom we have discussed and worked on these subjects in countries around the world, although they can seldom be cited directly. As is usually the case when it comes to discussing public policy, those who do the job are the real heroes of the story, and it is their accumulated experience and wisdom that in many ways provides the glue holding together our attempt to pull together theory and experience in this book. In some ways, we have written this book for our grandchildren, who may perhaps someday find here at least part of the answer to their questions about what, if anything, we really do. As always, however, we owe most to the always essential support of our respective life partners, Marilyn Bahl and Marcia Bird, for putting up with the (too many) years of constant travel and absence from family obligations that lie behind this book. We could not have done it without them, though they will – we suspect – be delighted to learn that we do not plan to write any more books. Roy Bahl Richard Bird

M4439-BAHL_9781786435293_t.indd 8

01/03/2018 11:08

To our grandchildren Roy – Margot, Thomas, Hadley and Carleigh Richard – Austin, Spenser, Jack, James, and Rose.

M4439-BAHL_9781786435293_t.indd 9

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 10

01/03/2018 11:08

PART I

Why Decentralization Matters

M4439-BAHL_9781786435293_t.indd 1

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 2

01/03/2018 11:08

1.  Fiscal decentralization 101 Fiscal decentralization is in vogue. Both in the industrialized and the developing world, nations are turning to devolution to improve the performance of their public sectors. (Oates, 1999, p. 1120)

This book is about fiscal decentralization and subnational government finance in low- and middle-income countries. Getting decentralization right matters. Many countries are already decentralized and not completely satisfied with the results. Many others are exploring at least the outer edges of these murky waters because urbanization is giving rise to important questions about how best to provide the public services needed by more concentrated populations and how best to finance these services. People everywhere seem to be looking for ways to have more involvement in their governance. Ensuring that provincial and local governments play a greater role in providing services resolves these issues to some extent. But there is a significant gap between what theory suggests is best practice and the fiscal decentralization found in most developing countries. Much of this book is our attempt to answer some key questions that inevitably come up in countries considering or attempting some form of fiscal decentralization. Many good studies of the success or lack of success of fiscal decentralization in various guises and forms already exist.1 Most such studies, however, focus on evaluating experience against a specific conceptual framework – a framework that seldom pays sufficient attention to the critical role played by the invariably context-specific and path-dependent ways in which fiscal institutions and administration, implementation strategies and political considerations come together. Perhaps the most important lesson one can learn from examining the practice of decentralization in developing countries is that there is no one best way to get it right. This is because ‘getting it right’ means designing a program that not only matches the diverse objectives of fiscal decentralization, which differ from country to country, but that can also be implemented in the specific and diverse (and usually changing and uncertain) conditions that exist at different times in different countries. On the other hand, experience shows that there are many ways to get decentralization wrong. What is often thought to be best international practice may turn out to be wrong for a specific country. Studying the anatomy of 3

M4439-BAHL_9781786435293_t.indd 3

01/03/2018 11:08

4

Fiscal decentralization and local finance in developing countries

decentralization failure can help us distinguish between theoretically bad (or at least seldom recommended) practices that may have been essential for some country to achieve its desired level and form of decentralization, and bad practices that are almost certain to make things worse in almost any circumstances. Similarly, studying successful decentralization experiences – cases in which countries have overcome different obstacles and succeeded in implementing some demonstrably successful type of decentralization – may provide useful lessons to others about how best to cope with their own problems. Since fiscal decentralization means different things to different people, we begin this chapter by setting out the basic definition we employ throughout this book. In the remainder of this introductory chapter, we discuss why countries decentralize – the main rationales offered for decentralization in both theory and practice and the main benefits and costs usually raised in discussing decentralization policy. We provide a brief introductory sketch – Fiscal Decentralization 101, as it were – of the arguments we set out in detail in this book. We turn in the next chapter to a review of what the existing empirical literature tells us about these matters.

DEFINING FISCAL DECENTRALIZATION In this book we understand decentralization in its ideal form as the empowerment of local populations by the empowerment of their elected local governments.2 The fiscal aspect of decentralization, which we refer to generally as fiscal decentralization, requires the central government to give subnational governments some power to make spending and financing decisions.3 Decentralization in this sense is often called political decentralization or devolution because decisions that would otherwise be made at the central level are devolved to some or all levels of government below the center: to the regional level (states or provinces); the local level (cities, towns, municipalities, or districts); and sometimes even to a fourth level (for example, districts within a city, rural villages).4 Unless fiscal empowerment (in terms of the right to make decisions about taxing and spending) is matched by political empowerment (so that these decisions are made by elected officials responsible to local citizens), theory, empirical studies and experience all suggest that devolution may not necessarily lead to good results.5 This is especially true when, as is not uncommon, more spending power than taxing power has been devolved. Fiscal decentralization is not simply fiscal devolution, but it is far more than the deconcentration of decision-making and service-delivery powers to regional or district offices of the central government. Administrative

M4439-BAHL_9781786435293_t.indd 4

01/03/2018 11:08



Fiscal decentralization 101 ­5

decentralization along these lines may make good sense when it comes to managing the delivery of central government services. Decentralization of decisions about such matters as the choice or design of infrastructure projects, for example, may properly be the responsibility of the provincial or district level office of a central government ministry in order to be sensitive to needs and demands particular to a region. Heads of schools or hospitals may similarly be charged with allocating the resources made available to them, generally subject to supervision from above with respect to both the honesty of their accounts and the results of their activities. In some instances, elected local government officials may be invited to the discussion as part of a planning committee or in some other capacity; but the deciding voice clearly remains that of the higher-level government that provides the funds supporting the activities in question. Although the dividing line between decentralization and deconcentration is clear in principle, the two approaches to service delivery are often not independent of one another in practice. Deconcentration does not empower the local population in any direct way; but sometimes, as in Indonesia in 2001 – when over 2 million central government employees were absorbed by local governments with no major disruption in service delivery – it was an important prerequisite to implementing a successful fiscal decentralization (Directorate General of Fiscal Balance 2012).6 Moreover, since local residents receive public services not only from local government expenditures but also from central government spending in the local area, combining the two (although seldom easy to do) may, as DeLog and LPSI Secretariat (2015) demonstrates, provide a more accurate picture of public service expenditures in the local area. Another form of decentralization often found is delegation, when a higher-level government contracts (implicitly or, less commonly, explicitly) with a lower-level government to deliver a particular service.7 This ‘principal–agent’ relationship allows the senior-level government (the ­ principal) to retain all decision-making powers, with the junior-level government (the agent) delivering the service. Much has been written about the inherent incentive problems with such agency arrangements owing to the asymmetric information problems arising when, as is usually the case, the putative agent knows much more about what it does and how it does it than the principal ever can. Little information is available about the effects that different forms of intergovernmental arrangements along these lines may have on the efficiency, efficacy and equity with which services are delivered.8 But there is no question that with delegation the main accountability for the service delivered is upward from the local government to the higher-level government. Because local governments usually have little or no discretion over the quantity or quality of the service or the way in

M4439-BAHL_9781786435293_t.indd 5

01/03/2018 11:08

6

Fiscal decentralization and local finance in developing countries

which it is provided, there is little or no local government accountability to the local population, and hence no real local empowerment. The bifurcation of political and administrative responsibilities forces local agents to be accountable downward to those to whom they provide services and upward to those who pay for the services they provide. Since he who pays the piper picks the tune, it is not surprising that such dual accountability seldom ends well for those at the bottom of the system. Another type of decentralization not included in our definition of fiscal decentralization is community-driven development (CDD), that is, the delegation of service delivery powers to community-interest groups. Although these community groups become parallel local governments to some extent, those who run such organizations are usually not elected and not explicitly accountable to the local population. The community-driven approach may enhance the probability of successful decentralization by providing a more effective lobby voice for local interest groups. It may help prevent elite capture – though in some instances it may itself be captured (perhaps by a different elite). It may sometimes be an effective way to deliver local public services, and may give a more effective voice to heretofore disenfranchised groups. These are all good things. But CDD is decidedly not a substitute for a representative local government, and may even crowd out traditional local governments (World Bank Independent Evaluation Group, 2008). On the other hand, some gray areas of fiscal decentralization are deliberately included in our discussion. Even when local governments are not elected their centrally appointed political leadership may act at least to some extent in the local interest. In China, for example, the central government rewards the successes of the provincial and local officials they appoint, and encourages them to compete with one another (Bahl and Martinez-Vazquez, 2006). Some have hailed such competition as a ‘market-preserving approach’ that has been an important factor in China’s development (Qian and Weingast, 1997). However, the Chinese system falls far short of one driven by local preferences, and has been criticized precisely because of the limited concern that many local officials have shown about local public services compared to their need to fulfill the wishes of the central authorities by focusing heavily on economic growth (Bahl et al., 2014; World Bank and Development Research Center of PRC, 2014). Accountability is mainly upward to the center and not downward to local residents. A second gray area relates to the very limited effective budgetary discretion that many local governments have, owing to limits imposed on their revenue powers and the unfunded expenditure mandates imposed on them by central governments.9 In some countries, even large elected subnational

M4439-BAHL_9781786435293_t.indd 6

01/03/2018 11:08



Fiscal decentralization 101 ­7

governments end up as little more than spending agents of higher-level governments. In Colombia, for example, although most expenditures for health and education appear in the budgets of the elected regional governments (departamentos), this is more a matter of delegation than devolution; most decision-making remains at the national level (Acosta and Bird, 2005). In Cambodia, local governments have no formal expenditure responsibilities, though they may undertake ‘permissive’ expenditures (Smoke and Morrison, 2011). Nonetheless, even in these cases local governments have some discretion and are to some degree accountable to the local population who elects them. Moreover, in many countries local governments could have more autonomy if they chose to do so.10 These ‘mixed’ cases are considered to some extent in this book because their experience may contain some important lessons about how low- and middle-income countries may perhaps transition more easily to full fiscal decentralization, as well as about the possible costs and benefits of such a transition.

WHY COUNTRIES DECENTRALIZE Real decentralization in the sense just defined is not easily achieved. It requires one level of government – usually the center – to give up some power to its ‘subordinate’ regional and local levels. Doing so is not likely to delight the hearts of those charged with carrying it out. Nonetheless, as mentioned earlier, a surprising number of developing countries are decentralized to some extent, though often more in form than reality. Some may have been created that way by their former colonial authorities or may have emerged from the fires of civil war or national struggle. Some may have been forced to follow this path to resolve political differences between regions or groups. Still others may have chosen to decentralize for a variety of administrative, technical and economic reasons. For whatever reason, many developing countries have in recent years decentralized government activities to some extent – often for reasons not explicitly spelled out – and with varying degrees of success. Some countries have invested considerable time, resources and political capital in designing and implementing their decentralization programs: establishing new secretariats or departments to initiate and guide the process; putting into place a legal foundation for local governments; redefining a host of fiscal institutions ranging from taxes to borrowing to expenditure management; and sometimes incurring heavy transition costs. Donors also have invested heavily in fiscal decentralization. Between 1990 and 2006, for example, the World Bank provided some form of lending and non-lending support

M4439-BAHL_9781786435293_t.indd 7

01/03/2018 11:08

8

Fiscal decentralization and local finance in developing countries

to 89 countries for decentralization initiatives, devoting about 8 percent of its resources (over $30 billion) to such efforts (World Bank Independent Evaluation Group, 2008).11 One reason so many countries and donors have been attracted to fiscal decentralization as a potential way to contribute to economic and political development is because it is not difficult to tell a wonderful story that can be sold by political leaders to various stakeholders. The story has to do with how people will become better citizens when they have more control over their public sector, and how the result will be better services, better governance, faster and more balanced economic development, and even nation-building. Of course, since circumstances differ widely from country to country (and from time to time even within the same country), different wonderful stories have been told in different places at different times. Russia’s fiscal decentralization as originally structured seems to have been mainly designed to head off separatist movements (MartinezVazquez et al., 2008).12 On the other hand, South Africa’s devolution of some decision-making power to a new set of provincial governments as well as to over 800 local level governments was arguably just the sort of empowerment one would expect in the aftermath of apartheid (Bahl and Smoke, 2003). However, as decentralization progressed, and especially as the political situation changed, both countries subsequently moved back in different degrees to more centralized structures. In contrast, when it reformed its constitution in 1988 after a highly unpopular centralization under the previous military dictatorship, Brazil went much further than either Russia or South Africa towards a sustainable decentralization. Despite many attempts to revise intergovernmental fiscal arrangements in recent years, the extent to which decentralization was enshrined in the Brazilian constitution and the way in which the political balance shifted has made it almost impossible to reverse the direction to any significant extent (Bird, 2012; Wetzel, 2013). Many claims have been made in the literature and in political discussions in many countries about such possible economic gains from decentralization as a faster rate of economic growth, greater public sector efficiency, improved service delivery, and greater democracy and nation-building. Opponents of decentralization often doubt the extent to which such good things result from decentralizing government activities, and assert that such bad outcomes as weaker stabilization policy and increased inequity, corruption and conflict are more likely. In the remainder of this chapter, we consider some of these arguments, and in the next chapter we examine the empirical evidence. Countries decentralize to differing extents and in different ways for several reasons. The economic rationales commonly mentioned mainly

M4439-BAHL_9781786435293_t.indd 8

01/03/2018 11:08



Fiscal decentralization 101 ­9

draw from the theory of fiscal decentralization first set out in detail in Oates (1972). This theory stresses the potential gains in terms of economic efficiency. However, this notion, central to the economic case for decentralization, is seldom raised in political debates. Instead, debates about decentralization usually stress grander but vaguer gains such as increased economic growth and nation-building.13 Other possible gains often cited include greater accountability, improved service delivery and perhaps more and more effective revenue mobilization. Unfortunately, as Treisman (2007), Mascagni (2016) and Lago-Peñas et al. (2016) discuss in detail, there is little evidence to support definitive conclusions with respect to any of these matters. Indeed, the safest generalization about the outcomes of decentralization is that one cannot safely generalize. How outcomes are affected by any particular decentralization measure depends so heavily on the specifics of its design, how well it is implemented and the historical, institutional, political and economic context within which it is implemented that those who seek simple, clear answers are bound to be disappointed. Nonetheless, in principle the basic efficiency argument for decentralizing public decisions – which is essentially the same as the case for market economics in general – is convincing. Decisions are best made when they are made by those in the best position to make them, which in most cases means those closest to the issue in question. Governments closer to the local or regional economy are thus presumably in a better position to facilitate economic development by deciding such matters as: the most appropriate regulatory environment for local business; the right infrastructure investments to make; the proper balance between taxation and user charges; and, in general, the fiscal environment best suited to support the local economy. The best choices are often those that are made locally. For some big cities, the main bottleneck to job growth might be traffic congestion, so the top priority may be investment in transportation infrastructure. For others, where heavy industry is important, the priority may be power and water. For those with significant in-migration from rural areas, it may be housing and slum control. Similarly, although those who do not live there may think that the biggest need in rural localities is for clean water or better education, many rural people and their local governments may be focused on more immediate needs and give priority to irrigation, electricity or farm-to-market roads (Fiszbein, 1997). In principle, people are more likely to get the package of public services they want, if not necessarily what others think they should want, under a decentralized system than under a centralized system. One result of central government supervision of the budgetary affairs of every subnational government may thus be poor public service outcomes, particularly ­

M4439-BAHL_9781786435293_t.indd 9

01/03/2018 11:08

10

Fiscal decentralization and local finance in developing countries

in large countries or those in which physical communications with the national capital are difficult. Consider the following anecdotal examples: ●●

●●

●● ●●

●●

In Russia, a country with 11 time zones, in the early 1990s the budgets of the 89 regional governments were approved and revised in Moscow, often after face-to-face negotiations (Martinez-Vazquez, 1994).14 In Colombia some years ago a request by a city for permission to borrow funds to fix a road required 65 official signatures and numerous visits to the national capital to secure approval (Bird, 1970). In Indonesia in the 1980s changes in bus routes in Jakarta reportedly required approval by the President of the country. In Papua New Guinea, a country created with an elaborate set of decentralized provincial governments – essentially large rural municipalities – well-founded fears of the low level of administrative capacity at the local level led to the creation of a central ministry that was supposed to monitor and control provincial finances. The most observable result was a complex and ineffective dual layer of administration (Bird, 1983).15 In mountainous Nepal, many of the 4,053 local governments are several days’ journey from Kathmandu.

In none of these cases does it seem credible that a few officials in a central ministry are going to be able (or willing) to do the best job at the least cost for every community. Indeed, in some countries many officials charged with supervising particular localities had never visited them and knew nothing about them.16 Size matters. China and India have populations in excess of 1 billion: China has 58,545 subnational governments,17 while India has 237,687;18 Brazil has a land area in excess of 8.4 million sq. km; and Indonesia is made up of more than 6,000 inhabited islands. Despite the centralization approach of Soviet-era government in Russia and most (not all) countries under Soviet influence until the late 1980s, most large countries have always had some form of decentralized governance.19 For example, although China has only one time zone for its vast territory, and is at least as controlled from the center in political terms as Russia, it has always been run in a much more decentralized fashion than Russia or most Central and Eastern European countries under Soviet influence (Wong and Bird, 2008). None of this is to say that small countries do not see advantages to fiscal decentralization. For example, Lao PDR (population 6 million) has been exploring decentralization options for a decade (Martinez-Vazquez et al., 2006a); Bhutan (population 0.8 million) included strengthened local

M4439-BAHL_9781786435293_t.indd 10

01/03/2018 11:08



Fiscal decentralization 101 ­11

governments in its vision for Gross National Happiness in its ninth fiveyear development plan; and several Caribbean countries have undertaken reviews of their local government structure (National Democratic Institute for International Affairs, 2004). As noted above, in countries big and small the economic argument for fiscal decentralization is straightforward (Oates, 1972). Assume that people’s preferences for government services vary for whatever reason  – religion, language, ethnic mix, climate, income level, economic base. ­ Assume further that to a considerable extent people have either sorted or adapted themselves so that those who live in the same neighborhood or region have broadly similar preferences for public services. If g­ overnments respond to these preferences, decentralizing public sector decisions to subnational governments will result in variations in the package of services delivered in different regions. People in each region will get more of what they want and less of what they do not want; and (in the absence of ­externalities) everybody will be better off – or, as economists put it, national welfare will be enhanced. A more centralized system with more uniform service provision would make people worse off, since even if they could move elsewhere to get the services they want, they would incur additional costs to do so. When the preferences of local people become more influential in affecting local government decisions, two important results ensue. First, in a democratic setting – and to some extent (as Chinese experience suggests) even when there are no local elections but politicians are for whatever reason sensitive to public sentiments – local officials become more responsive to the local population served for the quantity, quality and variety of services provided. Second, because people are more likely to get what they want, they are also likely to be more willing to pay for local services. Successful fiscal decentralization may thus, in principle, simultaneously resolve several common problems facing developing countries: increased revenue mobilization; improved accountability of elected officials; and more grassroots participation in governance. It may also, as Lewis (1967) noted long ago, lead to more learning and increased capacity development at the local level, thus weakening a constraint often said to make decentralization unfeasible. Some have argued that decentralization will also create incentives – essentially by ‘crowd-sourcing’ to more minds the task of how to deliver services within a budget constraint – that will increase innovation in decision-making about local public services (Oates, 1999). Since in most developing countries almost everyone can easily identify public services that can and should be better provided (that is, provided more in accordance with their preferences) this story is such a good one

M4439-BAHL_9781786435293_t.indd 11

01/03/2018 11:08

12

Fiscal decentralization and local finance in developing countries

that it is easy to believe that important welfare gains may result from decentralization even if they cannot be precisely measured.20 As we discuss in the next chapter, measurement is essential to good policy because decentralization may not be worthwhile unless the gains are sufficiently large to offset its costs. But even if the possible efficiency gains from fiscal decentralization are demonstrably important they may not be captured by subnational governments unless a substantial number of institutional conditions are in place. Ideally, for example, there should be an electoral system that gives an effective voice to the local population. There should also be a capable local administration reporting to a local council which has both significant expenditure responsibilities and significant local taxing powers and operates within the kind of well-designed and effectively implemented intergovernmental institutional structure we discuss in later chapters. In most middle- and low-income countries, few if any of these conditions are in place – and even if they are, success is not guaranteed. The political system may go astray. External crises may dominate the policy agenda. The local population may be too impatient to ride out the problems of the transition period. The central government may hinder rather than help subnational governments as they try to capture these efficiency benefits. Good stories do not necessarily lead to good policies. Democracy and Accountability The increased interest in fiscal decentralization since the mid-1980s has roughly coincided with a growth in democracy around the world. In some countries, democratic governance evolved over time with economic development and the development of competitive political parties (Mexico, Brazil). In others, authoritarian regimes were replaced by elected governments, with voters assuming more power and responsibility (Philippines, Indonesia). When fiscal decentralization has accompanied subnational elections, as in Colombia, it has become an important component of the democratic system. In 1980, fewer than half the countries in Latin America were classified as democracies, and in only three were mayors elected (Hausmann, 1998). In contrast, in 2010, mayors were elected directly by local citizens in every country in the region except Cuba and Guyana.21 Of course, the mere existence of elections does not necessarily mean strong local governments, as countries like Cambodia and South Africa show. In the so-called transition countries (mainly in Central and Eastern Europe), for instance, Kravchuk (2008) found only nine out of 29 in which local elections were critical to ensuring government accountability. Still, political decentralization appears to increase both local involvement and

M4439-BAHL_9781786435293_t.indd 12

01/03/2018 11:08



Fiscal decentralization 101 ­13

official responsiveness to local interests, and to lead to greater demands for fiscal decentralization (Ebel and Yilmaz, 2003; Manor, 1999). There are many problems in ensuring adequate accountability at the subnational level. Imperfect as the process is, open elections remain the best way we have to ensure that subnational politicians are accountable to the local population.22 Since such elections, especially at the local and regional level, are not the norm in much of the world – for example there are none in countries as important as China and Egypt – in practice, accountability usually rests on the extent to which local political leaders see their self-interest as being related to the satisfaction of the local population. Even then, what they do in terms of providing services is often more likely to reflect the wishes of the higher-level governments that appoint them than the preferences of the local population. In China, for example, the reward system for officials has long been heavily weighted toward economic development (Bahl et al., 2014). Different provinces take different approaches to economic development, so there is competition for jobs and investment; and perhaps, as Qian and Weingast (1997) have argued, the results may encourage growth. But no one knows, or asks, whether local people might perhaps prefer more in the way of social services.23 Even with full and free elections, accountability to local voters is of course not guaranteed. People may not have the information or inclination to push for more accountability. Elite capture may result in local governments being effectively accountable only to the agenda of a small group which may or may not map into local preferences (Bardhan and Mookherjee, 2000). Or local service delivery may be hostage to the rule-making and control goals of central ministries or their local representatives. The electoral process itself may be an impediment. For instance, elected leaders may be more accountable to the national party than to the local population. Strong party discipline may perhaps improve the quality of local governance; but whether better governance means a better match with local preferences varies from place to place (Hankla and Downs, 2010).24 Many of these constraints on downward accountability could be relaxed (Yilmaz et al., 2010). Institutional measures such as term limits, recall, local council oversight and requiring local council independence exist in some countries and may do some good. For example, term limits may protect against the capture of the local council by elites; on the other hand, they may also lead to discarding the important learning of incumbents. Recall and other ‘direct democracy’ devices (such as voting on specific issues) may curb abuses of power, but they may also intensify group conflicts or lead to fragmented and inconsistent policy outcomes. Some countries, such as Brazil, have introduced other ‘social’ accountability measures such

M4439-BAHL_9781786435293_t.indd 13

01/03/2018 11:08

14

Fiscal decentralization and local finance in developing countries

as public meetings, referenda and special citizen oversight groups, although again with mixed success.25 Accountability is also shaped by the way in which local government administration is structured. The structure most conducive to accountable decentralization is one in which the chief officers of the local government are accountable to their elected local body, and the local government controls employee pay rates and has the authority and responsibility to hire, promote and fire as well as to manage and evaluate employee performance (Sud and Yilmaz, 2013). Often, however (as in India), local chief administrators are appointed and rewarded by a higher level of government, so their primary allegiance is likely to be to those who control their careers. Even when employees are nominally under local government control they may, like teachers in some countries in Latin America, be more responsive to their unions and to the higher-level governments that establish their salaries and provide most of the funding. In Peru, for example, although regional governments have significant responsibility for health, education and transport expenditure, the central government imposes such strict expenditure mandates that the result is closer to deconcentration than real decentralization (Martinez-Vazquez, 2013). Many observers have noted that local governments in some countries often seem unwilling to exercise even those powers that they have. Critics have often focused on the poor performance of local governments as the principal explanation for failures in decentralization such as poor service delivery. Sometimes such criticisms are legitimate. Extended dependence on central charity is as likely to breed passivity and dependence in governments as in people. Indeed, sometimes local governments are too passive. As Breton (2004, p. 37) notes, local governments “cannot be empowered by others . . . If they are to become empowered, they must empower themselves.” This vision, that the upraised local fist must replace the outstretched local hand for decentralization to succeed, may be a bit extreme. Around the world some cities, regions and localities thrive in the same setting where others fail, so success to some extent likely follows initiatives from below rather than flowing effortlessly and evenly to all as bounty bestowed from above. Often, however, the basic problem is not so much that local governments are incapable of doing or unwilling to do more or better, but that the institutional structure within which they function – a structure essentially established by higher-level governments – creates incentives to make decisions that are subsequently viewed by some as bad. Central governments largely get the local governments they want, and deserve. What we conclude from all this is that the necessary conditions for successful fiscal decentralization usually include strong central government

M4439-BAHL_9781786435293_t.indd 14

01/03/2018 11:08



Fiscal decentralization 101 ­15

commitment to the objective, and strong central leadership and support in providing adequate resources and in managing the inevitable conflicts that emerge in any decentralization process. Prud’homme (1996, p. 357) may have overstated the case when he said that “where decentralization is needed (because central governments are corrupt and inefficient), it cannot be implemented. Where it can be implemented, it is not needed.” But he was basically right in stressing the central role of the central government in determining both whether and how effectively decentralization occurs in any country. We develop this theme in detail in later chapters. Service Delivery The argument that fiscal decentralization will improve public service delivery usually falls on receptive ears. The level and quality of services provided through more centralized systems in developing countries is often poor. Intuitively, it seems credible that more local control over expenditure decisions would make things better. Local governments are presumably better positioned to determine the right location for public facilities and to recognize local service needs and control the performance of local employees. In addition, if they are accountable downward for their actions they may have more of a vested interest in better local public services than do higher level governments, and therefore be more responsive to their constituents. For local residents to care enough about the quality of local services to hold elected officials accountable, however, the government may need to be responsible for important services such as, say, public utilities and primary education. If local governments are responsible for little more than such housekeeping functions as the maintenance of local parks and public buildings, people are unlikely be overly concerned about the quality of services delivered. They are much more likely to care if water is not clean, or even available, or if teachers do not bother to show up at the local school. When such highly visible results hit home, people are more likely to turn out at town hall meetings and elections, to refuse to pay for the services, and to protest visibly against service failure.26 Even China with its appointed local governments has a long history of such local protests at service failures. In response, although higher-level authorities have sometimes simply suppressed such dissent they have at other times reacted by replacing unpopular local officials and demanding changes in local policies.27 In some developing countries, especially federal countries, important functions are assigned to subnational governments, mainly those at the regional level. In many countries local governments do not have functions

M4439-BAHL_9781786435293_t.indd 15

01/03/2018 11:08

16

Fiscal decentralization and local finance in developing countries

that people consider significant to the quality of their lives. In Brazil, for example, state governments account for 26 percent of total taxation, while local governments account for only 6 percent of taxation. In India, although state governments account for 37 percent of taxation, local governments collect almost no taxes.28 In some unitary countries – Philippines, Indonesia and Colombia, for example – locally provided services are important; but in many others, such as Egypt, they are not. There are often accountability problems when the assignment of functional responsibility for services such as primary education to the regional or metropolitan area is too far removed from the local populations served. As we discuss in Chapter 5, the efficiency gains from decentralized service delivery are best captured when beneficiaries pay for the government services that they receive. For this to occur, subnational governments must be empowered to determine at least some tax rates and user charges. If subnational governments have such taxing power, local demands for accountability are likely to be stronger than if services are financed primarily by a transfer from the center. This accountability argument for subnational government taxation works best when there is equivalent accountability on the expenditure side. If expenditures are delivered under a delegation approach – where central line ministries make many or most decisions about who gets what in terms of service delivery (as is largely true at the regional level in Colombia, for example) – the case for independent local taxing powers is weakened (Acosta and Bird, 2005). It is hard to hold local governments accountable for the quality of services when they have little discretion over what they provide. The effective local accountability underlying the case for fiscal decentralization usually depends on highly specific local conditions that establish, influence and shape the degree to which different local groups are willing to and capable of playing a real role in local decision-making. Revenue Mobilization Another potential gain from fiscal decentralization may be an increase in the overall rate of revenue mobilization, reflecting the presumably greater willingness to pay for services more in tune with local preferences and perhaps also the potential comparative advantage of subnational governments in collecting certain taxes. Although the amounts involved may not be large, the potential revenue gain from decentralized taxation may nonetheless be significant for developing countries in which the average ratio of tax to GDP has been stuck at about 16 percent for three decades, with only about 11 percent of this amount being raised by subnational governments (Bahl and Bird 2008; Bahl, 2014, IMF, 2011).29

M4439-BAHL_9781786435293_t.indd 16

01/03/2018 11:08



Fiscal decentralization 101 ­17

Two questions may be asked about the possibility of raising more revenue from a more decentralized system. First, are central governments willing to allow local governments to impose and collect new revenues? Second, are local governments willing to take on the difficult political task of doing so? Higher-level governments may fear that increased regional and local government tax collections can be achieved only if the central government loses some of its own ‘tax room’ – for example, because taxpayers are willing to accept only a certain amount of tax burden on any one base.30 Or they may think that it is cheaper and more efficient for the central government to levy taxes and to finance subnational governments with transfers than for such governments to levy their own taxes.31 Of course, all governments are usually happier to spend ‘other people’s money’ in the form of transfers than to try to extract increased revenues from their constituents. Nation-Building Some countries appear to have moved in the direction of decentralization in order to achieve political stability. Chechnya, East Timor, Kosovo, Kashmir, South Sudan, Rwanda, Burundi . . . the list of territorially based ethnic groups that have taken up arms to varying degrees against the state is lengthening. Nor, as the cases of Scotland, Catalonia, Flanders and Quebec show, are such pressures confined either to lowincome countries or to outright civil war. It is not surprising that some countries have tried to pre-empt separatist pressures by decentralizing some activities. As World Bank (2000, pp. 107–8) puts it: “When a country finds itself deeply divided, especially along geographic or ethnic lines, decentralization provides an institutional mechanism for bringing opposition groups into a formal, rule-bound bargaining process.” This study went on to offer five specific political rationales for decentralization: 1. Decentralization may sometimes serve as a “path to national unity.” Two examples often cited are South Africa (Bahl and Smoke, 2003) and Uganda (Smoke et al., 2010). Of course, much has since happened in both countries. 2. Decentralization may in some instances “offer a potential political ­solution” to a civil war. Sri Lanka is a case in point, although decentralization has been less prominent on the political agenda since the central government defeated the insurgent Tamil north in 2009. The extent to which the peace might best be sustained by more effective decentralization continues to be a contested issue in Sri Lanka.32

M4439-BAHL_9781786435293_t.indd 17

01/03/2018 11:08

18

Fiscal decentralization and local finance in developing countries

3. Less drastically, decentralization may serve as “an instrument for deflating secessionist tendencies.” The examples often mentioned are Ethiopia and Bosnia and Herzegovina. The case of Bosnia is discussed in detail by Fox and Wallich (2007) and Ethiopia in Yilmaz and Venugopal (2011) and Prichard (2015). 4. Decentralization may attempt to achieve a similar aim by “conceding enough power to regional interests to forestall their departure from the republic.” The Russian Federation is an example (Treisman, 1999), though the course has been sharply reversed in this century as power has been reconsolidated at the center. 5. And, finally, the report suggested that decentralization may be used to co-opt “grassroots support” for central policies. World Bank (2000) cites Colombia as a relevant example, though for a quite different take on the Colombian case see Bird and Fiszbein (1998). A more appropriate example might be Bolivia, where a ‘municipalization’ program was adopted in the early 1990s, in part precisely to strengthen the national state by positioning it as the main lifeline of local development while at the same time countering the centrifugal tendencies that a more ‘federal’ decentralization program, focusing on the provincial level, would have fostered (Grindle and Domingo, 2003). For a fuller account of the Bolivian experiment, which succeeded at least for a time in making government more responsive to local wishes and needs, see Faguet and Pöschl (2015). Whether or not national unity is enhanced by fiscal decentralization is a complex, interesting and largely unanswered question.33 One reason it may be enhanced is because at root decentralization is all about citizen participation in government. How well the public sector works, and in particular the extent to which people have to pay for what they get and expect to get what they pay for, may be critical to establishing the kind of state legitimacy and state–society relations that constitute ‘state-building.’34 Many people in developing countries feel estranged from their governments: perhaps because they have been ignored for so long that they simply do not trust the government; perhaps because they feel that their voice is too small to count; perhaps because there is either no democracy or because, if there is, they have not yet learned to use it effectively. Whatever the reason, few participate in government decisions by voting or joining the debate about public policy in any way, sometimes because of fear of the consequences if the authorities do not like what they say.35 When people feel that they have little influence on governance, they may be content as long as their private well-being is improving, but they are unlikely to become actively involved in the process of governance. If one result of decentralization is

M4439-BAHL_9781786435293_t.indd 18

01/03/2018 11:08



Fiscal decentralization 101 ­19

more participation by more people in the governance process, the outcome may be, in at least some cases, more identification with both the local and the national governments.36 Another way in which decentralization promotes national unity may be by incorporating disaffected groups more effectively in the national framework. Those who live in different regions of the country, or even different neighborhoods in the same city, may want different things from their governments, especially if they constitute a relatively homogeneous group that differs in some way from others in the country or city. Such groups may not only want to be able to choose for themselves a package of services that best fits their demands and to deliver these services themselves; they may also wish to establish in some formal way their separate (usually ethnic) identity. Recent examples include the campaigns for regional autonomy in the largely Muslim Mindanao province in the Philippines as well as in Aceh province in Indonesia. The pressure for ethnic regions may become even stronger when control of natural resource wealth is at stake, as the case of Nigeria has long demonstrated. Of course, such regional or localized group pressures are not limited to countries that are ethnically diverse. Diversity in citizen preferences between urban and rural areas and in regions with different economic bases leads to some pressure for devolution of budgetary powers everywhere. Centralized systems, even if some service provision is deconcentrated or delegated, usually provide more regionally uniform service levels than more decentralized (devolved) systems, and many – perhaps most – people may consider this to be a desirable outcome. Others, however, may prefer more room for regional and local choice. On the other hand, in some cases, central governments may intentionally favor some regions over others, either to keep potentially troublesome regions calm or for reasons of national security (for example, in border regions). Sometimes they may achieve such differential results even if they do not do so intentionally, for example, by setting strict rules about not only what is done but also how it is done throughout the country to control input costs even though the result may be to produce different levels of service output in different regions. In general, however, when decisions are made centrally rather than locally there is less diversity in service levels across the country. One consequence of decentralization in such a system is almost always increased inequality when more decisions are made locally, as quickly became evident in much of central and eastern Europe with the breakup of the old Soviet bloc (Bird et al., 1995; Bahl, 1994). Similar issues often arise in the many developing countries which emerged from the colonial era with artificial boundaries and have since contained (and often restrained) ethnic divides within those boundaries. The result in some countries has

M4439-BAHL_9781786435293_t.indd 19

01/03/2018 11:08

20

Fiscal decentralization and local finance in developing countries

been civil wars (Nigeria, Syria) and continuing unrest and threats of secession (Turkey, Myanmar). One approach to controlling such problems has been ‘fiscal appeasement’ (Treisman, 1999) – the buying off of dissent in part by special concessions to local autonomy. Developed countries like Belgium, Canada, Spain and the UK have all trodden this path to some extent in recent decades.37 Sometimes such policies have also been followed in developing countries like Nigeria (after the civil war of the 1960s and the later unrest in the oil-rich south and the Muslim north) and Indonesia (particularly after Aceh and Irian Jaya threatened to follow the lead of East Timor and opt for independence). Similarly, Muslim Mindanao was given special standing as an autonomous region and some fiscal concessions in the Philippines (Wallich et al., 2007; Manasan, 2009). Even Russia, in addition to prolonged armed struggles with Muslim Chechens and others in the Caucasus region in the early 1990s, negotiated special treaties with disaffected regions that were rich in natural resources.38 Changing labels and paying money does not always work, however. Sometimes, national politicians pull back from making accommodations to ethnic regions because it makes it more difficult to build a governing coalition (Eaton et al., 2010). Sometimes, they do so in the belief that increased fiscal transfers might be used to finance insurgents.39 And sometimes, no matter what is done, countries fall apart, as happened in Czechoslovakia, Yugoslavia and the former Soviet Union as well as in Indonesia (East Timor) and Sudan (South Sudan) – and may well happen again in such other ‘failed states’ as Somalia and Libya. On the other hand, in Vietnam, Germany and Hong Kong reunification gave the ‘new’ regions special arrangements, including some autonomy or special financial treatment within the intergovernmental fiscal system. As noted above, however, such arrangements are sometimes criticized as harming rather than helping national unity. More autonomous decentralized governance may offer regional politicians the opportunity to develop a following, and hence potentially threaten the hegemony of the central government. Mexican history provides several instances of such regional revolts (Cline, 1962), which is one reason why Mexico remains wary of giving too much power to elected provincial governors. For similar reasons, federalism is an unpopular concept in Indonesia. The political impact of decentralization is thus far from clear in practice. Neither empirical research nor country case studies demonstrate that decentralization, asymmetric or not, is always likely to promote national unity (Vaillancourt and Bird, 2016). The available empirical evidence is thin and provides little support for the conjecture that decentralization avoids the breakup of nations (Lago-Peñas et al., 2016.) Perhaps the best recent appraisal of this complex issue is that of Sorens (2015), who

M4439-BAHL_9781786435293_t.indd 20

01/03/2018 11:08



Fiscal decentralization 101 ­21

c­ oncludes that fear of secessionist pressure is perhaps the main reason that so many developing countries have proved so reluctant to heed the arguments of economists (like us) who have urged the importance of decentralizing taxation power. Faced with what they see as a choice between getting decentralization right and keeping their often fragile countries together, most politicians – perhaps even most people – seem more willing to forgo some potential economic gains than to risk losing their country. All any honest advisor can perhaps do when decision-makers are unwilling for political reasons to do what seems needed to achieve the full economic benefits of decentralization is to be sure that the consequences of the decision are clearly understood. Throughout this book, for example, we stress the importance of decentralizing adequate tax power if decentralization is to work properly. This message has been hard to sell to national politicians and officials. The damage arising from the absence of this key link in the decentralization logic may be alleviated to some degree, as we discuss later, by introducing various small adjustments in the intergovernmental fiscal system. However, the interlocking marginal institutional alterations required are usually difficult to explain to those who demand a simple answer even when none exists. The diverse and complex reality of intergovernmental finance seldom permits achieving equally good results when the best solution is ruled out. A good understanding of the basic theory is essential. But even the best theory can only tell us the questions that must be answered, and not how to get the best results possible in any particular context. Designing and implementing a good fiscal decentralization system demands painstaking investigation and consideration of the specific context in question. To produce a clear answer, theory must often assume that a wide array of conditions hold which may not always be true for the case at hand. Without economics, however, we would have no sound framework within which to understand how the many forces at play in determining outcomes interact, let alone how to assess and evaluate the weight that must be attached to each given our policy aims. In this book, we try to set out the key elements of the analytical framework in a way that might help outsiders to understand and insiders to improve how decentralization may best work in any particular case. But we do not pretend to determine what any country should do, let alone guess what it might do. One lesson that emerges from our discussion is that different combinations of objectives and problems in different settings call for different solutions. As Litvack et al. (1998) said some years ago, if countries suffer from a whole range of problems – weak democratic institutions and processes, weak legal and regulatory systems, highly imperfect markets for land, labor and capital, poor information, weak financial systems, and

M4439-BAHL_9781786435293_t.indd 21

01/03/2018 11:08

22

Fiscal decentralization and local finance in developing countries

­ ontransparent fiscal systems – probably the most they can or should n attempt is very modest decentralization of a few clearly local services. If things are not quite that bad, it may perhaps be possible to construct feasible approaches to resolving each of the problems just mentioned, but none of the possible solutions are likely to be simple, most of them are long-term, and some may require fundamental (and difficult) changes in political institutions. In such cases, it is important not to claim that any specific reform, even one in the right direction, will yield the full benefits of fiscal decentralization. Anyone who thinks that it is simple to design and implement a fiscal decentralization program without substantial preparation and sustained effort needs to think again. The various costs and obstacles discussed in the next section can be overcome in principle, as we demonstrate later in this book. But it is of course much harder to do something well in practice than to stand above the fray and prescribe what others should do. Still, we think that many of the pitfalls and roadblocks that countries have encountered in their attempts to decentralize are less a reflection of the inherent problems with decentralization than of inadequate preparation. All too often, countries (and their advisers, outsiders and insiders alike) have placed undue reliance on the veracity of unproven generalizations about what is good and necessary for success. When coupled with the usually overoptimistic expectations about the extent and rapidity with which the expected gains will be realized – necessary though such hyperbole may seem for political acceptance – failure is all too often inevitable.

THE COSTS OF FISCAL DECENTRALIZATION In most low-income countries, the fiscal system is highly centralized with, on average, about 80 percent of government expenditures made by central governments. There are advantages to a centralized fiscal system and there are costs associated with fiscal decentralization. Several wellknown papers have focused on these costs and raised concerns about the “dangers of decentralization” (Prud’homme, 1995; Tanzi, 1996), although many of the arguments put forth were soon vigorously debated (McLure, 1995; Sewell, 1996; Spahn, 1997). We discuss here the principal disadvantages often associated with decentralization, including economies of scale, failure to deal adequately with externalities, lack of local capacity, stabilization concerns and corruption. While there are some risks on all these fronts, it turns out that most of the economic dangers attached to decentralization can be ameliorated – if not eliminated – by careful design and implementation.

M4439-BAHL_9781786435293_t.indd 22

01/03/2018 11:08



Fiscal decentralization 101 ­23

Stabilization Policy40 Most economists agree that macroeconomic stabilization policy is an issue best left in the hands of the central government (Musgrave, 1959). Fiscal decentralization makes the design and implementation of central stabilization policy more complex because it adds a layer of government actions that are not fully controlled by the center but that must be taken into account. Regional and local governments may in various ways make it more difficult for the central government to implement potentially important stabilization policies such as raising taxes or reducing public spending. Such constraints matter because many low- and middle-income countries encounter macroeconomic difficulties as a result of their vulnerability to commodity price fluctuations, crises of one sort or another arising from the international financial system, or simply poor management. For example, almost every country in the world, no matter how well managed, is affected by worldwide crises like that in 2009–2010, which reduced tax revenues almost everywhere. In countries in which intergovernmental transfers are tied by formula to central tax collections – as is the case, for instance, in Brazil, Colombia, Philippines, Pakistan, Mexico and Indonesia – transfers to subnational governments automatically declined so that subnational spending also fell, thus reinforcing the economic downturn. Because subnational governments are usually responsible for providing (labor-intensive) essential services, in most countries the pressure to offset their budget loss was intense. In the case of Mexico, for example, the federal government borrowed to create a stabilization fund to shore up some of the decline in transfers. In other countries, subnational governments protected their expenditure budgets by lobbying for discretionary grants, increasing local taxes and covering deficits with varying forms of short-term borrowing. In still others, local and regional governments were forced to bite the bullet and cut their expenditures significantly.41 Even without a worldwide recession, central governments in many developing countries have trouble controlling the size of the annual deficit. The conventional approach to resolving this problem is some combination of an increase in central taxes and a cut in central expenditures. The combination that is finally chosen depends to some degree on how intergovernmental fiscal transfers work. If part of any central tax increase automatically flows out to subnational governments in the form of intergovernmental transfers, taxes need to be increased by even more to meet any central government deficit reduction target. On the other hand, if transfers are discretionary, cutting them is often chosen as a convenient way to pass on part (or all) of the cost of deficit reduction to the subnational governments. In this

M4439-BAHL_9781786435293_t.indd 23

01/03/2018 11:08

24

Fiscal decentralization and local finance in developing countries

and other ways, decentralized countries face different and sometimes more difficult problems in designing and implementing stabilization policy than do more centralized countries. Subnational Government Borrowing A related problem that has been much discussed is the danger of default on debt by state and local governments, and subsequent bailout by the central government. When Brazilian state governments defaulted on debt in the early 1990s, it precipitated a national fiscal crisis and forced central government intervention (Ter-Minassian and Craig, 1997; Rodden, 2003; Dillinger et al., 2003). The states correctly perceived that they faced a ‘soft’ budget constraint – a concept discussed further in Chapter 5 – because they would be bailed out by the center. As a result, many over-borrowed and overspent relative to their revenue inflow, and then had to be rescued by the central government. Subsequently, a fiscal responsibility law was enacted to impose discipline on the states – as well as on the federal government – to prevent such problems in the future (de Mello, 2007; Rezende, 2007). Somewhat similarly, in Argentina provincial government borrowing from banks run by the provinces went unchecked in the 1990s. By the time the situation led to a similar bailout by the central government in 2000, many provinces had committed more than 60 percent of their intergovernmental transfers as repayment guarantees (Webb, 2004; Braun and Webb, 2012). Examples like this do not mean that subnational government borrowing is inherently bad. In fact, as we discuss in Chapter 4, such borrowing is often both desirable and efficient when it comes to financing investment in long-lived infrastructure. The problems mentioned above arise because regional and local governments are too often not held to a hard budget constraint by the market (which believes they will be bailed out by the central government) and not subject to a proper regulatory framework that ensures they can repay the loans they secure without being bailed out. In the end, therefore, the real problem is not so much that subnational governments are prone to borrow more than they should, but that the central government has not placed sufficiently credible controls in place to ensure that they will not. The central government creates a ‘moral hazard’ by standing ready to provide a bailout for local governments that over-borrow. Alas, like many people, few governments seem able to withstand temptation long when there are no consequences from sinning. The simplest way to avoid such problems, though not the best, is to prevent subnational governments from borrowing at all. In Egypt, for example, 98 percent of government revenues are raised by the central

M4439-BAHL_9781786435293_t.indd 24

01/03/2018 11:08



Fiscal decentralization 101 ­25

government, which also makes most government expenditures (86 percent) (Smoke, 2013) In such circumstances, it is simple to centralize all borrowing, with the proceeds being spent directly by the central government or assigned to regions as capital grants. A less centralized, and arguably better approach is to permit subnational governments to borrow through a centralized agency. The most decentralized approach is to establish a clear central regulatory framework and to require all subnational borrowing to follow the rules.42 We return to this issue in later chapters. Efficiency Fiscal decentralization will lead to a different package of public services than will centralization. For example, while national investment spending is more likely to focus on infrastructure with both regional and national benefits such as large irrigation projects, national (interstate) roads and power, local governments are likely to focus more on programs that benefit their constituency, which may bias them toward social development projects. But, as we discuss in Chapter 4, subnational governments are responsible also for much growth-facilitating investment, for example, on transportation networks. Some subnational ‘consumption’ expenditure on education or health may also do more for growth than such national ‘investment’ expenditures as the construction of a new presidential palace or a highway to the favored summer resort of political leaders. Central officials often express concerns about the capability of regional and local officials to plan and execute projects. Sometimes these concerns are warranted. If this is a problem, however, the best response is to do more to build capacity at the subnational level, not for the central government to take over everything. Similarly, while it is clearly correct to say that local governments have little incentive to take spatial externalities into account, the answer is not for the central government to eliminate local decision-making power; rather, as we discuss in Chapter 7, it is to design a transfer system to provide the needed incentive by ensuring that those who make local decisions face ‘prices’ that motivate them to take such externalities into account. The larger the role regional and local governments play in establishing and maintaining growth-facilitating physical and human infrastructure, the more important the central government’s role as coordinator and facilitator of such investment becomes. Decentralization does not mean that the central government’s role in development policy necessarily becomes smaller. But it does mean that the nature of that role changes substantially from doing things itself to ensuring that others have the incentives and ability to do them, and that they in fact do so – ‘steering’ not ‘rowing’ as Osborne and Gaebler (1992) put it.43

M4439-BAHL_9781786435293_t.indd 25

01/03/2018 11:08

26

Fiscal decentralization and local finance in developing countries

Concerns about local capacity are by no means confined to investment projects. Anyone who examines decentralization in any developing country quickly learns that the major central government reservation is often that the capacity of local governments to deliver services is too weak to risk devolving responsibility for important services to them. Sometimes such concerns may be put forward primarily in an attempt to maintain control and power at the central level. Sometimes, however, such concerns are valid and well advised, often particularly in the poorer (and most needy) regions of a country. This is one reason experts often suggest that any decentralization should be asymmetric, assigning expenditure responsibilities (and revenue powers) to subnational governments as and when they are able to carry them out effectively.44 We discuss later why such advice seems seldom to be well received. The combination of economies of scale, the superior ‘know-how’ of central agencies and the commonly observed tendency to deliver services in the same way in all parts of a country is often taken to mean that centralized provision will deliver services at lower unit costs than decentralized provision. This argument is not always correct. There are often important diseconomies of scale. Some public services require local factor inputs (for example, land and labor) that can best be managed locally, and the delivery of many services requires familiarity with the local area to deliver the service in an efficient way.45 Examples include refuse collection, traffic control and probably primary education. Other services such as secondary education may perhaps be better provided within an area large enough to allow scale economies to be captured but small enough for local voters to have some significant say in how such services should be delivered and financed. A more decentralized structure may also mean more duplication of services and more government employment. A large part of the problem is often unskilled human capital. In Luanda, Angola’s capital, for instance, as recently as the 1980s, 29 percent of civil servants had only primary school education, and only 7 percent had a degree from an institution of higher learning, with the situation being presumably even worse outside the capital city (Management Systems International, 2008). Better-qualified people are more likely to be drawn to the central government, where their opportunity for advancement appears to be greatest and they are paid more. Moreover, when central officials have been doing a job for a long time, whether it is collecting taxes or delivering specific services, they are presumably further up the learning curve than their counterparts at lower levels of government. And, of course, they are also unlikely to be willing to give up their jobs or to transfer to lower levels of government. Stories of service delivery failure by newly empowered local governments

M4439-BAHL_9781786435293_t.indd 26

01/03/2018 11:08



Fiscal decentralization 101 ­27

are not difficult to find. For example, Uganda embarked on an ambitious program of fiscal decentralization in the 1990s without adequately preparing local governments to assume their new responsibilities. The ­resulting failures in financial management and service delivery significantly ­weakened the case for further fiscal decentralization (Steffensen, 2006; Smoke et al., 2010). In Sierra Leone, local councils were given authority to recruit their own technical/professional staff, following guidelines set up by the Local Government Service Commission. However, the skill level of newly recruited staff was much below that of the deconcentrated central officials who previously had the posts in question because more qualified individuals could not be attracted by local governments with low rates of remuneration and uncertainties about career advancement duties (World Bank, 2009). On the other hand, as Jibao and Prichard (2015) show, some cities in Sierra Leone managed to do much better than others despite operating in the same environment. Generalizations about local government performance even in very poor countries always need to be viewed in the context of the underlying situation. In countries with stronger subnational taxes and considerably more experience with local administration, such as Brazil, some states reportedly have tax administrations at least as good as the central administration, and subnational governments have a comparative advantage in administering most subnational taxes (Pinhanez, 2008). In others, for example Colombia and South Africa, the level and quality of administration in such large cities as Bogotá and Cape Town is good, and these governments have had fewer problems in attracting skilled employees. Moreover, local administrative capacity can sometimes be quickly improved. In Indonesia, for example, a major decentralization effort in 2001 was largely successful because some 2 million central government employees who were already involved in providing the services in the local area on behalf of the line ministries were transferred to the jurisdiction of local governments. The result was that the quality and cost of these public services – now delivered under a decentralized system – did not suffer greatly (Hofman and Kaiser, 2004). As this experience suggests, one route to successful decentralization may sometimes be to begin with deconcentration – provided that those already experienced in running the services in question are kept in place at least to some extent (Bahl and Martinez-Vazquez, 2006a). Another route is to identify areas where subnational governments have some comparative advantage in principle, and then to begin by devolving greater responsibilities first to those who are demonstrably most competent to deliver them properly. As mentioned earlier, this approach will initially produce an asymmetric system in which the more advanced (and usually more urban) areas will have more expenditure and revenue

M4439-BAHL_9781786435293_t.indd 27

01/03/2018 11:08

28

Fiscal decentralization and local finance in developing countries

raising ­responsibility than others. Whether for this reason or because there are strong traditional, constitutional or political requirements to treat all regions or localities equally, asymmetry may prove unacceptable, even though equal treatment may produce very unequal results when applied uniformly to the usual local government mix of large metropolitan governments, smaller cities and towns, and villages and settlements varying widely in terms of poverty and remoteness. For many countries, and many parts of many countries, the ideal approach may not be either complete centralization or all-out decentralization, but some form of partial decentralization. Chapter 8 considers one important aspect of this question – the financing of large cities – in more depth. Yet another approach to the local capacity issue may be to accompany better training and preparation for local officials with a better central information system about what is really going on out there. ‘Big data,’ carefully used, can provide valuable information. It can serve not only as a basis for monitoring (and to some extent influencing) those decentralized activities where financing is still mainly national but also as a better starting point for the development and execution of decentralization policy in general.46 The capacity of subnational governments to deliver services has matured significantly in recent years in many low- and especially middle-income countries. More affordable technology, improved education of more skilled staff and better employment opportunities have all contributed to this improvement. Experts such as accountants, engineers and valuers are usually in short supply, and management systems in many local governments remain weak. But, particularly in the larger urban areas of many developing countries, local governments are much better than they were a few decades ago, which is one reason that decentralization is more likely to be effective and efficient in big cities (see Chapter 8). However, even countries where such improvements have been slow to develop have moved ahead with decentralization. In India, for example – where an important constitutional amendment in 1992 for the first time gave local governments an important role in the federal system (Rao and Singh, 2006; Bahl et al., 2010a) – there is still much to be done before that role becomes reality (Ahluwalia et al., 2014). In Brazil, where local governments have had an important constitutional role since 1988, their budgetary positions have been strengthened (Rezende, 2007), although again much more remains to be done (Ter-Minassian, 2015). Even when, as in the cases just mentioned, subnational governments have an important role in service delivery, their capacities to assume these responsibilities may be very different. For the larger urban governments and for regional governments the economic case for more fiscal ­responsibility is usually clear. For smaller and more rural local govern-

M4439-BAHL_9781786435293_t.indd 28

01/03/2018 11:08



Fiscal decentralization 101 ­29

ments, however, prudence may often suggest a more modest approach with respect to transferring responsibility for service delivery (perhaps allowing smaller local governments to take on more when they can do so effectively). As we discuss in more detail in later chapters, an asymmetric approach to decentralizing responsibility for public service delivery is generally ­advisable – though perhaps not always politically feasible. Corruption The early students of fiscal decentralization were concerned that it may lead to a greater rate of corruption. Prud’homme (1995) and Tanzi (1996), for example, suggest that the ‘closeness’ between elected local politicians and the local political power structure breeds corruption. It is not difficult to think of other reasons why decentralization might lead to more corruption. For example, decentralization often means less and weaker central monitoring and control, and hence more opportunity to steal. Increased direct contacts between the public and lower-paid local public officials may lead to more bribery and corruption. Local citizens may have little experience or knowledge of how to monitor and discipline local politicians and officials. Such problems may perhaps be most serious with respect to infrastructure where there is more latitude for fraud, bribery, embezzlement and patronage – although, as usual when it comes to inherently unrecorded activities like corruption, the evidence is far from clear.47 For example, although Mauro (1995) argued that corruption is more likely to raise infrastructure spending (higher unit costs), Tanzi and Davoodi (1997) argued that corruption will lower infrastructure spending (fewer projects will be undertaken). The empirical evidence that corruption costs are greater under a more decentralized system is neither clear nor convincing (Shah, 2016). Estimates are inherently difficult because data are scarce and because the conceptual model is not easily worked out.48 The common perception of high local corruption may perhaps reflect its greater visibility. Corruption at the central level may be much greater though less obvious. Even if smaller in scale, however, local corruption may often be damaging to building trust in government. At whatever level it occurs, as Estache (2006) notes, corruption is a symptom of a deeper underlying problem – the lack of political commitment and accountability. As Bardhan and Mookherjee (2000) show, simply financing local infrastructure through user fees (as discussed further in Chapters 5 and 8) rather than local taxes or intergovernmental transfers will reduce corruption, no matter how poorly local democracy works. Of course, when politicians and officials gain much of their income from exploiting their monopoly power to grant licenses, bestow contracts or

M4439-BAHL_9781786435293_t.indd 29

01/03/2018 11:08

30

Fiscal decentralization and local finance in developing countries

provide services, they are unlikely to give up such power (and income) without a fight. It is seldom easy to design or implement feasible solutions to long-established corruption networks short of a fundamental revision of the relationship between state and citizens – even though doing so is itself likely to be an essential element of any such revised relationship.

CONCLUSIONS Fiscal decentralization is not a cure-all for governance and service delivery problems. In theory it can help, and some countries have turned this theory into practice with good results. But most low- and middle-income countries are unlikely to be able to realize these potential benefits immediately and fully, and even those that could do so may not be able to muster and sustain the necessary political will. It is thus not surprising that despite the burst of literature and discussion about fiscal decentralization in recent decades, there is not all that much evidence of effective fiscal decentralization on the ground in low- and middle-income countries. Based on the International Monetary Fund’s (IMF’s) Government Finance Statistics – the only comparable (though limited) data source available – the subnational government share of public expenditures in developing countries has not experienced a significant uptick over the last two decades.49 There may have been a great deal of progress with fiscal decentralization in more subtle ways that would not necessarily show up in the subnational government expenditure share – for example, by removing central mandates on expenditures, replacing conditional with unconditional grants, and more local cost recovery from user charges – but we are not aware of any systematic evidence to this effect. The strengthening of subnational government finance may perhaps be likened to buying more of a luxury good. Richer countries find fiscal decentralization more affordable because their economies are more stable, regional wealth disparities are less pronounced, the infrastructure is mostly in place, and the administrative capacity of their subnational governments is stronger. The big battles about which level of government will do what and how it will all be financed are in the past. The story is very different in most low- and middle-income countries where the political costs of decentralizing are sometimes considerably higher. Central bureaucracies are usually firmly entrenched and difficult to move. Even at the local level, the tradition of central control often seems inviolate. In most countries, local voters – those who in principle would benefit most – are either impotent or do not understand that decentralization can be a desirable and sustainable component of a sound development policy. It is not easy to evaluate how successful decentralization has been –

M4439-BAHL_9781786435293_t.indd 30

01/03/2018 11:08



Fiscal decentralization 101 ­31

however ‘success’ is measured – in any country, let alone to evaluate that success in statistical terms. Data deficiencies and the complexity of the issues often make comparative statistical studies less informative than careful country case studies, although one must also be cautious in drawing generalizations from even the best-structured comparative case studies. One lesson that does emerge clearly from most case studies, however, is that the design of decentralization efforts has often been flawed, that implementation is complicated, and that it almost always takes a long time for any impacts to show up clearly enough to be evaluated in any meaningful way. Nonetheless, fiscal decentralization, if done right, can be an important component of a sustainable development strategy in many countries. Although countries usually embark on decentralization for essentially political reasons, they do so in better and worse ways from a developmental perspective. Our aim in this book is to provide some guidance to those involved in such activities by reviewing the evidence to date and considering in some detail what theory and experience suggest are the best ways to go when it comes to the never-ending task of adjusting governance structures to changing circumstances in ways that best satisfy what people want their governments to do. Summing up, decentralization as we consider it here is essentially a way of ensuring that people get what they want from government. Done correctly, it can be beneficial. But it is neither easy nor cheap to do in most circumstances, so it is worth spending considerable time and effort in working out exactly what it is you want to do, why you want to do it, and how you can best do it. It is also important to set up the appropriate institutional arrangements within which decentralization has a reasonable chance to be successful, and to do this well before engaging in the process – for it is a process, and an ongoing and evolving one at that. Given the complexity of the exercise, it is not surprising that few developing countries have moved very far in the direction of the kind of decentralization discussed here. However, many have begun to move down this path, and it is important that the considerable efforts already made and in train are not wasted. Decentralization can be rewarding from a developmental perspective in many ways. But doing it right requires deeper and more sustained efforts than seems generally to have been realized. It is not surprising that to date experience in many countries has not been all that good. Each country at each point in time is a unique case. But all face some common problems, and there are usually a variety of ways – often already tried elsewhere – in which they can attempt to deal with these problems. Our hope is that this book may perhaps help some to avoid common pitfalls and offer some clues about how countries may do better in decentralizing in the specific circumstances they face.

M4439-BAHL_9781786435293_t.indd 31

01/03/2018 11:08

32

Fiscal decentralization and local finance in developing countries

NOTES   1. In addition to many good regional and country studies, the more general studies include Bahl and Linn (1992), Dillinger (1994), Tanzi (1996), Ter-Minassian (1997), Litvak et al. (1998), Bird and Vaillancourt (1998), Bardhan and Mookherhee (2006), MartinezVazquez et al. (2006b), Smoke et al. (2006), Boadway and Shah (2009), Eaton et al. (2010) and Bahl et al. (2013).   2. This wording embodies two strong assumptions. The first is that good governments are those that do their best to fulfill the wishes of the people whom they govern – an unspoken assumption in most analysis of public sector economics. The second is that the principal way we know to ensure that governments are accountable to their people – at least periodically and to some extent – is through democratic elections. The first assumption is of course more a wish than an expectation in many countries. The second is contrary to the reality found in many developing countries. We return to both points later in the book.   3. In federal countries, such as the United States and Canada or Brazil and India, the division of spending and taxing powers between levels of governments is specified to v­ arying degrees of detail in the constitution, and not easily subject to legislative change. We discuss some issues that may arise in federal countries as a result of such ­constitutional requirements, but generally assume here that we are dealing with countries in which the extent of decentralization is largely a matter for the central government to decide.   4. To make things simpler in this book, we often refer to ‘local governments’ instead of using such more accurate but cumbersome expressions as ‘subnational’ or ‘sub-central’ or ‘regional and local’ governments. When relevant to our story – for instance, in federal countries in which some subnational governments have different constitutional powers (for example, regional governments in some countries may have essentially the power of a ‘central’ government with respect to the structure and powers of lower levels of government within their territory) – we distinguish between the different levels of ‘local’ government. In this book, however, we do not treat the special case of federal countries in detail: on this, see Slack and Chattopadhyay (2009), Bizioli and Sacchetto (2011), and Slack and Chattopadhyay (2013). For further discussion of the difference between federal finance and fiscal federalism, see Bird and Chen (1998).   5. In a well-documented and reasoned book, Treisman (2007) argues that no one can yet make solid generalizations about decentralization. We reach much the same conclusion in our review of the empirical evidence in Chapter 2. As Treisman concludes, however, decentralization can be good policy if it is done ‘right’ – that is, right in terms of the prevailing objectives, conditions and constraints – and if the gains are worth the costs inevitably involved in restructuring institutions. We agree, and see this book as an attempt to provide clearer guidance to would-be decentralizers seeking to do the best they can in the conditions they face.  6. Of course, deconcentration does not always work so smoothly. At one time in the Philippines, for example, nurses at some hospitals were employed by three different levels of government – central, provincial and local, with each category receiving different wages for essentially the same work. Much the same was true with respect to teachers in Colombia in the early 1990s (World Bank, 1996), although in this case the rationale was less orders from above than attempts from below to offset the effects of higher-level controls. Such situations are more likely to yield unhappy providers than good services.   7. See Spahn (2015) for an interesting discussion of such ‘contract federalism.’ For discussion of several interesting examples of intergovernmental contracting in Colombia, see World Bank (1996).   8. As Congleton (2015, p. 139) says, with asymmetrical information “decentralization is not an exogenous feature . . . but rather an endogenous result of ongoing negotiations over the assignment of central and local authority.” For a good appraisal of this issue

M4439-BAHL_9781786435293_t.indd 32

01/03/2018 11:08



  9.

10.

11.

12. 13. 14. 15.

16.

17. 18. 19. 20. 21. 22.

Fiscal decentralization 101 ­33 in the general context of federal arrangements, see Bednar (2009), and for an interesting early treatment, see Cremer et al. (1995). For an example, see World Bank (1996). A particularly egregious instance in Colombia occurred when the central government mandated that every local government must provide free telephone access to the central government so that residents could register complaints about their local government. In some remote rural areas the cost of complying with this mandate exceeded the entire municipal budget. While fiscal decentralization is about the devolution of budget autonomy, the extent of such autonomy and how it works are obviously related to the number, size and diversity of local governments operating in a country, irrespective of their fiscal powers, as is discussed with some care in Lago-Peñas and Martinez-Vazquez (2013). The dollar amount of support is not easily estimated because it requires prorating loan amounts between the decentralization component and other targets of support, allocating general overhead spending to decentralization, and even defining what is or is not a decentralization project. For example, should an urban transportation project that would empower a local government to deliver bus services be classified as a ­‘decentralization’ project? For a more recent study covering a broader range of donor assistance to larger urban areas, see Kharas and Linn (2013). The linkage between separatist movements and fiscal decentralization is discussed later in this chapter, and in more detail in Bird and Ebel (2007) and Vaillancourt and Bird (2016). We take up the discussion of the relationship between decentralization and growth in Chapter 2. As noted earlier, Russia has seen many changes, first towards decentralization then back to a more centralized structure: for glimpses of this changing picture, see, for example, Treisman (1999) and Martinez-Vazquez et al. (2008). For example, for three months no provincial monthly financial reports, which are supposed to be reviewed by the central ministry, were received. On investigation, it turned out that the reports had arrived but that procedure required their receipt had to be officially logged before they passed to the review section. This did not happen because the official responsible for logging incoming reports was on leave and, on his return, did not bother to catch up with the backlog. No one noticed. Sometimes in some countries the only way a central official could visit certain regions was with a military escort. It is not surprising that at times the only contact such regions had with the central government was when someone (and their escort) came to collect taxes. Regional conflict is sometimes handled quite differently. During Colombia’s ­prolonged armed conflict, for example, some municipalities under rebel control regularly received most of the national transfers to which they were legally entitled. China has 151 prefectures and 185 prefecture-level cities; 1,903 counties and 279 county-level cities; and 56,000 townships, towns and city districts. India has 3,609 urban local bodies and, in rural areas, 474 zilla parishads, 5,906 panchayats samithis and 227,698 gram panchayats. For a review of the very different ways and extent to which decentralization occurred in different parts of Soviet-controlled central and eastern Europe after the late 1980s, see Bird et al. (1995). One well-documented example is that of Madhya Pradesh state in India, where teacherabsentee rates were much more effectively monitored by local communities than by the state government (McCarten and Vyasulu, 2004). http://www.citymayors.com/government/mayors-americas.html (consulted February 2, 2016). Even when elections exist and are free from corruption, how accurately and effectively they reflect local preferences depends on many context-specific factors, including: the nature and role of political parties; the structure of the electoral system; the rules determining eligibility to vote; and the extent and nature of voter participation. The extensive literature exploring these and other problems in establishing even an imperfectly functioning democratic electoral system cannot be further explored here, however.

M4439-BAHL_9781786435293_t.indd 33

01/03/2018 11:08

34

Fiscal decentralization and local finance in developing countries

23. For a limited study at the village level that briefly explores this issue, see Bird et al. (2011). 24. For some interesting takes on how government structure can get in the way of accountability, see Eaton et al. (2010). 25. For an interesting account of Brazil’s experience with ‘open budgets’ as well as other approaches to greater budgetary transparency, see Khagram et al. (2013). 26. The role of the media in informing people about the activities of local government is also important in the accountability process. While this issue seems to have been relatively little studied, one study of the local editions of three widely circulated newspapers in rural Kerala province (India) found that the space allocated to news about local government was only a little over 3.5 percent of the total space available (Sethi, 2005). 27. See, for example, the interesting discussions of rural political unrest in China in Bernstein and Lü (2003) and O’Brien and Li (2006). 28. Tax data are from various sources and various dates (Bird, 2012). 29. Even this is probably an overstatement because many countries do not report the finances of their local governments, and because some of the “taxes’ attributed to subnational governments are really, as we discuss in the next chapter, intergovernmental transfers in disguise. 30. Bahl and Cyan (2011) found some evidence of such ‘crowding out’ in Organisation for Economic Co-operation and Development (OECD) countries but not in developing countries. 31. As we discuss further in Chapter 5, such arguments often confuse economic and administrative efficiency (Bird, 2015). 32. The Sri Lankan case is not discussed further in this book. Moore (2017) provides an interesting look at recent trends in public finance in Sri Lanka but, interestingly, does not even mention the issue of decentralization. 33. For further discussion, see Bird and Ebel (2007) and Vaillancourt and Bird (2016). 34. See, for example, the seminal studies in Brautigam et al. (2008), and especially the recent detailed exploration of Ethiopia, Ghana and Kenya in Prichard (2015). 35. One of us once protested some arbitrary decisions made by a state-appointed tourist ‘guide’ in the Soviet Union, only to be quickly told by fellow tourists (all from Sovietdominated countries) that “the only safe thing to do is to keep your head down and your mouth shut.” Alas, this is still sound advice for both visitors and residents in all too many countries around the world. 36. There is some evidence from developed countries that smaller sized governments lead to increased political involvement (Pommerhene and Schneider, 1983). In early surveys in the US, respondents consistently identified local government as the level of government in which they have the greatest trust (ACIR, 1994). A similar finding emerged from a survey in Colombia (Acosta and Bird, 2005). 37. See the case studies in Bird and Ebel (2007) and Vaillancourt and Bird (2016). 38. However a few years later, when Russia again became a much more centralized system, these treaties were recalled (Martinez-Vazquez et al., 2006c). 39. Interestingly, sometimes national governments have deliberately continued to pay transfers to regions under rebel control both to keep lines of communication open and to, as it were, pay Danegeld (a term from early British history, when bribes were paid to persuade Danish invaders to refrain from armed assault). 40. More detailed discussion of this subject may be found in Bahl and Linn (1992), Prud’homme (1995), Ter-Minassian (1997), Tanzi (1996) and Spahn (1997); an excellent recent overview is Boadway and Shah (2009). For an empirical analysis, see Wibbels (2000). 41. For a set of interesting case studies, see Eccleston and Krever (2017). 42. For a good discussion of regulatory frameworks for subnational government borrowing, see Liu and Waibel (2010). See also Van Ryneveld (2006). 43. For an interesting application to local government in two OECD countries, see Barlow and Robler (1996).

M4439-BAHL_9781786435293_t.indd 34

01/03/2018 11:08



Fiscal decentralization 101 ­35

44. For extensive discussion of such ‘asymmetric federalism,’ see Bird and Ebel (2007) and Congleton (2015). As Acosta and Bird (2005) discuss, Colombia seems to have been fairly successful in decentralizing most education and health services to the regional level asymmetrically, with the central ministries remaining more directly involved in the less-advanced regions. Colombia’s experience with its earlier decentralization of most responsibilities for most local water services to municipal governments also appears to have been generally successful, though again significantly more so in some parts of the country than in others (Granados Vergara et al., 2008). 45. Even in industrialized countries, there is evidence that the ‘optimal’ size delivery unit for many local public services is in the range of 20,000–40,000 people: see the studies summarized in Slack and Bird (2012). This issue is discussed further in Chapter 3. 46. Indonesia and South Africa are two examples of countries that significantly updated their database on subnational government finances in the aftermath of fiscal decentralization. At the individual level, India’s introduction of a national biometric identification system obviously has potentially major implications for improving service delivery at the local as well as the national level. 47. The fear that ramping up local infrastructure spending is likely to increase corruption is by no means confined to developing countries. In Canada, for example, when a major federal program was launched in 2016 to spend billions on local infrastructure across the country the federal Competition Bureau felt it necessary to double its training workshops for public procurement officers “on how to identify and prevent collusion and corruption” (Curry 2016). 48. For reviews of this literature, see Martinez-Vazquez et al. (2007), Boadway and Shah (2009) and Shah (2016). 49. We measure decentralization here as the subnational government share of total government expenditure in the country – that is, subnational government expenditures in the numerator, and total central plus subnational government expenditures in the denominator. The limited utility of this measure is discussed further in the next chapter.

M4439-BAHL_9781786435293_t.indd 35

01/03/2018 11:08

2.  Has decentralization worked?* The ultimate outcome relies on a host of factors that vary from country to country. . . . Getting from ‘it worked there’ to ‘it will work here’ requires many additional steps. (Rodrik, 2015, p. 24)

In Chapter 1 we discussed the benefits and costs of fiscal decentralization, noting that how decentralization plays out in practice depends heavily on how the program is structured and implemented, with both aspects being shaped to a considerable extent by the characteristics of the country. In this chapter, we look at what the empirical evidence tells us about why countries adopt fiscal decentralization and about what its impacts have been.1 We begin by discussing the measurement of fiscal decentralization. What exactly are we trying to measure? How well can we do so with the available data? Next, we consider what answers the available data suggest with respect to two questions: why are some countries more decentralized than others? And what is the impact of fiscal decentralization on economic development, on the well-being of the population, and on the size and quality of government? The complexity of the issues and the mixed and variable quality of the evidence means that the answers to such questions inevitably remain less than crystal clear, but three conclusions emerge from this review. The first is that fiscal decentralization is likely to be more fully developed in larger and higher-income countries than in smaller and poorer countries. The second is that, as the epigraph to this chapter suggests, decentralization appears sometimes to have had some positive impacts in some countries – but not always or everywhere. And the third is that at least some of the common traps and obstacles leading to less favorable outcomes may have been avoided by following some of the guidelines we set out later in this book. Of course, this last conclusion assumes that a country not only wants to decentralize successfully but is also capable of doing what is necessary to do so.

MEASURING DECENTRALIZATION Measuring fiscal decentralization requires us first to decide how we should measure it and then to see how close we can come to this ideal with the data 36

M4439-BAHL_9781786435293_t.indd 36

01/03/2018 11:08



Has decentralization worked? ­37

available. Many empirical studies do this well. They carefully consider the normative question and note how the ideal differs from the measure used. However, other studies sometimes seem simply to have seized on some numbers they think provide a reasonable measure of fiscal decentralization; assumed the data adequately measure the normative concept they are attempting to measure; and then skipped forward to an interpretation of results that is often not well supported by the evidence provided. Most who work in this field, including us, have likely fallen into this trap at some point. It is not easy to settle on a single measure of fiscal decentralization because there are so many different dimensions of subnational government finance (Wasylenko, 1987; Martinez-Vazquez and Timofeev, 2009; Blöchliger, 2015). The right choice depends on exactly what it is that one is trying to measure. The definition of fiscal decentralization we offered in Chapter 1 – the empowerment of local populations through the empowerment of their elected local governments – suggests that a good measure should encompass both the share of government expenditures or revenues that are administered through subnational governments and the amount of discretion that subnational governments have in deciding how they will spend and tax.2 In fact, however, most existing comparative studies have simply measured decentralization by the IMF’s Government Finance Statistics (GFS) data, that is, by the share of spending that shows up in the budgets of local and regional governments (see Box 2.1). A few studies have used a more refined index that accounts for the discretionary power that subnational governments have over revenue decentralization (Ebel and Yilmaz, 2003; Stegarescu, 2005) but only very recently has similar attention been paid to the question of how much power they have over their expenditures. Another problem is that the subnational sector is usually made up of multiple levels of government. Many studies of decentralization use an aggregate measure of all provincial and local government spending as a proxy for governance at the local level. Doing so allows one to have a comparable sample because expenditure responsibility between regional governments and local governments may be divided in very different ways in different countries. However, lumping large and small local governments and regional (provincial or state) governments into a single subnational unit with very different compositions in different countries makes it hard to interpret the results of any analysis of the impact of specific decentralization measures, or to discuss in any meaningful way the effect of moving governance closer to the people. Not only is fiscal decentralization “notoriously difficult” to measure (Blöchliger 2015, p. 631), owing to its many dimensions and the many different institutional settings within which it takes place, but also different

M4439-BAHL_9781786435293_t.indd 37

01/03/2018 11:08

38

Fiscal decentralization and local finance in developing countries

BOX 2.1  COMPARATIVE CROSS-COUNTRY DATA The best comparative fiscal information available that includes developing countries is in the IMF’s Government Finance Statistics Yearbook (GFS), an annual compilation that provides data on the amount of taxes, expenditures and transfers that show up in subnational government budgets. The IMF goes to great lengths (and expense) to gather these data and to report them in a comparable way. It is a huge task to update the GFS on a regular basis and to include all countries for which adequate data can be gathered.* The coverage and availability of this data source makes it unsurprising that GFS data has been used in most comparative studies of fiscal decentralization. For this, we are all in the debt of the IMF. But the measure we get from the GFS leaves out some important information. Notably, it provides little information on the degree of discretion that subnational governments have in making expenditures or raising revenues. For example, an expenditure of $1000 determined solely by an elected local government and the same amount mandated for a specific use by the central government are reported in the same way by the IMF if both are recorded as local budget expenditures. Similarly, a tax whose rate and base are determined by a higher-level government is often reported in public accounts exactly like a tax where the subnational government can freely choose the tax rate (and perhaps even the base). In Germany, for example, where subnational governments have little power to determine the tax rate or base (Spahn and Föttinger, 1997), the GFS classifies a significant percent of total national taxes as raised by subnational governments. Tax sharing arrangements differ from country to country, and it is not always easy to know just how to interpret them with respect to the degree of subnational control. In Argentina, for example, the provinces must agree to any changes in the tax sharing (co-participation) agreement, while in Mexico subnational governments have no control over such arrangements (OECD, 2017). On the expenditure side, similar questions arise with respect to intergovernmental fiscal transfers (grants). Again, however, GFS data do not address this issue and lump conditional and unconditional grants under the same heading. GFS data provide valuable information on subnational finance for many countries, but only qualified inferences can be made from such data with respect to the extent and structure of fiscal decentralization (Litvack et al., 1998). On the revenue side, when GFS data are compared with an OECD (1999) data set that builds in a measure of discretion, they clearly overestimate the actual degree of revenue decentralization (Ebel and Yilmaz, 2003; Stegarescu, 2005). In a similar comparison, Saavedra (2010) found systematic overestimation of both revenue and expenditure decentralization in GFS data. Coverage is another problem, though the rate of coverage varies by year. Taking GFS (2009) as an example, out of 115 countries that could be classified as developing (or transitional), data on subnational government finances were reported for only 49 countries. No subnational data were included for such major countries as India, Indonesia, Brazil, Nigeria and Pakistan, often because the country did not track subnational fiscal data or failed to report it in a way that fit into the GFS format. This is not surprising, given that the GFS format was originally not designed to be used in analyzing subnational government accounts (Levin, 1991), but only for central governments. In many cases only central government statistics are included in the annual GFS data.

M4439-BAHL_9781786435293_t.indd 38

01/03/2018 11:08



Has decentralization worked? ­39

Recognizing some of these problems, at one point the World Bank set out to develop a new series of qualitative decentralization indicators, but subsequently abandoned this effort. Some useful work has been done by the OECD (as discussed in this chapter), but coverage remains limited. At present, there is no good internationally comparable data set when it comes to analyzing fiscal decentralization. The only data set that is broadly consistent over time is the GFS, so that is what we use in this chapter, supplementing it with information from some of the many country studies conducted by the IMF, the World Bank and others to fix some inconsistencies and broaden the sample, though no doubt at some cost in terms of comparability. Note:  * Recently, the International Centre on Taxation and Development (ICTD) made a major effort to clear up some inconsistencies and gaps in the GFS data base, and some recent studies have made good use of this data base. Unfortunately for researchers concerned with subnational finance, however, the ICTD data cover only central governments. See www.ictd. ac/datasets/the-ictd-government-revenue-dataset for the latest version of the ICTD data base and Prichard et al. (2014) for the original study.

measures are really needed for different purposes. There is no one-sizefits-all answer to the question of how to measure fiscal decentralization. As is always true in policy-oriented economic discussion, differing circumstances mean that the question can seldom be formulated in precisely the same way in any two countries. Considerable effort must be devoted to determining both the simplest meaningful model that can be applied to the cases included in the study and to working out how close one can get to approximating the appropriate measures needed to answer the question at hand. Expenditures The empowerment definition of fiscal decentralization that we presented in Chapter 1 leads us toward trying to measure the share of total ­ government expenditures that is determined by elected local governments. An algebraic definition of expenditure decentralization is straightforward: Let a 5 the percentage share of subnational government expenditures over which the subnational governments have discretion; LEj 5 the direct expenditure of the subnational governments in country j (including the expenditure of revenue from intergovernmental transfers); and CE j 5 the direct expenditure of the central government in country j (excluding intergovernmental transfers to subnational governments).

M4439-BAHL_9781786435293_t.indd 39

01/03/2018 11:08

40

be

Fiscal decentralization and local finance in developing countries

Then an index of expenditure decentralization (DE) for country j would DEj 5



     aLEj (2.1) LEj 1 CEj

where 0 ≤ a ≤ 1.3 Many empirical analyses (based on GFS data) implicitly assume that a 5 1, which means that all subnational governments have complete discretion to make budget decisions for the range of functions assigned to them. If they do, and only if they do, the share of expenditures administered through subnational government budgets is an accurate indicator of fiscal decentralization. Since in reality a is always between 0 and 1 – varying with the strictures of conditional grants; the extent to which expenditure mandates are imposed by the central government; whether local officials are appointed rather than elected; and the extent and nature of budgetary supervision and control exercised by the central government  – this assumption overstates fiscal decentralization in all countries, with an error that varies from country to country. At the other extreme, with a 5  0 for all local governments, there is no expenditure discretion. Subnational governments are simply spending agents of the higher-level government, and there is effectively no fiscal devolution.4 An attempt to measure expenditure decentralization by adjusting for the discretionary powers of the subnational government was made for some OECD countries by Bach et al. (2009), who distinguished five areas in assessing the degree of ‘spending autonomy’ of subnational governments: ●● ●● ●● ●● ●●

policy autonomy: how much control do they have over what they do and how they do it? budget autonomy: how much freedom do they have in determining how much to spend on various services? input autonomy: do they control, for example, wages and employment? Are they free to outsource services? output autonomy: how much control do they have over standards and service criteria? monitoring and evaluation autonomy: how much control do they have over evaluation and to whom do they report?

Using these rules for identifying local autonomy, they developed indicators of decentralization for primary and secondary education and public transport in several OECD countries. Three major conclusions emerged from this pilot study:

M4439-BAHL_9781786435293_t.indd 40

01/03/2018 11:08



Has decentralization worked? ­41

1. There are considerable differences between such ‘spending power indicators’ and simple subnational expenditure ratios in different countries: for example, Portugal’s local governments have far less control over education expenditure than the expenditure numbers suggest, and much less than other countries such as Germany. 2. A perhaps more surprising conclusion is that the difference between power and spending ratios is often greater between functions within the same country than it is between countries. 3. Finally, the ‘spending power’ of the subnational level is sometimes lower than the simple spending ratio suggests, as is generally the case with respect to education, where considerable controls are often exercised from above. In other cases, however, the opposite is true – for example, with respect to public transportation, where local governments may exercise power over the delivery of services without paying for them. While no such studies appear yet to have been carried out for developing countries, the conclusions would likely be similar owing to the greater variation in the capacity of local governments to deliver services and the higher level of centralization in most such countries. The bottom line here is that to sort out the expenditure discretion issue requires a detailed comparative case study by function and country. As OECD (2016, p. 147) correctly notes, “gauging spending power entails detailed assessments of each policy area’s regulatory environment and intergovernmental fiscal frameworks.”5 Local and regional spending is shaped by a complex political, economic and administrative system. To understand how decentralization works in any country, let alone to improve it, one must first understand in detail many aspects of the specific institutional context. There is no short cut. Only through detailed study of the country can one deal adequately with the many issues involved in determining how much discretion local and regional governments have. Such a study will almost certainly find, as did the OECD study cited above, that the extent and nature of discretion vary not only from function to function but also among regions and localities, with the nature and extent of variations depending on many different factors.6 Understanding expenditure decentralization in a country is thus both difficult and time-consuming. But such study is needed to understand what is going on sufficiently well to be able to discuss in any useful way how it might be performed better or reformed.

M4439-BAHL_9781786435293_t.indd 41

01/03/2018 11:08

42

Fiscal decentralization and local finance in developing countries

REVENUES As in the case of expenditures, comparative analysis of tax decentralization can be hazardous if the limitations of the data are not kept in mind. For example, only when effective tax rates are determined by subnational governments can such taxes be considered local taxes (Bird, 2011). Taxes that are assessed and collected by the national government and then redistributed to regional and local governments, although included as subnational revenues in many data sets, are clearly not local taxes. Such taxes are more properly treated as an intergovernmental transfer even when they are returned to the areas in which they are collected (Bahl and Linn, 1992; Bahl, 1999; Martinez-Vazquez et al., 2008). An index of tax decentralization in the jth country (DFj) can be defined as

DFj 5

     bjTj (2.2) Tj 1 CTj

where CT 5 total revenues raised from all central government taxes in country j; Ti 5 revenues raised from subnational government tax sources in country j; and bj 5 the percentage share of subnational government taxes over which the subnational governments in the country j have discretion. The term on the right side of equation (2.2) is the percentage of total national tax revenue that is raised at the discretion of the subnational ­governments. The tax discretion coefficient (bj) refers to the degree to which the subnational government can control the level of revenue raised. If bj 5 1, then the subnational government has complete control, as is true for many subnational government taxes in some OECD countries. As noted above, where bj 5 0 it is not a local tax at all but an intergovernmental transfer (as are most provincial and local government taxes in China, for instance). The value of b will lie between 0 and 1, though the specific ways in which subnational governments can exert discretion over the tax base and rate vary widely from country to country. In the United States, it is close to 1 and in China it is close to 0. Although the extent of local revenue-raising discretion is one of the most important and meaningful indicators of the extent of real fiscal decentralization, no country in the world regularly measures such a tax discretion coefficient, and no internationally comparable data exist for developing countries.7 One reason is simply because it is difficult to measure the degree of tax discretion: see, for example, Box 2.2,

M4439-BAHL_9781786435293_t.indd 42

01/03/2018 11:08



Has decentralization worked? ­43

BOX 2.2 INDICATORS OF TAX AUTONOMY OF SUBNATIONAL GOVERNMENTS 1. Tax rate and base   a.  Determined by the subnational government (sng).   b.  Set by sng but with approval of higher-level government. 2. Tax rate   a.  Determined by sng with no restrictions.   b. Set by sng but within upper and/or lower limits set by higher-level government. 3. Tax relief   a.  Sng is free to grant tax relief in any form.   b.  Sng may grant tax exemptions.   c.  Sng may grant tax relief but only in the form of tax credits. 4. Tax-sharing   a.  In which sngs determine the revenue split.   b. In which the revenue split can be changed only with the consent of sngs.   c.  In which the revenue split established by legislation.*   d. In which the revenue split is determined annually by higher-level government. 5. Other cases in which higher-level government sets the rate and base of sng tax. 6. None of the above categories applies. Note:  * Split may be changed by higher-level government, but not annually. Source:  Adapted from Blöchliger (2015).

in which various possibilities are arrayed broadly in descending order of tax autonomy. Even this categorization does not capture all possibilities: for example, subnational governments may be able to decide whether or not to tax a particular base. In Pakistan, for instance, provincial governments have the option of levying a sales tax on services. Though all have done so and both the tax rate and base are determined by the provincial government, the provincial laws were written by the federal government and simply adopted by the provinces (Bahl et al., 2015). A few other examples may

M4439-BAHL_9781786435293_t.indd 43

01/03/2018 11:08

44

Fiscal decentralization and local finance in developing countries

be useful to underline the extent to which assessing the level of autonomy with which local governments determine the revenue they collect depends on close understanding of the entire governance and fiscal structure of each country: ●●

●●

●●

●●

Subnational governments may have the authority to set tax rates for some taxes. In most low- and middle-income countries, however, subnational governments are only authorized to set rates within limits (for example, with respect to property taxes in the Philippines or Malaysia). How much discretion this gives to a subnational government depends on how binding the rate limit is and whether the intergovernmental fiscal system provides incentives to set the rate at a particular level.8 When subnational governments can grant reliefs, as in the case of Brazil, the level of the revenue raised can be significantly affected. In other cases, however, costly exemptions (for example, of government property) may be established by national legislation or even in the constitution. Although the subnational government may not set the rate or base of a tax, it is sometimes responsible (at least to some extent) for assessment and collection, and may thus influence the effective tax rate and the level of tax collection. A celebrated example of this was in Russia, where the regional governments had no legal taxing powers but the local tax office of the central government was to some degree subordinate to the provincial government (Bahl, 1994; Kurlyandskaya, 2005; Martinez-Vazquez et al., 2008). Similarly, although the rate and base of the business tax in China was set by the center, the tax was administered at the local level and the local tax bureaus used this power to influence revenue collections with ‘backdoor’ collection methods (Cui, 2011; Bahl, 1999). Should failures to collect the tax according to the law, or reductions in the effective tax rate due to tax rebates be considered as ‘local government taxation’? Subnational governments may be allocated certain tax bases by the national constitution, thus prohibiting others from taxing those bases. In India, for example, the constitution says only the central government may tax production and only states can tax the sale of goods and specified services – a provision that has considerably complicated the structure of consumption taxes and made the introduction of a value-added tax difficult (Rao, 2009; Bird 2015a).

M4439-BAHL_9781786435293_t.indd 44

01/03/2018 11:08



Has decentralization worked? ­45

DETERMINANTS OF DECENTRALIZATION Why have some countries made more use of subnational government budgets and service delivery systems than others? How does the observed pattern square with what might be expected from the basic economic theory of fiscal decentralization as set out in Oates (1972) or the ‘secondgeneration’ theory of Weingast (2009)? Is there a benchmark indicating an average level of decentralization, or something else that may provide a useful norm against which to evaluate country policies? Questions such as these can be answered – to the extent they can be answered at all – only by comparative cross-country studies. Good fiscal decentralization requires careful analysis and consideration of the unique situation in each country. But even good case studies lack the comparative dimension needed to put matters into perspective and to provide some idea of how one might reasonably benchmark performance. Cross-country comparative studies can help us understand more about the principal determinants of decentralization and its impacts in different settings, and thus help in framing and appraising decentralization in any particular country. Countries often use the results from comparative analysis to benchmark their own progress relative to other countries, and sometimes even to establish a ‘yardstick’ against which to measure improvement.9 Local politicians seldom admit it, but emulating successful policies from other countries is a common stimulus for reform. It can be easier to sell something to the public if there is evidence that it has worked well (or not worked well) in other, comparable countries.10 An example we discuss further in Chapter 6 is the feasibility of adopting area-based property tax assessment. Comparative analysis is often used as the basis for broad assessments of the impact of fiscal policies on decentralization outcomes – for example, assessing the extent to which unconditional intergovernmental grants stimulate or substitute for local tax efforts – or to appraise whether debt levels are out of line with those in comparable countries (Canuto and Liu, 2013). Comparative analysis may also allow analysts and political outsiders to raise questions about the policies that have led their own country to be an outlier, e.g., whether the low level of subnational government taxation in Indonesia has compromised the success of its decentralization program (Directorate General of Fiscal Balance, 2012). The Level and Determinants of Fiscal Decentralization The place to begin understanding fiscal decentralization in developing countries is with an analysis of how well it is entrenched as a public financing

M4439-BAHL_9781786435293_t.indd 45

01/03/2018 11:08

46

Fiscal decentralization and local finance in developing countries

Table 2.1  Fiscal decentralization indicators (average values for data available for 2000s) Subnational governments Country classification

% of total expenditures

% of GDP

% of total taxes

% of GDP

High income

27.8 (26) 22.1 (23) 20.4 (28)

13.9 (26) 8.3 (25) 5.9 (32)

22.4 (25) 22.7 (24) 12.1 (28)

6.4 (25) 4.9 (25) 2.4 (32)

Transition Developing

Source:  IMF, GFS data and author calculations.

strategy and why some countries use it more than others. The data will not let us answer either of these questions the way we would like – by including an indicator of the fiscal discretion given to subnational governments in our measures – but we can make a good start on both. On average, subnational governments in developing countries account for about 20 percent of total government expenditures, which is equivalent to about 6 percent of GDP (Table 2.1).11 This is well below the average level in developed countries. If these data could be adjusted properly for the discretion factor, our experience in many of the countries included in this sample suggests that the gap between developed and developing countries would be even greater. On the revenue side, regional and local governments in developing countries mobilize an amount equivalent to only about 2.4 percent of GDP, and the average subnational government share of total taxes is only 12 percent, or about one-half the share in high-income countries. Specification of a model to estimate the determinants of inter-country variations in the level of fiscal decentralization depends on the hypothesis being tested. Spending as a share of GDP is an appropriate measure to assess the relative importance of subnational governments in shaping economic and social development. On the other hand, if the focus is on their role in governance and finance, the share of total public expenditures is a more appropriate indicator. A similar issue arises in determining what to include in the dependent variable for revenue decentralization. Is the best measure local taxes or local ‘own source’ revenues? The true financial importance of subnational governments would seem to be measured better if user charges, licenses and other non-tax revenue were included.12 Such revenues are

M4439-BAHL_9781786435293_t.indd 46

01/03/2018 11:08



Has decentralization worked? ­47

often directed to certain expenditures, with any shortfall from these sources – such as deficits from local transportation operations – often being financed by local government taxes. On the other hand, including user charges will make the decentralization index more sensitive to the assignment of expenditure responsibilities. For example, electricity might be delivered by subnational governments in country A, and therefore included in local government accounts, but delivered by the private sector in country B. Comparing local government revenues in A and B would then be misleading if user charges were included, just as comparing the functional distribution of expenditures would be misleading if an adjustment were not made.13 The case for including licenses and other fees and charges is stronger because they are more clearly a substitute for taxes. Indeed, in some instances licenses or fees are simply taxes imposed under a different name to get around some legal restriction or to hide them from the public.14 Early studies of the determinants of fiscal decentralization were simple (OLS) regressions based on GFS data (see Box 2.1). More recent research uses panel data with more sophisticated estimation techniques to take endogeneity issues into account, and thus provide a better picture of the underlying true structural relationship. Still, cross-section, times series data on the budget outcomes for subnational governments in developing countries limit what these analyses can reveal. Almost all analyses admit to specification errors owing to omitted variables. Researchers cannot control for all relevant factors in estimating the determinants of fiscal decentralization. For example, research often focuses on the income elasticity of demand for local public goods, but says little or nothing about the different price elasticities of different goods in different countries. Variables that would adequately proxy cultural influences on the demand for spending and taxes are not usually available at the local level, and almost never in a way that is easily comparable across countries. There are no good measures of the size and quality of the public capital stock, and even the share of the population living below the poverty line is often not measured in a comparable way across countries. Clever techniques are used in some studies to cope with such problems; but one can always argue that a different – and of course better – specification of the model might have yielded different results. Empirical Evidence Most studies have focused on three determinants of expenditure ­decentralization – the level of economic development, the size of the country and the heterogeneity of the population:

M4439-BAHL_9781786435293_t.indd 47

01/03/2018 11:08

48 ●●

●●

●●

Fiscal decentralization and local finance in developing countries

A positive relationship is expected between expenditure decentralization and the level of economic development. The perceived advantages of fiscal centralization – for example, with respect to macroeconomic control and interjurisdictional equalization – are expected to diminish as the income of a nation increases. Sometimes, expenditure decentralization might be likened to a luxury good, the demand for which increases only after reaching some threshold in terms of the level of per capita GDP (Wasylenko, 1987; Bahl, 2008). The level of expenditure decentralization is also expected to be positively related to size in terms of either population or land area. One reason may be the high costs of managing a large country from the center – costs that may in part result from the limited access of local officials to the central government, which may in turn arise partly from poor transportation and communication infrastructure. If so, what seems to be a size effect may be in part a remoteness effect. Panizza (1999) argues that larger land areas reflect lower population density and a large ideological distance from the median voter. In such circumstances, decentralization might lower costs by empowering subnational governments to manage more of their own affairs. Greater heterogeneity in the population is expected to result in increased decentralization, both to reduce costs by matching what local governments do with what local people want (rather than what the central government thinks they should have) and perhaps by damping down secessionist pressures that might otherwise be fostered by visibly enforcing the ‘rule of the other.’15 On the other hand, as discussed in Chapter 1, governments decentralizing for this last reason may sometimes find they are encouraging the very thing that they are trying to suppress – more demand for yet more local self-sufficiency in governance.16

Most empirical research on this question has concentrated on the expenditure side of the fiscal equation and used GFS data.17 As a rule, the dependent variable is the share of total government expenditure that passes through subnational government budgets, a measure that implicitly assumes that all subnational governments have complete control over their expenditures. Some studies have used own source revenues as a percent of subnational government expenditures as a dependent variable.18 Different studies used different data sets (panel or cross-section), included different countries (a cross-section of developed countries, or one group or the other), and used different model specifications and different independent variables. On the whole, most found that the income effect and the size effect were significant.19

M4439-BAHL_9781786435293_t.indd 48

01/03/2018 11:08



Has decentralization worked? ­49

The higher the level of per capita GDP, the greater the share of government expenditure channeled through local governments (Bahl and Nath, 1986; Wasylenko, 1987; Letelier, 2005; Arzaghi and Henderson, 2005; Bahl and Wallace, 2005). Richer countries often have a longer tradition of elected local government, and arguably more demand for local control over the provision of local public services. Moreover, in many rich countries the balance has probably shifted from the need for national infrastructure to establish the framework of the nation and its economy to services such as education and health that are often provided largely at the regional and local government level. While the question of whether there is a threshold level of income for a successful decentralization is still open, Sepulveda and Martinez-Vazquez (2011) reach the interesting conclusion – one that differs from the increased centralization traditionally associated with the question of the development of the ‘welfare state’ in developed countries (Wilensky, 1975) – that fiscal decentralization tends to be associated with increased income equality, although only after the size of government reaches 20 percent of GDP. Most empirical studies have found that size matters. Larger countries (by population size or land area) tend to flow more expenditures through subnational government budgets than smaller countries. Pommerehne (1977) finds a strong and robust relationship between fiscal decentralization and population size. Panizza (1999) finds a size effect using the land area of the country as the explanatory variable. Bahl and Wallace (2005) show that either land area or population size works equally well as a determinant of the level of expenditure decentralization. However, the evidence for the hypothesis that a more heterogeneous population will lead to more devolution of expenditure responsibility is much less strong. Panizza (1999) finds some evidence that countries with more ethnic fractionalization tend to be governed with more fiscal decentralization, but the results are highly sensitive to the composition of the sample. Bahl and Wallace (2005) and Letelier (2005) do not find a significant relationship between the degree of expenditure decentralization and ethnic fractionalization. Gomez-Reino and Martinez-Vazquez (2013) explicitly test the hypothesis that ethno-linguistic fractionalization will lead to a greater number of local governments per capita, but find that countries with more diverse populations are less fragmented. As we suggested earlier, these results may in part reflect the fact that what may dominate is not ‘fractionalization’ but ‘fragmentation’ and the fact that the latter may go either way when it comes to decentralization, depending on the specifics of each case (Vaillancourt and Bird, 2016). Our own estimates – based on the augmented sample of low-, middleand high-income countries for the 2000s used in Table 2.2 – do not give

M4439-BAHL_9781786435293_t.indd 49

01/03/2018 11:08

50

Fiscal decentralization and local finance in developing countries

Table 2.2  Determinants of fiscal decentralization Dependent variable Constant Per capita GDP Population size Federalism dummy Transition country  dummy R-squared Sample size

Subnational government expenditures as % of GDP

Subnational government taxes as % of GDP

−2.71 (4.83) 0.39*** (6.67) 0.27*** (3.76) 0.74** (2.78) 1.03*** (4.70) 0.57 71

−3.60 (5.14) 0.39*** (5.26) 0.17 (1.91) 1.21** (3.40) 1.52*** (5.55) 0.50 71

Note:  OLS estimates. Data are country averages for the 2000s; t-values shown in parentheses and all continuous variables are expressed in logarithms. ** denotes significance at 0.05 level and *** at 0.01 level.

markedly different results from those summarized above. Using the expenditure and tax shares of subnational governments as the dependent variable (with no adjustment for discretionary powers), we find both population size and per capita GDP to be significant and positively related to the share of fiscal activity that flows through subnational government budgets. Expenditure decentralization is significantly higher in transition countries and in federal countries. About 60 percent of the inter-country variation can be explained. Various studies have tested other hypotheses. For example, countries that invest heavily in the military or feel threatened by neighbors in the region tend to be more centralized (Letelier, 2005; Bahl and Nath, 1986).20 Huther and Shah (1998) develop an interesting set of indices of good governance – measures of political transparency and voice, absence of corruption, social development and equality, and a favorable climate for stable growth – and find these indexes to be significantly and positively correlated with the rate of expenditure decentralization for an 80-country sample of industrialized and developing countries. However, the direction of causation between good governance and fiscal decentralization remains to be sorted out. Colonial heritage may matter. Arzaghi and Henderson (2005) find that countries with a French law tradition tend to be more centralized.21 Other studies have found that local governments developed

M4439-BAHL_9781786435293_t.indd 50

01/03/2018 11:08



Has decentralization worked? ­51

much more strongly in English than in Spanish colonies in the Americas for a variety of institutional reasons (Sokoloff and Zolt, 2006). No doubt many other factors deserve closer attention. For instance, many countries have taken a sectoral approach to fiscal decentralization, in particular by assigning a significant portion of health and education expenditures to subnational governments. In an interesting comparative study of 29 low- and middle-income countries, DeLog and LPSI Secretariat (2015) finds that, on average, about two-thirds of total health and education expenditures at the local level are made through vertical (centrally financed and controlled) programs. Studies focusing only on local government expenditures on these services may miss a large part of the story. Health differs from education for several reasons. For example, as Sokoloff and Zolt (2006) argue, the much greater local control over education in English than in Spanish colonies was a principal reason for the greater development and strength of local governments in the former. Education at the primary and secondary level lends itself much more readily to local control and finance than health services, where externality problems give rise to greater need for coordination, and scale and scope economies similarly point to more centralized involvement in service provision. Letelier (2005) found that as income levels rose, the subnational government share of health expenditures declined. There is much more work to be done on understanding and analyzing the sectoral patterns of decentralization. Other empirical research has focused on the revenue side: What determines the share of total tax revenues raised by subnational governments? As in the case of expenditure decentralization, the tax decentralization measure used in most studies implicitly assumes that all local government revenues raised are a result of local discretion, that is, b 5 1, in the terms used earlier. Empirical analyses of intercountry variations in tax decentralization have focused on two questions. One is whether tax decentralization follows the same determinants as expenditure decentralization. An obvious follow-up question is why some countries choose more ‘vertical balance’ in their systems than do others.22 The results here could be consistent with the ‘finance follows function’ gospel commonly found in the literature, although the causal linkage does not seem to have been subject to rigorous empirical testing.23 The second question that has been explored is whether central governments are more willing to let subnational governments levy and administer their own taxes as they become more capable of doing so. Increased local capability and economic development should go hand in hand. Not only do the expenditure needs of subnational governments rise with income levels, but so does their access to such productive revenue sources as taxes

M4439-BAHL_9781786435293_t.indd 51

01/03/2018 11:08

52

Fiscal decentralization and local finance in developing countries

on vehicles and real property, and perhaps also sales and payroll taxes (see Chapter 5). On the other hand, tax decentralization might be held back in developing countries if central governments fear that they may lose some revenue themselves through competition for tax bases and perhaps end up with insufficient revenue flexibility to achieve their macroeconomic and stabilization goals.24 Studies attempting to explain inter-country variations in tax decentralization have been less successful that those focused on expenditure decentralization. One reason is because of the strong role history and culture play when it comes to determining how we tax ourselves (Bird, 2015). Although both Germany and China are quite decentralized on the expenditure side, neither country gives any significant taxing power to its subnational governments. In contrast, Canada, Brazil and the US have decentralized both expenditures and revenue raising, and Spain seems to be following a similar path. As usual, there is much in this world that cannot be easily understood through cross-country statistical analysis alone. Most econometric analysis of cross-country data finds a significant positive relationship between the subnational government share of tax revenues and per capita GDP (Letelier, 2005; Wasylenko, 1987; Bahl and Cyan, 2011). Perhaps unsurprisingly, federal countries appear to devolve more taxing powers than do unitary countries, although this proposition has not been rigorously explored. Bahl and Cyan (2011) find some evidence that subnational government taxes might crowd out central government taxes, although only at levels of subnational government taxation well above those in most developing countries. Our own analysis of the determinants of subnational government variations in the ratio of tax-toGDP finds a positive income effect with a significantly higher effective tax rate in federal countries as well as in transition countries (Table 2.2). To mention one last study focusing on a narrow causality question, Bahl and Martinez-Vazquez (2008), using panel data and treating decentralization as endogenous, find that expenditure decentralization is a significant determinant of property tax effort. Although this result is broadly consistent with the conventional finance-follows-function explanation, it too is of course vulnerable to all the issues discussed above with respect to data used and model specification. There is much still to be learned about the determinants of tax and expenditure decentralization and the connection between them.

DOES DECENTRALIZATION WORK? Those who advocate fiscal decentralization often promise many good results. Opponents are equally quick to demonize the whole process. Of

M4439-BAHL_9781786435293_t.indd 52

01/03/2018 11:08



Has decentralization worked? ­53

course, much political discussion about such matters simply amounts to telling people stories, however unrelated to reality, that they can identify with and that will incline them to support (or oppose) whatever it is the story-teller is pushing.25 Unfortunately, when it comes to the potential impacts of decentralization even the best politicians can do little more than tell the story that they want people to think is correct because good research has not given us many unassailable conclusions about the impacts of fiscal decentralization. Still, there are some tentative conclusions from the research that can help inform the debate about the impacts of decentralization.26 In the balance of this chapter, we review some of that evidence, noting that even when the evidence seems solid there is not always agreement about the conclusions that should be drawn from it. We begin with a brief sketch of some important methodological issues that arise in estimating the impacts of fiscal decentralization. The key problem is how to isolate the effects of fiscal decentralization from the effects of everything else, and how to account for the very different ways that countries do decentralization. aLE Recent research has reduced DEj 5 some of these problems. For example, LEj 1 CEallow fiscal decentralization to more sophisticated estimation techniques be treated as endogenous; and some new panel data sets have been used in creative ways, permitting better specification. Still, few researchers are willing to bet the farm on their estimates of the impact of fiscal decentralization. The subject is inherently difficult bjTi because the impact is often indirect and difficult to sort out.27 DF To jtake = a topic of current interest: suppose, for Ti 1 CT example, that the goal is to estimate the effect of fiscal decentralization (FD) on income inequality (IE). For purposes of illustration, we might argue that the marginal effect has three components:

0IE 0IE 0G 0C = · · 0FD 0G 0C 0FD

where G 5 economic growth and C 5 corruption. Theory suggests that fiscal decentralization may increase income equality; but whether it does so or notTmay depend on the interaction between the 0EG c 0EG 0C (presumed, for illustrative purposes) = negative a b aimpact b of decentralization on y 0FD of 0C 0FD corruption, the negative impact corruption on economic growth and the positive impact of economic growth on income equality. (And, of course, it depends on a host of other factors that need to be controlled.) A reduced model that estimates only the gross relationship between fiscal decentralization and income equality often misses the underlying structure, and hence & & ) much of the real story. Similar with most impact analysis. GAPproblems 5 a ( Earise 2 R i i i

VS 5 M4439-BAHL_9781786435293_t.indd 53

a (GAP) CR

Ri 5 g1

x1i

G 1 g2

x2i

G

01/03/2018 11:08

54

Fiscal decentralization and local finance in developing countries

A general problem that confounds all research in this area is that we seldom have the data needed to make a good estimate of impacts. Neither good measures of fiscal decentralization, nor the right control variables, nor adequate data on the instrumental variables that might be used to take account of endogeneity in the estimation are easy to find. All interpretations of econometric results are thus subject to many caveats.28 We do not discuss these basic problems further here, except to note that one must be very careful in drawing conclusions from such research either in general terms or with respect to any particular case. We do  know something from this research, and it is worth knowing. But the last word on the impacts of fiscal decentralization has not yet aLE aLE been heard. Here we DE simply review briefly what has been learned so far DE j 5j 5 LELE CE 1 CE of fiscal decentralization: on about four important potential j 1jimpacts economic growth, the size of government, the quality of governance, and equality. Economic Growth

bjTbi jTi DFDF j= j= CT Timove 1 CTup the economic growth ladder and Low-income countries wantTto i1 become high-income countries, so it is no surprise that some advocates of fiscal decentralization assert that economic growth will be faster in more decentralized countries. It is difficult to find empirical support for this proposition, in part because any such effects are likely to be indirect. There 0IE 0IE 0IE0G0G 0C0C may, for instance, be 0IE effects on corruption, income = =of · decentralization · · · 0C0C0FD 0FD 0G0G 0FD inequality, technical 0FD efficiency, revenue mobilization and even conflict resolution that may in turn affect economic growth. Picking up on an earlier example, the impact of fiscal decentralization on economic growth might be measured as

Tc Tc 0EG 0EG 0EG 0EG 0C0C = a= a b ab a b b y y 0C0C 0FD 0FD 0FD 0FD

where EG 5 economic growth, FD 5 fiscal decentralization and C 5 corruption. If fiscal decentralization leads to lower levels of corruption, and less & & & & corruption is growth-enhancing, effect of fiscal decen(then GAP 55 E(i 2 R GAP Eithe 2i )Rmarginal i) a a tralization on economic growthi would be positive. This approach seems a i sensible way to explore why – though not how – fiscal decentralization may affect the rate of economic growth.29 Such indirect effects are complicated and not easily sorted out. For example, if decentralization also leads to an increase in revenue mobilization, would the net effect on economic ) ) a (GAP a (GAP growth be positive or VS negative? 5 5 The answer depends on the relative size of VS CRCR x1i x1i x2i x2i g2 g2 G G Ri R 5i 5 g1 g1 G 1 G1 x x x x aa aa & & Ri Ri M4439-BAHL_9781786435293_t.indd 54

01/03/2018 11:08



Has decentralization worked? ­55

the ­possible interaction effects, and is seldom either obvious or likely to persuade people who think those effects are different.30 The empirical test traditionally used to estimate the impact of fiscal decentralization on economic growth (essentially measured as increases in per capita GDP, though often referred to by the more multi-dimensional label of economic development) is relatively straightforward. The dependent variable is the level of per capita GDP or the growth rate in GDP. The independent variable of interest is fiscal decentralization (either the subnational government expenditure share or the subnational government tax share) and the control variables are other factors that affect the rate of economic growth. The hard part is to identify all the other relevant factors and to separate their impact on growth from that of fiscal decentralization. Unfortunately, the combination of our uncertainty about what leads to growth, the difficulty of dealing fully with the endogeneity issue and the absence of data is almost fatal to this endeavor. It is thus not surprising that such research has produced no firm conclusion. Among the broader cross-country studies that used the expenditure share as the independent variable, Martinez-Vazquez and McNab (2001) found a positive relationship with income level, as did Zhang and Zou (1998) and Lin and Liu (2000). However, other studies using roughly the same approach did not find a positive relationship (Davoodi and Zou, 1998; Rodríguez-Pose and Bwire, 2003; Rodríguez-Pose and Ezcurra, 2011). Studies measuring fiscal decentralization as the subnational government share of total revenues raised also found no positive relationship between the subnational government revenue share and economic growth (e.g. Woller and Phillip, 1998). Other research has studied the hypothesis that decentralization within a country will lead to a faster rate of economic growth; i.e., that regions that are more decentralized will show a higher rate of growth. Again, the results are mixed. In studies of Chinese provinces, for instance, Zhang and Zou (1998) found a negative relationship, but Qiao et al. (2008) and Lin and Liu (2000) – using a different specification – found a positive relationship.31 Similarly, Neyapti (2006) found that decentralization stimulated the economic growth of Turkish provinces, but Tosun and Yilmaz (2008) did not reach the same conclusion. Devkota (2014) found a positive relationship in a study of districts in Nepal. Boadway and Shah (2009) suggest that the empirical evidence is broadly supportive of a positive influence of decentralization policies on economic growth. We agree more with Martinez-Vazquez and McNab (2001) and Breuss and Eller (2004) that the empirical evidence is not convincing, for at least two reasons. First, the effects of fiscal decentralization have not been adequately distinguished from all the other factors that may

M4439-BAHL_9781786435293_t.indd 55

01/03/2018 11:08

56

Fiscal decentralization and local finance in developing countries

affect economic growth. Apart from the obvious point that the complex relationship between decentralization and economic growth is not easily picked up by a few control variables in an econometric model, there is also the important underlying question of whether fiscal decentralization stimulates economic growth, or vice versa. Second, as we have discussed earlier, fiscal decentralization is not a simple concept, and the effects on growth may depend not only on the level but also on the structure of decentralization. Moreover, such effects may take a long time. As LagoPeñas et al. (2016, p. 11) conclude, “it should be no surprise that studies using different data, estimation techniques and specifications and different definitions of decentralization as well produce diverging results. Clearly, the results are not robust.” Cross-section analyses like those reported above probably are not the best way to tease out long-run effects. Salmon (2012) suggests a better way to proceed may be to develop hypotheses about the way in which decentralization may affect economic performance in particular circumstances, and then to test them against individual countries in which those circumstances prevail. In fact, as Baskaran et al. (2016) note in a meta-analysis of research on decentralization and growth, studies of single countries often find a positive effect on growth, perhaps because they are conducted within a common institutional framework and do not suffer from the demonstrated sensitivity of the estimates in cross-country regression models to the choice of control variables. On the other hand, even country case studies have trouble separating the effects of decentralization from everything else. No doubt someone may figure this all out someday. The hypothesis that fiscal decentralization may be growth enhancing may seem intuitively right, but the evidence is scant and the jury is still out. Research may eventually confirm that local choices about local public finance can make a positive contribution to local economic development and that, as economic development proceeds, voters will demand more of it. Urbanization should hasten this process. The economic impacts of fiscal decentralization are likely to be indirect and require the right supporting cast of policies, and will be difficult to uncover. But empirical modeling and data are improving, and more solid results may perhaps soon be available. The Size of Government Does fiscal decentralization lead to a larger subnational government sector? Some authors (e.g. Prud’homme, 1995) have argued that it does because service delivery by subnational governments will be costlier. Others have argued that decentralization may make the government sector

M4439-BAHL_9781786435293_t.indd 56

01/03/2018 11:08



Has decentralization worked? ­57

smaller. Musgrave (1983), who argues that redistribution should remain a central government responsibility, suggested that total government expenditures would decline with decentralization because subnational governments, operating in (as it were) an open economy, would choose less redistribution spending than would the central government. A different approach, arguing that the growth in government (Leviathan) is led by bureaucrats who act to maximize the size of government in their own self-interest, leads Brennan and Buchanan (1980) to conclude that if firms and individuals are mobile, then fiscal competition within a decentralized system would induce governments to hold down their level of taxes, and therefore the size of their expenditure budgets, unless they had access to intergovernmental transfers or to taxes whose burden can be exported. The empirical question raised in this context is whether developing countries that assign more spending and taxing responsibility to subnational governments have a systematically larger or smaller public sector. Most analyses of the relationship between government size and decentralization specify the dependent variable as total government expenditures as a percent of GDP. The explanatory variable of interest is generally either the subnational government expenditure share of GDP to test for the direct effect on fiscal decentralization or the subnational government share of taxes if the goal is to test the Leviathan argument. As in the case of the relation between decentralization and growth, the results are mixed. Some studies of OECD (high-income) countries have found evidence that revenue decentralization is associated with smaller government size (Rodden, 2003a; Jin and Zou, 2003; Feld et al., 2003), while others have not (Oates, 1981). In developing countries, Prohl and Schneider (2009) find strong evidence that fiscal decentralization dampens the growth in government size, based on pooled data from 20 industrial and nine developing countries. However, Oates (1985) finds no relationship in his subsample of developing countries.32 Using a much larger sample of developing, transition and industrial countries, Martinez-Vazquez and Yao (2009) find that the level of total government employment is stimulated by expenditure decentralization but is not significantly related to revenue decentralization, perhaps because intergovernmental transfers tend to offset the competition effect of fiscal decentralization on expenditure. Stein (1998) also found a positive effect of expenditure decentralization on government size, using data from both OECD countries and a sample of Latin American countries. A related question is the relationship between decentralization and the size of the ‘shadow economy.’ In principle, decentralization may reduce the level of ‘shadow’ activity (broadly, activity not recorded in official statistics) by increasing tax morale (Torgler et al., 2010) and by increasing

M4439-BAHL_9781786435293_t.indd 57

01/03/2018 11:08

58

Fiscal decentralization and local finance in developing countries

the ability of (more local) officials to observe such activities (Dell’Anno and Teobaldelli, 2015). On the other hand, if local governments are less efficient (Treisman, 2000) or engage in a ‘race to the bottom’ with respect to business taxes and regulation (Brueckner, 2003), decentralization may make shadow activities more attractive (Prud’homme, 1995). Several studies find that fiscal decentralization tends to be associated with a smaller shadow economy (Buehn et al., 2013). However, causation may run the other way to some extent, with a larger shadow economy making decentralization less effective because of problems in coordinating policies across more governments. Using different measures of fiscal decentralization (expenditures, revenues, employment), Goel and Saunoris (2016) find that decentralization is more effective in reducing the shadow economy when it is relatively small than when it is large. The case for expenditure and tax competition driving down the size of government in developing countries is probably clearer at the metropolitan level. As numerous studies in developed countries suggest, firms and (to a lesser extent) households do indeed seem to shift location when an alternative local jurisdiction offers a tax/service package that better fits their preferences (Tiebout, 1956).33 However, the assumption that local governments have sufficient fiscal autonomy to compete with one another does not hold in most low-income countries. Moreover, many firms and families in such countries may be more tied to particular jurisdictions owing to the limited availability of public services and housing elsewhere. While there is some evidence that intra-metropolitan competition is emerging in countries such as Colombia (Bird, 2012a), most developing countries seem still to be far from a Tiebout world. But stay tuned: things are changing rapidly in this respect in some countries. The Quality of Governance The core idea of fiscal decentralization is good governance. If local populations are empowered by the vote, they can hold local officials accountable for the quality of services delivered.34 A more decentralized system would thus seem to imply more budget transparency, less corruption and generally better public services. If local voters are properly empowered, they may insist on institutional reforms often associated with good ­governance – such as accountability to local citizens, adherence to a rule of law, open elections and an independent media. Increased inter-jurisdictional competition, although it may in some instances result in smaller governments (as mentioned earlier), may also lead to better outcomes as so-called ‘yardstick competition’ leads to improvements in terms of the quality of public services delivered per tax dollar paid.35

M4439-BAHL_9781786435293_t.indd 58

01/03/2018 11:08



Has decentralization worked? ­59

Decentralization may thus improve the quality of governance by moving political decisions from top-down central decisions to the more bottom-up local arena. However, if citizens do not have a voice in subnational government decisions or are insufficiently informed or interested to vote or get involved in the discussion, fiscal decentralization may simply shift control from the central bureaucracy to a local elite with perhaps even less interest in delivering good services to the masses (Bardhan and Mookerjhee, 2006; Treisman, 2007). Elite capture and the fostering of clientelism (serving favored groups) as a result of decentralization have been demonstrated in a number of empirical studies, such as Araujo et al. (2006) on project choice in Ecuador, Juul (2006) on Senegal, and De and Nag (2016) on water supply and drainage in India. A good general discussion may be found in Khemani (2015). However, as Faguet and Pöschl (2015) stress, similar problems abound at the central level in many of the same countries. It is not easy to determine if the net social outcome is better or worse with decentralization when the answer usually depends on many countryspecific factors. An additional problem that has been noted by many (Bahl and Linn, 1992; Prud’homme, 1995) is that subnational governments – especially the smaller and poorer ones – may be unable to deliver services efficiently even if everything else permits better matching of budget expenditures with preferences. Again, however, the available evidence on experience with decentralization in low-income countries does not clearly tell us whether the impact on service delivery is positive or negative.36 Fiscal decentralization may, or may not, lead to better governance. As usual, disentangling the evidence is not simple. Measures of output in the public sector have long been criticized as conceptually unsatisfactory (Burkhead and Miner, 1971). Such proxies for output as expenditures, school test scores and infant mortality rates, or input measures such as the number of public employees are commonly used. The World Bank has done useful work in constructing indexes based on an assortment of measures of factors thought to be associated with good governance at the national level – such as the rule of law, lower levels of corruption, citizen participation and so on.37 Many studies have subsequently made good use of this work. Empirical work in this area must first determine the dependent variable to be explained. There is no definitive answer about the best measure, and different studies have used different indicators of good governance. Some interesting insights have emerged. De Mello and Barenstein (2001) find that countries with a greater subnational government expenditure share score significantly higher in terms of rule of law, political stability and government effectiveness. Huther and Shah (1998), who study the relationship between an index of the quality of management and four measures

M4439-BAHL_9781786435293_t.indd 59

01/03/2018 11:08

60

Fiscal decentralization and local finance in developing countries

of the quality of governance (citizen participation, government efficiency and the lack of corruption, human development and the equality of the income distribution), find a strong positive relationship with expenditure decentralization. Kyriacou and Roca-Sagalés (2011) find a positive effect of fiscal decentralization on the quality of governance where the latter is measured as the average of indexes of corruption, rule of law, regulatory quality and government effectiveness. Studies focusing on the effects of fiscal decentralization on corruption as an indicator of good governance have sometimes found that a larger subnational government expenditure share tends to be associated with a lower level of corruption (Fisman and Gatti, 2002; de Mello and Barenstein, 2001; Ivanya and Shah, 2011). However, Treisman (2000), using political decentralization rather than fiscal decentralization as the independent variable of interest, finds a greater perception of corruption in decentralized systems. In a later review of the evidence, Treisman (2007) stressed the complexity of the problem and notes some conflicting evidence, but concludes again that decentralization alone seems not to reduce corruption. More recently, Nelson (2013) reports that a more fragmented structure of local government is associated with more corrupt behavior, and Fan et al. (2009) reached a similar conclusion. As with the problem of elite capture mentioned earlier, clear and definitive answers are hard to find with respect to whether decentralization increases or reduces corruption, even in a single-country study. The outcome depends on the combination of where we start (the initial conditions), how exactly we measure decentralization and corruption, and how we control for other changes between the preand post-decentralization periods that might have affected the amount of corruption at all levels of government. Much the same can be said with respect to the effects of decentralization on the quantity and quality of public services delivered to local people, although there is a little more evidence of beneficial outcomes than with respect to such general issues as corruption, perhaps because it is easier to obtain decent proxies for outcomes with respect to the provision of services such as health and education. Treisman (2007) concludes that the evidence of increased public sector efficiency is inconclusive, a conclusion that was reinforced for OECD countries by Ahmad et al. (2008). However, Adam et al. (2008) almost simultaneously found a strong association between decentralization and generalized measures of public sector efficiency for OECD countries, while Barankay and Lockwood (2006) – using a panel regression of data for 20 years for Swiss cantons – found a positive relationship between decentralization and educational attainment. Many studies in developing countries have touched on these same issues in varying ways. For example, Olken (2008) finds that Indonesian villagers

M4439-BAHL_9781786435293_t.indd 60

01/03/2018 11:08



Has decentralization worked? ­61

were happier when they had a voice in choosing development projects than when they did not, even if their wishes were in the end not met by the projects chosen. Iregui (2005) found decentralization in Colombia to be positively associated with improved provision of health and education. Bossert (2015) finds similar results for health in some other developing countries, but stresses the importance of understanding the decision space within which local decisions about provision are made and the difficulty of implementing the sort of random control trials (RCTs) now often taken as the gold standard of empirical testing when it comes to assessing the impact of decentralization.38 Sow and Razafimahefa (2015), employing the quite different approach of stochastic frontier analysis (SFA) to assess the effect of decentralization on the efficiency of public service provision, find that it appears to lead to improvements – but only when three key conditions are satisfied: the political and institutional environment is adequate; there is sufficient local control over expenditure (decision space) to make a difference; and there is a sufficient degree of revenue decentralization. If these conditions are absent, they conclude, fiscal decentralization may worsen rather than improve the efficiency of public service delivery. Another stream of research considers the question of whether decentralization leads to the development of social capital. Classical thinkers held social capital to mean “trust, concern for ones’ associates, a willingness to live by the norms of one’s community and to punish those who do not” (Bowles and Gintis, 2002, p. 419). De Mello (2011) studies the relationship between fiscal decentralization and social capital formation defined in terms of attitudes towards the importance of having a voice in government decision. His dependent variable is the answer to the question about respondents’ views on the importance of having a say in government decisions, as reported in the World Values Survey (WVS). He finds that respondents living in federations/decentralized countries are more pro-voice than those living in unitary/centralized countries. More recently, Lago-Peñas et al. (2016) read research in this area as consistent with showing that decentralization may encourage more collective action, interaction and, ultimately, social capital, and that it increases trust in government as well as in other political institutions. All this is interesting, if inevitably a bit vague; but it does not move us closer to what we would most like to measure in terms of the quality of governance – namely, whether people who live in more decentralized systems get more of what they want. However, such studies do suggest that how happy people are with government depends not only on what the government does but also on whether they feel that their views have been taken into account, as Olken (2008) noted in a study in Indonesia. What little evidence there is on this comes from limited citizen surveys in individual

M4439-BAHL_9781786435293_t.indd 61

01/03/2018 11:08

62

Fiscal decentralization and local finance in developing countries

countries. Encouragingly for those who think decentralization is usually a good thing, at least some such studies suggest that the services delivered may match citizen preferences more closely under decentralized systems.39 As usual, however, the evidence is hardly solid enough to persuade those with different views. Equity Fiscal decentralization can have important effects on the distribution of real incomes in many ways. It can lead to a greater focus on poverty alleviation through upgraded public services in urban areas. Services such as safe drinking water, better sanitary facilities, better local health care and access to education for poor families have the potential for significantly increasing the standard of living and the real incomes of poor families. However, because fiscal decentralization usually has the most marked effects on the most advanced localities it tends to increase the disparity in fiscal resources between urban and rural areas, and may worsen the gap in incomes between the rich and the poor. There is considerable variation from country to country, and the overall distributional effect of any decentralizing policy is affected by a myriad of country-specific policies. Still, one question that almost always comes up is whether fiscal decentralization is likely in the end to improve the distribution of income or alleviate poverty.40 Two policy paths might be explored in this connection. The first is to determine if subnational governments can be more effective than central governments in focusing service delivery on poor families. Local and regional governments are often responsible for such essential services as safe drinking water, sewerage, local health clinics and primary education, and in many countries are charged with partial responsibility for slum upgrading. If the poor are targeted by local governments and given access to these services, the result could do much to reduce poverty (Bahl and Linn, 1992). Indeed, lower-level governments would seem to have a comparative advantage in identifying targets of opportunity for poverty reduction because of their more intimate knowledge of the structure of poverty in their area. But this assumes that they are interested in doing so, which has not always been the case – for example, when those in charge see the poor people in question are ethnically distinct or simply not ‘one of us.’41 The second path is to ask if fiscal decentralization can be structured to reduce the disparities in resources available to subnational governments in low- and high-income regions or municipalities. As Bird and Rodriguez (1999) argue, whatever the effects of decentralization may be on personal income distribution, it is also important to assess its effects on

M4439-BAHL_9781786435293_t.indd 62

01/03/2018 11:08



Has decentralization worked? ­63

interjurisdictional equity. The devolution of more taxing powers to local governments tends to increase the fiscal advantage of the rich places, but if structured correctly it can also increase the tax price in urban areas (Bahl, 2013). Moreover, as most policy analysts would argue, a decentralizing scheme that devolves significant revenue-raising powers to local governments should include an equalizing intergovernmental transfer that protects the provision of public services by poorer jurisdictions.42 But this is often easier said than done. As we discuss in Chapter 7, although it is often difficult to find a politician or senior bureaucrat who speaks out against equalization in intergovernmental transfer systems, it is equally difficult to find developing countries that do a good job in this respect. Do countries with more decentralized fiscal systems have more or less income inequality or more or less poverty than those that are more centralized? The answer, as usual, is that it depends on the structure of the decentralization, on whether the effects of all other poverty-related programs can be factored out, and whether enough time has lapsed to enable income effects to be realized. Also as usual, the evidence on outcomes is mixed. Crook and Sverrisson (2001), for example, found a positive relationship between decentralization and inequality in studies of several developing countries. On the other hand, Sepulveda and Martinez-Vazquez (2011) in a panel study of a large sample of developing and transition countries find decentralization to be associated with a lower concentration of poverty, though only when the size of government is greater than 20 percent of GDP – a threshold that rules out most low-income countries. Yao (2007) reached a similar conclusion in a large cross-section study. Von Braun and Grote (2000) and Lindaman and Thurmaier (2002) used cross- country analysis with the UN’s Human Development Index (HDI) as the dependent variable, and concluded that decentralization improves equity. Several studies focusing on individual countries and using various measures of the quality of public goods have found no positive association between decentralization and poverty reduction (Enikolopov and Zhuravskaya, 2003; West and Wong, 1995; Azfar and Livingston, 2002). On the other hand, a study of Bangladesh (Galasso and Ravallion, 2005) found a positive impact on poverty reduction, as does an interesting recent study by Basurto et al. (2017). Some evidence thus suggests that fiscal decentralization is associated with improvements in the real incomes of poor families and reductions in the disparity between the rich and the poor. On the other hand, some countries seem to have decentralized without the resources, the capacity or perhaps the will to pursue equity goals. We do not know much about all of this. Any significant effects of decentralization on income distribution and poverty seem unlikely to show up quickly and are unlikely to be picked

M4439-BAHL_9781786435293_t.indd 63

01/03/2018 11:08

64

Fiscal decentralization and local finance in developing countries

up in comparative studies using the broad brush of fiscal decentralization as the key independent variable. What seems to matter more than the level of decentralization are the objectives of those who control government decisions at different levels of government and the institutional structure within which intergovernmental fiscal arrangements operate in different countries. Whether it is better to tackle poverty alleviation directly through central government programs – as Bird and Rodriguez (1999) suggested for the Philippines and Bahl et al. (2010) found in a study of West Bengal state in India – or more indirectly through fiscal decentralization is by no means a settled issue. For example, if decentralization does stimulate economic growth and better governance, the indirect effects on reducing income inequality and poverty may be much greater than direct redistributive actions at the local level alone. Or it may not, since once again the evidence to date leads to no clear conclusions. But this discussion need not end on a negative note because, as we discuss in Chapter 7, at least with respect to problems with interjurisdictional disparities we know how to use the fiscal tools – local expenditure assignments, local revenue assignments and equalizing transfers – to deal with the problem.

CONCLUSIONS One of our five-year-old granddaughters recently said about her experience in school: “I don’t know everything, but I know a lot.” Our conclusion on how well fiscal decentralization has worked is pretty much the same. Most of the big questions are still unanswered, but the research to date has nonetheless taught us much, both about what a well-designed and well implemented fiscal decentralization might do and how we can at least begin to figure out what happens as a result. Those who make policy are always looking to future outcomes. But their expectations of what a particular policy measure may do rest at best on their understanding of what similar measures have done when applied elsewhere – usually in somewhat different circumstances. Treisman (2007) concluded in an earlier review of much of the literature discussed in this chapter that it is difficult to draw firm conclusions from these studies. For every study that reports one result, another seems to qualify it or even outright refute it. This is not surprising since there is no clearly correct, measurable definition of fiscal decentralization, and estimation problems abound. As an old joke has it: to a (good) economist the only good answer to any question is always “it depends.” Nonetheless, we have learned quite a lot about the most critical elements affecting outcomes, and we now better understand the complexity of the issues. In

M4439-BAHL_9781786435293_t.indd 64

01/03/2018 11:08



Has decentralization worked? ­65

their excellent recent review of the literature, Lago-Peñas et al. (2016, p. 27) conclude that: “overall, there are reasons to be optimistic about a net positive impact of decentralized systems having been introduced all over the world in the past several decades, especially when those decentralization processes have been well-designed and implemented.” Not everyone is so optimistic. For example, Mascagni (2016, p. 25) concludes that “the evidence on political and economic outcomes is mixed and often disappointing.” We agree that the overly ambitious expectations aroused by political discussions in which proponents often promise more than even the best decentralization program could possibly deliver are bound to be disappointing. But this does not mean that we think we have learned little or that decentralization cannot ‘work.’ What we think we have learned, and have still to learn, may be summarized in a few stylized points, most of which are developed more fully in later chapters: ●●

●●

●●

There is a strong association between the level of fiscal decentralization and the level of development, with larger countries tending to decentralize more quickly. We still have not sorted out whether decentralization leads to faster economic growth, or whether economic growth leads to more demand for fiscal decentralization. Nor do we know whether fiscal decentralization at the margin leads to a faster rate of economic growth. We think we know that, for a developing country, the best way to develop a stronger subnational government fiscal regime is to put into place the institutions needed for good decentralization as the occasion arises, and then to deal with the inevitable roadblocks as they occur. ‘Big bang’ reform may sometimes work, but as a rule there is more to be said for ‘muddling through’ (or, as we prefer, gradual incrementalism). Even the biggest of big bangs inevitably requires decades of subsequent incremental adjustment before it works properly, quite apart from the continuing need to react to changes in the economic and institutional environment.43 But it is important that all the pieces of the fiscal decentralization strategy are in the plan even though the implementation is gradual. Decentralization may, as theory suggests, result in better governance and higher levels of citizen satisfaction. But this requires that countries adopt such key policies on the fiscal side as decentralized expenditure discretion, local government taxation and increased citizen voice in governance. Countries that do such things are more likely to succeed with decentralization than are countries that fail to do so. Since no single decentralization measure can adequately

M4439-BAHL_9781786435293_t.indd 65

01/03/2018 11:08

66

●●

●●

●●

Fiscal decentralization and local finance in developing countries

capture all relevant factors, it is no surprise that cross-country econometric analysis seldom yields persuasive results. Research on the effects of decentralization on the size of government and the quality of the services that it delivers (or on its effect on corruption) has yielded mixed results, to some extent again reflecting the inherent problem in comparative studies that implicitly assume that the many and varied institutional arrangements usually lumped together in the econometric jar as ‘observations’ are drawn from the same underlying population. A critical issue that needs to be more carefully considered is the effect of decentralization on fiscal disparities and income distribution. More work needs to be done on why fiscal decentralization might worsen or improve distributional outcomes. The design and level of interjurisdictional equalization through intergovernmental transfers, the equity effects of decentralized tax policies and expenditure assignments, and the extent and effectiveness with which public investment is targeted to poor neighborhoods are all part of this complicated story. More work needs to be done on empirical tests of these hypotheses. A key to better research in this area is better data. Many developing countries are doing a better job every year in pulling together the information base necessary to analyze the impact of fiscal decentralization policies. But there is much more to be done, and it is important to do it because the best chance for sorting out the impacts of decentralization is almost certainly in the context of single countries. More careful and better-structured case studies of the impact of specific policies in different circumstances are needed to build up a sufficient body of knowledge about the impacts of fiscal decentralization. Cross-country analyses also are important; but to get more out of such studies, some group or institution must step up to the task of providing a more comprehensive and complete series on subnational government public finances. Some good steps have been made in this direction with the longstanding contribution of the IMF (GFS), with new efforts in OECD countries and even to some extent for some developing countries, particularly with respect to cities in recent years, but again there is much more to be done.44 The World Bank and the IMF would seem to be the best candidates for this job, although neither seems currently to be concerned with establishing a better data base for comparative analysis of subnational finance. We discuss this point again in Chapter 9.

M4439-BAHL_9781786435293_t.indd 66

01/03/2018 11:08



Has decentralization worked? ­67

NOTES  *  Fatma Romeh and Yared Seid provided valuable research assistance for this chapter.   1. We refer here to many studies, but there is now so much published literature on this subject (not to mention a mountain of country and agency reports) that we certainly do not claim to have found it all. Any who wish to pursue specific topics discussed will find extensive additional references in many of the works cited here.  2. An early attempt to quantify this approach is Boex and Simatupang (2008), which found that the ‘empowerment gains’ from decentralization were small; however, as the authors emphasized, this conclusion could only be considered extremely tentative given the conceptual complexity of the question and the limitations of the data. As OECD (2017) shows, even in the relatively advanced Latin American context, both problems remain serious.  3. To make things simpler, we assume there is only one expenditure function and one discretion coefficient.   4. Even in this case, owing to the informational advantage of the agent there may often be some degree of discretionary control at the subnational level.   5. As Rodrik (2015, p. 72) puts it, “when the relevant data cannot be forced into succinct rules without sacrificing too much relevance . . . economic science advances by expanding its collection of useful cases.”   6. For a good example of the importance of such highly specific local factors as leadership, see the analysis of a local tax reform in Sierra Leone in Jibao and Prichard (2015).   7. The OECD has done much in recent years to make the concept operational through a series of pilot studies in many of its member countries (OECD 1999, 2016). OECD (2017) contains a pilot study for several Latin American countries.  8. For example, Blöchliger (2015, p. 621) notes that “the alleged tax autonomy of Norwegian municipalities is a bluff, since the overall intergovernmental framework provides all municipalities with a strong incentive to set tax rates at the legal maximum.” More generally, as we discuss later in Chapter 7, most ‘equalization’ transfer systems provide incentives for local governments to impose at least average tax rates.   9. As an example, note the frequent mention of the World Bank (annual) Doing Business indicators when countries are trying to attract foreign investment. The ‘yardstick’ idea was first formulated in Besley and Case (1995). 10. Almost never, however, is sufficient attention paid to the importance of all the other relevant features in place in the successful country: see also the epigraph of this chapter. 11. See Box 2.1 on data. For reasons discussed there, these comparisons are based on GFS data augmented by information from other sources to increase the number of developing countries in the sample and, especially, to include some very large countries not in the GFS data base for certain years. As described below, we think that the additional data we have added are largely comparable to the GFS data. The countries where no data on subnational government finance are available in the volumes of GFS that we used include (for certain years) India, Indonesia, Brazil, Nigeria and Pakistan. The ‘transitional’ countries – those emerging from the former Soviet bloc during the 1990s – are grouped separately because of their very different intergovernmental structures (see e.g. Bird et al., 1995). 12. If user charges and so on are included in the numerator they should also be included in the denominator. Each case is different. For example, some non-tax revenues like the production royalties that go to some Colombian regional governments should arguably be included for some purposes because even though various conditions are imposed on how such revenues may be spent, the regions still have considerable discretion in how they use these resources (Bird 2012a). On the other hand, most intergovernmental transfers (including the share of central royalties that is transferred to all regional governments in Colombia) as well as foreign aid should be excluded. 13. For example, even locally owned public utilities are not included in local government

M4439-BAHL_9781786435293_t.indd 67

01/03/2018 11:08

68

14.

15.

16.

17.

18. 19. 20. 21. 22.

23.

Fiscal decentralization and local finance in developing countries accounts in the US (Ebel and Wang, 2017), but they usually are in Canada (Slack and Tassonyi, 2017). Another difficult question is how to treat debt finance. As discussed in later chapters, the view we take here is that debt is not a source of revenue, but rather a method of finance (unless of course the higher-level government stands at the ready with a bailout to avoid any default) since it must be repaid by locally raised revenues or intergovernmental transfers. Consequently, we do not include borrowing in the denominator of the revenue decentralization measure. One well-known colonial model, Britain’s ‘indirect rule,’ in a sense worked round this problem (and minimized administrative costs) by utilizing existing power structures where possible. As Ali et al. (2015) suggest, in the post-colonial era one result is that it has been more difficult to build effective nation-states in anglophone than in francophone Africa. Interestingly, Soifer (2015) argues that the historical development in Colombia of major regional population centers that were largely independent of the national capital weakened the development of effective central government in that country relative to others in Latin America. Several studies have measured ‘fractionalization’ on ethnolinguistic and religious scales (Mauro, 1995; Annett, 2000; Alesina et al., 2003). These numbers have sometimes been used as measures of heterogeneity in determinants studies, perhaps in part because they are available. As Bird and Ebel (2007) argue, however, when it comes to decentralization what matters more than fractionalization is ‘fragmentation’ – that is, the extent to which ethnically, linguistically or religiously different groups are located in different areas. Several years ago, Frey and Eichenberger (2004) suggested an interesting proposal for a possible restructuring of governance into a system that would in principle permit geographically separate but otherwise cohesive groups to be democratically self-governing within a larger ‘national’ (or European Union) framework. However, no one seems yet to have taken up this idea. One of us presented the results from an empirical analysis of fiscal decentralization at a conference, and was challenged from the floor for using such inappropriate data on subnational government expenditures. The challenger was a senior official from the IMF’s Fiscal Affairs department. Both buyers and sellers, it seems, are well aware of the problems with the data. A 2008 review of 26 studies in this area found that one-third of them measured the dependent variable as the subnational government expenditure share, while one-fourth used the own source financed share of local expenditures (Baskaran et al., 2008). A good review of a wide range of empirical studies focusing on the three basic hypotheses may be found in Letelier (2005). This hypothesis has been much explored in the historical literature: for an interesting application to Latin American development, see Centeno (2002). For an alternative explanation, see note 15 above. We discuss the concept of vertical balance later in the book; see also Bird (2006) and Bahl and Wallace (2007). Of course, this question is not one that comes up only in developing countries. It has, for instance, been much discussed with respect to developed federations: see e.g. May (1969) and Bird (1986, 1994). Some influential scholars (e.g. Peacock and Wiseman 1967) have argued that causality may go the other way, while Musgrave (1969) suggested that taxes and expenditures tended to move together in response to other factors. For a review of this earlier literature, see Bird (1970a). Subsequently, many papers have explored the causal relationship between taxes and expenditures at the national level in different countries – for two examples, see Owoye (1995) on G-7 countries and Cheng (1999) on Latin America – sometimes finding causality flowing one way, sometimes the other, and often concluding that changes in the levels of both taxes and expenditures are jointly determined. Bahl and Linn (1992) reviewed several studies in developing countries that found some evidence consistent with displacement effects. Interestingly, a recent study of tax-expenditure causality at the subnational level (Garcia 2012) surveying a number

M4439-BAHL_9781786435293_t.indd 68

01/03/2018 11:08



24. 25.

26. 27. 28.

29. 30.

31.

32. 33. 34. 35.

36. 37.

Has decentralization worked? ­69 of similar earlier subnational studies (mainly in the US) concluded that taxes appeared to lead expenditures in Spanish regions – functions following finance, so to speak. Vertical tax base competition is a form of vertical fiscal externality, as discussed in e.g. Dahlby (1996). Essentially, this is what Akerlof and Shiller (2015) pejoratively label ‘phishing for phools’ – that is, the sort of manipulation and deception that has been long and successfully employed by those seeking to separate fools (or, if one prefers, the rationally ignorant) from their money (or votes). For surveys of the literature on this subject, see Martinez-Vazquez (2011), Lago-Peñas et al. (2016), Mascagni (2016) and Treisman (2007). The general argument is developed in Martinez-Vazquez and McNab (2001) and demonstrated theoretically in Brueckner (2006). Some question the ‘power’ of econometric results on statistical grounds: for example, Ioannidis et al. (2016, p. 28) assert that “empirical economics has low power and much residual bias . . . the typical economics result reported by any single study is not very credible, and its magnitude needs to be reduced, typically by half or more, rather than taken at face value.” As Deaton and Cartwright (2016) discuss, even the latest and best methods of assembling data, such as ‘randomized control trials,’ are difficult to carry out properly, and often equally difficult to interpret. None of this means we should not do such work; but we should be very clear about what we are doing, and why, and very careful about how we interpret the results. For such investigations of the indirect effects of decentralization on growth, see: on corruption (Martinez-Vazquez and McNab, 2001); on employment (Martinez-Vazquez and Yao, 2009); and on income inequality (Sepulveda and Martinez-Vazquez, 2011). The idea that decentralization is a route to faster economic growth lies behind the local economic development (LED) approach in which the role of local government is seen to be as a promoter, facilitator and coordinator of local and regional development activities, taking on such tasks as: identifying promising sectors for local economic development; developing appropriate infrastructure and a suitable regulatory environment; providing information to potential investors about the local economy; and connecting local firms to potential suppliers and markets (Lennon and O’Neil, 2003). Case studies of such programs have so far not provided convincing evidence that LED programs are superior to more centralized approaches to stimulating economic growth. The Chinese case is particularly interesting because some (Qian and Weingast, 1997) emphasized the key role of local and provincial competition to grow faster (largely because growth was the major ‘success indicator’ by which the central party rated local officials). As Brandt and Rawski (2008) show in detail, this was indeed one of many factors explaining China’s great economic transformation; however, we can say nothing definitive about how much it mattered in the big picture. Lindert (2004) argues plausibly that the effects of decentralization on spending may be different at different levels of development, sometimes promoting spending and in other instances decreasing it. To mention only one recent study providing some supporting evidence, see Tassonyi et al. (2015). Alternatively, they may perhaps, as in China, appeal over the head of local officials to those in the official (or party) hierarchy for relief. See such studies as: Rincke (2008) on local innovation; Terra and Mattos (2015) on education in Brazil; Li and Zhang (2015) on China; Capuno et al. (2015) on health in the Philippines; and Bossert (2015) on health in several countries, including Colombia and Morocco. For good discussions of the literature on fiscal decentralization and the quality of governance, see Kyriacou and Roca-Sagalés (2011) and Lago-Peñas et al. (2016). See http://info.worldbank.org/governance/wgi/index.aspx#home for a full description of these indicators.

M4439-BAHL_9781786435293_t.indd 69

01/03/2018 11:08

70

Fiscal decentralization and local finance in developing countries

38. As Deaton and Cartwright (2016) stress, one must of course also be careful in using and interpreting RCT studies. 39. See, for example, Campos and Hellman (2005) on Indonesia; Azfar et al. (2001) on the Philippines; Faguet (2004) on Bolivia; and Acosta and Bird (2005) on Colombia. 40. It should be remembered that success in reducing poverty does not necessarily reduce interpersonal income inequality as usually measured. 41. As Wilensky (1975) and others have argued, it was not by chance that the growth of the welfare state in most advanced countries coincided with the increased concentration of revenue and power at the central government level. 42. We take up this question in more detail in Chapter 7. 43. This conclusion is based mainly on our own field experience as well as on case studies like those cited at the beginning of Chapter 1. It is developed further in the last chapter of this book. 44. See, for example, OECD (2016, 2017) as well as www.oecd.org/tax/federalism/oecdfis​ caldecentralisationdatabase.htm. See also such webpages as www.uclg.org.

M4439-BAHL_9781786435293_t.indd 70

01/03/2018 11:08

PART II

Decentralizing Expenditure

M4439-BAHL_9781786435293_t.indd 71

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 72

01/03/2018 11:08

3. Expenditure assignment and management Adequate funding is one of the three things that are essential if democratic local government is to work well. The other two are adequate powers and reliable accountability mechanisms. (Manor, 2013, p. 1)

We discuss the question of funding in depth in Part III of this book. This chapter and the next are about the other two pillars of good local government mentioned by Manor. Most of the present chapter focuses on the powers of local government: if one defines effective fiscal d ­ ecentralization as empowering local people to get what they want, then expenditure assignment – the determination of the functions for which local governments are responsible – in effect defines what that means. In the latter part of this chapter, we consider some ways in which local governments should organize and manage the expenditures with which they are charged in order to live up to this potential in as effective, efficient and accountable a way as possible.1 The next chapter explores some of the same issues in more depth with respect to the surprisingly important role that regional and local governments play in building up the infrastructure that affects both how well people live and the productivity of the public and private sectors. The standard approach to fiscal decentralization essentially follows the adage that ‘finances follow functions’ by first determining which level of government should in principle be responsible for which functions. New things – the threat of terrorism, regulating the internet and aerial drones, concerns about congestion and pollution, and so on – come along from time to time to send governments back to the drawing board. But, with respect to most government functions, the dice have already been thrown: just about everything that government does is already, for better or worse, allocated in some way. However, no country ever gets it completely right. The question usually faced is what changes seem needed to improve governance and whether the benefits of any change are sufficient to offset the inevitable costs.2 We begin our discussion of this issue by setting out the basic theory and then considering why it often proves so difficult to get it right in developing countries. 73

M4439-BAHL_9781786435293_t.indd 73

01/03/2018 11:08

74

Fiscal decentralization and local finance in developing countries

Expenditure assignment is inevitably a political issue because responsibility for expenditure programs carries with it the power not only to affect the well-being of various population groups but also to control the bureaucracy and resources needed to deliver public services. Local and regional officials and politicians, like the people they are supposed to serve, are eager to have more control over spending. On the other hand, central (and regional) politicians and officials are threatened by the devolution of expenditure responsibility because it means a reduction in their control over budget resources, over those who are directly in charge of service delivery and over which citizens end up getting what services. Many other political factors may also come into play. For instance, some central officials may be transferred to subnational levels, possibly losing status and prospects, if not necessarily salary; others may lose seniority and influence with the central bureaucracy. Politicians too may lose and gain: for instance, increased subnational control over spending may end up strengthening potential opponents of the central government. One way or another, changing existing expenditure assignments may, like all changes in the status quo, prove to be painful and costly. Such changes are unlikely to be undertaken unless there are sufficiently large net social gains to be able to provide at least some compensatory offset to those who end up as losers. We begin this chapter with a review of the normative criteria that can be used to decide who does what when it comes to dividing up public sector activities among levels of government. Even if a country does manage to change its status quo expenditure assignment in any significant way – never an easy task – circumstances keep changing, so a well-run intergovernmental fiscal system must also change with the times.3 We then discuss some issues that have led to problems with fiscal decentralization in a number of countries, including: the lack of clarity in expenditure assignment; the lack of expenditure autonomy; the distributional question; the role of non-governmental (civil society) and informal institutions in service delivery; and, of course, politics. We conclude the assignment discussion by suggesting a few key recommendations and implementation rules that might be helpful to those charged with monitoring and improving how things are done, before turning to a brief review of some of the key issues in expenditure management.

WHAT DO WE MEAN BY EXPENDITURE ASSIGNMENT? A government is assigned responsibility for delivering a service when it is legally required to do so and the relevant expenditures are included in

M4439-BAHL_9781786435293_t.indd 74

01/03/2018 11:08



Expenditure assignment and management ­75

its budget.4 Such assignments may be set out in broad or narrow terms in the constitution, particularly in federal countries. They may be prescribed in detail or in broad terms in a central government law dealing with local governments or budgetary matters. Sometimes, a function may be assigned to local governments on a permissive basis (they can do it if they choose to do so) or may simply be taken on as a task of local government because someone in charge decides that it needs to be done and that a regional or local government is the one to do it. One way or another the expenditure assignments observed in practice are heavily influenced by history. Often no one seems consciously to have decided that this level of government (or, perhaps this particular government at a certain level) should do it; rather, it is simply what they do and are simply assumed to have always done.5 No wonder that outside experts almost always remark that expenditure assignments in most countries are ‘murky.’ One way of assigning expenditures would be to set out a list of functions and then to allocate each one to some level (or levels) of government, based on one’s understanding of the basic concepts emphasized in the traditional theory of fiscal decentralization. These concepts might include the geographical area within which the benefits and costs are considered to impact on welfare; the size and scope of externalities; scale economies; distributional considerations; and the impact on other policy goals considered to be relevant to the case at hand. An example of this approach, described in Table 3.1, seems at first glance to provide quite a clear guide to what should be done by whom.6 To determine how expenditures should be assigned one needs to deal with many different considerations, some but not all of which are referenced in Table 3.1: ●● ●● ●● ●● ●● ●● ●● ●●

Who decides which government does what? Who decides how much is to be spent on an activity? Who decides how the funds are to be spent? Who finances the activity? Who takes responsibility for producing and delivering the service? Who regulates how the services are provided? Who monitors what is done? Who decides to act if something goes wrong, and what actions can they take?

Adding still more complexity is the fact that even the simplest governmental function usually encompasses a wide variety of separate sub-functions, each of which may be treated separately with respect to each of the string of decisions noted above.

M4439-BAHL_9781786435293_t.indd 75

01/03/2018 11:08

76

M4439-BAHL_9781786435293_t.indd 76

01/03/2018 11:08

R, L R, L L R, L R, L N, R, L N, R N, R, L N, R, L N, R, L N, R, L N N, R, L N, R, L N, R, L N

Water and sewers Solid waste Fire protection Police Parks, recreation Roads Natural resources Environment Education Health Social welfare Transfer payments Taxation User charges, fees Regulation Borrowing

L L L R, L R, L N, R, L N, R, L N, R, L R, L R, L R, L N N, R, L N, R, L N, R, L N, R, L

Provision, administration L, P L, P L R, L R, L, P R, L, P N, R, L, P N, R, L, P R, L, P R, L, P R, L, P N N, R, L N, R, L N, R, L N, R, L, P

Production, distribution

Mainly local benefits Mainly local benefits Mainly local benefits Mainly local benefits Benefits vary in scope Benefits vary in scope Benefits vary in scope Externalities vary in scope Externalities; transfers in kind Transfers in kind Redistribution Redistribution – – May include subsidization –

Comments

Source:  This table is inspired by and to some extent adapted from Boadway and Shah (2009, pp. 134–5), but the present authors are solely responsible for its contents.

Notes: N 5 national; R 5 regional; L 5 local; P 5 private or non-governmental. Functions best exercised by national government – foreign affairs, defense, immigration, monetary and fiscal policy, maintenance of internal common market and resolving interregional conflict – are not included here. Some functions (e.g. environment) may have an important international dimension. Others (e.g. natural resources or health) may be allocated to one level or another by a constitution.

Policy, standards, oversight

Function

Table 3.1  Assigning subnational expenditure responsibilities



Expenditure assignment and management ­77

To take a simple example, providing perhaps the most basic of public services – ensuring a sufficient level of public security so that people do not live in constant fear of losing their lives or property to bandits or thieves – is both conceptually and in a practice a surprisingly complex task, whether considered in political, economic or technical terms. Preventing and dealing with crimes against people, crimes against property, regulating traffic and dealing with accidents, and handling the many other activities with which police forces are often charged (from dealing with fires in some cases, to administering local bylaws and administering fines and court and even correctional facilities in others) are in many ways quite different functions, calling for different resources and organizational structures.7 Much the same can be said of most common regional and local government activities: health, education, housing, water and sewerage, and such ‘street’ functions as lighting, cleaning, and refuse removal and disposal. There are many possible ways to assign every function and even subfunction. Each arrangement has different implications when it comes to such possible goals of decentralization as: increasing the accountability of local officials; providing stronger incentives for local revenue mobilization; encouraging fiscal discipline; and improving the efficiency and equity with which scarce resources are deployed to provide services. The textbook model in which a subnational government has full responsibility for a function, full discretion in deciding how it will be delivered and enough own source revenue to cover the cost is seldom found in any country.8 For example, consider the case of primary education, which Americans often assume is a function that is and should be assigned to local governments. In many countries, education is largely financed by central (or state) government transfers to lower-level governments: examples include countries as different as Canada (Kitchen and Auld, 1995) and Colombia (World Bank, 1996). In both these cases, central (or, in the Canadian case, provincial) funding is so firmly entrenched that the amount of funding for education that local authorities receive has proven to be secure. In other cases, however (such as Pakistan), the amount that subnational governments receive to spend on education ebbs and flows with changes in the level of central government revenues and in the willingness of the center to transfer funds to the local level. When local governments can blame inadequate funding on other levels of government, their accountability to the local population is weakened. Box 3.1 further discusses the question of accountability and education. Even when local governments clearly have formal responsibility for delivering a service, and the flow of funds is secure, their real spending autonomy may be limited. In Colombia, for example, although regional governments have the formal responsibility to provide education, the

M4439-BAHL_9781786435293_t.indd 77

01/03/2018 11:08

78

Fiscal decentralization and local finance in developing countries

BOX 3.1  ACCOUNTABILITY AND EDUCATION Sokoloff and Zolt (2006) argue that one important factor explaining some aspects of the divergent developmental paths in the United States and Latin America may be that providing education was from the beginning a ‘bottom-up’ exercise in the US, with the result being the development of much stronger and more fiscally independent local governments. More recently, Mbiti (2016) has explored the puzzle of why the considerable increase in spending and in enrollment rates in education in many developing countries in recent years has not so far yielded much in terms of raising the low level of learning. He attributes this largely to the low level of accountability characteristic of the centralized education system found in developing countries. In principle, those who provide educational services should be accountable to parents (and children). In practice, parents can seldom hold providers accountable either by taking direct action against them (e.g. by withholding funds) or through such indirect actions as voting for governments that will hold them more accountable. The governments that pay the bills and control the system are usually far away from the local school and often have only tenuous control over what (if anything) happens to the resources supposedly being used to educate children. In some countries, private schools, which often pay their teachers less but hold them more to account than do public schools, provide some competition that in some instances has increased the information available to parents and their influence on schooling (Mbiti, 2016). Additional resources alone have not solved the problems of public (state) schools. Sometimes the additional funds simply disappear and never reach the education sector at all: a study of Uganda by Reinikka and Svensson (2004) found that almost 80 percent of the funds were diverted for political purposes and never reached the schools. In other cases, increased grants to schools have been offset by reductions in expenditures by parents, as Das et al. (2013) show happened in Zambia and India. In still others, additional resources end up mostly as increases in teacher pay. Unsurprisingly, this makes teachers happier but, as de Ree et al. (2015) show for Indonesia, increasing teacher pay substantially had no discernible impact on either teacher effort or student learning. Educating children properly is an inherently complex and heterogeneous task. No simple rules or solutions can resolve the many different problems uncovered in different countries. But some problems can be solved. For example, the widespread lack of teacher accountability – ranging from substantial absenteeism to simple incompetence – can be dealt with by giving local parents some control over teachers, as was done in Madyha Pradesh, India (McCarten and Vyasulu, 2004) by improving school inspection systems and by better teacher training, as well by such more controversial and often contested (especially by teachers’ unions) approaches as tying teacher incentives to student outcomes as, for example, in a recent pilot program in China described in Loyalka et al. (2016). More can be done to make the size and nature of resource flows going to each school clearer both to all those involved in the process and to the intended beneficiaries themselves – the students, and especially their families. Importantly, more could also be done to ensure that those finally responsible for how those resources are used – the schools themselves – have much more

M4439-BAHL_9781786435293_t.indd 78

01/03/2018 11:08



Expenditure assignment and management ­79

discretion than they now have in most countries to be able to allocate their resources: in Kenya, for example, most schools reported in 2013 that they had no discretion at all in how they used the resources directed to them (Mbiti, 2016). If schools have more discretionary spending authority, as a recent study in Indonesia suggests (Pradhan et al., 2014), it is important they are explicitly accountable to elected school management committees that are themselves linked to local government bodies which both have some resources in the game and some authority to respond to (and influence) how schools behave. As Mbiti (2016) notes, decentralizing education alone is unlikely to be enough to ensure improved accountability in providing education. However, doing so may not only strengthen local government; if done well, it may also help improve educational outcomes, and hence contribute doubly to the potential ‘state-building’ role of decentralization.

power to hire, fire, promote and compensate school teachers is largely not under their control, but is instead determined at the national level by negotiations with the teachers’ union.9 In addition, regional governments can and do affect the allocation of teachers to schools in different localities and also, if they choose to do so, hire additional teachers whom they pay directly from their own funds – as may municipal governments using municipal funds.10

THEORY AND NORMATIVE RULES A common question that comes up when thinking about expenditure assignment is whether there are clear normative guidelines that tell us which is the right level of government to provide a particular service. The answer is that there are some rules that provide guidance, but that they need to be applied with judgment and with such common country conditions as severe capacity constraints on both the revenue and the expenditure side of local budgets kept firmly in mind. Different countries may follow the same rules but make very different choices when it comes to who does what: there is no one best expenditure assignment that suits all. The economic literature on this topic focuses on the goal of achieving economic efficiency and has largely been written (implicitly) for a country like the United States in which there is democracy, a tradition of local governance, a mobile population and a high rate of literacy and media involvement. It is not surprising that this literature suggests that there is no ‘one size fits all’ solution. What may be more surprising is that the conventionally suggested assignment rules turn out to be useful guides to policy-makers in low-income countries and are similar to what most ­countries do. Most of the problems we note in this chapter arise when

M4439-BAHL_9781786435293_t.indd 79

01/03/2018 11:08

80

Fiscal decentralization and local finance in developing countries

trying to implement any assignment regime in any country, and are not limited to low-income countries. The Decentralization Theorem The basic rule of efficient expenditure assignment is to devolve each function to the lowest level of government consistent with its efficient performance. In the economics literature, this idea is expressed in the decentralization theorem developed in the classic study by Wallace Oates (1972).11 When there are inter-local variations in tastes and costs there are potential efficiency gains from assigning responsibility for public sector activities to the lowest level of government consistent with its efficient performance. The underlying assumption is that if local decision-makers whose primary responsibility is to satisfy local voters are responsible for deciding what services are provided, to whom and in what quantity and quality, then people are more likely to get what they want and overall public welfare will be enhanced.12 However, a key phrase included in our Chapter 1 summary of the decentralization theorem – “consistent with its efficient performance” – is critical. Assignment to lower levels of government may not always improve national welfare for several reasons.13 Five such reasons may be mentioned. The first two – externalities and economies of scale –are rooted in economic theory. The third – the inadequate capacity of lower levels of government to provide services effectively and efficiently – is more likely to be a problem in lower-income countries, and is one reason why such countries tend to be more centralized. We discuss the capacity question in this and other contexts later. The other two reasons are more political in nature. One is that the relative fragility of many central governments in developing countries may make them reluctant to devolve significant authority to potential rivals at any level. The other, emphasized in the traditional literature on the development of the welfare state (Wilensky, 1975), is that when there is reason to suspect that local elites with little interest in serving the interests of anyone except themselves might capture local governments, more ‘service to the people’ may be achieved by centralizing rather than decentralizing power. Externalities When the delivery of a service leads to impacts on households that reside outside the boundaries of the jurisdiction responsible, lower-tier governments will not spend enough on the function in question from the point of view of society because they will consider only local benefits and costs in making budgetary decisions. Some of the impacts external to the local

M4439-BAHL_9781786435293_t.indd 80

01/03/2018 11:08



Expenditure assignment and management ­81

decision-making process (externalities) may be negative, as when a town sewer empties into a river that is the source of water for another town downstream. Other ‘spillovers’ (as spatial externalities are often called) may be positive, for instance when students who are educated in one community move to another. Whether negative or positive, the existence of externalities in the form of interjurisdictional spillovers means that social welfare is unlikely to reach as high a level with purely local provision as it would if the service had been assigned to either the regional or the national level. Concern with such spillover effects and their potential distributional impact is one reason that so much of the development of social security systems and other components of the welfare state took place at the national level in higher-income countries during the twentieth century. Another is the high degree of migration out of rural areas that in many developing countries has often led to the creation of large, informal settlements in and around cities, and made it all but impossible to restrict expenditure benefits only to taxpaying voters in local and regional governments. The same factors sometimes lie behind the failure of local governments in low-income countries to recover costs with user charges or to impose local taxes even to the limited extent they are free to do so, as we discuss below. Another reason, as we also discuss later, is that local provision of basic services constitutes a more essential component of redistributive policy in low-income countries than in higher-income countries. Close examination of local public services suggests that almost all of them have some externality component. For example, neighborhood parks and cleaner streets clearly have primarily local effects. However, since the creation of such amenities also tends to improve life in the city or town as a whole, they may also make it more attractive for businesses to locate there, and those living in neighboring communities may also benefit as a result.14 As a practical matter, estimating the benefit (or cost) zone for each function is a difficult and complex task, and such zones are likely to differ widely not just from function to function but also between sub-functions (for example, traffic control vs. crime prevention), of which there are often 100 or more in an urban budget. Consequently, probably the most analysts can do is to identify those services where spillover effects are most important, estimate the size of such effects and estimate the extent to which they can be internalized by shifting responsibility to a higher decision level (for example, a metropolitan region) or compensated for in some way (financially through a conditional grant, or perhaps by the provision of some offsetting service). Such analysis seldom yields definitive answers. Even so, the answers produced are likely to be better than decisions reached without evidence. For instance, in several countries around the world, analysis of

M4439-BAHL_9781786435293_t.indd 81

01/03/2018 11:08

82

Fiscal decentralization and local finance in developing countries

the spillovers related to water and sewerage provision has led to the creation of special authorities and districts.15 Sometimes, the external benefit or cost zone is so great that only central government responsibility will do. Examples include activities that benefit or burden the entire population – such as defense and scientific research, and public health services like vaccination. In some cases, there may be an overwhelming national purpose in maintaining standards, for example, with respect to higher education or perhaps old age pensions. In larger countries, regional governments such as provinces or states may be the right choice for delivering such services as inter-municipal roads, watershed management and higher education. In some circumstances, however, externalities may be internalized more efficiently by assigning functions to a lower level such as a district or metropolitan regional entity of some sort.16 Economies of scale The second reason for assigning a function to a higher tier of government is the presence of economies of scale in the delivery of a service. If a service can be delivered by a higher-level government with the same quality but at a lower unit cost than it could be delivered by a lower-tier government, there are production efficiencies to be had by shifting decisions to a higher level. However, economies of scale can also be captured by expanding the serviced population or expanding the territory serviced through such methods as annexation, consolidation or the creation of a special district with a broader service area. Economies of scale (or size) arise for two reasons. The first is technical. Substantial capital investment may be required to lower the per capita cost of delivering the service, as in the case of public utilities or mass transit. Spreading large fixed costs over a larger user base both ensures a fuller level of utilization and presumably lower unit costs and prices for the service. In addition, a larger local government unit may be in a better position both to attract a more skilled workforce and to achieve such distributional goals as coordinating service delivery to even out service level disparities among local areas.17 For example, a larger school district can better afford special education services and may be better linked to the regional provision of other child services. In addition to enabling larger capital investment (mass transit), more productive capital–labor substitution might be possible (refuse collection) and the provision of more specialized services (education, fire protection) may become feasible. Another source of lower unit costs associated with scale may arise from the pecuniary economies achieved by purchasing larger quantities of inputs (e.g. school books, medical supplies). An important example of such economies is the

M4439-BAHL_9781786435293_t.indd 82

01/03/2018 11:08



Expenditure assignment and management ­83

considerable savings realized by national insurance plans that purchase large quantities of drugs and thus can negotiate substantial discounts from suppliers.18 Unfortunately, there is not much hard evidence about economies of scale in the provision of local government services. One reason is problems with the methodology used to measure cost and output (Byrnes and Dollery, 2002). In most studies of scale economies, for instance, per unit government expenditures are used as the measure of cost, and population is used as the proxy for size or scale. Population is not the best measure of the scale of operations for a local public service. For example, two cities might expand their populations by the same amount, but the one with the larger geographic area might not realize the same unit cost reductions as the one with the higher population density. Other factors that affect the gains from increased scale include the concentration of school-aged children and elderly, distance to a water catchment area, and topography and climate. An additional problem is that scale economy calculations often do not factor in the use of local services by non-residents. Questions may also be raised with respect to the measurement of cost: expenditures are not a good proxy because they reflect not only costs but also the quality of services delivered, and perhaps wasteful expenditures. Methodology problems aside, research in this area is not convincing, even in high-income countries. Byrnes and Dollery (2002), for example, reviewed research on economies of scale in the UK and the US, and concluded that only 8 percent of the studies covered found any evidence of economies of scale in local government, compared to 39 percent that found no statistical relationship between per capita expenditure and population size, and 24 percent that found evidence of diseconomies of scale. The remaining 29 percent found evidence of U-shaped cost curves, that is, with costs rising up to some point and then decreasing. Studies that analyzed specific services (e.g. fire, housing) have also shown mixed results. Perhaps the most striking result that has emerged from studies in developed countries is that there is no strong evidence of economies of scale once localities exceed surprisingly low population levels. Studies in both Canada and Finland, for example, found little evidence of economies of scale in large municipalities. Found (2012) analyzed economies of scale for fire and police in 445 municipalities in Ontario, Canada from 2005 to 2008. He found that fire services exhibited U-shaped costs with a costminimizing population of approximately 20,000 residents. Police services also exhibited U-shaped costs with a cost-minimizing population of about 45,000 residents. In Finland, Moisio et al. (2010) reported the results of studies of the effects of municipal mergers on per capita expenditures to

M4439-BAHL_9781786435293_t.indd 83

01/03/2018 11:08

84

Fiscal decentralization and local finance in developing countries

be mixed, with the biggest cities showing relatively low cost efficiency with respect to basic welfare services. Other studies in Finland that focused on specific municipal services (health centers and schooling) found the optimal size of the municipality to be somewhere between 20,000 and 40,000 people (Moisio et al., 2010). Owing to the difficulty of separating the pure effects of scale on government costs from the effects of everything else it would be wrong to conclude from such results that per capita expenditures do not fall as the scale of local government operations increases. But it does seem implausible that bigger is always better when it comes to the size and operating scale of local service units.19 No doubt there is considerable evidence of the cost functions of various local services in developing countries hidden in project design documents, engineering cost studies, case analyses for individual governments and similar sources; but, like Fox and Gurley (2006) in their useful review of this subject, we know few readily accessible sources for such data. The evidence we have suggests that there is no simple answer in terms of scale (and related) economies to the question of what level of government should do what. As Shoup (1969) noted long ago, the average and marginal costs of providing even the most basic local government services vary not only with the service concerned but with such other factors as: the quantity and quality of service provided; the size of the population served; the area covered; the density of population; the physical characteristics of the terrain and of the structures served; and even the variation in the demand within the population served. It is thus difficult to reach conclusive answers as to what services should be provided by what government on the basis of the scanty and case-specific empirical evidence we have. While there are surprisingly few studies of this question in developing countries, one recent study provides evidence of the importance of economies of scale with respect to school consolidation in parts of China, owing in part to the heterogeneity in both local preferences and in local productivity. Even after a decade when over 50 percent of primary schools were closed, many small schools continue to exist. Interestingly, however, the small schools that remain tend to be those with the highest productivity, while the lower-productivity schools that were consolidated appear to have gained sufficiently in terms of economies of scale to produce the result that, on average, student performance has improved (Ma, 2016).20 Technical studies of Indian metropolitan areas indicate that the cost of delivering basic services is 30–50 percent cheaper in metropolitan areas than in sparsely populated areas, and the cost of delivering a liter of piped water is about 50 percent cheaper because cities can leverage common supply depots and cut distribution costs (McKinsey Global Institute, 2010). Another interesting study of economies of scale in water and

M4439-BAHL_9781786435293_t.indd 84

01/03/2018 11:08



Expenditure assignment and management ­85

sanitation services in 14 developing and transitional countries for which comparable data existed (including data on the quality of service) found a similarly mixed picture (Nauges and van den Berg, 2008).21 Presumably, utilities where returns to scale are constant are minimizing costs. However, this study found that only 35 percent of the utilities covered operated with constant returns to scale, while 62 percent – including most that produced less water than those at constant returns to scale – showed increasing returns to scale and the remaining few utilities (usually large and in higherincome countries) had decreasing returns to scale, with much the same picture shown for utilities within a country as between countries. Even the economic efficiency with which water flows over a dam and through the water system is not simply a matter of building to the right scale. The probability that a utility is operated at minimum unit cost depends not only on the volume of water produced but also on how well the country’s economy works in general, as evidenced by such measures as the degree of investor protection, the cost of enforcing contracts and perceptions of corruption (Nauges and van den Berg, 2008). No doubt there is much more evidence about scale economies to be had from such studies but, as usual, we are a long way from having a solid body of evidence on even such basic issues as water supply. Even if significant scale economies did exist for most services provided by subnational governments, one cannot conclude that the national (or a higher-level regional) government should take over. Often, there are alternative ways of organizing service delivery both in rural areas and in large urban areas such as metropolitan regions – for example, through contractual arrangements between local governments or with higher-level governments or with private contractors, although these arrangements do not always give satisfactory results (Wetzel, 2013; Bahl, 2011). Sometimes size economies can be captured with delivery by a regional government on a metropolitan area basis (intra-urban bus services, water supply). Sometimes, regional agencies deliver local public services directly in urban areas, as is done, for example, by parastatals (state-owned enterprises/ SOEs) in India. Sometimes in less densely populated areas some other services (universities, mental hospitals, trunk roads) may be best delivered, especially in large countries, by regional governments. However, we seldom know with precision the exact population size, land area or other characteristics of the supplying unit needed to provide services at least cost. Even if we do have a good idea of the cost function it may not be feasible to operate at the right scale, for example, because of the high transaction costs of making the necessary contractual arrangements. Still, as in the case of externalities, it is better to base decisions about expenditure assignment on evidence about scale economies to the extent possible.

M4439-BAHL_9781786435293_t.indd 85

01/03/2018 11:08

86

Fiscal decentralization and local finance in developing countries

From Theory to Practice The assignment rules suggested by decentralization theory, although far from precise, are both reasonable and helpful. If one asks in almost any country why this level of government does that and another level does something else, officials and experts almost always mention such efficiency issues as those just discussed. However, the actual way functions are divided among levels of government is seldom exactly what the theory would seem to suggest. In some instances, the departures are quite ­radical – for instance, with respect to social insurance in China and local public services in jurisdictionally fragmented metropolitan areas. As Boadway and Shah (2009) correctly note, few services provided by governments are pure public goods in the traditional sense of being both ‘joint in consumption’ (that is, everyone in the relevant provision area gets the same service) and ‘non-excludable’ (that is, they get it whether they want to or not). In poorer countries, where such free-riding phenomena as illegally tapping into water and power sources are prevalent, the line between public and private services is often blurred. What is clear, however, is that, just as the relevant service area for each service usually differs, the fact that few public services are pure public goods means that they cannot be neatly compartmentalized into local, regional and national goods. Another reason what is done in practice is seldom exactly what theory seems to prescribe is that, as just discussed, externalities and scale economies are easier to talk about than to measure. In addition, the question of efficiency in resource use that dominates economic analysis is not all that matters. Suppose central decision-makers determine that increasing the vocational component of primary education and moving away from traditional schooling would improve the productivity of the workforce and hence make everybody, both at local and national levels, better off – but local people and the local government strongly prefer traditional education. The potential productivity gain may perhaps be estimated. However, because it is not clear how to measure the value locals place on traditional education there is no persuasive way to compare the costs and benefits of the two approaches, let alone to determine precisely who would gain and who would lose. In a democratic, decentralized system such issues are somehow decided by voters and their representatives, often reflecting intangible feelings about preferences and judgment. In many low-income countries – in which literacy rates are low, public interest in government budgets is not high and government decisions are seldom transparent – local interests may have little say. Primary education may stay with local governments either because of explicit local political decisions or because central politicians decide that the productivity gains from centralization

M4439-BAHL_9781786435293_t.indd 86

01/03/2018 11:08



Expenditure assignment and management ­87

are unlikely to be large enough to override local preferences and take control of primary schools. Or the opposite may happen. There is in either case no explicit way anybody can compare the welfare costs and benefits of choosing one way or the other. Other concerns such as reducing fiscal disparities and encouraging infrastructure investment in accordance with the aims of the national government also frequently shape decisions on assigning (or reassigning) expenditures. If higher-level governments want to ensure that externalities and scale economies are adequately taken into account when subnational governments make policy decisions, they have many other fiscal instruments at their disposal to influence local decisions, such as the conditional grants and expenditure mandates discussed later. In general, if one is not quite sure about which government should do what, it is probably better to utilize a more flexible instrument like a conditional grant rather than set out a rigid constitutional allocation of functions to a level of government. Mistakes in designing grants are easier to fix than mistakes in designing constitutions.

THE PRACTICE OF EXPENDITURE ASSIGNMENT The data presented in Chapter 2 indicate that about 20 percent of total government expenditures in developing countries show up in regional and local government budgets. But there is considerable variation around this average. For example, subnational governments in Colombia account for 32 percent of government spending but those in Thailand account for only 9 percent. Even in countries in which subnational governments account for a significant share of government spending, wide variation among regional and local governments is common. For example, the richest county in China spends about 48 times as much per capita as the poorest county (Dollar and Hofman, 2008). Governorates (district governments) in the frontier provinces in Egypt spend twice the per capita national average compared to those in Upper and Lower Egypt, which spend only two-thirds the national average (Martinez-Vazquez and Timofeev, 2011). Are citizens more likely to keep local governments accountable if those governments are responsible for a larger share of total government expenditures? In 11 out of the 77 countries for which data are reported in GFS/ IMF in the 1980s, the subnational share is less than 5 percent. Per capita spending is also minuscule in many countries. For example, rural local governments in West Bengal, India, spent an average of only US$3 per capita in 2005 (Bahl et al. 2010). Some years earlier, Stren (1992) noted that, in per capita terms, spending in even the largest cities in many ­developing

M4439-BAHL_9781786435293_t.indd 87

01/03/2018 11:08

88

Fiscal decentralization and local finance in developing countries

countries seldom amounted to more than 1–2 percent as much as in cities of similar sizes in developed countries. If subnational governments spend little and have little control over what they spend, it is not surprising that citizens in many developing countries take little interest in reforms intended to hold elected local officials more accountable for the quality of public services that they deliver. There is little solid evidence about such things as scale economies and externalities, and wide variation in the importance of subnational governments across and within countries. Nonetheless there is a rather surprising degree of consensus about the division of expenditure responsibilities between central and subnational governments in many countries. Viewed from a theoretical perspective, most countries seem to get it right according to the sort of normative schema set out in Table 3.1. For example, such obvious national public goods as defense, foreign affairs and the like are usually assigned to the central government, and local goods like refuse collection to local governments, as has often been noted (Shah, 2007). Many case studies show this pattern: examples include Russia (MartinezVazquez et al., 2006c), South Africa (Khumalo and Mokate, 2007) and India (Rao, 2009). Life is seldom neat, however, and very often more than one government is involved in a number of functions, with higher levels usually taking on an oversight role (and sometimes even the principal responsibility) for service delivery for functions where substantial economies of scale or nationally important externalities are thought to exist. Sometimes the patterns followed deviate from the norm. For instance, China assigns responsibility for health insurance and for old age and disability pensions to subnational governments (Bahl and Martinez-Vazquez, 2006; Bahl et al., 2014). On the other hand, although subnational governments are mainly responsible for primary schools in many countries, basic education is provided by the central government in the Philippines (Manasan, 2009). In Colombia, as noted earlier, all three levels are involved in education, with most funding from the central government (which also sets standards and salaries for teachers and students) being channeled through regional governments which are responsible for administering the system, and local governments being charged with school maintenance. In Mexico, at the time of the devolution of the education function to states, it was decided that the negotiation of future compensation of ‘federalized’ teachers would remain between the federal government and the unions. State government teachers would be governed by negotiation between each state and the union. Currently, 21 states have both federal and state teachers, while 11 states have only federalized teachers, so different states report spending very different amounts on teachers (Revilla, 2012).

M4439-BAHL_9781786435293_t.indd 88

01/03/2018 11:08



Expenditure assignment and management ­89

Expenditure assignments do change, and sometimes in very big ways. In Colombia, for example, the present educational structure evolved from one established after a major national teachers’ strike in the 1960s led to the replacement of the previous regionally dominated educational structure. Something similar had occurred earlier with respect to policing when a decade of substantial regional and national violence beginning in 1948 led to the nationalization of the previously subnational police forces. As in these cases, expenditure reassignments sometimes arise from crises which call earlier arrangements into question. However, they may also result from more deliberate and studied reforms. The backbone of the Mexican health care system is federal-sponsored social security organizations. However, with the initiation of a conditional grant program for health care, state government responsibilities in this area have increased significantly. In other cases, such as Indonesia’s decentralization reform in 2001, assignment reforms were packaged in a ‘big bang’ approach in which, although some important functions were assigned to the center – national defense, international relations, justice, police, monetary policy, development planning, religion and finance – everything else was devolved to local governments. Unfortunately, the failure to assign specific functions to the regional governments became a source of confusion and, arguably, some inappropriate assignments (Hofman and Kaiser, 2004; Directorate of Fiscal Balance, 2012). How Much Autonomy Is Enough? The assignment of an appropriate level and mix of expenditure responsibility to subnational governments is a necessary condition for fiscal decentralization. However, to make decentralization effective, subnational governments must have sufficient autonomy to decide on how much to spend on a service and how to deliver it. There is no hard and fast rule about how much autonomy is enough; and indeed, as discussed in Chapter 2, there are not even accepted guidelines about how autonomy should be measured. About all one can say is that the idea is to give sufficient discretion to elected local officials so that they are held responsible for the quantity and quality of services delivered. In practice, perhaps the best way to think about this issue is to approach it from the opposite direction and, as the decentralization theorem would suggest, ask not how much autonomy local governments should be given, but rather when there are sound reasons for limiting their spending autonomy. Central (and regional) governments almost always limit the fiscal choices that governments lower down in the legal hierarchy can make. Sometimes, such limits make good economic sense, for example, when local actions

M4439-BAHL_9781786435293_t.indd 89

01/03/2018 11:08

90

Fiscal decentralization and local finance in developing countries

may create significant externalities. Sometimes, however, limits may be intended simply to maintain central control – to make it clear who is in charge and to maintain the importance of central officials; or they may be more paternalistic in nature (‘father knows best’). One common limitation on the freedom with which subnational governments can spend is to maintain central control over the number, qualifications and – most ­frequently – the salaries of local government employees. Other limits may be imposed on the composition of local expenditures both through mandates (legal requirements that certain expenditures be made) and through rules requiring, for example, that a certain share of total expenditures should be directed to some function (often education). A more indirect approach to the same end is to provide funds to local governments in the form of conditional grants. Two questions should be asked about all controls limiting local spending. First, is the gain in national welfare sufficient to offset the welfare losses to which restrictions on autonomy presumably give rise? Second, might it be better (more efficient and effective) simply to move responsibility for the expenditures in question to the higher-level government? Although these questions seem seldom asked, let alone answered, many countries have nonetheless imposed a variety of controls and restrictions on the autonomy with which local governments can spend even funds they raise themselves, let alone funds they receive in the form of central government transfers. We look briefly here at three of the most common methods of limiting and directing local government expenditures. Direct Controls on Expenditures The simplest method is to impose direct controls. In many developing countries, public employee compensation at the local level is set by the central or regional government. The number of employees may also be fixed. In both Colombia and Mexico, for example, subnational governments have no control over the wage level, and very little control over the number of teachers hired (World Bank, 2009a). For a labor-intensive activity like education, such a mandate may dramatically limit the ability of local governments to establish expenditure priorities, introduce programs, recruit employees or reward better performers. When such rules prevail, the service may be delivered at the local level and recorded as a local expenditure, but in fact the local government is essentially little more than an agent for a higher-level government. In some cases, although a higher-level government establishes wage and salary levels and the number of employees that must be employed by local governments they may not pay the mandated costs. The results are inevita-

M4439-BAHL_9781786435293_t.indd 90

01/03/2018 11:08



Expenditure assignment and management ­91

bly bad. Local governments may end up raising taxes wherever they can get away with it, cutting other spending, seeking larger transfers from higher level governments (either directly or through such ‘backdoor transfers’ as unrepaid loans) or simply running deficits and hoping to be bailed out. In India, for example, a national Pay Commission is constituted every tenth year to recommend pay increases for central government employees. State governments generally follow the commission recommendations for their own employees both because they feel pressured to do so and because it has become customary. The predictable result is often a significant deterioration in state government finances.22 In other cases, such as Mexico, when the education transfers to states are less than the amount necessary to cover centrally determined teachers’ salaries, the balance is usually covered from unconditional central grants. In some cases, however, as in the Congo (DRC), although the central Ministry of Education determines teachers’ salaries, it appears that increases have been largely funded because allocations were mainly based on existing liabilities (Kaiser et al., 2011).23 Mandated Budget Shares The usual argument for placing restrictions on the composition of subnational spending is to support national priorities – for example, to expand access to education or to expand spending on infrastructure. Brazil’s 1988 constitution requires subnational governments to spend at least 25 percent of their revenue on education, and central government regulations require 60 percent of education expenditures to be earmarked for wages and salaries. Vietnam’s subnational governments are required to spend 20 percent of their budgets on education (and an additional 2 percent on science and technology). In China, subnational government spending on agriculture must increase as fast as total revenues (Mountfield and Wong, 2005). It is unclear whether any of these restrictions is effectively monitored. A less admirable interpretation is suggested by the fact that a common restriction imposed from above is on the number and salaries of local officials. In the Philippines, for example, the local government code imposes a cap on personal service expenditures by local governments (Manasan, 2009). This kind of limit suggests that central officials may see local officials (and their constituents) as slow learners who do not know what is best for them and are likely to make bad decisions unless constrained by a paternalistic hand. Such limitations are often justified as reactions to media stories of local vote buying, nepotism and elected officials not only abusing their power for private gain but also living it up with public revenues. Another justification may be that people tend to think of administrative expenditures as inherently unproductive. In some

M4439-BAHL_9781786435293_t.indd 91

01/03/2018 11:08

92

Fiscal decentralization and local finance in developing countries

countries, it appears that central ministries may have tried to protect their own programs (and bureaucracies) by imposing budget mandates on local governments, as reportedly happened in Uganda in reaction to a move toward more autonomy in local government budget decisions (Smoke et al., 2010). If one thinks that those closest to a problem are most likely to know what it is and to be able and willing to deal with it in the most efficient way, such budget restrictions will impose a welfare cost. But whether or not local governments are truly trying to serve their people as best they can, mandated expenditure controls are often not binding. Sometimes limits can be escaped just by changing names – as in Colombia, where the requirement that a certain percentage of local expenditure must be spent on investment was rendered almost meaningless by defining spending on, for example, education and health as ‘social investment.’ Similarly, in Mexico teacher salaries were simply relabeled to get around the mandate. Anyway, it may not matter much what the central government says because it is unable to monitor what local governments do. The rules may be window dressing included in a law simply to look good or to send a signal to lower-level governments; or, the rules may sometimes be remnants of an earlier time that have not yet been removed from the law. Whatever their origin or intent, countries that impose unfunded expenditure mandates negate many of the potential benefits of decentralization. Central line ministries themselves often are badly run, with payrolls and administrative costs equal to or greater than those of the lower governments that they tell to behave better than they do. Does ‘father know best’ even when he breaks the rules himself ? Central governments with so little confidence in local choices about public services that they are unwilling to put any trust in the potentially beneficial results of empowering local governments and people would seem better advised to stick to a more centralized system with at most some tightly specified and monitored delegation of powers to local governments. Of course, some mandates are found in even the most decentralized of countries. The political case for taking this road is clear. The central government that imposes such a rule can take credit for cutting back what seem to be low-priority expenditures or unnecessary outlays: no cost to themselves, and political points and good publicity for cracking down on local abuse – what politician could resist? Sometimes there really are flagrant abuses, and central governments should step in, particularly when the money being wasted comes from the center in the first place.24 Even though not all local government officials are corrupt or inept, some are. Both elected mayors and appointed local officials have been known to waste public money on hiring incompetent relatives, giving contracts to

M4439-BAHL_9781786435293_t.indd 92

01/03/2018 11:08



Expenditure assignment and management ­93

friends (or to those willing to pay enough in bribes), or slowing construction on a local project to draw out a bribe. Rules against such behavior or expenditure limits reducing the amount that can be wasted on such things may sometimes improve efficiency in the delivery of local public services. But a better way to deal with such problems is not by blindly imposing rules, but rather by better monitoring of what is going on and establishing an appropriate system to deal with abuses through administrative review and penalties, and, if necessary, even criminal proceedings (Shah, 2016). While central governments often do not know in much detail what is actually going on ‘out there’ in the real local world, it is not hard to think of a few guidelines that can be helpful in controlling potential local spending abuses: ●●

●●

●●

●●

Local autonomy should only be limited where there is a strong justification for doing so and a feasible way to capture the intended benefits can be devised. Ideally, both ‘sunshine’ and ‘sunset’ rules should be applied: sunshine requires that the evidence supporting the limitation in the first place as well as regular reports on its effectiveness should be made public; sunset means that limits should be imposed for a defined period with a thorough review process required for renewal. The same rules should apply to tax preferences and conditional grants. If fiscal decentralization is the objective, subnational governments should be allowed to decide what level of administrative support is needed to deliver services, with the decision as to whether they are adequate or not being left up to local voters who (it is assumed) can throw the rascals out – or at least make their remaining days in office so unpleasant that they will leave of their own accord. Unfunded mandates imposed on local governments should be disallowed by law. This was done, for example, in Ukraine (MartinezVazquez and Thirsk, 2011) and is a well-established principle in Denmark (Kurachi, 2015). Countries that decentralize must expect some bad local government behavior, especially during the possibly prolonged transition to the new system. However, when problems occur – for example, some locality introduces a particularly bad law or regulation, or some mayor is particularly egregious in hiring his relatives – they should be dealt with on a case-by-case basis rather than with a sweeping law or regulation imposing this or that limit on local autonomy everywhere.

Such guidelines may seem rather idealistic and impractical to many with experience in much of the developing world. This approach implicitly

M4439-BAHL_9781786435293_t.indd 93

01/03/2018 11:08

94

Fiscal decentralization and local finance in developing countries

assumes that real decentralization is a key policy goal and that many of the other aspects needed to make its achievement possible – stronger own-source revenues, well-structured transfers, adequate capacity-building support, effective information and monitoring systems, and a sound legal framework – are already in place or on the way in a functioning democratic state. Almost every developing country falls short of this prescription in some ways, and some do so in all. Even so, it is often useful to take a much more skeptical view of the myriad ways in which central governments in so many developing countries attempt to control and shape what local governments do. If countries really think their local governments are as ineffective as the proliferation of such controls implies and are so convinced that the central government knows so much better than subnational governments do what they should do and how they should do it, why bother with local governments at all? Conditional Grants Another way to impose limits on local government autonomy is through a grant which the recipient is required to spend for a specific purpose. Conditional grants are much like the funded mandates mentioned earlier. The intended result is usually that the recipient government will spend more on the designated activity than it otherwise would.25 Such grants may improve rather than reduce efficiency provided they reduce the price in terms of the local resources required to provide the service by just enough to (in effect) pay for the additional benefits the activity in question generates for those located in other jurisdictions. Unfortunately, it is difficult to be this precise because we seldom have good estimates of how large such externalities are. It is equally hard to estimate the effect a subsidy of a certain size will have on the provision of the service in question because the extent to which local budgetary decisions react to such inducements varies from locality to locality and from service to service. Because revenues are fungible – a dollar is a dollar, no matter what label it carries – we have no way of knowing how much would have been spent in the absence of the grant. Moreover, to obtain a conditional grant, local governments must usually incur compliance costs that again may vary from program to program and place to place. In China, for example, there have been over 200 different earmarked grant programs, each of which is monitored by a controlling central line ministry which has a strong bureaucratic interest in maintaining its programs (Bahl et al., 2014; Qiao and Liu, 2013; World Bank, 2013). A conditional grant is a less onerous restriction on subnational government autonomy than a fully funded spending mandate because in many

M4439-BAHL_9781786435293_t.indd 94

01/03/2018 11:08



Expenditure assignment and management ­95

countries localities can choose not to accept the grant. Even if they do accept it (and they almost always do in low-income countries), in some instances the grant conditions may be weak enough to allow the recipient government to minimize the displacement effect on other revenues. In others, of course, the conditions (and their enforcement) may be so strong that they in effect end up recentralizing a nominally decentralized activity. In Uganda, for example, several ministries have induced local governments to adopt new approaches to service delivery by imposing new conditions on intergovernmental transfers and crowding out previous unconditional grants to the point where local spending autonomy has been significantly reduced. Unsurprisingly, some observers conclude that fiscal decentralization has been weakened in Uganda (Smoke et al., 2010). Similarly, in Tanzania the central government also controls subnational government expenditures in its priority areas through a set of sector-specific grants earmarked to fit central requirements. Constraining local autonomy through mandates and conditionality may make sense in some cases, particularly when the effect is to expand externality-generating activities or reduce those that generate costs for others. But such measures do little or no good when they are imposed primarily to protect existing political or bureaucratic interests, as sometimes seems to be the case. Even when such restrictions are imposed in reaction to widely perceived abuses at the local level, they are seldom preferable to such alternatives as making the central (or state) government directly responsible for providing the service in question or establishing and implementing an administrative review structure that can and does deal with abuse on a case-by-case basis. One of the key elements in empowering local authorities is to let them make mistakes – and to pay the price for doing so, either by solving the problems themselves or by being replaced by someone who will. A properly designed system with hard budget constraints (see Chapter 5) and an appropriate division of central and sub-central functions with governments being effectively accountable to those they are presumed to serve does not need cumbersome and inefficient mandates and limits, and not all that much in the way of conditional grants. The proliferation of such ‘band-aid’ solutions as mandates and conditional grants neither reaches nor corrects the basic problems.

ACCOUNTABILITY In principle, elected government officials at all levels should be accountable to their citizens (voters) for their actions. Accountability in this sense

M4439-BAHL_9781786435293_t.indd 95

01/03/2018 11:08

96

Fiscal decentralization and local finance in developing countries

is the public sector equivalent of the bottom line in the private sector. Fiscal decentralization is essentially about changing the accountability of subnational government officials from reporting up to higher levels of government to reporting down to their citizens. Doing so is never easy in any country, even those which have formally embraced political decentralization. Nilekani (2008, p. 480) sees limited accountability in India as being largely a result of “an impenetrable bureaucracy [that] protects the elected minister from the often spiky concerns of citizens.” Even if local officials are elected – something not yet true in many low- and middle-income countries – effective downward accountability is not guaranteed. For example, the electoral system may not be open or fair, as when the selection of candidates and the voting process is captured by a small elite group or an ethnic minority, or is largely dominated by party allegiance at the national level. Such structural features as term limits for elected officials and an effective national electoral system including full registration of voters and supervision of the electoral process may reduce such problems to some extent, but few countries have succeeded in cleaning up local electoral voting everywhere. To hold local elected officials accountable to local people for the quality of services delivered (and the taxes collected) the budget process must be transparent and reported to citizens on a regular basis. Elected officials need to be tied visibly to the fiscal decisions that they make, whether this is done by the media, in government reports, or aired in town meetings or on the internet, or in all these ways. Uganda’s practice of making the procedures and rules for reporting corruption law clearer to people may have a perceptible effect on accountability (Deininger and Mpuga, 2004). Another approach is to establish more transparent and comprehensible formal systems for budgeting and financial reporting, as we discuss later in this chapter.26 Voters cannot throw the rascals out unless they know who the rascals are and what they have done. People can become more directly involved in evaluating the quality of public services in a number of ways – such as public hearings, community score cards, citizen report cards and public-interest litigation. The use of citizen report cards supported by well-developed fact-finding (and with the results well publicized in reports on the outcomes) can be an especially effective way of measuring citizen satisfaction with public services (World Bank, 2004).27 However, as one of the pioneers in developing and fostering such demand-side approaches to good governance noted with respect to experience in India, although a few good outcomes have been observed, this approach has not been very systematically applied or studied; but it seems likely that “the impact of demand-side initiatives is unpredictable and may vary with time, the context, and complexity” (Paul, 2014, p. 257).

M4439-BAHL_9781786435293_t.indd 96

01/03/2018 11:08



Expenditure assignment and management ­97

The idea is good; but much remains to be learned about what to do and how to do it in order to make downward accountability a reality in the circumstance of most developing countries.28 For people to take regional and local governments seriously, those governments need to be responsible for services that affect their lives and, as Manor (2013) says in the quote that launched this chapter, to have some real control and influence over how well those services are provided. This is one reason why hiring, firing, promotion, and wages and salaries of local officials should be determined by the governments in question and not controlled by the central bureaucracy. Similarly, as we discuss further in Part III of this book, for people to care much about local governments they need to know that these governments are ‘theirs’ – that is, that they are responsible for raising a significant amount of their own funding from revenue sources over which they have visible (e.g. rate-setting) control. People will pay more attention to how well local officials do their job when they know that they themselves are paying for much of what the officials are doing, just as they are more conscious of the tax price of public services when they can perceive the taxes in question. Given the lack of autonomy and the visible ineffectiveness of local governments in some low- and middle-income countries, it is not surprising that some have suggested various alternatives to tackling the difficult task of improving those governments. However, as suggested by the discussion in Box 3.2, none of these suggestions seem to do the job. There is no short cut to improving local governments.

CLARITY, CONCURRENCY AND UNBUNDLING One reason it is often difficult to assess local government accountability and performance is that there are no clear measures or standards for either of these concepts. Such commonly used measures as the World Bank’s measures of good governance are based almost entirely on perception and largely ignore important questions of process. When it comes to considering how well subnational governments perform, it is essential to take explicitly into account what Taylor (2016) calls the ‘embeddedness’ of local governments. Such governments never stand alone. They are always embedded in broader economic, societal and institutional contexts that shape to a large extent how and how well they perform. These contexts determine both the capacity and the autonomy of local governments, and hence largely define how and to what extent they are accountable and to whom, and often also how well they perform their assigned functions. Anyone who looks closely at local governments in any country soon learns

M4439-BAHL_9781786435293_t.indd 97

01/03/2018 11:08

98

Fiscal decentralization and local finance in developing countries

BOX 3.2  ALTERNATIVES TO LOCAL GOVERNMENTS Conflicts between traditional local governance institutions and informal local government structures sometimes give rise to concern. As Mohmand (2016) stresses, however, a key role played by the informal local governance institutions found in many developing countries – such as informal panchayat (village) councils in India, chiefs and traditional authorities in many parts of Africa, and the neighborhood associations found in many countries – is to intermediate between citizens and state authorities, including formal local government structures. One way such groups may do so is by supplementing what the formal governments do in terms of providing services. To the extent informal governance institutions play this latter role, they may, as Mohmand (2016) argues, be a way to ensure that at least some of the more marginalized groups in local society become incorporated into the system, thus strengthening decentralization.1 Others seem to have given up the struggle to improve local governments, and suggested that alternative civil society (non-government) institutions offer a preferable way to deliver (or at least to influence the delivery of) local public services. The World Bank has labeled this movement ‘community-driven development’ (CDD). A simple definition of CDD is an approach that gives communities control over the decision-making, management and use of development funds (Wong and Guggenheim, 2005). A CDD defined in this way is in effect an alternative to the formal local government. Unsurprisingly, this approach has given rise to considerable controversy.2 The community organizations that are the focus of this approach are generally concerned with addressing the unmet needs of the poor in small rural communities. They are often governed by an elected local committee, have significant autonomy in service delivery, and may be aligned with either the central or the local government. Their funding usually comes from outside and is dedicated for a project, often a small-scale development project. The idea is attractive: satisfying local needs and preferences by moving government decision-making to a grassroots level that gives beneficiaries more voice. Often, it is argued that simply electing a conventional local government will not serve because local governments have been captured by an elite group.3 Outside donors often like this approach because they can bypass the cumbersome local government bureaucracy, and therefore simplify the project delivery and monitoring process. Other advantages of CDDs are that they may contribute to nation-building and enhance the accountability of elected government officials by stimulating citizen interest in governance. Small rural projects often get lost in the shuffle in planning and service delivery by subnational governments or deconcentrated line ministries. Sometimes the cost of delivery may be lower and more local matching resources may be drawn out: in one Philippine case, for example, CDDs were shown to deliver infrastructure projects at costs 25–56 percent less than similar projects delivered by other public agencies (Wong and Guggenheim, 2005). CDD leadership may also be more able to explain projects to the community constituency, and may be able to implement more effective monitoring than can a more general government agency. An obvious criticism of many of these possible gains is that they largely depend on who leads the CDD and how. Do the leaders truly represent ‘users’? Do

M4439-BAHL_9781786435293_t.indd 98

01/03/2018 11:08



Expenditure assignment and management ­99

potential beneficiaries have more voice in such committees than they would have in a local government election? As Manor (1999) notes, and Marcesse (2016) shows can happen, whatever the intention of the funders, it may well be that the local leaders are in fact effectively appointed by government officials or are from the local elite. As Smoke et al. (2010) note with respect to civil society organizations in Uganda, the poor may well end up once again marginalized. Mansuri and Rao (2004, p. 1) argued that community driven projects “are dominated by elites, and both targeting and project quality tend to be markedly worse in more unequal communities.” The World Bank (2005, p. xv) urged caution: “Given the mixed and limited evidence of the impacts of CDD projects . . . the Bank should approach future . . . projects, particularly CDD, with greater care.” We agree. Notes: 1. Much the same process occurred over the last two centuries in North America as many immigrant groups were gradually assimilated into urban political life (Banfield and Wilson, 1963). 2. Documents taking this approach may be found in such sources as www.worldbank.org/en/ topic/communitydrivendevelopment. More critical views may be found in such sources as Manor (1999), Mansuri and Rao (2004, 2012), King (2013), and Mogues and Erman (2016). 3. For an example of how such a group can control even policies intended directly to serve poor households, see the analysis of India’s National Rural Employment Guarantee Scheme (NREGS) in Marcesse (2016), as discussed further in Box 3.4.

that many of their apparent defects are less the results of any inherent local incapacity than reflections of the institutional structure in which they must operate. For example, as already mentioned, the assignment of expenditures between different levels of government is almost always murky. It is seldom entirely clear who is to do what or, for that matter, just what ‘what’ (or sometimes even ‘who’) means. One result of this lack of clarity may be duplication in the delivery of some services; another may be no delivery at all. Almost always there is little coordination among agencies or governments that provide related services. Sometimes such duplication signals waste that can be reduced. Sometimes, however, what seems to be duplication may provide a degree of useful or even necessary redundancy in a complex and changing environment (Landau, 1969). Care needs to be taken in generalizing about such matters, and perfect clarity is probably an unattainable goal. Still, given the grossly inadequate level of local service provision in most developing countries, it is obviously important to eliminate unnecessary duplication where possible and to rectify service shortfalls due to missed assignments. Clarifying the lines of responsibility and accountability with respect to who does what is therefore important. But doing so is difficult in most countries because the legal and institutional arrangements that govern expenditure assignment are seldom easy to change, especially when they

M4439-BAHL_9781786435293_t.indd 99

01/03/2018 11:08

100

Fiscal decentralization and local finance in developing countries

are enshrined in the constitution. Change is also difficult because the delivery of public services is generally complex. It is not always clear in theory at which level services should be provided, and some governments at regional and local levels have considerably less capacity than others. Change is also difficult because the whole question is essentially political. Changing the distribution of assignments, even if the changes make the system more rational, risks turf wars among ministries as well as among central government officials and politicians, between them and local ­politicians and officials, between different levels of subnational government, and among different regions of the country. As with the striking case of the assignment of responsibility for pensions to city and county governments in China (Bahl et al., 2014), once assignments are in place they often become sacrosanct to the point where the political and economic risk of change, even in a centralized regime, may be too high to do anything. The trouble often begins simply because the law is not clear about exactly who is responsible for doing what. Sometimes, as in Albania’s Local Government Act, there is a list of exclusive and shared functions but no real clarity about functional responsibilities (Dafflon, 2011).29 Or the law may say something about this but it is hard to know what it means. In Egypt, for example, Martinez-Vazquez and Timofeev (2011, p. 401) note that: The closest resemblance to a formal assignment of expenditure responsibilities can be found in Law 43 of 1979 on Local Administration. However, the references to specific functions are scattered all over the text of this law . . .. Given the vagueness of the law, it is impossible to map formal expenditure assignment by level of government.

Even in Indonesia, which made expenditure assignment the centerpiece of its intergovernmental fiscal reform in 2001, instead of assigning specific expenditure responsibilities to each level of government, only the functions of the national government were identified (national defense, international relations, justice, police, monetary policy, development planning, religion and finance), with everything else being devolved to the subnational governments. Presumably, and perhaps understandably, those supporting the new law deferred the potentially difficult and controversial expenditure assignment issues in order to get on with implementation before the political opposition could get organized. The result, however, was considerable confusion and, arguably, some inappropriate assignments (Hofman and Kaiser, 2004). Haste may sometimes be necessary to get laws in place; but the result may often be substantial waste in the years – and decades – before the holes left in the initial legislation are patched. Even setting expenditure assignments out in the law may not produce

M4439-BAHL_9781786435293_t.indd 100

01/03/2018 11:08



Expenditure assignment and management ­101

clarity. Constitutions and budget laws often have not only an exclusive list of functional responsibilities for central and subnational governments but also a concurrent list, i.e., a set of responsibilities that can be taken up by more than one level.30 In practice, responsibility is often divided, with the lower level responsible for delivery of the service, and policy, standards and oversight being the responsibility of the higher level (see Table 3.1). Sharing responsibilities like this allows higher-level governments to play a role in guaranteeing some standards of service. But it also often results in confusion about who is responsible for what, and sometimes leaves the door open for any level of government to introduce or kill programs. In Karnataka state in India, for instance, a study found 24 programs for drinking-water supply as well as 72 related programs for the welfare of tribal people, with five levels of government involved in the provision of those services – and a significant failure to deliver the service (World Bank, 2007). Too many cooks may not just spoil the broth but may also consume so much of it themselves that there is little left for the people they are supposed to serve. Another reason why matters become easily confused is because many expenditure functions are inherently complicated. A case in point is primary education, a service comprising many sub-functions that are usually unbundled with different levels of government taking responsibility for different subcomponents. In India, decentralizing the control of primary school teachers to rural local governments in Madhya Pradesh state – including the power to hire and fire – resulted in one of the lowest rates of teacher absenteeism in the country (McCarten and Vysalu, 2004). Some confusion must be lived with, but life can and should be made simpler for people and for local governments than it now is in many countries. This is a problem that cannot be resolved from below. The central government (or in some countries regional governments) must lead the way because the first necessary step is to clean up the sort of legal ambiguity and confusion noted above. The second step is to clarify the issues arising with concurrency by unbundling functions sufficiently to identify intergovernmental responsibilities more clearly. In the Indian case mentioned above, for instance, the sub-functions of curriculum design, textbooks, the development of learning materials and monitoring and evaluation continued to be assigned to higher levels of government. Unbundling may sometimes give rise to further confusion, however. In Peru, for example, although the sub-functions assigned to regional and local governments are described in great detail, there remains considerable overlap between the two levels as well as with the central government (World Bank, 2010; Martinez-Vazquez, 2013). Sometimes, such issues are best dealt with one by one as they come up. In other cases, however, a

M4439-BAHL_9781786435293_t.indd 101

01/03/2018 11:08

102

Fiscal decentralization and local finance in developing countries

more holistic approach may be required. For example, to respond to needs in different neighborhoods, health clinics may require different treatment capabilities, medicines, ambulances and physical facilities, and getting the right mix may be more difficult if each is the responsibility of different levels of government. One way to address such problems is with formal intergovernmental cooperative contracts that make the division of responsibility clear. But for such arrangements to make implementation easier they need to be detailed, well understood and enforceable contracts about exactly who is responsible for doing what, when and how. Such arrangements must obviously be made on a case-by-case basis and may require considerable negotiation.31 Some developed countries have extensive experience along these lines (OECD, 2006; Spicer, 2015). There are also examples of workable arrangements in low-income countries, for instance, for solid waste collection and disposal in the metropolitan areas of Manila and Buenos Aires (Nasehi and Rangwala, 2011; Llach, 2011). As Dunn and Wetzel (2000) note, the failure to establish effective coordination was a general problem with fiscal decentralization in transition countries, with one major problem being the inability to establish a workable way to settle the inevitable conflicts that arise in contracting arrangements.32 A very different approach is simply to recentralize control over subnational expenditures, something that is obviously easier to do in the absence of local democracy – as the relative ease and success of recentralization in Russia (De Silva et al., 2009) and China (Bahl, 1999) shows.33 As noted earlier, moves to recentralize were apparently less successful in Uganda (Smoke et al., 2010). The opposite approach is formally to assign some concurrent functions (or sub-functions) to lower-level governments and to live with the consequences. As we show in later chapters, if the fiscal structure is established correctly – with hard budget constraints for lower-level governments that have sufficient access to revenues under their own control, and adequate accountability to those they tax for how well they use the funds – the logic of fiscal decentralization suggests that the outcome should be better public sector services than any conceivable centralized alternative (provided the central government takes primary responsibility for dealing with distributional issues). This is roughly how it works in decentralized developed countries like Australia, Canada, Germany, Switzerland and the United States – although each of these countries takes a different position with respect to the distributional issue.34 Pakistan and Indonesia are obviously very different from the countries just mentioned. While Indonesia’s wholesale shift of nearly all functions to the local government level appears to have gone fairly well, Pakistan’s experience in 2012 – when it assigned all concurrent functions to the

M4439-BAHL_9781786435293_t.indd 102

01/03/2018 11:08



Expenditure assignment and management ­103

provincial governments and eliminated several central ministries – seems to have been less successful. One reason for this move was to reduce total government outlays. However, few cost savings were realized because many of the redundant central government employees were not absorbed by the provincial governments (as had been envisaged), and many others were reassigned to other central ministries, and thus remained on the central payroll (Ghaus-Pasha and Pasha, 2015). Despite such experiences and the probability that there will always remain a certain degree of murkiness in expenditure assignment when there are multiple levels of government, countries should keep a close eye on problems that arise in this respect and deal with them as best they can before serious damage is done. One way to do so may be to have a periodic review of how well the ‘who does what’ issue of expenditure assignment is being handled and how it may be improved. Colombia, for example, moved a long way toward a more decentralized structure in its 1991 constitutional reform. Unusually, however, it provided for periodic reviews of the new structure, with the first taking place in 1995 (World Bank, 1996). Over the last quarter century, these reviews have considerably clarified and improved the system, although it still, of course, falls far short of getting it right in all respects (Bird 2012a). But then the same could be said with respect to every country. As the world turns, so must the ways countries govern and finance their public sectors.35

THE DISTRIBUTIONAL ISSUE One of the most difficult normative issues arising in expenditure assignment is the extent to which regional and local governments in low-income countries should play a role in shaping the distribution of income. Traditional economic arguments suggest that central governments have a comparative advantage when it comes to distributional issues (Musgrave, 1959). If a subnational government finances pro-poor services and transfer payments with progressive taxes and direct levies on business activities, higher-income families and capital investment may migrate to other jurisdictions. Meanwhile, lower-income families would be attracted (or held) by the better public services available. The result would be a fiscal balance that is not sustainable: an oft-cited example is the New York City fiscal collapse in the 1970s.36 Although the rural poor in developing countries frequently migrate to cities to take advantage of job opportunities and better services, there is relatively little evidence of how location-sensitive capital and labor are to small local fiscal differences. The well-off are unlikely to leave behind the

M4439-BAHL_9781786435293_t.indd 103

01/03/2018 11:08

104

Fiscal decentralization and local finance in developing countries

better infrastructure, public services and amenities in the big cities, and businesses are likely to want the competitive advantages of cities stressed by Glaeser (2011) and others. But some businesses may leave if congestion and increased security problems increase the cost or erode the quality of life for employees and clients, particularly if they can move to a lower-tax suburb and still have all or most of the advantages of a large metropolitan area, as Bird (2012a) suggests has occurred to some extent in Bogotá. The locational impact of local public finances seems to be more a problem for metropolitan areas than for the country at large (as discussed in Chapter 8). The real issue about the distributional role of subnational governments in developing countries is related less to how they tax than to how they spend. One reason why local expenditures may have important distributive impacts is simply because such essential services as water supply and sewerage, public transit, and sometimes education and health services are within their control. Another is because lower-level governments should have more detailed knowledge about the local economy, and hence a comparative advantage in identifying beneficiaries and tailoring services to meet local needs. In part for this reason, local governments sometimes explicitly share responsibility for redistributive transfer payments with central governments.37 Safety net programs for poor families and individuals can be provided in many ways, such as cash transfers (sometimes dependent on meeting certain conditions such as enrolling children in school), public employment programs (as in rural India) and housing assistance. As a rule, most financing for such programs comes from higher-level governments. Because there is a national policy objective (poverty alleviation), there is a strong case for centralization – although some programs may still be best administered at the subnational level (see Box 3.3). One reason for centralization may be to ensure uniform national standards rather than risk wide variations based on the views held in some areas about the importance of protecting the poor – perhaps particularly those from a minority ethnic group. Another issue, discussed further with respect to intergovernmental transfers in a later chapter, is the need for risk pooling to deal with such problems as higher unemployment in declining regions. A frequent argument is that only the central government has sufficiently strong administrative capacity to manage some aspects of these programs. Regional provision of such basic services as health and education is sometimes contested because it is thought that it will exacerbate regional disparities, although the evidence on this point is far from clear.38 The main argument for involving subnational governments in distributional policy is that they presumably know more about the specific needs of their constituency, provided they define their constituency to include all who reside within their borders.39 This argument applies most clearly

M4439-BAHL_9781786435293_t.indd 104

01/03/2018 11:08



Expenditure assignment and management ­105

BOX 3.3  SUBNATIONAL REDISTRIBUTIVE PROGRAMS In China responsibility for social security and unemployment compensation is assigned to the lowest levels of subnational governments. Provincial and especially local governments may be the right ones to identify needy households and administer distributional programs, but the extent to which they can or should finance them is problematic. Obviously, it is the poorest communities that are least able to finance such programs, so one result of assigning this responsibility to lower levels of government has been increased disparities in funding and growing arrears in poorer regions (Bahl et al., 2014; Dabla-Norris, 2005). A similar funding issue arose in Russia in the 1990s when the central government mandated that subnational governments implement new welfare programs but provided no funding for them. The obligation to pay monthly child benefits and subsidies for disabled persons led to a steep increase in the subnational government share of social protection spending – and a considerable increase in the size of their deficits (Martinez-Vazquez et al., 2006). Responsibility for India’s central schemes for rural public employment and income maintenance programs is shared. The gram panchayats (village governments) administer this program with central government grants that pass through their budgets and with state governments covering 25 percent of the cost (Bahl et al., 2010a). Local governments have some discretion in deciding on the type and location of public works projects to be carried out, with the distribution of the funds across states and the eligibility rules determined by the central government (Manor, 2013). Somewhat similar programs exist in Bangladesh, Argentina and several other countries. In Chile and Brazil, the central government finances an upgrading of urban infrastructure when slum upgrading is part of a municipal development plan, with grants to local governments that implement the projects.* Cash programs for poverty relief have been growing in popularity in Latin America in recent years. Most programs of this type are administered directly by central governments. However, Brazil’s Bolsa Familia program is a cash transfer for low-income families where beneficiaries are selected by municipalities (Medeiros et al., 2008). Note:  * For a useful recent review of slum upgrading and its financing in developing countries, see Freire (2013).

to programs designed to deal with individual problems such as disabilities, or intended to be conditional on recipients behaving in certain ways – e.g. actively looking for employment, attending school or visiting a health clinic. With the wider extension of information and banking access even in some of the poorest parts of the world it is increasingly easy to handle simple cash transfers centrally, as is the common practice in most developed countries. One variation of this approach that has received a great deal of attention is India’s massive National Rural Employment Guarantee Scheme. This ambitious program – which Manor (2013, p. 2) calls “the largest downward

M4439-BAHL_9781786435293_t.indd 105

01/03/2018 11:08

106

Fiscal decentralization and local finance in developing countries

transfer of funds to democratic local government, ever, anywhere in the world” – was launched in 2005 and is discussed briefly in Box 3.4. An important distributive role for subnational governments is to provide such essentials of a better life as public security, safe drinking water, sewage and drainage, housing, primary education and local health clinics to poor neighborhoods and poor people. An ambitious transfer program in India – the Jawaharlal Nehru National Renewal Mission (JNNURM) – was focused on urban infrastructure investment and poverty relief.40 The main elements of the poverty component of this program were the provision of shelter, basic services and civic amenities. Somewhat like the rural employment program discussed in Box 3.4, the JNNURM was funded by a combination of transfers from the center and by state and local government resources, with implementation being led by the state and local governments. In practice, the state and local share of resources was slow in coming, as was progress with the project (Pethe, 2013).41 Perhaps the most common approach aimed at protecting the standard of living of poor families found in low-income countries is to establish a program of subsidies on items that weigh heavily on the budgets of poor families, although such subsidies are seldom targeted only at the very poor. The primary targets for subsidy policy are: basic foodstuffs (which in poorer countries, as in richer ones, are often subjected to lower taxes and sometimes explicitly subsidized through the budget); housing rents (which are also sometimes openly subsidized); and items like transit fares and utility rates that are often provided at prices well below cost either to targeted groups or to everyone. As shown in Box 3.5, such subsidies often apply to services provided (and in some instances largely financed) by subnational governments. The usual justification for such subsidies is that they address the needs of low-income families for essential consumption. As many studies have shown, however, such supply-side subsidization seldom delivers the promised goods.42 Not only does much of the benefit accrue to the non-poor, but also when utilities charge users less than the cost of the services provided the result is often that they are unable to extend access to services or to maintain service quality adequately, with the poor being the most likely to lose out on both accounts. Moreover, since costs are not covered even for the inadequate services provided, financing must be found somewhere: either deficits expand continually or some level of government raises taxes (or cuts spending) elsewhere to cover the utility’s deficit. The net allocative and distributive outcomes of distorting both the operating decisions of the utility and the budgetary decisions of the funding governments are seldom clear but are most unlikely to be all positive, since complexity and o ­ bscurity is seldom conducive to good decision-making at any level. Subsidy policy may often be justified on distributive or externality grounds, but whenever

M4439-BAHL_9781786435293_t.indd 106

01/03/2018 11:08



Expenditure assignment and management ­107

BOX 3.4 INDIA’S RURAL EMPLOYMENT GUARANTEE SCHEME The Mahatma Gandhi National Rural Employment Guarantee Act became law in 2005. This program was unique not only in terms of its scale but also in its ambitious design. In contrast to most transfer programs it was explicitly intended to be essentially demand-driven and to finance projects proposed and developed by local communities. In addition, the substantial funds transferred through this program were intended to be delivered not to such intermediaries as local governments or private contractors but directly to the intended recipients – the rural workers who actually worked on the projects. Two recent evaluations of how well it has worked so far appear at first glance to present very different views, but on closer examination do not differ substantially. Manor (2013) basically concludes that this program largely succeeded in channeling vastly increased funds to India’s thousands of rural villages. Over US$10 billion of the $20 billion spent on the National Rural Employment Guarantee Scheme (NREGS) between 2006 and 2010 flowed to elected rural local councils, and over 60 percent of the total went to pay workers’ wages. Moreover, in many instances the program worked in ways that achieved its aim of increasing both the transparency of government operations and the extent to which the intended primary beneficiaries, poor rural workers, were involved in the process. Many projects were conceived locally to meet local needs as more people became active in response to the greater availability of funds, and millions of local workers (more than half of them women) were employed to deal with such local priorities as repairing and building roads and wells that had long been neglected by higher-level governments. In many cases workers were paid what they were supposed to be paid, on time and without fuss, which is not the way the public sector has usually worked in rural India. But not all went well. As Manor (2013) notes, money is not enough to empower the rural poor. They also need access to sufficient power to enable them to use the funds as they wish. In addition, everyone involved in channeling all this money around has to be publicly accountable for what they do. Unfortunately, matters did not always work out as intended in either of these respects. State governments did not always flow the full share of funds to local governments as intended, and state officials who were required to approve projects sometimes overrode local wishes; sometimes they had to be paid off to provide a needed approval or other document; sometimes they simply did not act; and they were in almost every case completely unaccountable to local people for what they did or did not do.* Many local councils were ineffective, with decisions essentially being made by the chairpersons, some of whom were not immune to temptation; some were ineffective and overwhelmed by the need to deal with the masses of reports and documents they now had to deal with, and some continued to serve mainly the local elite group they represented. A subsequent detailed examination of how the NREGS functions in one of the poorest states shows that at least in this state officials at almost every level of government seem to have managed to continue traditional informal practices through which they skimmed funds from the public sector even when confronted with modern accounting and technology intended to block such corruption. The

M4439-BAHL_9781786435293_t.indd 107

01/03/2018 11:08

108

Fiscal decentralization and local finance in developing countries

result was that those at the end of the process – rural workers – still have little real say about who does what (or in some instances, anything, is done) and did not receive much of the funds that flow through this system (Marcesse, 2016). Soon after the program was launched, a major ‘transparency mechanism’ intended to check corruption in the form of siphoning funds from the system before they reached workers was introduced. This stipulated that workers must be paid not in cash but only through bank accounts – an approach intended also to expand the extent to which the rural population was involved with the formal financial system. Money could still be stolen by e.g. overbilling for other components of projects, but this requirement made it harder to steal. However, clever crooks can always find a way: in one area studied by Marcesse (2016), for example, frequent power cuts – a common problem in India – rendered dependence on inadequate rural banking facilities so problematic that the only way to enter the data required to generate wage payments was by using a local cybercafé. However, the café happened to be run by a former official who managed to introduce many discretionary ‘mistakes’ (diverting funds into various official pockets as well as his own) into a system intended precisely to reduce the undue exercise of such discretion. While the improper diversion of funds is less under the NREGS than under many other transfer programs in India, there is nevertheless always some slippage (and corruption) in programs that inevitably depend on the transmission of credible information across an array of bureaucrats, each of whom may have something to gain by distorting the flow of information in ways that benefit them – usually at the expense of the intended beneficiaries. Note:  * Manor (2013, p. 8) labels the failure of the NREGS to incorporate transparency mechanisms covering levels above the local council as a serious defect. He is right; but experience suggests that attempts to improve performance at the working level in any organization almost invariably suffer from this same problem, namely, the failure to take a systemic view and to understand the ways and extent to which local performance is embedded in a larger system.

possible such policies should be implemented by directing aid to the intended beneficiaries and not to suppliers. Mandating that locally controlled utilities charge below-cost prices to some and above-cost prices to others is unlikely to help the poor much (though it can perhaps be a way to tax the rich), and may be a recipe for fiscal disaster. We return to this topic when we discuss user charges more generally in a later chapter.

MANAGING SUBNATIONAL FINANCES It is incumbent on local governments to steward resources effectively, manage relationships with higher levels of government, properly account for public funds to citizens and creditors, and maintain the fiscal transparency and strengthen the accountability that is the foundation of the social contract between citizens and local government. (Habitat III Policy Paper Framework, 2015)

M4439-BAHL_9781786435293_t.indd 108

01/03/2018 11:08



Expenditure assignment and management ­109

BOX 3.5 SOME EXAMPLES OF SUBSIDIZING BASIC SERVICES ●

In India, some states subsidize the cost of power with much of the benefit flowing to the agricultural sector, which pays less than 10 percent of the cost of supply (Monari, 2002): in 2000, such subsidies amounted to 1.4 percent of GDP (World Bank, 2004). Although the agricultural sector is also heavily subsidized when it comes to irrigation, the cost of this subsidy element can only be calculated indirectly. One estimate is that on average for the period 2004–2008 its cost in four southern Indian states alone was almost US$600 million (Palanisami et al., 2011). ● Like many cities, Budapest provides a significant continuing subsidy to the city transport company. Bucharest similarly subsidizes its city transport company (which covers only 60 percent of costs), and also provides operating subsidies to the heating sector (Bahl, 2011). In Mumbai, in 2011, about 45 percent of the cost of the bus system was subsidized (Cropper and Bhattacharya, 2012). ● In a sample of Asian countries, water tariffs were found to cover about 85 percent of operating costs and to make no contribution to the cost of capital, with the subsidy for sanitation being even larger (Peterson and Muzzini, 2005). In a broader survey, fewer than 40 percent of water utilities worldwide imposed tariffs sufficient to cover short- and long-term costs; and none in Africa, Eastern Europe or Central Asia imposed tariffs sufficient to cover basic operating costs (Le Blanc, 2007).

Important as public financial management (PFM) at the subnational level is, all we can do here is provide a brief sketch of the subject.43 We begin with an outline of some key components of a good expenditure management system, noting some problems common to many low-income countries. We emphasize the role the central government needs to play both in establishing an intergovernmental structure in which good subnational PFM can thrive and in providing sufficient support and oversight to ensure that it does.44 It is not an easy task to implement a sound PFM system for subnational governments. The aim of PFM is expenditure control and fiscal discipline, that is, defining a resource envelope for the government and staying within it. Doing so successfully is especially problematic for subnational governments because their revenue limitations are generally determined largely by the rules for distributing intergovernmental transfers, and their control over expenditures is are limited by mandates, conditional grants and sometimes by non-controllable wage budgets. Accountability, fiscal discipline and controlling corruption require governments at all levels to have clear and properly implemented budgetary and financial systems and procedures. Budgets, financial reports and

M4439-BAHL_9781786435293_t.indd 109

01/03/2018 11:08

110

Fiscal decentralization and local finance in developing countries

audit outcomes should be comprehensive, comprehensible, comparable, verifiable and public. In addition, and importantly, public sector resources should be applied as efficiently and effectively as possible to achieve desired public outcomes. Adequate and appropriate procedural norms are important means to the end of achieving desired outcomes, but outcomes are what really matter. Proper public expenditure management requires: (a) adequate control of the total level of revenue and expenditure; (b) the appropriate allocation of public resources among sectors and programs; and (c) procedures to ensure that governmental institutions operate as efficiently as possible (World Bank, 1998). To achieve these aims, as stressed earlier in this chapter, local and regional governments need sufficient authority to manage their expenditure budgets. They also need not only the capacity to do so efficiently and effectively, but also an adequate and appropriate level of support, guidance and monitoring from the central government as well as an intergovernmental fiscal structure that makes it possible for them to operate effectively. This list of requirements for a perfectly managed subnational fiscal system is not one that any country, let alone relatively poor and institutionally underdeveloped countries, is likely to be able to fully satisfy. As so often in life, however, it is most important to strive to do all these good things precisely in the places where they are hardest to accomplish, because wasting public resources imposes the highest social costs where there are the fewest such resources to waste. Whether one’s primary aim is to grow the economy, to build a stronger nation-state or to share the wealth more fairly, it is usually more important to spend well than simply to tax more. Improving budgetary processes, carrying out better analyses of proposed programs and projects, and managing, controlling and reviewing public expenditures are seldom glamorous or politically saleable activities. But progress on all these fronts at the subnational level is as essential an ingredient for success at the local and regional level as it is at the national level – though it may be a mistake to try to do too much too fast.45 Basics First A recent IMF study on expenditure control (Pattanayak, 2016) notes that several key ideas have dominated much of the recent discussion of managing expenditures at the national government level: ●● ●●

an increased focus on ex ante controls over commitments rather than simply ex post controls over payments; a shift from controlling only cash expenditures towards controlling accrued liabilities as well;

M4439-BAHL_9781786435293_t.indd 110

01/03/2018 11:08



Expenditure assignment and management ­111 ●●

●● ●●

more devolution of responsibility for routine expenditures to operating agencies, with central control moving towards a risk-based approach; stronger reliance on internal and external audit; and emphasis on greater transparency and accountability to the legislature and the public.

This is good counsel. At the subnational level in many developing ­countries, the lack of systematic tracking and control of commitments and e­ xcessive, and redundant, controls over payments are common ­ problems. More fundamental to the building of the social contract between c­itizens ­ and local governments mentioned in the epigraph to this chapter is the emphasis on greater transparency and accountability. The possibility of ­devolving more management responsibility to the operating level should be given more consideration at the local level, as we mention with respect to ­education in Box 3.1. Nonetheless, it would be a serious mistake for most subnational governments in developing countries to try to move too quickly down any of these paths, especially perhaps to move from emphasizing control over cash to a more accrual-based accounting system.46 Accrual budgeting and accounting is clearly superior in terms of managing the full resource implications of policies and programs. But it is a serious mistake for subnational governments to try to run before they can walk. Most low-income countries developed budgetary and accounting systems that focused on controlling inputs rather than outputs, in part because that was all they could do. But they were also generally more interested in building trust in government to ensure that money was not stolen than to ensure that it was put to the best possible use. Tight control over cash outlays is a basic essential that must be in place before a country can or should venture down such exciting new roads as the sort of accrual (or modified accrual) accounting systems now being gradually adopted by subnational governments in a few developed countries. “Basics first is best practice!” as a well-known American budgetary expert has said (Schick, 2012). We agree. Budgeting Ideally, good regional and local budgeting will be part of a medium-term expenditure framework (MTEF). This can help ensure the proper financing of investment projects and reduce the scope for short-term political manipulation of budgets – for example, to expand pre-election public employment in an unsustainable fashion.47 An essential first step in this direction is to put sound budgetary and financial procedures in place (Vaillancourt, 2006).

M4439-BAHL_9781786435293_t.indd 111

01/03/2018 11:08

112

Fiscal decentralization and local finance in developing countries

An MTEF provides an essential baseline for measuring and monitoring the budgetary impact of policy changes (World Bank, 2004, 2013). Another requirement for fiscal discipline at the subnational level is that the amount of revenue expected from intergovernmental transfers is predetermined and is not subject to political renegotiation during the budgetary year. As we emphasize in later chapters, local and regional governments must not be able to depend on central government bail-outs of such imprudent financial decisions as unsustainable borrowing or expenditure increases. Efficient and effective subnational spending is too often made difficult by a variety of rules imposed in part to curb expenditure abuses by regional and local governments and in part, perhaps, to ensure that those governments will spend in ways that are consistent with the objectives of higher-level governments.48 In some countries, for example, many transfers to localities are supposed to be spent on infrastructure investment, in part to prevent them from simply expanding local payrolls. Such rules are never fully effective because money is fungible, and there is usually some substitution of transfers for own-source revenues.49 Even to the extent such earmarking succeeds in its stated objective of stimulating local investment, there is little to be said in its favor. It distorts local preferences, exacerbates perverse incentives already found in the local finance system, imposes new operating and maintenance expenditures, and often connects revenue sources with expenditures in illogical ways (Bird and Jun, 2007). Like the related process of mandating subnational governments to spend in accordance with central preferences rather than their own, such earmarking reduces their ability to manage expenditures even when they have the will and capacity to do so. Like central budgets, regional and local budgets should be comprehensive, accurate, periodic, authoritative, timely and transparent. The budget law should be uniform and clear, and it should be enforced. Moreover, expenditures should be subject to external audit to ensure that the law is followed. The central government has several important roles to play in all of this. For example, it needs to ensure that the rules are in place and enforced. While subnational budgets should not require prior central approval, the central government should establish a good framework budget law with which local and regional governments must comply, as well as requiring their expenditures to be subject to a satisfactory audit. Budgets, final expenditures and audit results should be transmitted promptly to the central authorities and made public in some acceptable fashion. Good Financial Management The central government needs to ensure that adequate institutional arrangements are in place to assist local governments in maintaining allocative

M4439-BAHL_9781786435293_t.indd 112

01/03/2018 11:08



Expenditure assignment and management ­113

and operational efficiency. Even in an ideal world in which local officials are well trained, completely honest and strongly motivated to do the right things, managers at all levels need adequate and accurate information on the effectiveness and social outcomes of the programs for which they are responsible. This might be provided through information on the projected revenues produced by properly designed user charges (as we discuss later), or perhaps through participatory interaction with clients at both the budgetary and implementation stages (Vergara, 1999). Moreover, local governments should have strong incentives to respond to these signals, for example, by facing a predetermined spending limit which can be altered only if they can ‘sell’ more services that their client groups are willing to pay for. Finally, local and regional governments need to know that all their financial actions are subject to external audit.50 The sustainability of public management improvements can be further enhanced if there are institutional mechanisms that bind public officials to various rules and constraints. These transparency and accountability mechanisms can impose implicit costs (sanctions) on politicians and bureaucrats for violating rules, and this can make their commitment to the rules credible (Campos and Pradhan, 1996). A strong budgeting and financial system along these lines satisfies two essential requirements of good government. First, it establishes the basis for financial control. Second, it provides reasonably accurate, uniform and timely financial information. However, even the best set of financial procedures does little to ensure that scarce public resources, even if properly spent and accounted for according to law, are spent in the best possible way. Nor can even the best-enforced set of budgetary procedures ensure that aggregate fiscal discipline is adequately maintained by all governments at all levels. In the absence of such systems, however, favorable outcomes in any of these respects is even less likely. The best way to achieve operational efficiency on the expenditure side of the budget may be to allow local governments – or even line managers such as school heads (see Box 3.1) – significant discretion to reallocate funds among inputs or perhaps even across budgetary periods. This could lead to a shift in emphasis from input controls – hiring so many persons at such and such a wage or renting so many square meters of space – to such output controls as providing health care of a determined quality to some number of persons within a specified period. This approach carries with it some risks, however; and it is by no means certain that many, or perhaps any, local governments in most developing countries are or soon will be able to take such risks.51 However, there is already considerable experience in some countries with such ways of possible client monitoring of spending as client surveys, participatory budgeting, performance budgeting and user financing (Burki and Perry, 1999; Khagram et al., 2013). A positive

M4439-BAHL_9781786435293_t.indd 113

01/03/2018 11:08

114

Fiscal decentralization and local finance in developing countries

feature of a sounder and more transparent expenditure management system is that it makes it possible to experiment further with such matters and to improve the services that local residents get for the taxes they pay. Developing countries should begin to go down such paths, albeit with caution and after getting basic financial controls in place. The risks of going wrong when doing so may sometimes be substantially reduced by piloting such ideas first in a few better-run local governments rather than making big bets on being able to do it right the first time on a national basis.52 Good PFM is not the answer to ensuring that local and regional (as well as national) governments spend as well as possible. As a rule, standard prescriptions for better expenditure management like those touched on briefly in this section need to be implemented gradually and incrementally, and usually adapted to some extent to fit local conditions and to cope with the changing environment. Nonetheless, any country considering decentralization, including countries that have begun to improve PFM at the national level, should pay more attention to improving it at the subnational level than most seem to have done so far.

CONCLUSIONS The devolution of expenditure responsibilities to local governments is an important component of fiscal decentralization. Rarely if ever do countries take a step back and rethink the efficacy of their expenditure assignments. Since everything is usually assigned somewhere, what is really being considered usually are possible reassignments, each of which will create possible losers as well as gainers. Even if there is a clean slate, the best solution is seldom obvious: the economic (efficiency) issues are seldom measurable; the distributional (equity) issues are value-laden; the implementation (administrative) issues are usually not clear; and, of course, the political issues are both complex and difficult.53 Although optimal expenditure assignments cannot be easily defined, good practices can. Some guidelines on expenditure assignment are suggested by theory and experience, and countries seem often to have followed at least some of them: ●●

●●

Assign (or reassign) expenditures before determining how they are to be financed. Many public services are best financed by those who benefit from them. Recognize that it is unlikely that even the best-intentioned expenditure reassignment initiative will get it right the first time. Good strategy requires periodic reviews, with such reviews including

M4439-BAHL_9781786435293_t.indd 114

01/03/2018 11:08



Expenditure assignment and management ­115

●●

●●

●●

assessment of the clarity of assignments, the buy-in from the various governmental units involved, and the impacts of all limits on subnational government fiscal autonomy as well as an appraisal of the effectiveness of selected expenditure programs. A good review should also consider whether some functions might be better privatized wholly or in part (an issue we discuss further in the next chapter with respect to infrastructure). A performance evaluation of the public expenditure system like those in World Bank public expenditure reviews can also be valuable (Pradhan, 1996). Do not create unfunded mandates and, where possible, reduce or eliminate funded mandates on subnational government spending in favor of reassignment of functional responsibilities or the devolution of new local revenues. Make as few functions as possible concurrent: assign them to one level of government or another, and unbundle functions as needed to do so. When complete separation is not feasible, pay close attention to structuring sound inter-local contracting and agreements as well as establishing clear procedures for resolving intergovernmental conflicts. Poverty alleviation is often an important task of government. Subnational governments may have two key roles in delivering the goods in this respect: they can be encouraged to target poor neighborhoods in improving access to local public services; and their informational advantages can be used to improve the implementation of centrally funded programs. But they should generally not be expected either to implement direct transfer programs or to provide subsidized services from their own resources.

Finally, the expenditures assigned to subnational governments need to be managed properly. Good public finance management – the dreary but necessary task of preparing, executing and monitoring budgets and expenditures – is every bit as important for regional and local governments as it is for national governments. Even in low-income countries in which spending by central governments often falls short of satisfying many of the conditions just mentioned, subnational governments are likely to be subject to more severe scrutiny. One reason is simply because what such governments do or do not do, and how well they do it, tends to be more visible and understandable to citizens. A less laudable reason is because politicians and officials at the national (and perhaps regional) level, especially those who lose power when decisions are decentralized, are often happy to single out and criticize subnational missteps. Subnational governments are generally hard-pressed to increase

M4439-BAHL_9781786435293_t.indd 115

01/03/2018 11:08

116

Fiscal decentralization and local finance in developing countries

r­ evenues, so it is even more important for them to spend what they have as well as possible than it is for the central government. Taxing (or being given) more is often a less effective way to improve public services than spending better. But spending better usually receives less attention because it is more difficult for people to tell if governments are doing a good job on the expenditure side of the budget. And, of course, it is always easier for governments at all levels to blame their poor performance solely on the inadequacy of revenues. As Boxes 3.1 and 3.3 illustrate, however, simply pouring more money into the leaky and ill-aimed funnel that is all too often the subnational expenditure management system is unlikely to improve how expenditures are actually managed: that difficult task needs to be tackled directly.

NOTES   1. ‘Effectiveness’ refers to the quantity and quality of the services provided, ‘efficiency’ to minimizing the cost of providing those services; and ‘accountability’ to the ways and extent to which local governments may be held responsible to the people they are supposed to serve. Each of these terms contains multitudes of shadings, only a few of which are covered here.   2. In federal countries, as Breton and Scott (1978) note, the problem is always whether and how to reassign expenditure functions, since invariably they are already assigned in some way or other by the constitution. History, if not the constitution, creates the same situation in all countries.   3. Institutionalizing a process to deal adequately with future needs is thus an essential part of delivering a feasible and sustainable decentralization package, although we do not explore this important topic in depth in this book.   4. Providing a service does not mean that the service must be either produced or delivered by the responsible government itself, since it may contract with another public agency or a private firm for production and/or delivery. Indeed, it is not even essential (though, as discussed later, it may be desirable) that the responsible government finances the provision of services assigned to it.   5. We assume in most of this discussion that all jurisdictions at a certain level (regional, local) are treated similarly. In fact, as we discuss further in Part IV, this is almost never the case – and indeed should not be the case. However, we leave aside for now the issue of asymmetrical fiscal federalism.   6. Boadway and Shah (2009), whose similar table we have condensed and adapted as Table 3.1, are of course well aware of the limitations of this approach: their discussion of this subject in Chapter 3 of their book is – like the book as a whole – well worth reading.   7. For example, an early study of police ‘productivity’ at the local level in a Canadian province (Scicluna et al., 1982) suggests two conclusions. First, it is surprisingly difficult to figure out just what ‘policing’ is and to measure how effectively it is done. It took several years to collect data (much of which had to be proxied to some extent) and to estimate a reasonable multiple-input, multiple-output productivity function. Although the results were not particularly robust, they clearly demonstrated that the only productivity measure then used to determine the allocation of resources to and within the policing function (using uniformed personnel as the only input) was highly misleading. Second, this analysis (and many others like it) was completely ignored when over the next few decades most policing was reassigned from the local to the regional level. Much has been

M4439-BAHL_9781786435293_t.indd 116

01/03/2018 11:08



 8.

  9.

10. 11. 12.

13.

14. 15. 16. 17.

18.

19.

Expenditure assignment and management ­117 learned about policing, and data are much better than they were 30 years ago. But even today in most countries (even the most developed) who does what, why and how seems still largely to be determined at best by the same (almost) ‘rule-of-thumb’ system as the misleading single-factor efficiency measure mentioned above. To return to the Canadian policing example, there are federal, provincial, regional (multi-municipality) and municipal police forces, as well as a wide array of specific public sector security agencies (from transit police to national intelligence agencies) and an enormous variety of private sector policing agencies which in total employ 50 percent more people than the public sector police (see www.statcan.gc.ca/pub/85002-x/2008010/article/10730-eng.htm). Within the public sector, in some provinces, the federal police also act (under contract) as provincial police, and often as municipal police; in others, provincial police may contract to act as municipal police; in addition, there are many other contract arrangements between police forces at different (and the same) levels with respect to many specialized services. In Mexico, such an arrangement is enshrined in law. Interestingly, at least in Canada’s largest province (Ontario), although formally teachers must negotiate with local school boards and about half of school funding comes from local property taxes, in practice disputes are almost invariably settled at the provincial level by direct negotiations between the teachers’ unions and the provincial government. The province consistently tries to assert that the responsibility is local, but neither teachers nor parents are fooled. The main direct responsibility of local municipalities in Colombia is to look after school maintenance. Two useful embellishments and updates are Oates (1999) and Oates (2008). A well-known example of this in practice is the principle of subsidiarity adopted by the European Union, which “rules out Union intervention when an issue can be dealt with effectively by Member States at central, regional or local level” (see www.europarl. europa.eu/ftu/pdf/en/FTU_1.2.2.pdf, accessed July 25, 2016). Breton et al. (1998) provide an interesting comparison of this principle and the decentralization theorem. The technical reasons discussed in the text below are, for instance, clearly recognized in the formal statement of the EU’s subsidiarity principle quoted in the previous note, which goes on to say that intervention from above is justified when “the objectives of the proposed action cannot be sufficiently achieved by the Member States . . . [or] the action can . . ., by reason of its scale or effects, be implemented more successfully by the Union.” There may also be ‘fiscal externalities’ arising from how expenditures are financed (Dahlby, 1996), but we leave these for discussion in later chapters. See, for example, the discussion of approaches to improving sewerage systems in developing countries in Van Dijk (2012) and the discussion of alternative ways to provide water in OECD (2009). The special problems of metropolitan governance and finance are discussed further in Chapter 8. Slack and Bird (2013), for instance, identify the equalization of service provision as perhaps the major gain from the creation of a metropolitan government for the Toronto region. Bahl and Campbell (1976) argue that equalization was the major obstacle to the creation of metropolitan government in the US, where there are no metropolitan area-wide governments, and that economies of scale were not likely to be captured. Even in Canada, where provincial medical plans already purchase many drugs for hospital use, a strong financial case may be made to extend the scope of ‘pharmacare’ beyond the elderly (who are largely presently covered under such schemes in most provinces) to the population as a whole. Morgan et al. (2013), for example, estimate that extending the coverage of the health plan to cover all drugs would lower not only total spending on health but also public sector spending, owing to the combination of the lower purchase prices that they argue (based in part on evidence from other countries) could be negotiated with larger purchasers and the offsetting reduction in currently fiscally subsidized private insurance and medical costs. Economies of density may also come into play (Bel, 2013). For example, a study of

M4439-BAHL_9781786435293_t.indd 117

01/03/2018 11:08

118

20. 21. 22. 23. 24.

25. 26.

27.

28. 29.

30.

31.

Fiscal decentralization and local finance in developing countries annexation that analyzed 952 US local governments (with at least 10,000 people) that annexed other municipalities between 1992 and 2002 found efficiencies from increasing the size of the local government area – but only if the annexation was accompanied by higher densities (Edwards and Xiao, 2009). The authors found that service delivery and administrative efficiencies were achieved with high-density developments but compromised with low-density developments that were spread out and costly to serve. An earlier study found that service costs were generally higher in low-density areas (Downing, 1973). As Box 3.1 notes, however, there is still much we do not understand about the links between how schools are organized and financed and educational outcomes. However, as the authors stress, the periods for which data were available were not the same for all countries, and the quality measures used are not strong. After implementing the 2001 commission recommendations, for instance, the aggregate state government deficit rose significantly (see Andrew Young School of Policy Studies, 2005). As Box 3.1 indicates, however, even if the pay arrives on time, there is no guarantee that any education will ensue. In a particularly egregious case that one of us once encountered, the head of a territorial government was found to be using the educational grant paid to the region to educate his own children in Paris rather than to help the poor rural population for whom the money was intended. Similar illegitimate diversions of funds intended for local education have been observed on another continent in Uganda (Reinikka and Svensson, 2004). Of course, such abuses are seldom limited to local politicians and officials. How much spending is increased depends on many factors, as we discuss further in Chapter 7. It is difficult everywhere for those directly affected by the decisions taken by local governments to understand the fiscal conditions shaping those decisions. Even in the well-established and data-rich local government system of the US, for example, a prominent research organization has recently found it necessary to undertake a major effort to provide a new comparative, visual, web-based tool to make it more feasible for nonexperts to identify and compare what is going on in their city compared to others, and to be able to understand how fiscally healthy (or unhealthy) it is: see www.lincolninst. edu/news-events/news-listing/articletype/articleview/articleid/546264/lincoln-instituteof-land-policy-launches-campaign-to-promote-municipal-fiscal-health). The situation is no better elsewhere in the developed world (Bird and Slack, 2015). A citizen report card is based on information collected via a random-sample survey questionnaire concerning citizen perceptions about public services. Media reporting of the results is critical. A community scorecard uses information collected from focus groups to gain perceptions about the quality of public services provided. Citizen report cards are more commonly used in urban areas, and community scorecards in rural areas. See also the extensive experience and literature on participatory budgeting discussed in Khagram et al. (2013) and the discussion of prescriptions for better success in World Bank (2004, pp. 185–91). Problems about who can do what with respect to public finances are the grist of constitutional courts in all developed federations also, as even a glance at the lengthy discussions in Bizioli and Sachetti (2011) of the problems on the tax side in countries from India to Argentina and Germany to Canada shows. There are many variants of the legal status for expenditure assignments. In some countries, the responsibilities of third-tier local governments are explicitly stated on a concurrent list (India), while in other countries this matter is left to the provinces (China). In Nigeria, there are also residual functions that are not mentioned in the federal or concurrent list, and which may be taken up by states (Ekpo, 2007). In Cambodia, commune and district governments have no formal expenditure responsibility but may take up functions on a permissive basis. Similar issues arise with respect to public–private arrangements concerning infrastructure, as discussed further in the next chapter.

M4439-BAHL_9781786435293_t.indd 118

01/03/2018 11:08



Expenditure assignment and management ­119

32. Wetzel (2013) discusses some of the successes in resolving such problems over the last few decades within the São Paulo Metropolitan Region of Brazil, but notes that the task of coordinating the interests and activities of 39 municipal governments within the region and that of the state of São Paulo is still far from finished. 33. In the case of Russia, its original decentralization of functions in 1992–93 was more a matter of simply shifting the central deficit downward than any intention to reassign functions as part of a strategy of decentralization (de Silva et al., 2009). 34. For an early detailed comparative analysis of these countries, see Bird (1986); a more recent capsule update of the Canada–US comparison is Bird and Zolt (2015). Interestingly, as Bird and Tassonyi (2003) stress, the much more controlled and centralized way provincial–municipal financial arrangements operate in Canada does not seem to produce very different results than the highly decentralized and uncontrolled federal–provincial financial structure in that country. 35. Because rectifying constitutional mistakes is usually more difficult than rectifying mistaken laws, one suggestion for countries contemplating decentralization is not to be too precise in specifying how to structure the fiscal system in the constitution. To take an example from the revenue side, when India’s original constitution said that taxes on services were allocated to the states and Papua New Guinea’s that its provinces could impose retail sales taxes, no one was thinking how difficult such provisions would later make it for the country to adopt a value-added tax. 36. As Gramlich (1976) and Bahl and Duncombe (1991) show, another important factor underlying the New York crisis was the ability of city employees to obtain substantial wage increases despite declines in private sector incomes and employment in the city. Some years later this pattern was again emphasized as leading to a fiscal crisis in another large American city – Philadelphia (Inman, 1996). 37. Sometimes such responsibility is thrust upon them without being accompanied by the necessary funding – something that has happened in places as different as Ukraine (Martinez-Vazquez and Thirsk, 2011) and Ontario, Canada (Slack, 2007). 38. Although this point is not discussed in depth here, it is worth noting that evidence for Canada (Zhong 2010) as well as for Spain and Italy (Costa-Font and Turati 2016) suggests that regionalizing health care was actually equalizing. The latter study finds that this result occurred not simply because of transfer payments specifically intended to produce this result, but also because – exactly as such students of decentralization as Oates (1988) have suggested – one result of regionalization appears to have been more policy innovation and more diffusion of new ideas than under the previous, more centralized system. See also Chapter 7 and Stossberg et al. (2016) for further discussion. 39. One of us once had a discussion with a city mayor during which he insisted his government not only knew who needed help but was also delivering it to the poor without exception. When his attention was directed to a large group of very poor people (from a minority ethnic group) clustered outside the main entrance to the city hall, his response was a dismissive, “Oh, those people: they are not worth helping.” 40. For discussion of this program, see Ministry of Urban Development (2011), Ahluwalia et al. (2013) and Mohanty (2014). 41. A major reform of transfer programs replaced the JNNURM, with some changes in design, with several smaller programs (see Institute on Municipal Finance and Governance, 2015). 42. For an early review of such studies, see Bird and Miller (1989). Recent studies show no great changes in either the limited distributional impact or the distorting allocative effects of such subsidies: see, for example, the survey of water pricing in Le Blanc (2007) and the review of urban transit subsidies in Serebrisky et al. (2009). 43. There are many good introductions to the subject: see, for example, the International City/County Management Association (ICMA) for a US perspective (Bartle et al., 2012) and United Cities and Local Governments (UCLG) for a more international perspective (Farvacque-Vitkovic and Kopanyi, 2014). A good link to other sources and recent views is the Public Financial Management blog, http://blog-pfm.imf.org.

M4439-BAHL_9781786435293_t.indd 119

01/03/2018 11:08

120

Fiscal decentralization and local finance in developing countries

44. To mention a simple example, subnational governments should perhaps have a later starting date for their budget year than the central government because they cannot accurately determine their budgets until they know the size and nature of the transfers they are likely to receive from the latter. 45. Harberger (1989, p. 27) once noted that the lessons he had learned from experience with tax reform in developing countries were “not exciting – more like ‘how to be a good public accountant’ than ‘how to be a star in the movies or in the opera or on the football field’.” As Harberger would have predicted, one of the key lessons the present authors have learned from experience with public finances around the world is that developing countries often need good public accountants more than they need tax reformers. 46. There is much to be said for accrual accounting when governments are up to it; but, as Schick (1998) emphasizes, there are also many possible pitfalls when overly ambitious reforms are attempted. Well-run metropolitan governments in some countries (see Chapter 8) may perhaps follow this path with some success, but most local governments are likely quickly to encounter serious problems. 47. Of course, unless there is an adequate MTEF at the central level, it may be difficult to require one locally. On the other hand, given the importance of local governments in many countries, it is increasingly difficult to have an adequate MTEF at the central level in any case without explicitly incorporating aggregate local revenues and expenditures into the budgeting and planning exercise. 48. Brazilian states are constitutionally restricted in the extent to which they can reduce expenditures, even those funded solely from their own revenues (Rodden, 2003). Such limitations are clearly not conducive to good expenditure management. 49. As Mintz and Smart (2006) note, budgetary constraints imposed from above are more likely to be effective in poor than in rich localities, since the latter’s larger economic base makes it easier to escape such constraints. 50. We discuss in a later chapter the extent to which it may be appropriate to condition receipt of central transfers to local and regional governments on their compliance with the expenditure management system established by the central government. 51. For example, there must not only be full accountability by clearly identifiable decisionmakers with respect to all expenditure decisions to reduce the possibility of fraud, but also an information system sufficient to ensure that any bad behavior is quickly noticed, and a strong enforcement system to ensure that it is appropriately punished. Few countries satisfy such conditions. 52. This is how China has proceeded in recent years with respect to many of the changes it has introduced in its financial and tax systems. Although it is often difficult in democratic systems to treat some local governments differently than others, more asymmetric fiscal federalism may nonetheless be necessary not just to see what innovations may be worth generalizing, but also to produce relatively equal results in the very unequal conditions often prevailing in different localities and regions. See Bird and Ebel (2007) for a fuller discussion of asymmetric fiscal federalism. 53.  Once, in a transitional country that had decided to decentralize some expenditure functions, one of us was asked by the minister in charge for advice about which of three alternatives being considered seemed best: to decentralize to three regions (roughly, ancient kingdoms), to 20 or so districts (roughly the former communist structure) or to 500 or so municipalities (ranging from the very large capital to some tiny rural settlements). The basic economic theory of decentralization turned out to be of little help in answering even such a ‘clean slate’ question as this: try this exercise in any country with which you are familiar.

M4439-BAHL_9781786435293_t.indd 120

01/03/2018 11:08

4. Decentralizing and financing infrastructure The infrastructure need is enormous, but so is the possibility of building the wrong infrastructure. . . . The public sector is often ill-suited to manage such projects, because of both a lack of technical expertise and because of the short time horizons that often influence political leaders. (Glaeser and Joshi-Ghani, 2014, p. 18)

In developed countries infrastructure investment is commonly decentralized, with even important national projects such as ports and airports being locally managed and largely locally funded (Bel and Fageda, 2009). In the European Union (EU), for example, the share of subnational governments in public investment in economic infrastructure is 60–70 percent in the older member states and 40 percent in the newer member states in southern and eastern Europe, and even higher in social investments such as schools and hospitals (Kappeler et al., 2012). In low- and middle-income countries, one would expect less decentralization because much of the basic infrastructure is not yet in place. There is great variation in the extent to which responsibility for infrastructure services is assigned to regional and local governments in developing countries, but on average the share is lower (Alm, 2010). Vinuela (2016) shows that the average subnational share of total capital formation for the 2001–2010 period was only 38 percent for non-OECD countries compared to 64 percent for OECD countries. However, expenditure on capital formation accounted for only 8 percent of subnational revenues in the OECD countries compared to a rather astounding 24 percent in developing countries (Vinuela, 2016). Are subnational governments in developing countries more reliant on intergovernmental transfers which direct their budgets toward capital spending; are they more burdened by spending mandates that result in more capital spending; or do they have a higher marginal propensity to spend their revenues on infrastructure because they need infrastructure more? How one answers such questions is key to understanding whether increasing local autonomy and revenue-raising powers will lead to a more or less efficient package of capital spending. But getting an answer is difficult because subnational governments are so very different in their needs, capacities and preferences, as well as in the decisions they make about 121

M4439-BAHL_9781786435293_t.indd 121

01/03/2018 11:08

122

Fiscal decentralization and local finance in developing countries

investment spending. In some countries, local and regional governments contribute significantly to national investment levels. In fact, as Table 4.1 shows for Peru, the amount spent on infrastructure by subnational governments may even be greater than the amount they raise from own sources – an observation that suggests the role played by transfers in financing their investment, a point to which we return later. The key role played by subnational investment in countries like those shown in Table 4.1 has sometimes been questioned. In Brazil, for example, Afonso et al. (2005, p. 9) argued that “these regional governments do not have the competence to concede, regulate or carry out functions in the majority of actions and services that are classified as infrastructure related (with the exception of sanitation).” Others have questioned the implications for stabilization policy of decentralized investment in a time of increasing budgetary pressure (Fernández Llera, 2012). It is not generally clear whether expenditure responsibility for infrastructure has been over-assigned or under-assigned to subnational governments, or whether the right expenditures have been devolved. The answers depend  on a  range of issues: how local and regional governments are structured and governed; the nature of their spending and revenue mobilization powers; borrowing restrictions and practices; the degree of autonomy in management; the design and implementation of intergovernmental transfer systems; and the accountability of governments at all levels. While these are all questions that we address elsewhere in this book, this chapter focuses on a number of specific theoretical and practical issues related to the extent to which responsibility for infrastructure is devolved to the subnational sector, particularly those relating to structural issues rather than day-to-day management issues.1 Management issues are important, but even the best management can do little if the basic institutional structure of local governance and finance is badly flawed. In this chapter, we define ‘infrastructure,’ set out briefly the normative principles that could guide the assignment of responsibility for delivering infrastructure services, and note how the practice observed in most countries deviates from what theory appears to suggest. We then turn to the question of the extent to which theory and practice might be better reconciled. In the last sections of the chapter, we provide an overview of the practice and possibilities for financing increased infrastructure spending at the subnational government level.

M4439-BAHL_9781786435293_t.indd 122

01/03/2018 11:08

123

M4439-BAHL_9781786435293_t.indd 123

01/03/2018 11:08

1.3 1.9 1.4 2.4

Argentina Colombia Chile Peru

1.1 1.3 0.2 1.2

Subnational government investment (% GDP) 84.6 68.4 14.3 50.0

Subnational as share of total (%) 30.8 15.9 19.6 17.4

General government taxes (% GDP) 5.6 2.9 1.5 0.8

Subnational own revenues (% GDP)

18.2 18.2 7.6 5.0

Subnational as share of total (%)

Sources:  Investment (average acquisition of fixed capital on accrual basis) data for 1995–2006) from de Mello (2012); revenue (for 2008) from Gomez Sabaini and Jimenez (2012).

General government investment (% GDP)

Country

Table 4.1  Public investment and decentralization in Latin America

124

Fiscal decentralization and local finance in developing countries

WHAT IS INFRASTRUCTURE? The essential characteristics of investing in infrastructure are acquiring physical assets with a long life, and prolonging that service life with adequate support for operating cost and maintenance. Public sector infrastructure is the built capital stock that is owned by the government and that generates a flow of benefits over a long but finite life. Highways, bridges, schools and sewerage treatment plants are examples. Public infrastructure investment is often taken to mean the same thing as government investment, and measured as the total amount of capital expenditures by governments. However, it is usually the source of funding, not the formal ownership, that determines how investment is classified in national and most public accounts.2 Since countries differ in how they structure and fund ‘public’ institutions, intercountry comparisons can be tricky even if all concerned seem to follow the same rules. Moreover, the intrinsic value of a capital asset is the capitalized net flow of benefits to the population as a whole, even though this is seldom directly measurable. The better the operation and maintenance of an asset, the higher this benefit flow presumably is. But this may also be determined differently in different countries. Both the public capital stock in countries and the benefit flow generated may differ significantly from country to country depending on the choices made for capital investment. Even within the ‘pure’ government sector capital formation falls into several distinct categories. One way to think of it, for example, is in terms of the functions served by such investment: (1) redistribution (housing, recreation, social protection); (2) public goods (defense, environment, order and safety); (3) hospitals and schools (health and education); and (4) traditional public works projects. As Alegre et al. (2008, p. 26) note, it is the last of these categories that “has the most direct economic impact by reducing firms’ production and transaction costs.” The other three categories have more indirect effects on productive efficiency, e.g., by improving the quality of life in the community. The most relevant government investment from the perspective of its immediate economic impact is thus what is sometimes called ‘network’ investment (e.g. transportation and communication) rather than ‘point’ investment (e.g. hospitals, administrative buildings), even though the former is also obviously geographically specific and the latter may be part of a network from some perspectives (for instance, primary health centers feeding into hospitals or elementary schools into secondary schools). We focus here mainly (though not exclusively) on this narrower definition of infrastructure investment, i.e., the construction, operation and maintenance of the long-lived physical assets required to deliver such specific public services as land t­ransportation (highways,

M4439-BAHL_9781786435293_t.indd 124

01/03/2018 11:08



Decentralizing and financing infrastructure ­125

roads, streets, bridges and ancillary services such as street lighting, street cleaning, signage, etc.), potable water (supply, distribution), wastewater management (sewerage, disposal), and solid waste collection and disposal (including hazardous waste). Another point to keep in mind when studying data about ‘investment’ (or capital formation) is that, although the figures used often include the acquisition of physical assets with a useful life longer than one year, as well as improvements or rehabilitation that extend the life of the asset (Jacobs, 2009), as a rule they do not include maintenance expenditure needed to make the asset functional during its life or the operating expenditure necessary to actually use the asset. The line between ‘maintenance’ and ‘rehabilitation’ – like that between ‘operation’ and ‘maintenance’ – is hard to draw; but the fact is that not only are both operating and maintenance expenditure (O&M) essential for the delivery of public services, but both are also usually considerably larger over the life of an asset than its initial capital cost (Estache, 2010). Assigning responsibility for maintenance to one level of government and responsibility for new investment to another, as is often done, thus creates a serious moral hazard. This practice is almost guaranteed to yield less than optimal results.3 A broken water pipe, even if acquired only yesterday, is not one that is going to deliver the right amount of water in the right quality to the right places.

DECENTRALIZATION AND INFRASTRUCTURE There is no one best way to divide the responsibility for infrastructure expenditures between central and subnational governments. Yet the issue is of great importance because investible resources are scarce and developing countries cannot afford to waste them. Will a given project generate sufficient external benefits to justify overriding local preferences? Will the central government have sufficient knowledge to know how best to design and implement it in local conditions? Will local or regional governments have access to the resources needed to finance the project, the capacity to implement it, and the resources and inclination to maintain it properly? Will local politicians have too high a discount rate to choose adequate spending on maintenance, and will central politicians be too swayed by self-interest to devolve responsibility for infrastructure when devolution is warranted? These are among the many questions that need to be answered when considering the decentralization of infrastructure (Bird, 1995). Some normative rules can be helpful in deciding whether to entrust particular projects to higher or lower levels of government; but in general both levels need to be involved, although concurrency in the sense of

M4439-BAHL_9781786435293_t.indd 125

01/03/2018 11:08

126

Fiscal decentralization and local finance in developing countries

shared responsibility is perhaps less likely with respect to infrastructure than other activities because many infrastructure projects are expensive, capital-intensive and characterized by spillover effects.4 In some cases, externalities are so great that a project must be national (e.g., a port facility or a railroad) or regional (e.g., a hydroelectric plant or an interprovincial highway in a large country). The case for decentralizing infrastructure to subnational governments rests on two pillars: matching expenditures with local preferences and making use of local knowledge of local conditions. Although the economic base, the existing quality of infrastructure services, population density and local preferences may all differ substantially from place to place, in all instances better local information may nonetheless result in better project selection, better implementation, and better operation and maintenance of physical infrastructure.5 The extent to which local investment captures these benefits depends upon how responsive the local government is to local preferences and the extent to which it has the capacity and resources to respond. As Faguet and Pöschl (2015) note, there are three ways in which decentralized local governments may prove more democratically responsive to local preferences: because better information is available to them; because they are sufficiently accountable to local citizens to induce them to make good use of this information; and because they are responsive to competitive comparisons with the performance of other governments, which Besley and Case (1995) call ‘yardstick competition.’ Information and local accountability have been shown to play important roles in, for example, Colombia (Fiszbein, 1997), Bolivia (Faguet, 2004) and Indonesia (Arze del Granado et al., 2008). Box 4.1 presents a more anecdotal example of its potential importance. Recently, Capuno et al. (2015) find that municipal competition improves local performance in the Philippines, though only when mayors are not subject to term limits (that is, they can be re-elected and are thus more likely to pay attention to local wishes). Similarly, Li and Zhang (2015) find evidence that area-based competitions (ABC) – in which local governments in a particular region directly compete against each other for some award offered by a higher government – have led to significant innovations. In another study, Zhang and Chen (2007) show that competition at the provincial level in China has encouraged a higher level of investment in infrastructure. Local demand for local control of infrastructure spending may be high because the benefits are more readily identified by voters than is the case for many other government outlays. Neighborhood residents are more likely to get excited about better water supply or a new street lighting system than they are about an increase in public employee wages, the preparation of a new master plan or improved tax administration. This

M4439-BAHL_9781786435293_t.indd 126

01/03/2018 11:08



Decentralizing and financing infrastructure ­127

BOX 4.1 INFORMATIONAL ADVANTAGES OF LOCALIZING INVESTMENT DECISIONS An interesting illustration of the informational advantages of local participation is the Ahmedabad Bus Rapid Transit Service (ABRTS) in India. This project was surprisingly successful in its first years for a number of reasons (National Institute of Urban Affairs, 2011). First, its designers paid close attention to ­experience elsewhere with similar projects, both successful (e.g. Bogotá) and unsuccessful (e.g. Delhi). Second, in both the design and the implementation phases, extensive efforts were made to consult not only with customers (actual and prospective) but also with such stakeholders as those who operated competitive modes of transport. As a result, substantial efforts were made to turn competing into complementary (feeder) modes. Moreover, close attention was paid to the specific urban context in which the system was to operate by establishing routes that would reduce congestion, and by minimizing conflict about religious sites. In addition, an extensive pilot operation was carried out to test the system, which led to significant improvements. For example, the initial trial buses had metal seats which – perhaps unsurprisingly in the Indian context – turned out to be too hot for comfort and were therefore replaced before the system was launched. Involving local people in all stages of the project cycle appears to have contributed substantially to the success of this project, not just by securing their agreement and approval but also by actively involving them in identifying and initiating projects (what most needs to be done and where), in project design (exactly where and in what way), and even to some extent in execution and operation (including, perhaps, using local labor where feasible).

kind of anecdotal reasoning leads some to the conclusion that decentralized infrastructure responsibility should improve allocative (economic) efficiency by attuning provision more closely to local preferences. As an example, a study in Colombia found that giving responsibility for water supply to local governments rather than public enterprises increased the quality of the output (Granados Vergara et al., 2008). Others have suggested that decentralizing the responsibility for infrastructure may result in poverty reduction, for example, by improving the access of the poor to markets, work and improved welfare (Jalan and Ravallion, 2002; Majumder, 2012). Poor people in an isolated mountain village may, for instance, rationally choose to employ increased resources in improving market roads to improve their basic economic position quickly rather than investing in primary education or whatever else planners in the capital city may think is best for them in the long run (Fiszbein, 1997). Of course, one may still raise questions about the relative effectiveness of such programs as compared to, say, conditional cash transfers (McCord, 2012). Is infrastructure spending higher in more decentralized countries? Some

M4439-BAHL_9781786435293_t.indd 127

01/03/2018 11:08

128

Fiscal decentralization and local finance in developing countries

evidence suggests that it may be (Estache and Sinha, 1995; Boadway and Shah, 2009). An econometric study in Europe found that decentralizing revenue authority to subnational governments increased their investment in economic infrastructure, without reducing their ‘social’ investment (Kappeler et al., 2012). On the other hand, Vinuela (2016) suggests that decentralization leads to lower levels of infrastructure spending in developing countries as a result of their generally lower level of accountability and institutional capacity. Potential efficiency gains from decentralization seem sometimes to be captured, but there are many reasons why this may not happen. For example, the local population may not have a vote which gives them at least an indirect way to voice their preferences. They may also not have any creditable threat of exit. In China, for example, although subnational governments have considerable discretion in making infrastructure decisions, local political leaders are appointed and are accountable upward. Infrastructure accomplishments have been remarkable in China, but it is not clear how much of this is driven by local preferences (Dollar and Hofman, 2008). For example, one study of some 500 villages indicated that relatively little attention was paid to local preferences in making decisions about local expenditures (Bird et al., 2011). Even when people can vote, local decision making may be captured by powerful interest groups within the community, such as a particular ethnic group, local politicians or community organizations that may represent only themselves, or it may be captured by central line ministries.6 The process of selecting and designing capital projects seldom seems to be weighted in favor of the preferences of the local population (Peterson and Muzzini, 2005), and of course the procurement process itself may in any case be corrupt and inefficient.7 The scale of government may be too far removed from the scale of the services provided to be efficient. In large regions or cities, the benefits from small, much-desired local capital investments are often all but invisible to those outside the neighborhood. On the other hand, when local governments are too small they may simply be unable to deliver the service level that their people want. Another problem is sometimes what may perhaps be called ‘expert capture.’ Experts, usually in the capital city, often favor the latest ‘bells and whistles’ (state-of-the-art) investment, which may be well beyond what local people want or need. Or they may be intent on making the capital city a showcase for the world and concentrate on wide boulevards, gardens and modern buildings. Local people may prefer a new sports field to something they may need much more, such as investment to prolong the life of a water supply system. But political capture may result in even worse outcomes, such as statues of political leaders or roads to nowhere (or to

M4439-BAHL_9781786435293_t.indd 128

01/03/2018 11:08



Decentralizing and financing infrastructure ­129

their summer palaces). One way to overcome some obstacles to capturing potential efficiency gains from decentralizing infrastructure may be to give sub-municipalities within large urban areas sufficient decision-making power and budgets to provide some small-scale projects (sports fields) at the neighborhood level, and to give a stronger voice to neighborhood populations in the selection and design of larger capital projects (such as fixing the water supply system). Examples of this approach may be found to some extent in some of the barangay governments in the Philippines (Masagca et al., 2009), as well as the extensive use of locally based benefit funding of investment projects in some Colombian cities (Bird, 2012a). Other ways to build in community input are through a bottom-up approach to project identification, as in Cambodia and rural India (see Box 3.4), as well as by monitoring client satisfaction with infrastructure services and making the results public. Such signals of consumer satisfaction are important to matching service delivery results with preferences and willingness to pay for these services. Using surveys to gain some information on citizen perceptions about the quality of services delivered can sometimes be an important input to identifying new infrastructure needs and maintenance needs, and to stimulating the demand for better services (Peterson and Muzzini, 2005; Paul, 2014). Finally, the capacity problems hampering small, poor communities may be overcome in some instances by taking a more asymmetric approach to decentralizing responsibility for infrastructure decisions. If local governments are assessed (preferably in some objective fashion) as not being sufficiently qualified to take on a given task, the next highest level of government would do so. Alternatively, local governments could be empowered to enter into various forms of cooperative agreements to make up for their limited capacity, or some functions might be better supplied by a special district of some kind. We discuss these possibilities in the context of metropolitan areas more fully in Chapter 8. Scale Economies Economies of scale can be critical in determining what level of government should be responsible for infrastructure investment. When it comes to services such as railways and national trunk roads, small governments cannot really provide such services in a cost-effective way. For many capital projects, however, the right level when it comes to taking advantage of economies of scale (and scope) remains an open question (Fox and Gurley, 2006). A related question is whether one level of government can deliver the same quality of infrastructure services as another. A study in Ontario, Canada, for example, reports that 20 percent of treated water is lost in the

M4439-BAHL_9781786435293_t.indd 129

01/03/2018 11:08

130

Fiscal decentralization and local finance in developing countries

distribution system before it reaches consumers (Herstein, 2012). Such leakages through technical failures (as well as simple theft) are often much larger in developing countries and may perhaps be reduced if providers are more directly responsible and responsive to consumers, although there seems to be little evidence that bears directly on this point. The scale of the technology used to produce and maintain services also needs to be considered in making appropriate investment decisions. Anyone who has spent much time in the rural areas of developing countries has seen all too many instances in which central governments (or, often, foreign donors) have installed water systems or built schools that cannot be properly maintained without expensive and often unavailable expert help, rather than simpler pumps or schools that can be built with local labor and materials and maintained by local people. Sometimes the right (feasible, effective and economically efficient) solution might be to outsource, either to higher-level governments or to private firms, the design, production, procurement and perhaps even installation and maintenance of particular facilities – provided, of course, that local people have sufficient direct input into the process to ensure that adequate attention is paid to local conditions and preferences (for an urban example, see Box 4.1). The relevant scale and size factors will vary considerably from country to country and from project to project. Many regions and cities in China and India, for example, are bigger than most countries in the world. Some of the variations in ‘who should do what’ when it comes to decentralizing infrastructure are described in Table 4.2. As was shown in Table 3.1, not much ends up as purely locally provided, although local and regional governments can play important roles in delivering a number of other important services, mainly in terms of the ‘last mile’ delivery to households (or businesses). But even in these cases, coordination is important. Even in the most decentralized developed countries, for instance, subnational governments are usually required to follow central regulations and standards (Frank and Martinez-Vazquez, 2016). In many low- and middle-income countries, metropolitan governments have local government status and have boundaries and populations of a size that can deliver economies of scale. The main message of Table 4.2, however, is that “the key to decentralization is flexibility in accommodating a wide range of conditions” (Andres et al., 2016, p. 43). Many factors – economic, technological, environmental and political – enter into how countries decide to carry out these activities; and in the end, as always, the decision (although conditional on many of the factors noted in Table 4.2) remains essentially political. Although the evidence is not conclusive, it seems clear that, just as there are information advantages from decentralization that can improve project

M4439-BAHL_9781786435293_t.indd 130

01/03/2018 11:08



Decentralizing and financing infrastructure ­131

Table 4.2  The economics of decentralizing infrastructure Sector

Level

Determinants

Arrangements

Water: catchment

Central/regional

Externalities

Water: distribution

Local*

Scale, efficiency

Sanitation

Local*

Sewage treatment Solid waste  collection Solid waste  disposal

Local* Local

Efficiency, externalities Externalities Efficiency

Line ministry/ utility Local utility or agency Local entity

Regional

Externalities, scale

Roads: highways

Central/regional

Roads: rural

Local*

Scale, externalities** Efficiency, externalities**

Local entity Local Metro agency, local associations Central Regional, local

Note:  * Central and regional governments can/should regulate. ** Roads are club goods, i.e. they are ‘non-rival’ until congested but users can be excluded. Source:  Adapted from Andres et al. (2016, Table 2.2).

quality as well as possible cost savings from using local labor and materials and avoiding the bureaucratic costs of managing a project from the center, at least some subnational governments in most countries are likely to have sufficient size and skill to capture economies of scale for some functions (Peterson and Muzzini, 2005). This is most likely for the regional level and metropolitan areas as well as for some large cities. As always, the situation varies from country to country as well as within countries. For example, World Bank (2009a) reports that subnational government spending for roads in Colombia is expensive and wasteful.8 However, the underlying problem in this case is not decentralization but rather the unclear nature of the initial assignment of functions and the confusing way in which roads are financed, which results in continuous disputes about which level of government is responsible for what when it comes to roads. When, as in Colombia, all three levels of government are involved in financing an activity but the two subnational levels receive most of their funding from the national government through several different channels, not only is the ownership of (and hence responsibility for) roads confused, but the diverse and volatile

M4439-BAHL_9781786435293_t.indd 131

01/03/2018 11:08

132

Fiscal decentralization and local finance in developing countries

set of funding sources also makes it exceptionally difficult either to invest sensibly or to maintain investments properly once made (Bird, 2012a). Externalities When the benefits (or costs) of an infrastructure project spill over to households that reside outside the boundaries of the jurisdiction, lowertier governments will underspend (or overspend) because they will not account for impacts on non-residents. Some infrastructure may be better provided by either the regional or even the national level because a larger population and service area is needed to internalize the external effects. For national trunk roads, for example, the predominantly national nature of the benefits and the need for cross-country coordination make decentralizing such infrastructure unlikely to be sensible. One way to deal with spillover problems is thus to assign responsibility to a higher-level government, or perhaps to a special district or agency covering a larger service area. This approach makes sense when externality costs are so great that they overwhelm the welfare gains that may arise from the informational advantages of smaller local governments. A second approach is to unbundle expenditure programs into subfunctions, separating those with external costs and benefits from the others: for example, refuse collection might be assigned to the municipal governments but solid waste disposal to a regional government or a special district. Yet another approach may be for central governments to use intergovernmental transfers – usually conditional grants (as discussed later) – to induce subnational governments to correct for underspending on services with large externalities (e.g. vaccinations for communicable diseases). Alternatively, when local governments are unable to handle the externality (or perhaps the management) aspects of an activity, the gap might perhaps be filled by creating special-purpose public enterprises that operate on a local basis. In some Indian metropolitan areas, for example, certain local infrastructure services are delivered by single-purpose metropolitan special districts, largely to address the externality and scale economies issues (Pethe, 2013). However, since these parastatals are owned and controlled by the state government, they are often not very responsive to local preferences (Mohanty et al., 2007; Mohanty, 2014). Another option is contracting for services, either to a private firm or to another local government. While arranging some form of contracting between larger and smaller local governments for service delivery is an attractive alternative in principle, in practice it has turned out to be difficult to sustain such arrangements, even in such well-governed and homogenous developed countries as Denmark and Finland.9

M4439-BAHL_9781786435293_t.indd 132

01/03/2018 11:08



Decentralizing and financing infrastructure ­133

Managing and Maintaining Infrastructure Proper maintenance of infrastructure can markedly enhance the flow of services from capital assets and extend their useful life. But maintenance is often problematic under a decentralized system for several reasons (Fox and Murray, 2016): ●●

●●

●●

●●

The local revenue base available to maintain the capital stock is not adequate. Subnational governments raise only about 2.5 percent of GDP from own-revenue sources in developing countries (see Table 2.1). Ingram et al. (2013) estimate that the required annual maintenance costs for urban infrastructure in developing countries is equivalent to about 2 percent of GDP.10 Local politicians may ignore maintenance and favor the higher public profile offered by new infrastructure. Or they may prefer to satisfy demands for wage increases or meet some other need considered more pressing. Such behavior may be quite sensible from their perspective. Although the media are quick to report the results of a major break in a sewer line or some other disaster, there is seldom the same coverage when a repair is completed as when opening a new facility. It is usually more politically attractive – and often simpler – to obtain new capital financing from higher levels of government than to secure more local tax revenue or impose user charges to cover O&M costs. Sometimes, as in the Colombian road case mentioned above, there is confusion over which level of government is responsible for maintenance. When everyone involved can plausibly blame someone else for problems, accountability is weak. Higher-level governments often impose paternalistic rules about maintenance schedules and materials in ways that sometimes amount to unfunded mandates, almost always complicate local management of matters, and generally reduce willingness to pay at the local level. Despite such problems, however, centrally imposed maintenance requirements, construction standards and inspections are usually desirable.

Fortunately, it is not difficult to think of arrangements that can overcome many of the problems just mentioned. Unfortunately, all solutions seem to require initiatives by the central government, thus underlining the point stressed earlier that unless a central government supports effective decentralization and creates the conditions that make it possible, it is unlikely to happen. If countries want to do decentralization right with

M4439-BAHL_9781786435293_t.indd 133

01/03/2018 11:08

134

Fiscal decentralization and local finance in developing countries

respect to managing infrastructure – or indeed, with respect to infrastructure activities more generally, or almost anything else – they should follow the five rules set out below: rules that are deceptively simple to state but, it seems, devilishly hard for many countries to follow.11 First, limit interference from upper-level governments to concerns about externalities, and perhaps the regular monitoring of maintenance and other activities. Mandates requiring that a certain minimum share of the budget should be spent on investment or economic development, or that only a specified maximum share can be spent on personnel, or even that a specified detailed maintenance schedule should be rigorously followed are not easily enforced and seldom yield useful results. Conditional grants may be a better way to go, as we discuss later; but again, they are unlikely to improve matters much if the result is that a large proportion of local funding (from all sources) ends up being effectively earmarked for particular uses. If the central government has so little trust in what local governments do that it wants them to simply act as its agents – doing what the central government wants, when it wants and how it wants – why pretend that local authorities have any autonomy at all? As we noted in Chapter 1, if decentralization does not empower local governments with some degree of autonomy – including the right to do some things wrong from the viewpoint of the center – it means little in terms of improving either efficiency in the use of scarce national resources or producing desired outputs more effectively. Second, in most countries regional and larger urban local governments need more and freer access to a larger revenue base than they now have – for example, to provide the needed funding base for infrastructure operations and maintenance. We develop this theme further in Part III of this book. Third, close the ‘back door’ to infrastructure finance that is all too often provided by higher-level governments. In Mexico, for example, as much as 10 percent of intergovernmental transfers to subnational governments are given during the fiscal year in the form of ad hoc (non-budget) grants to subnational governments (Bahl and Sethi, 2012). This discourages local revenue mobilization, may distort capital expenditure choices and reduces the transparency of the system of intergovernmental transfers. Much the same arguably occurs in India when states receive ‘implicit’ transfers in the form of ‘loans’ from public financial agencies that are seldom repaid (Rao and Singh, 2006). In Hungary, the central government actually assumed all the debt obligations of its thousands of small towns in 2012, extending the offer to towns of all sizes two years later (Werner, 2016). Many towns had financed expansions with bonds denominated in foreign currencies, which meant that the devaluation of 2008 vastly increased debt service costs; so this drastic step may have been the only way to avoid the complete collapse

M4439-BAHL_9781786435293_t.indd 134

01/03/2018 11:08



Decentralizing and financing infrastructure ­135

of local public finances and services. However, saving people from the consequences of their own bad decisions (to fund infrastructure investments by borrowing euros or Swiss francs) establishes a bad precedent and provides little incentive to behave more sensibly in the future. We discuss further below the importance of establishing as firm a central government stand as possible against the inevitable arguments for bailouts. Fourth, central governments must establish good ‘framework’ laws (e.g. on tendering and on intergovernmental agreements) and upgrade both local capacity to deal with such issues and central capacity to monitor and evaluate outcomes – not simply to detect corruption but, more importantly, to learn from experience and, also importantly, to adjust the system as necessary to improve outcomes. The issue here is similar to one that we raise later in this chapter with respect to public–private partnerships (PPPs) for infrastructure projects. For such activities to be productive, the national government needs to pay close attention not only to improving its own regulatory framework and capacity but also to supporting and monitoring how well things are being executed at the operating level, whether that means by local governments or by private contractors. One good place to start in many countries might be to assemble an inventory of the state of the infrastructure that is already in place as a base point for future developments. At the other end of the process, it is especially important to ensure that local governments can and do appropriately monitor the state of their own infrastructure, as well as how well it is being maintained. It is all too easy for the value of even the most substantial initial effort in investing in critical infrastructure such as water systems to be quickly destroyed by inadequate O&M which carries on to its dismal end because no one is really watching what is going on.12 Finally, as discussed earlier, it is often both possible and desirable to ‘unbundle’ infrastructure services so that those sub-functions not characterized by clear advantages of centralization might be assigned to subnational governments. Costs may be reduced by disaggregating infrastructure expenditures into components and making assignments on a basis of comparative advantage. For example, while perhaps technical specifications required to ensure water quality might be a matter of national concern, the construction of major water supply and sewage lines (like interurban highways) may best be handled at the regional level, and local distribution lines might be the responsibility of local governments. Unbundling can thus do much good. Since too much unbundling may sometimes make things worse rather than better, however, it is worth discussing this point a bit more. The most common way to think about unbundling is as disaggregating a function or project into components that are characterized by significant externalities or significant economies of scale and those that are not. This

M4439-BAHL_9781786435293_t.indd 135

01/03/2018 11:08

136

Fiscal decentralization and local finance in developing countries

may provide a guide to how delivery of a function may best be assigned to different levels of government (see Table 4.2). Of course, dividing up responsibility in this way gives rise to the problem of how to coordinate service delivery when more than one level of government is involved – a problem very clear with respect to water in the Mumbai area, for example (Pethe, 2013). An alternative approach to unbundling expenditure responsibilities is to do so by activity, with responsibility assigned on grounds of comparative advantage. One can think of three levels as in Table 3.1 (Boadway and Shah, 2009): (1) responsibility for policy, standards and oversight; (2) responsibility for provision (financing) and administration; and (3) responsibility for production and administration. With respect to primary and secondary education, for example, the national Ministry of Education may be responsible for determining national standards to be met by schools, teachers and students, while regional (state) governments have the primary responsibility for financing education, and local (municipal) governments (or, getting closer to the action, even school authorities) are in charge of delivering education services. In Chile, for instance, although municipal governments deliver both basic health services and primary and secondary education, these services are largely financed by the national government, which is also responsible for ensuring that minimum national standards are met in all municipalities. World Bank (2007) illustrates how one may consider unbundling service responsibilities for rural local governments in India by cross-classifying activities against sectors and services. Another way to think about this question with respect to infrastructure is to focus on how best to assign each phase of the project cycle: (a) planning and appraisal; (b) selection, budgeting and financing; (c) procurement and supervision of implementation; (d) operation and maintenance; and (e) ex post evaluation. Frank and Martinez-Vazquez (2016a) stress that it is not only important to get assignment in this sense right, but that it is equally critical to strengthen coordination both vertically (across tiers of government) and horizontally (across jurisdictions and agents, public and private, involved in the process) all along the cycle. To a considerable extent, no matter how an infrastructure project (or, more generally, any expenditure function) is unbundled, coordination is the key to success. Frank and Martinez-Vazquez (2016a) distinguish three ways of strengthening coordination. The first and weakest is what they call ‘signaling.’ The example they give is earmarking, which they note is difficult to track and enforce (see also Bird and Jun, 2007). The second is to attach ‘conditionalities’ (like those often found in the conditional grants we discuss later) such as targets with respect to timing, coverage, quality and operational effectiveness. And the third, and strongest, is through explicit contracting

M4439-BAHL_9781786435293_t.indd 136

01/03/2018 11:08



Decentralizing and financing infrastructure ­137

with both the objectives and each party’s commitments being clear. Of course, how effectively any of these methods work depends not simply on how well they are designed but also on how well they are implemented, by whom and with what consequences for non-compliance.13 We discuss these contracting issues a bit more later in this chapter with respect to PPPs, and in later chapters with respect to conditional grants and earmarking. Summing up, unbundling is a rather paradoxical strategy. Dividing responsibilities by sub-function allows governments to address spillovers; dividing them by activities may allow capture of comparative advantages; and dividing them by phases of the project cycle may facilitate effective management. But going down all three paths at once could create a maze that may prove difficult for any country to navigate and may give rise to some insuperable coordination problems. For example, rural local government welfare programs in Karnataka state in India are overlapped by eight central government programs and 47 state government programs (Rao et al., 2004). Thus something is likely to be lost in efficiency terms through reduced attention to local preferences, reduced competition in provision, and probably reduced monitoring and oversight. Moreover, a simpler and clearer system is not always better since each country, each locality, and each service and project may call for a different approach. Considerable experimentation may be needed before hitting on the right combination when it comes to building infrastructure or providing particular services in any context. Although what fits best in one instance may not easily carry over to the next, much could be learned by letting people try out different ways of doing things. Those looking for simple answers may not always get the right answers, just as those looking for the right answers are unlikely to find simple ones. Capacity Many subnational governments in developing countries have little capacity to design, build and operate infrastructure, and are unlikely to be able to outsource the needed skills efficiently and effectively. On the other hand, the heterogeneity of the local government sector in most countries means that one should be careful in making generalizations. Often, some larger (and even some smaller) regional and local governments may do as good or better a job in delivering services as the central government does or can do. Over the last few decades there has been a marked improvement in the quality of local management in developing countries, particularly in some of the larger cities (World Bank, 2009b). Still, it is undoubtedly the case that many local governments in developing countries remain weak vessels when it comes to delivering infrastructure projects – or, often, public services in general – efficiently and effectively.

M4439-BAHL_9781786435293_t.indd 137

01/03/2018 11:08

138

Fiscal decentralization and local finance in developing countries

While there are many reasons for the existence and persistence of this problem, to a considerable extent it simply reinforces a message we deliver many times in this book: countries usually have the local governments they want and deserve. Subnational politicians and officials, like those at the central government level, respond to the incentives with which they are faced. If those incentives discourage initiative and reward inefficiency and corruption it is no surprise that local governments are often corrupt and/or inefficient. Given appropriate incentives (in terms of heightened expectations of improved services from their constituents, accountability to those constituents, and access to resources for which they are politically responsible to those who provide them) even very small local governments have sometimes demonstrated significant improvements in administrative capacity within a relatively short time. For example, Fiszbein (1997) noted in an early evaluation of decentralization in Colombia that when they received more resources after a constitutional revision, some – not all – municipalities took steps to improve their capacities.14 Some, for example, improved the skills of local officials through competitive hiring; some shared the services of professional staff with neighboring municipalities; and some undertook more training of municipal employees. Some municipalities also improved their capacity to carry out effective infrastructure projects. One, for example, privatized road maintenance; another put private developers in charge of the construction of urban roads; others introduced computers to monitor water and sanitation services, shared equipment with others and directly attempted to improve their capability to better manage municipal projects. While by no means all local governments did such things, one result of the increased local engagement and ‘ownership’ was that surveys found most respondents trusted the local government more than the national government to deliver the goods and services they wanted (Fiszbein, 1997). Many of the smaller municipalities concentrated on roads, education and water works. These priorities may not always have been what the central officials previously in charge of these areas thought was most important; but since these were the needs local people perceived, these were the needs that at least some of the newly empowered and responsive local governments attempted to meet. Although over the succeeding decade of political turmoil and civil strife much of this initial positive response appears to have faded away (Chaparro et al., 2005), even a decade later surveys continued to show that most Colombians were happier to pay local than national taxes, presumably at least in part because they felt they were, so to speak, getting more for their money (Acosta and Bird, 2005).

M4439-BAHL_9781786435293_t.indd 138

01/03/2018 11:08



Decentralizing and financing infrastructure ­139

Assigning Infrastructure Despite the many differences among developing countries, nearly all are plagued by expenditure assignment problems with respect to infrastructure. Sometimes the problem is that functions are assigned to the wrong level and present arrangements cannot be sustained. Sometimes the expenditure responsibilities are unclear, with the result that service delivery systems are harmed. And, sometimes there are just too many cooks in the kitchen. The central government assigns responsibility to the local government but retains some control over delivery with mandates; local governments use back-door approaches to local legislators and line ministries to get around funding constraints; and inter-local cooperative agreements, PPP arrangements and special-purpose districts add more layers of complication. The problems vary substantially from country to country, but existing arrangements are usually so well entrenched that there often seems to be no apparent solution. There is no silver bullet that solves these problems. However, there are some ways to make them more tractable, although most require central governments to approach matters differently than they usually do. Theory tells us that an appropriately structured fiscal decentralization can lead to a higher quality of infrastructure services than will a fully centralized system if conditions are right.15 If the assignment of expenditure responsibility is correct and if the capacity to deliver services is in place, consumer-voters will be empowered to express their preferences for infrastructure services, services will be delivered with appropriate technologies and at lower costs, there will be more willingness to pay, and public facilities will be better maintained. Almost regardless of capacity, however, as countries develop, more decentralization of infrastructure spending seems likely. Many of the problems that arise with decentralized investment can only be addressed through measures like those discussed earlier in this book as ways to improve decentralization in general – such as more transparency, increased local autonomy on the revenue and expenditure sides of the budget, and improved accountability of elected subnational government officials. However even the best intergovernmental fiscal institutions supporting decentralization cannot resolve all the problems of coordinating infrastructure spending and management. Earlier, we suggested that the best approach to resolve such problems may often be what is sometimes rather pejoratively labeled ‘muddling through’; although, as Hirschman and Lindblom (1962) noted long ago, perhaps the process may be more positively thought of as purposeful incremental gradualism – that is, doing the best one can to improve particular situations while not losing sight of the final objective of improving the lives of people.

M4439-BAHL_9781786435293_t.indd 139

01/03/2018 11:08

140

Fiscal decentralization and local finance in developing countries

Most low-income countries fall short both in terms of the general context within which decentralized infrastructure decisions are made and the willingness and capacity of their governments to do better. Nonetheless, a few guidelines that would perhaps improve outcomes provided they are followed in a way appropriate to the particular country context may be offered: ●●

●●

●●

A comprehensive approach needs to be taken to considering and fitting together all the components necessary to make the infrastructure decentralization work.16 The basic case for such decentralizing infrastructure is that local control is more likely to meet local needs and that, ideally, local users are both willing and able to pay the costs of the benefits they receive. It follows that an essential requirement for assigning infrastructure decisions to local governments is that local governance institutions ensure in an accountable way that those who benefit and pay are also those who decide what is done. All the improvements possible in, say, procurement and financial management systems will do little good if the final decision responsibility is assigned to the wrong level of government. To the extent feasible, public sector functions should be sufficiently unbundled to ensure that it is clear to all who is responsible for what and who is accountable to whom for what. Clarity in assignment must be matched by accountability, in terms of both political democracy and transparency of operation, as well as by authority in terms of both the ability to manage expenditures and to determine (within limits) revenues. Experience suggests that it is especially important that asset ownership be made clear and that the owners are responsible for maintaining and operating costs. Local governments must be adequately accountable in political, administrative and financial terms both to those they are supposed to serve – their residents – and to those above them in the governmental hierarchy who may be responsible for developing policy, regulating how it is carried out, and often for financing the activities of such governments. Such ‘dual accountability’ is never easy to design or to implement, and always requires walking a fine line. On one hand, unbundling schemes such as allowing higher-level governments to set standards and require creditable compliance tests call for upward accountability by local governments. So do conditional grants. On the other hand, too much expenditure mandating by higher-level governments can take away the local autonomy that is the key advantage of fiscal decentralization of infrastructure. As we have stressed throughout, local governments must ultimately be accountable to their constituents, and mandates and conditionalities

M4439-BAHL_9781786435293_t.indd 140

01/03/2018 11:08



Decentralizing and financing infrastructure ­141

●●

●●

●●

imposed from above should be held to the test of accommodating externalities. Some specific reforms in the design and management of the project cycle can help the efficiency of capital spending. As Frank and Martinez-Vazquez (2016a) argue, reforms of separate components of the project cycle can be carried out independently and can be effective. For instance, improvements in the assignment of and oversight responsibility for planning and appraisal might lead to better project selection. Changes in procurement can reduce corruption concerns.17 Operations and maintenance policy can benefit from adhering to a maintenance schedule, considering new institutional arrangements to carry out the work (e.g., contracting), and perhaps in some cases even earmarking appropriate revenue sources for O&M (Fox and Murray, 2016).18 Most local governments in developing countries will need to raise more revenue if they are to take on more responsibility for delivering infrastructure services. In particular, the larger urban local governments need to be given access to stronger tax bases, and encouragement to reform the tax base to which they already have access (namely the property tax). User charges must be ratcheted up significantly so that they approach cost recovery levels. Incentives to do these things must be great enough in magnitude to offset the natural aversion of local politicians to tax increases. The reform agenda might include closing off the option of ad hoc capital transfers to cover infrastructure financing by local governments, perhaps the provision of incentives (e.g. rewards) for increased local effort, and to some extent also conditional grants that are designed with a mixture of clearly specified conditions and local matches. All these approaches are discussed further later in this book. Finally, as we develop further in Chapter 8, there is a strong case for an asymmetric approach to decentralizing infrastructure delivery and finance. Most large urban areas have more capacity to plan, deliver and finance infrastructure services than do smaller and more rural local governments. Large cities should have more responsibility for delivering infrastructure as well as for financing infrastructure with own-source revenues. Rural local governments, on the other hand, will continue to rely heavily on vertical programs of higherlevel governments. Such a regime should improve the productivity of infrastructure spending. However, depending on the weight assigned to equity concerns, it would also likely require close attention to improving capacity in rural areas, and to the design of other policies related to attaining equalization and poverty alleviation objectives.

M4439-BAHL_9781786435293_t.indd 141

01/03/2018 11:08

142

Fiscal decentralization and local finance in developing countries

FINANCING INFRASTRUCTURE A major obstacle in the way of assigning more responsibility for infrastructure to local governments in developing countries is that most regional and local governments lack the financial, managerial and technical ability to do the job. Two solutions seem possible. The first is to kick it all upstairs to a higher-level government. The second is to work away at the long, grinding process of building up local governance and financial structures and managerial capacity. Since it is almost always quicker and easier to give money and to outsource tasks (e.g. via PPPs of various sorts) than to build up the institutional structure needed to enable local people to do the job themselves, the short horizons within which aid agencies and national governments usually operate mean that they often choose the first solution mentioned. In Bangladesh, for example, the national Local Engineering Department essentially centralized local investment. Doing so may have improved the technical efficiency of projects. It was also usually enthusiastically supported by local governments who were glad to be freed of the need to tax their own people. The result, however, was that local autonomy was significantly undermined, with several dire consequences. First, because local people have little or no ownership in the assets bestowed on them from above, they tend to run them down rather than maintain them properly. Second, whether they behave this way or not, they have no incentive to do maintenance right. Third, what the central government bestows may not be what local people really want in any case. And, finally, of course so long as outsiders are willing to pay, the same sad cycle is likely to continue. A similar system of centrally controlled infrastructure investment existed in Indonesia for three decades after 1970. But when it was replaced with a decentralized system in 2001 so that the responsibility for choosing and implementing infrastructure projects began to shift to elected local governments, local infrastructure investment increased and was more in accordance with local preferences (Chowdhury et al., 2009). Financing local capital projects through subnational government budgets can enhance efficiency if done in such a way that local officials are fully accountable to local residents for the quality of services delivered. Elected provincial and municipal governments pass this test, but special district governments that are governed by an appointed board might not. There is also an equity consideration. Rural local governments and some smaller municipalities do not have the capacity to follow the financing rules and are likely to require significant subsidy. The same may be true of most higher-income local governments, at least until broad-based taxes are devolved to subnational governments.

M4439-BAHL_9781786435293_t.indd 142

01/03/2018 11:08



Decentralizing and financing infrastructure ­143

Subnational governments in developing countries have access to three sources for financing capital expenditures: local government tax and nontax revenues, user charges and intergovernmental transfers.19 In theory, the rules for financing infrastructure projects should be the same as for current expenditures (set out in Chapter 5); that is, the costs of local benefit services should be recovered with own source revenues and the spillover benefits should be recovered with intergovernmental transfers. Repayment of debt and private partnership costs also should fall on beneficiaries, and be paid from these same sources. More generally, we can say that to the extent that subnational government revenues are raised according to the benefit principle so that at the margin those who benefit from local infrastructure investment are those who finance it, the financing of infrastructure will be efficient (McLure, 1998). In contrast to this picture, however, much of the practice in developing countries has steered away from a benefit approach to capital financing. User charges rarely cover even operating and maintenance costs, and ownsource tax and non-tax revenues do not make much of a contribution to capital costs in most countries. The fear of over-borrowing by subnational governments, illustrated by several highly visible Latin American experiences in the 1990s (Dillinger et al., 2003), has resulted in efforts to control excessive borrowing by subnational governments, but there has been no cry for increased local government revenue mobilization to cover debt servicing. Discussion of investment financing – whether at the local, national, international or even the academic level – in recent decades has instead focused on such issues as privatization, concessions, franchises and various types of PPPs, with only secondary attention to various forms of capital grants, subsidized lending from higher-level governments and foreign donors, and almost no attention to developing action plans for significant increases in local revenue mobilization. Nor has more than token attention been paid to ensuring that how projects are financed is consistent with reaching the presumed core objective of getting the right projects in the right places at the right time.20 Arguably, a key element in achieving this goal is to ensure that users should pay for what they get, to the extent that it is essential that they do so (that is, allowing for externalities) and feasible for them to do so (that is, allowing for distributional concerns).We discuss these points further in the next chapter. This normative view about infrastructure finance from local resources may seem strict, but numerous viable revenue options may be used for this purpose (Bahl and Bird, 2008). Significant increases in subnational government revenue mobilization through good (local benefit) taxes and charges are feasible in many countries. Often, however, higher-level governments in

M4439-BAHL_9781786435293_t.indd 143

01/03/2018 11:08

144

Fiscal decentralization and local finance in developing countries

developing countries are not willing to give up productive revenue sources such as income or general sales taxes to subnational governments. Already strapped for revenues, central governments are hesitant to introduce competition for their tax bases in the large urban areas where most central government taxes are raised, even though these are areas where needs are great and the capacity to deliver services is best. The result has been stagnation in revenue mobilization in most developing countries (Bahl, 2014). One should not underestimate the problems in shifting to more local financing of local infrastructure. All too often, when subnational governments do get the green light on taxation, they not only face a limited choice of instruments but are also in effect encouraged to implement badly structured taxes that have undesirable efficiency effects (Martinez-Vazquez, 2013a). Regional and local governments themselves are often hesitant to push for new taxing powers precisely because it would ultimately force them into a position of greater accountability to their constituents. They prefer to claim increasing shares of national transfers (or foreign aid) rather than face their residents with the need to pay higher (and more visible) taxes. Central governments may equally prefer to promise and even provide larger transfers, which carry with them some power to continue to influence and control subnational spending, than to bestow more taxing power on regional and local governments. Carrying this line of reasoning further, one might argue that the order in which governments at all levels prefer different sources of finance for infrastructure appears to be almost the exact inverse of the order that theory suggests. Basically, as we next discuss, theory tells us that user charges, benefit taxes and other local taxes are the best ways to finance local investment (including debt repayment). There are basically two appropriate roles for intergovernmental transfers. First, for governments that do not have the capacity to finance the minimal acceptable level of public expenditure (including infrastructure investment) at normal levels of tax effort, an equalization transfer might be appropriate. The second role for intergovernmental transfers (including subsidies to local borrowing) is an ‘efficiency’ transfer to compensate for the externality bias that distorts some local decisions. We discuss both these approaches in detail in Chapter 7. User Charges For most services that can be priced, and for utilities in particular, user charges are the obvious and ideal source of finance (Bardhan and Mookherjee, 2000). Beneficiaries pay, and when such charges are set to attain full cost recovery, debt service and maintenance can be supported.

M4439-BAHL_9781786435293_t.indd 144

01/03/2018 11:08



Decentralizing and financing infrastructure ­145

User-financing of infrastructure ensures that those who benefit are willing to pay for what they receive, which remains one of the very few ways that one can ensure that providing a particular service is worth what it costs. All this is good, and we and others have said much about it elsewhere.21 However, two big problems have held back the financing of infrastructure through user charges in developing countries. The most commonly mentioned problem is that many people are simply too poor to pay much: “the payment capacities of consumers are severely strained,” as Andres et al. (2016, p. 48) delicately put it. By definition, most people in poor countries are poor. But not all are, and all too often distributional concerns have resulted in charging prices that are too low to cover operating and maintenance expenditures, let alone all costs, and that turn out to benefit mainly that part of the local population that is able to access and utilize such services. Less than full cost recovery – often even less than operating cost recovery – is the rule rather than the exception in developing countries (Estache, 2010). An all too common result of this (presumably) well-intentioned policy is that the bulk of the public subsidy is financed through regressive taxes, inflationary deficits and reduced public service levels. Some of the revenue shortfall ends up being financed by low-income families and some is dealt with by making it more difficult to provide access to such essential services as basic water and sewage facilities to those unable to pay ‘connection fees’ – that is, the poor – as well as by failing to maintain the services adequately, so that the quality goes down for everyone while at the same time inadequate maintenance often leads to huge losses through theft. Poor communities dependent on ‘private’ water supply (e.g. tanker trucks) frequently pay much more per unit of water than they would if on the public system, often with no guarantee of better quality. Many schemes to overcome such problems have been put forward over the years, but as yet few low-income countries have made a serious attempt to price public services efficiently while taking adequate account (for example, through ‘lifeline’ or other tariff structures) of the need to provide a basic level of service to those who are unable to bear the full cost.22 As Spratt (2012) notes, not only the UNDP (2006) but also the World Water Council (2006) and the OECD (2009) all conclude that better water pricing is both necessary and equitable. Unfortunately, no one seems to be listening. The second problem is inertia. Most politicians see their constituencies as being resistant to change, especially to change that costs them something. When the cost is highly visible, like the need to put more money in the fare box, resistance is even stronger. Even if subnational governments accept the principle that they should charge the right prices for such utility services as water, sewerage, irrigation, transportation and power, they are

M4439-BAHL_9781786435293_t.indd 145

01/03/2018 11:08

146

Fiscal decentralization and local finance in developing countries

unlikely to find it easy to do so. People who have been paying a (wrong) low price for something are unlikely to accept easily a higher (right) price. People who are now receiving (often poor) service do not see why they should have to pay more when they do not believe that they will, someday, receive something better for their money.23 Politicians do not want to make current service recipients unhappy by making them pay, and it is not easy to sell ‘pay for what you get’ to people who have long seen others not paying. While there is some experience that suggests neighborhood groups and small villages can and have sometimes agreed to finance small infrastructure works that will benefit them directly (Bird, 1995a), there is little or no evidence that anyone has been able to devise and implement any simple way to rectify the inefficient, inequitable and ineffective fee structures now found in most developing countries. As politicians know, the easiest way to make up for inadequate user charges is with an intergovernmental transfer financed from some distant and hence largely invisible source. One way developing countries can escape from the corner into which most have painted themselves with respect to properly charging for public services may be to focus on self-financing of infrastructure in large cities. Increased levels of user charges may have a better chance of acceptance in large urban areas, where people are better educated and have higher incomes and the business sector is vocal about the need for better infrastructure services. As we discuss further in Chapter 8, floating metropolitan area local governments on their own bottoms with respect to financing infrastructure projects may work best if accompanied by increased local autonomy with respect to expenditure budgets, taxing powers and reduced intergovernmental transfers. Another possible way out may be if a windfall like a commodity boom or a large grant from abroad permits sufficient new investment to improve services so that those asked to pay more can be persuaded they are getting more for their money. But few governments that want to stay in power – and they almost all do – are likely to be willing to diminish the glow of such good fortune by telling people that from now on they must pay more for what they get. Even if a country did begin to charge users full cost prices for new infrastructure, it would likely find it harder to so with respect to something like a new water facility than for a new road or bridge even if there were already some charges for water and none for roads. People can understand why it may cost something to drive on a better road, but they are unlikely to be willing to pay more for what looks like the same water for which someone else (on the old system) pays less. Moreover, while people can choose to take the new road and pay or the old one for free, they seldom have any choice when it comes to water supply. We return to some of these issues in the next chapter.

M4439-BAHL_9781786435293_t.indd 146

01/03/2018 11:08



Decentralizing and financing infrastructure ­147

Local Taxes and Fees24 The property tax can be an efficient source of revenue for financing infrastructure. In many cases, it finances infrastructure that enhances property values, and its revenue potential is large. Even if infrastructure is financed by other means, the property tax base will be enhanced. Bahl and Martinez-Vazquez (2008) estimate that the property tax yields revenue equivalent to 2.2 percent of GDP in high-income countries but less than 0.7 percent in developing countries. When the family of taxes on real property is considered – transfer taxes and various forms of betterment levies – the revenue potential is even greater. Importantly, few higher-level governments want to levy property taxes in any case, perhaps because they recognize that the tax is best levied by local governments or perhaps because they recognize that the tax is unpopular and would likely cause them more trouble than the relatively meager revenue it yields would be worth. Property taxes in low- and middle-income countries seldom finance much infrastructure, and often do not yield enough to pay for even basic facility maintenance. The political unpopularity of the property tax – e­ specially of the economically more justified residential property tax – often leads to a neutering of the tax base, sometimes by higher-level governments that take credit for new exemptions while leaving local governments to deal with covering the revenue costs. Moreover, property tax administration is usually weak and always expensive (McCluskey and Franzsen, 2013; Bird and Slack, 2004). Subnational governments may also use property-based taxes to recapture the increases in land value (betterment) attributable to public investment. Some countries have gained substantial revenue from such taxes (Alm, 2010; Smolka, 2013). Some (Brazil, Colombia) have aggressively used such ‘value capture’ methods in connection with infrastructure finance, but in most it remains a largely underused source. Even in the best of cases, however, a successful value recapture scheme requires a great deal of technical and political investment and may take decades before any significant success is achieved (Bird, 2012a). Regional and metropolitan level governments sometimes mobilize significant revenues with broad-based subnational taxes. In Argentina and Colombia, for example, subnational governments use variants of turnover (gross receipts) taxes, while Brazilian states impose value-added taxes and Brazilian municipalities impose a gross receipts tax on most services. All these levies provide general revenues and are not earmarked for infrastructure, though they doubtless contribute to operating and maintenance costs. In India, the octroi, a duty levied on goods entering the city for

M4439-BAHL_9781786435293_t.indd 147

01/03/2018 11:08

148

Fiscal decentralization and local finance in developing countries

sale, was until very recently a primary source of revenues for Mumbai. Although often criticized for its undesirable economic effects and its weak administration, this levy helped finance some infrastructure cost (directly or indirectly). Most of the taxes just mentioned have been widely criticized as archaic levies that impose large economic costs; but the fact that they persist suggests strongly that not only are they politically acceptable but also that the revenue they yield is considered worth the economic cost. We return to this issue in the next chapter. Motor vehicle taxes would seem a logical way to finance infrastructure in the form of roads and mass transit. Vehicle and driver licenses, fuel taxes, parking fees and tolls and so on can all be seen to some extent as benefit taxes. This tax base is growing in most countries. In India, for example, the number of motor vehicles was 100 times larger in 2004 than in 1951, and has continued in recent years to increase at an even faster rate (High Powered Expert Committee, 2011). In many industrial countries, fuel taxes contribute to a road fund which is earmarked for highway construction and maintenance. However, this practice is less common in developing countries where fuel taxes tend to flow into general central government revenues. Even in countries where such taxes are imposed by subnational governments, as in Colombia and Turkey, the revenues are seldom earmarked solely for transportation-related purposes. Although fuel taxes are generally paid to governments at the distributor level, the proceeds may be allocated to localities on the basis of shipment records, and the tax is of course actually paid by users as part of the retail price at the pump. Most other forms of motor vehicle taxation (registration, licenses, parking) tend to be unduly neglected in low- and middle-income countries. We discuss this issue in more detail in Chapter 8. Some non-tax revenues may also be earmarked for infrastructure finance, although the link between beneficiaries and burdens is less clear. An important example is the extent to which Chinese metropolitan governments have engaged heavily in land sales (long-term leases) as a method of mobilizing resources to finance infrastructure. For all local governments in China, land leases accounted for more than 30 percent of subnational government revenues and over 8 percent of GDP in 2013 (Wong, 2013; Bahl et al., 2014). Land sales have great advantages, namely the revenue potential and the low political cost (at least in China) of raising money this way. But even in China there are drawbacks to paying for local infrastructure through this means, such as the sensitivity of land revenues to the real estate cycle and the riskiness of land value collateral for loans. In addition, the fact that dealing in land in a strongly rising market produces such ‘easy money’ appears to have led to some overspending by local government budgets, and has likely led them to underestimate the opportunity costs

M4439-BAHL_9781786435293_t.indd 148

01/03/2018 11:08



Decentralizing and financing infrastructure ­149

of converting land to urban use. Moreover, commune-owned agricultural land is an exhaustible resource. A spending binge financed by land sales must inevitably come to an end.25 Nonetheless, other countries (such as India) where urban land available for development is generally scarce should examine closely whether all government-owned land is being put to its highest and best use (Rao and Bird, 2011, 2014). Sometimes specific revenue sources are earmarked for infrastructure development. In Peru, for example, where natural resource revenues are distributed on a derivation basis, they must be spent for economic development purposes (Canavire-Bacarreza et al., 2015). One aim in doing so was in part to replace the ‘heritage’ of the region that was presumably lost as a result of mining or other extractive activity, an argument that may have some merit. However, even leaving management issues aside, such earmarking provides little incentive for efficient spending, and it can result in greatly increased regional disparities in periods when commodity prices are booming (Martinez-Vazquez and Timofeev, 2016). Other countries sometimes earmark significant portions of general intergovernmental transfers to investment, apparently to restrain the feared dissipation of such transfers in current expenditures.26 It is far from clear that such earmarking achieves its objective. Indeed, some evidence suggests that reducing earmarking is more likely to improve than to distort the allocation of local resources. The theoretical argument is simply the standard decentralization theorem: that allowing funds to be spent according to local tastes results in superior allocative and distributional results. Empirical evidence in Norway strongly supports this argument (Borge et al., 2012), but of course one may doubt the relevance of Norway’s experience to the many developing countries in which ‘elite capture’ of local governments seems more likely. However, studies in both Bolivia and Colombia also point in this direction (Faguet, 2005).27 A careful study of the EU finds that revenue decentralization tends to increase subnational infrastructure investment (Kappeler et al., 2012). Earmarking local revenues for infrastructure may make sense when good benefit reasons (efficiency, equity and management) support doing so. Small projects with well-defined benefit groups are most likely to meet these conditions. In other cases, however, such earmarking may distort local preferences, exacerbate perverse incentives to build new rather than maintain existing infrastructure, and connect specific revenue sources with specific expenditures in ways that lack both economic and political logic. Owing to the fungibility of money – one dollar is just like another dollar – it is not usually possible to determine the extent to which earmarking revenues tends to increase expenditures on an activity rather than simply replace revenue from other sources that would have been spent in the same way.

M4439-BAHL_9781786435293_t.indd 149

01/03/2018 11:08

150

Fiscal decentralization and local finance in developing countries

Intergovernmental Transfers Because regional and local governments in most developing countries usually have limited own resources and little access to private capital markets they generally rely heavily on grants (or subsidized loans) from higher-level governments to carry out costly public works. The effect of transfers on infrastructure finance depends largely on how the transfers are structured. As discussed further in Chapter 7, most countries structure their transfer systems using some combination of three approaches. The first is unconditional grants, some of which may be used to maintain the public capital stock and to repay debt, though they are not directly tied to either. In the eyes of lenders, the security associated with unconditional grants as the base for repayment is not strong. Some countries attempt to get around this by making provision for these transfers to cover debt service costs by providing for an intercept arrangement or allowing these transfers to be pledged to repay debt, as has been done in Mexico (Revilla, 2012). This arrangement in effect earmarks grants to service the debt on capital projects. Lenders tend to like it because it guarantees that they will get paid even if the recipient subnational government wastes the money; but central governments are seldom keen, perhaps because they would prefer lenders to have a direct interest in requiring more disciplined budgeting and better execution from borrowers. A second approach is to make grants to local governments explicitly conditional on their expenditure for a specified capital purpose, often with no matching arrangement and in practice often with little supervision, either before or after the money flows. Sometimes, such conditional grants are distributed on a formula basis; sometimes they take the form of cost reimbursements. The appropriate design of such transfers raises several questions. Is one aim to rectify imbalances in the distribution of existing infrastructure (Ahmad and Searle, 2006)? To what extent are such transfers intended to improve economic efficiency – as is presumably the case with respect to most economic infrastructure such as transportation? Or are they intended to be equalizing, e.g. by providing at least a minimum standard of such public services as education and health by providing hospitals and schools in certain areas (Josie et al., 2008)? Whatever the intended goal, transfers intended for capital spending that include fiscal capacity as a determining factor – an element in the design of many grants, as we discuss further in Chapter 7 – should be expanded to encompass not just tax capacity but also the ability to access credit markets (Herrero-Alcalde et al., 2011). A third channel is more directly tied to infrastructure budgets. One approach is to establish a block grant for infrastructure services on a

M4439-BAHL_9781786435293_t.indd 150

01/03/2018 11:08



Decentralizing and financing infrastructure ­151

matching basis, as in India, where a large federal grant for urban infrastructure development and slum upgrading was allocated to cities on a matching basis (High Powered Expert Committee, 2011; Ahluwalia et al., 2014.). The program was introduced in 2005, and while it succeeded in focusing increased attention on urban infrastructure issues, implementation progress was slow (Rao and Bird, 2014) and the program was replaced in 2016 by a new system. South Africa makes use of a more formal municipal infrastructure grant, designed primarily to improve services in poor neighborhoods (van Ryneveld, 2007). In Brazil, as in other countries, ad hoc grants frequently support specific capital projects. If done in conjunction with increased local revenue mobilization and the imposition of at least cost recovery levels of user charges, any of these approaches to intergovernmental transfers may produce more efficient delivery of infrastructure services by local governments. Most could be improved, however. Conditional grants, for instance, can be significantly improved by requiring subnational governments to prepare an adequate investment plan and an adequate maintenance plan, as well as an appropriate user charge policy. The governments receiving such transfers could be selected by a systematic process that pays attention both to need and capacity factors, and to the economic evaluation (cost–benefit analysis) of the project in question. For all but the biggest and most advanced regions and cities, adequate technical assistance needs to be made available to permit them to develop plans, arrange financing, manage construction and operate the facility (or to contract out its operation) in the most efficient possible fashion. To ensure accountability as well as good outcomes, the execution and operation of the grant-aided work should be monitored and evaluated, with periodic progress reports, field inspections and formal evaluations of outcomes, including consumer surveys. Achieving much along the lines just mentioned may be a counsel of perfection in the conditions of many developing countries. Nonetheless, as the early experience with ‘municipal development funds’ demonstrated (Davey, 1988), if such conditions are not satisfied, the results of capital transfers (and subsidized loans, which can be thought of as a combination of a grant and a loan) are unlikely to live up to expectations. Subnational infrastructure finance when local fiscal autonomy is low and many localities are poor is inevitably dependent on transfers from higher-level governments. These transfers may be efficiency enhancing if they are conditional grants designed to correct for under-spending on services characterized by externalities, although few such transfers are found in practice. To design such a grant requires identifying the optimal level of spending for a particular function that takes account of social as well as local benefits and also allows for the elasticity of demand for the infrastructure service.

M4439-BAHL_9781786435293_t.indd 151

01/03/2018 11:08

152

Fiscal decentralization and local finance in developing countries

In practice, every element of such a formula in most low-income countries can usually be little more than a guess, so it is not surprising that most infrastructure grants bear little if any resemblance to the ideal externalitycompensating subsidy of theory, as we discuss further in Chapter 7. Borrowing Issuing public debt is like taxing the future. (Winer, 2016, p. 11)

The ‘golden rule’ of public sector borrowing at the subnational level is that finance from this source should be used only for capital formation.28 Borrowing is not an independent source of revenue. Loans must be repaid from user charges, from subnational taxes or from intergovernmental transfers (i.e. taxes imposed by higher-level governments). The rationale for borrowing is not to obtain additional revenue, but rather to smooth out the difference between benefits/revenues coming in and expenditures going out. Borrowing is a way subnational governments can attract the large amount of funds necessary for financing the construction phase in a way that is both economically efficient (and equitable) because the debt can be repaid (from current revenues) during the life of the project when the project is presumably providing sufficient benefits to local residents to enable them to bear the cost of servicing the debt. Expenditure on infrastructure investment is inevitably ‘lumpy’ so the principal rationale for borrowing is to reduce intergenerational transfers of benefits and costs. By matching payment for the infrastructure with the time pattern of consumption of the asset, governments defer payments until the corresponding returns arrive. Debt is thus at best a potentially efficient way to arrange payment for the purchase of public assets that have a long life. At worst, it may prove to be a fiscal time bomb that may blow up not only subnational but even national finances if badly mishandled. Sometimes, of course, subnational governments may roll over (extend) loans and postpone repayment further into the future, thus making it the problem of some successor government. But users and taxpayers must still pay at some point. The most essential element of a sound subnational borrowing program is thus a sound subnational fiscal structure in terms of adequate access to ‘own revenues’ backed by a well-designed and stable intergovernmental transfer system, within an institutional structure capable of dealing appropriately with any problems that might arise. Until countries have all these elements in place there is often good reason to be skeptical about the efficacy of subnational government borrowing. Reckless borrowing by subnational governments may lead to problems, as students of local public finance have long argued (Prud’homme, 1995;

M4439-BAHL_9781786435293_t.indd 152

01/03/2018 11:08



Decentralizing and financing infrastructure ­153

Tanzi, 1996). The usual problem is that the stream of local government revenues is not large enough to sustain repayment but lenders are nonetheless willing to go forward with the project because they anticipate some form of bailout, with the central government stepping in to service subnational debt if necessary. This well-known moral hazard problem has led to overborrowing and to some form of bailout of subnational governments in Brazil, Argentina and South Africa, and more recently in China (de Mello, 2007; Wong, 2013). Many countries attempt to control over-borrowing by subnational governments through various forms of fiscal responsibility legislation (Liu and Webb, 2011), though these programs have met with varying degrees of success, as we discuss below. One approach to infrastructure finance is for local governments to borrow directly from the private market through bond issues, as is done in the US and some European countries, or direct borrowing from private banks. Another may be, as in the province of British Columbia in Canada, to pool local government debt through a provincial Municipal Finance Authority (MFA), which secures funds at lower costs both by centralizing administration of the marketing and management of municipal borrowing for capital projects and by ensuring that even the smallest municipality’s borrowing is supported ‘jointly and severally’ by all the municipalities in the province (except for the city of Vancouver, which has opted out of the MFA). None of these approaches involve any subsidization of local borrowing. Other approaches involve the imposition of a specialized agency to serve as an intermediary between small local governments and prospective sources of funds. When the ultimate source of funds is the public sector, such arrangements in the past have suffered from several problems. Most obviously, loans may end up being made primarily for political rather than economic reasons: those with the best political connections get the most money. A second problem is that loans may be used mainly to service existing debt or to roll over existing loans, which means in effect that what is being funded is current spending.29 At the extreme, this may lead to a soft budget constraint which removes any incentive for subnational governments to be disciplined in their borrowing by guaranteeing that the central government will, in the end, bail them out. That this is the case in many countries is indicated by the fact that credit agencies seem often to rate all subnational governments in a country, regardless of how well they are run or what their financial state may be, at the same risk level as the central government. Sometimes, such central agency loans are secured by servicing the debt through an ‘interception’ of another central transfer payment (Philippines) or an advance deposit in the central bank to cover debt service (India). Several other approaches that have been used to package local borrowing in various countries are discussed in Box 4.2.

M4439-BAHL_9781786435293_t.indd 153

01/03/2018 11:08

154

Fiscal decentralization and local finance in developing countries

BOX 4.2  PACKAGING LOCAL BORROWING Municipal Development Funds (MDFs) About 60 developing countries have some form of MDF, a financial intermediary that can pool funds from various sources and then on-lend them to local governments to finance projects or serve as ‘bridge’ funds to secure private financing. Usually, the central government bears the default risk, but in some cases (e.g. Colombia) private banks bear the risk, with the central bank standing by as the last-call source of liquidity. Some MDFs crashed and burned in earlier days (Davey, 1988), but in more recent years a few in South Africa and elsewhere appear to have been successful in providing a way for local governments to access credit markets (Freire, 2013). Local Investment Corporations (LICs) China has made considerable use of LICs to finance its exceptionally rapid development of urban infrastructure in recent years. Until recently, subnational governments in China were not allowed to borrow without explicit permission from the central government. The obvious impracticality of this requirement in a country of China’s size and complexity given the rapid pace of development led to the creation in 1992 of special investment corporation, funded from the local budget and authorized to borrow. Once it became clear that the central government would accept the use of LICs to issue debt on behalf of municipalities, and to use the expected revenues from land leases as collateral for these loans, the problem of finding the financing for capital projects was essentially solved (Bahl et al., 2014). Soon not only public infrastructure facilities such as highways, subways, airports, schools, etc. but also even private developments around the country were being financed through LICs. The volume of LIC financing expanded enormously and concern about the sustainability of the system, given the high dependence of local government finance on rapid increases in revenues from land development, became more marked (Wong, 2013). Social Investment Funds (SIFs) In contrast to MDFs, SIFs are not intended to finance local government investment as such, but rather to support specific national policies such as delivering clean water or primary education. They were often established as a way of delivering substantial amounts of foreign aid to specifically targeted objectives. It is thus not surprising that for the most part they have not operated through or even in conjunction with local government systems, but have instead operated in a parallel fashion to achieve national objectives rather than to satisfy local priorities. Although substantial efforts were sometimes made to integrate local participatory planning processes into the SIF system – as in the case of the community-driven developments (CDDs) discussed in Box 3.2 – this was usually done without building any links with the formal local government structure, and was sometimes in direct opposition to it. Finally, as has often occurred with some MDFs in the past, many SIF investment decisions seem to have been driven largely by political rather than economic concerns (Romeo and Smoke, 2016).

M4439-BAHL_9781786435293_t.indd 154

01/03/2018 11:08



Decentralizing and financing infrastructure ­155

Many countries have imposed rules and regulations intended to control borrowing by subnational governments. A few developed federal countries such as the United States and Canada impose no central controls at all on borrowing by the regional governments (states and provinces respectively) that are the component entities of the federation. Even in these countries, however, substantial regulatory control over local borrowing is always (in Canada) or sometimes (in the US, depending on the state constitution) imposed on local government borrowing. And such controls are almost always imposed on all subnational governments to some extent in most other countries, especially in developing countries – even those like Brazil, Argentina and India that are federations. As noted above, there are good reasons why subnational governments should borrow to finance the acquisition of some long-lived assets. But there are also potentially serious problems in giving them a free hand to do so, especially in countries with relatively weak institutional and intergovernmental fiscal structures and a record of subnational default. Some countries deal with the issue by simply prohibiting any borrowing by subnational governments. More commonly, however, countries attempt to preserve fiscal discipline by imposing ex ante rules and/or ex post controls with the aim of preventing unwarranted borrowing and possible insolvency at the subnational level. The primary reason such controls are considered necessary is because there is at least implicitly a soft budget constraint in many countries, and hence a ‘moral hazard’ that subnational governments will not borrow responsibly because they (and those from whom they borrow) expect that the central government will, in the end, bail them out. The problem with bailouts is that once subnational governments learn that there is a way around the hard budget constraint, they are likely to continue overspending on infrastructure.30 A good borrowing framework can head such problems off without closing the borrowing window for those local governments that have the greatest infrastructure gaps. The most common form of regulation is to impose ex ante controls or rules that limit subnational borrowing, combined with sanctions for non-compliance (Box 4.3). Ex post regulation may consist of administrative or judicial sanctioning processes. Many countries combine various elements of these different approaches. For example, in Canada provincial borrowing is solely subject to market discipline, while municipal borrowing is regulated by the provinces, usually through rule-based regulations such as debt limits (Bird and Tassonyi, 2001). A recent empirical examination of an unbalanced panel for 57 countries at various income levels for the 1990–2008 period found that, although countries with deeper financial markets and lower subnational debts were more likely to use the ‘softer’ approaches of market discipline and cooperative regulation, none of these

M4439-BAHL_9781786435293_t.indd 155

01/03/2018 11:08

156

Fiscal decentralization and local finance in developing countries

BOX 4.3  REGULATING SUBNATIONAL BORROWING Ter-Minassian and Craig (1997) distinguish four types of ex ante controls: 1. Market discipline (i.e., private lenders are willing to lend), as for provinces in Canada and for states in the United States without any restrictions; and in a number of other countries subject to some mild restrictions – e.g. for capital projects only or borrowers must have a balanced (current) budget. 2. Rule-based regulation (e.g. debt ceilings, debt-repayment capacity, the ‘golden rule’ of borrowing only for capital projects, expenditure limits such as a balanced budget requirement, etc.), as illustrated by the debt ceilings established for Brazil’s states and municipalities and the various fiscal targets of the 2000 Fiscal Responsibility Law, or the ‘traffic light’ rules established in Colombia in 1997, with those with fiscal indicators lower than ‘green light’ level being permitted to borrow only after adopting an agreed adjustment program. Many countries have different rules of varying stringency for different categories of subnational government. 3. An administrative approach (e.g. direct central approval required for any loan), as in the United Kingdom, Korea and some other countries, or central government guarantee required for external borrowing as in Peru. 4. A cooperative approach in which central and subnational governments jointly negotiate and agree on some rules (e.g. Australian Loan Council). In Switzerland, cooperation is required directly from citizens, since borrowing by state and regional governments (under the golden rule only) must be approved by a popular referendum.

regulatory approaches appeared to have any significant direct effect on subnational fiscal sustainability (Martinez-Vazquez and Vulovic, 2016). This result is certainly not conclusive. However, it is yet another indication of the point made by the discussion in Chapter 2: we understand much less than we should about many important aspects of regional and local finance in developing countries. In general, bundling subnational borrowing – for example, by having regional entities borrow for smaller local authorities – likely often makes sense on cost grounds. Similarly, packaging subnational borrowing in some form, such as a national infrastructure bank, may make sense when financing flows to projects that, although the responsibility of subnational governments, are of national significance – e.g. as part of a national network of roads or electricity distribution. In most cases when such institutions exist in developing countries, however, loans from public financial institutions have been extended on subsidized terms and are in effect a form of matching grant. Although one must hope that most countries have learned from the many unfortunate past experiences with such ‘soft’ funding, there

M4439-BAHL_9781786435293_t.indd 156

01/03/2018 11:08



Decentralizing and financing infrastructure ­157

is no magic institutional way to avoid the hard necessity of first establishing subnational finances on a sound and sustainable basis so that at least some of them – probably the largest and better off – may, when financial markets are sufficiently developed, be creditable candidates for private sector financing. When subnational governments have an important role in providing infrastructure, most of them need both technical and financial support to play that role adequately. Equally, it is important to ensure that financial resources do not arrive in ways that distort the incentives facing subnational decision-makers in ways that may damage both the provision of infrastructure and, more broadly, governance at the subnational level.31 We discuss these issues in detail in Part III of this book. Public–Private Partnerships Much the same can be said with respect to the interest in public–private partnerships (PPPs) since the mid-1990s. During this period, increased private involvement was encouraged to increase the efficiency of service provision and to provide badly needed resources to support urban infrastructure investment. Klein (2012) argues persuasively that both justifications need to be qualified. The additional private sector resources that flow to infrastructure through such schemes in effect come from user charges, and could also be realized by local governments if charges were sufficiently large to recover all costs. Moreover, while private sector expertise may often be a step up from local government capacity, it is usually more expensive. There are no free lunches when it comes to infrastructure finance. In addition, there are questions about how and to what extent PPPs should deliver services – for example, full privatization with various degrees of regulation or some form of contracting for operation. The build-operate version is often especially attractive to resource-poor governments with short political horizons because it offers a way to get facilities built without incurring highly visible government debt. As yet, however, PPPs have not lived up to expectations in terms of how much this financing vehicle has added to urban capital financing in developing countries (Annez, 2007; Alm, 2010).32 Moreover, less than 10 percent of the investment has been in the high-priority water/sewer sector, and an even smaller share has taken the form of full or partial privatization (Ménard, 2013). Most PPPs have been focused on the more obviously and easily profitable energy and telecommunications sectors (Klein, 2012). Relatively little has been attracted to the higher-priority water and sewer sectors, mostly because of pricing risks. In some instances where private capital has been attracted to these areas (e.g., in Bolivia, Venezuela and Argentina) the result has been conflict-laden and generally considered a failure, ­sometimes

M4439-BAHL_9781786435293_t.indd 157

01/03/2018 11:08

158

Fiscal decentralization and local finance in developing countries

(as in Bolivia) ending with the effective re-nationalization of projects initially carried out by PPPs. The initial refusal of governments to impose appropriate prices (user charges) to cover costs may have been one reason the PPP route was attractive; but when regulators were similarly unwilling to let private operators cover their costs (including normal profits), they bailed out, leaving taxpayers to pay not only the costs they had originally dodged but also the additional costs attributable to failed privatization. A World Bank study (Garcia-Kilroy and Rudolph, 2017) reports that 68 percent of the 1,700 PPP projects in Latin America between 1990 and 2013 were renegotiated, on average one year after the project award. Lunch sometimes costs more when more people are at the table. Opinions differ sharply with respect to both the merits of PPP arrangements and how best to design and implement them. In an assessment of the practice, Merk et al. (2012) conclude that a critical feature (if a PPP is to minimize project costs) is that the contract should be ‘global’: that is, that a single contractor is responsible for managing the whole project in order to reap full economies of scale and scope and provide maximum incentives to invest and innovate. In contrast, Siemiatycki and Friedman (2012) – considering urban transit projects in which a major issue is how to allocate ‘ridership (demand) risk’ – argue that ‘unbundled’ contracts that exclude facility operation are often preferable because they are likely to attract more competitive bids and to lower the cost of private sector borrowing (see also Box 4.4). The contradictory nature of these recent assessments emphasizes the importance of the specific context and institutional and regulatory settings (e.g. the relative development of financial markets and who sets user charges) as well as the detailed specifics of PPP contracts in determining outcomes.33 The question of who bears what risks lies at the heart of public–private partnerships. Many observers, like the IMF, have often expressed concern about the macro-fiscal risk that PPPs may be used to circumvent budget constraints by hiding the real fiscal costs of providing infrastructure services and building up hidden contingent liabilities, future fiscal deficits and growing debt.34 There are many other risks that may be allocated in different ways between the public and private sectors. For the private sector, for example, there are risks that the regulatory framework and/or pricing commitments could change and cause delays in the project. Annez (2007) and Ingram et al. (2013) argue that the inherent riskiness of urban investments in water and sanitation is the main constraint to increasing the flow of private capital to this sector. There is a weak record of full cost recovery, and often an unwillingness of local governments to stand behind the kinds of tariff levels and regulatory arrangements that would be necessary to attract private investors. Who is ultimately responsible for the

M4439-BAHL_9781786435293_t.indd 158

01/03/2018 11:08



Decentralizing and financing infrastructure ­159

BOX 4.4  FINANCING RAPID TRANSIT: TWO EXAMPLES Even when a transit project is carried out entirely by the public sector, regulatory and financial factors are critical. China, for example, has in recent years carried out huge urban infrastructure investments through what may perhaps be thought of as a Chinese variant of public–private partnership (PPP), where both parties to the contract are really ‘public.’ To illustrate, the extensive Beijing metro system, like many other projects in China, was financed primarily by bank loans to a local investment corporation – the Beijing Infrastructure Investment Corporation – created and controlled by the local government (Su and Zhao, 2006). Since the national government strongly encouraged local governments to carry out such investments, and banks to finance them in doing so, presumably the banking sector in effect considers such loans to be guaranteed by the state. In addition to this implicit subsidy, loans to local investment corporations (LICs; see Box 4.2) have generally received an explicit subsidy in the form of an interest rate about 10 percent lower than the normal rate of long-term debt. Most local borrowing is directly serviced from local revenues, which are in turn highly dependent on revenue from leasing land. Such revenue is highly sensitive both to property values and to the amount of land sold: in 2011, for instance, Beijing’s revenue from this source decreased by 36 percent from the 2010 level. Even if one assumes that the whole rapid transit system is optimally designed, constructed and operated, current plans to expand the metro system further in the next few years with a flat-fare policy that already requires an annual operating subsidy (and will need an even larger one in an expanded system) make it likely that the sustainability of the financing model used to build not only the Beijing metro but also much of the extremely impressive development of urban infrastructure in China in recent years is an example of what Wong (2013) calls ‘riding the tiger.’ It may soon require careful reconsideration and in all likelihood substantial ­adjustment  – unless perhaps the urban real estate sector can realistically be expected to continue to boom at pre-2010 rates for the next few decades. To take a quite different, and perhaps more widely applicable example, the Ahmedabad Bus Rapid Transit System (ABRTS) – financed in large part (35 percent) by a national urban development program (the Jawaharlal Nehru National Urban Renewal Mission, JNNURM) and in smaller part (15 percent) by the state of Gujarat (with the remaining 50 percent coming from local sources, including a dedicated urban transport fund) – was developed and implemented through nine separate PPP arrangements negotiated between the special public corporation (Ahmedabad Janmarg Limited) created for the purpose and various private providers (see Box 4.1). These arrangements covered everything from major system investment (bus stations, corridor, flyovers, as well as buses) to housekeeping and parking (National Institute of Urban Affairs, 2011). The resulting system has both substantially improved people transport in the city and, like the somewhat similar earlier TransMilenio BRTS in Bogotá, won national and international acclaim as a model that seems worthy of emulation elsewhere. Some aspects – notably dedicated funding sources for urban transit systems – have already appeared in several North American cites in recent years. Vancouver, for example, collects fees and taxes that are dedicated to a regional transportation authority, as do Chicago, Los Angeles and Salt Lake City (Institute on Municipal Finance and Governance, 2012).

M4439-BAHL_9781786435293_t.indd 159

01/03/2018 11:08

160

Fiscal decentralization and local finance in developing countries

project – central government, subnational government, a public enterprise owned by one or the other level of government? Who really controls it, and how? Who is ultimately supposed to pay for the costs of the project: the government with general public funds or some earmarked revenues, or by collecting or allowing the private sector to collect user fees? Is the government supplying additional support to the private partner through guarantees, tax amnesties or in some other way? PPP contracts must deal with these questions and many more. Sometimes, no matter what the contract may say, there is what Pethe (2013) calls a ‘trust deficit’ between public and private sectors that has sometimes resulted in episodes like the Bolivian re-nationalization mentioned above.35 It is not surprising that the World Bank (2009b) found few positive results in efforts to attract private financing of municipal services. There also is a risk that the services provided may not be what the public wants. Another downside risk is that the private partner will fail, or insist on re-contracting, and the public sector will be forced to take on the obligation in full. How successful such arrangements are from the perspective of either partner depends very much on the details of exactly how the contractual arrangements are structured and how the risks are shared.36 Given the weak institutional capacity of subnational governments in many developing countries, it seems unlikely that they will have a strong hand in negotiating such contracts. The Indian High Powered Expert Committee (2011, p. 101) puts it well: “Weak governments cannot rely on private agents to overcome their weaknesses nor can they expect to make the best possible bargains for the public they represent.”

NOTES   1. For an extensive recent discussion of this subject, see Frank and Martinez-Vazquez (2016). This chapter draws on our contribution to that book (Bahl and Bird, 2016).  2. The extent to which investments by publicly owned companies such as utilities are recorded as government (rather than corporate) ‘investment’ varies in different countries and is often rather fuzzy: for a careful discussion of the how this question is treated in the United States, for example, see Ebel and Wang (2017).   3. The Indonesian INPRES grants, which were abolished with the decentralization reform in 2001, were earmarked for projects determined by the central government. It is not surprising that there was limited enthusiasm to operate and maintain these projects at the local government level (Brodjonegoro and Martinez-Vazquez, 2005).   4. There are of course ways around such problems, such as the intergovernmental transfers and private contracting that are discussed later.   5. There are exceptions: for example, very small countries may have need for only one full-service university, or national defense considerations may require more central government involvement in airports.   6. Berry (2009) provides a detailed account of the extent to which special-purpose local governments in the US have been ‘captured’ by such groups as developers, other

M4439-BAHL_9781786435293_t.indd 160

01/03/2018 11:08



  7.

  8.   9.

10. 11. 12. 13. 14.

15.

16.

Decentralizing and financing infrastructure ­161 business interests and public sector workers. The history of municipal infrastructure development around the world is replete with similar stories of elite capture: see, for example, the account in Briggs (1996) of city development in the UK in the nineteenth century, when most investment in infrastructure was financed by (subsidized) private firms. Bardhan and Mookherjee (2000) provide a classic account of elite capture of local governments; this theme, and corruption at the local level, is further discussed in some of the contributions in Bardhan and Mookherjee (2006). As recent scandals in Brazil have again emphasized, corruption continues to be rife when it comes to local public works contracts in Latin America as in other developing regions – and not just there (Curry, 2016). For a useful recent review of the rather indecisive empirical literature on decentralization and corruption, see Shah (2016). Box 3.4, for example, discusses how an important attempt to give more weight to local preferences in rural India was, at least in some places, thwarted by local elite interests and corrupt officials. To some extent, as Lewis-Faupel et al. (2016) show with examples from India and Indonesia, technology – in this case electronic procurement – may reduce not only corruption but also delays, and improve outcomes. However, as another Indian case discussed in Box 3.4 shows, ingenuity may still sometimes defeat technology. For a more broad-ranging review of the provision and maintenance of rural roads, see Ellis and Menendez (2016). For a more extensive discussion of such problems in OECD countries, see e.g. Kim et al. (2010). Charbit and Gamper (2016) also discuss the coordination problem in such countries, noting some successful experiences in France and Canada. They stress the extent to which success depends on explicit contractual arrangements, good information and the establishment of a voluntary ‘partnership’ rather than a mandatory solution ‘imposed from above.’ An earlier study for the 2005–2010 period (covering rural and urban areas in developing countries) estimated future maintenance needs in developing countries to average about 3.3–3.5 percent of GDP (Estache, 2006). We state these rules here with respect to O&M, but of course many of them apply more widely, as discussed later in this chapter and elsewhere in the book. For an interesting example of the importance of monitoring O&M see Hu and Ebel (2016) on the water sector in Albania. As Charbit and Gamper (2016) note, success may also depend on the extent to which the process involves both parties and is voluntary rather than mandatory. Fiszbein (1997) notes that his study cannot be considered to be representative of the general effects of Colombia’s initial moves toward decentralization in the early 1990s as the municipalities studied were selected precisely because they were responding in effective ways to the increased funds and freedom that had been bestowed upon them. If conditions are not right – as Bardhan and Mookherjee (2000) argue, and as Juul (2006) illustrates for Senegal in the late 1990s – the result of decentralizing authority to spend or to tax to local levels may be to reinforce existing clientelistic and patronage networks and to weaken accountability and trust. This point can be overdone, however. For example, the World Bank has developed a highly logical and coherent system for infrastructural investment projects which ­identifies eight ‘must have’ features of an effective public investment management system (Rajaram et al., 2010). Observing, correctly, that few national governments in low-income countries are likely to have either the resources or capacity necessary to ‘have’ all these features, a subsequent study developed an interim ‘stopgap’ approach that would be less demanding but might still do a good part of the job (Marcelo et al., 2016). The ‘piloting’ experiences (in Vietnam and Panama) with the proposed ‘multi-criteria decision analysis’ approach to project prioritization set out in the latter document are interesting. However, an approach focusing on the development of an overriding technical and centralized approach to project selection is unlikely to mesh easily with the overwhelmingly political questions that arise from the inevitably ­decentralized effects of

M4439-BAHL_9781786435293_t.indd 161

01/03/2018 11:08

162

17. 18. 19.

20.

21. 22. 23.

24. 25. 26. 27.

28. 29. 30.

31. 32. 33.

Fiscal decentralization and local finance in developing countries most infrastructure investment decisions, even when the decisions are not or cannot (or even should not) be made at some level below the central government. Procurement changes may have other beneficial effects also, as noted in Lewis-Faupel et al. (2016). As we discuss later, such earmarking is essential when infrastructure (including debt service) is financed by user charges. Borrowing and public–private participation schemes (PPPs) are not sources of revenue but financing methods and ways of risk sharing, as discussed below. Countries may also receive capital transfers from external donors. As Estache (2010) notes, aidfinanced infrastructure is especially important in some countries in sub-Saharan Africa. However, we do not consider this source further here: for a recent review, see Kharas and Linn (2013). Bird (2005) suggests that the conventional cost–benefit analysis often used to rank project desirability is sometimes misleading because it does not correctly take into account the different marginal costs of funds (MCF) from different sources of funding. Although this question is not discussed further here, interested readers are directed to Dahlby (2008) for a thorough discussion of the MCF concept. User charges are discussed further in the next chapter. For earlier and more detailed discussions, see e.g. Bahl and Linn (1992), Bird (1976, 2001), Bird and Tsiopoulos (1997) and Martinez-Vazquez (2013a). For an early review, see Bird and Miller (1989). Unfortunately, as UNDP (2006) documents for a number of countries, it appears that little has changed for the better since that paper was written (see also Le Blanc, 2007). For a discussion of the resistance to user charges in high-income countries, see Bird (2017). Those who feel this way may not be all that wrong since user charge increases may be needed simply to cover the (increasing) cost of doing business: e.g., increased gasoline prices for the buses, increased cost of chemicals for the sewerage system, increased wages for the employees of the water company, etc. This subject is discussed further in Chapters 5 and 6. An interesting discussion of this issue is in World Bank (2012). In Colombia, the pernicious effects of such earmarking are mitigated to some extent by the fact that ‘investment’ is interpreted to include so-called ‘social investment’ in health, education and so on. A less positive outcome emerged in a study of state-level data in India, which found decentralized investment to be notably less productive (Asthana, 2003). Similar divergent results emerge in recent study by Vinuela (2016), which found that decentralization in Argentina both increased public investment and better aligned it with local needs, while in Mexico results were less favorable in both respects. She attributes the difference mainly to the greater local autonomy over revenues in Argentina. Often, however, some in-period borrowing is permitted in order to smooth cash flow over the budget year. When central transfers are used as collateral, as in Mexico, this has been seen as a problem (Sutherland et al., 2005). This tendency is encouraged by the fact that the current and prospective future beneficiaries of infrastructure services do not face (and do not think they will face) the real price – a point we return to in the next chapter, in which we also discuss the hard budget constraint in more detail (see also Rodden et al., 2003). For a study emphasizing the importance of appropriate incentives, fiscal and ­otherwise, with respect to establishing and sustaining sound decentralized governance, see Faguet (2011). IMF (2014) estimates that less than 25 percent of infrastructure investment has been added by private investment. These points are underlined in a critical review of privatizations in developing countries which concludes – echoing again one of the main lessons 50 years of observing development projects should have taught us all – that ‘one size does not fit all’ (Gasmi et al., 2011; see also Estrin and Pelletier, 2015).

M4439-BAHL_9781786435293_t.indd 162

01/03/2018 11:08



Decentralizing and financing infrastructure ­163

34. IMF and World Bank (2016) is a manual on how to assess the fiscal risk of PPP projects. See also the World Bank’s PPP Guide (World Bank, 2014). 35. For a good brief review of the growing literature on the importance of ‘trust,’ see Graser and Robinson (2016). 36. For detailed exploration of the structuring of PPP arrangements, see Engel et al. (2010). For a skeptical view of the range of opportunities to exploit such possibilities, see Menard (2013). PPPs may also give rise to problems if the private contractor does so well out of the deal that the strong public reaction against the government for ‘selling the family silver’ induces the government to make popular but bad decisions.

M4439-BAHL_9781786435293_t.indd 163

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 164

01/03/2018 11:08

PART III

Financing Local Government: The Key to the Puzzle

M4439-BAHL_9781786435293_t.indd 165

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 166

01/03/2018 11:08

5. Financing local and regional government If local bodies are to play any significant part in economic and social development, they must clearly have access to adequate finance. If they are both to act responsibly and to show initiative, some, not negligible, part of this control over resources must be independent, in the sense that local councils are free to choose the rates and their service charges. (Ursula Hicks, 1961, p. 277)

No matter what local and regional governments should do in theory or are required to do according to the constitution – or what central governments or local residents would like them to do – what they actually do in practice is largely shaped by the level and nature of their financial resources. The reason is simple: unlike central governments, even the richest and largest regional or local governments cannot print money.1 Subnational governments must either raise funds through taxes or fees or obtain intergovernmental transfers. They can also borrow, but this does not remove the need to raise funds from other sources; it simply moves the need to do so to another time because loans must be repaid.2 How much local and regional governments spend is ultimately determined by how much revenue they have. Spending money well and sensibly – the subject of the previous two chapters – is not always easy. But raising it correctly may be even harder, and the politics is downright fierce. Experience around the world suggests that the toughest nut to crack in building a successful fiscally decentralized structure in low- and middle-income countries has been the implementation of an adequate and appropriate subnational government revenue structure. How to do so is the subject of Part III of this book. We begin here with a simple framework setting out the case for imposing a hard budget constraint on subnational governments. We then discuss some of the factors shaping the size and scope of local and regional government activities in different countries, how these factors and the different objectives of decentralization policies can and should affect the policies adopted, and the guidelines that emerge about how to do it right. The balance of this chapter and the next then considers in more detail some of the revenue sources that may be raised by local and regional governments. The last chapter of Part III takes up the question of the appropriate design and 167

M4439-BAHL_9781786435293_t.indd 167

01/03/2018 11:08

168

Fiscal decentralization and local finance in developing countries

level of the intergovernmental transfers that in practice tend to dominate local finance in developing countries. We make no apology for devoting much of this book to revenue issues. Not nearly enough attention has been paid to getting local and regional revenues right. To illustrate, one of us once spent several weeks in a country working on local revenues as part of an international mission. On the last day there, purely by chance he encountered some other experts from the same institution who were helping the country develop a reform of the educational and health sectors. Since all schools and most hospitals and health centers in this country were run by local governments, he asked them how their proposed reforms were to be financed. Oh, they said, we have not thought about that. Nor, as it turned out, had the government. One reason for the failure to communicate between those concerned with local spending and those concerned with local revenues is that governments generally operate through a set of ‘silos’ (departments, ministries or agencies) that are responsible for different activities. Often, they do not work well with one another. It is only when the bills have to be paid, that is, when the revenue and expenditure sides of the budget must be fitted together, that the need to view government activity as a whole rather than a disparate set of parts has to be recognized. This is one reason why the Ministry of Finance so often plays a dominant role in economic policy. Finance matters – a lot; and it matters even more at the subnational government level where, as noted above, governments cannot print money. It seldom makes sense to think of reforming expenditure policy at any level of government without considering how it is to be financed. Similarly, it makes little sense to raise revenue without knowing how it going to be spent.3 The first step to bringing the two sides of the budget together at the subnational level is to get the totals right. A simple way to approach this issue is set out in the next section.

THE BENEFIT MODEL OF LOCAL GOVERNMENT FINANCE Subnational governments may be thought of in economic terms as enterprises engaged in providing a package of local public services within a given geographical area.4 This is both a considerable oversimplification and a useful starting point. Charles Tiebout (1956) started this line of thinking by treating localities as competing firms. However, in his model local governments sold only pure public goods enjoyed equally by all residents. In the real world, many goods and services provided by local governments are not ‘public’ at all (non-excludable and for joint consumption)

M4439-BAHL_9781786435293_t.indd 168

01/03/2018 11:08



Financing local and regional government ­169

but rather ‘private’ because individuals benefit, and in doing so reduce the benefits available to others – for example, by increasing congestion.5 To the extent publicly provided goods are privatized in use, they should be paid for by those who benefit directly from them. The first rule of local finance should thus be whenever possible, charge. It is both equitable and efficient for the direct recipients of services – whether residents, businesses or properties (whoever owns or occupies them) – to pay for what they get. It is also important that the correct (roughly, marginal cost) price – which is sometimes difficult to determine and seldom easy to implement – is charged, because only then will the right amounts and types of service be provided to the right people, that is, those willing to pay for them.6 It is not likely that all the financial needs of any local government can be met from user charges. Nonetheless, it is useful to think through what a strict benefit model of local finance might look like. The basic requirement for efficient and effective local government can be described simply in terms of the correspondence or matching principle. Expenditure responsibilities should be matched with revenue resources.7 Similarly, the ability to raise local revenue should be matched by as full local political accountability as possible. This requires not only effective local democracy in the form of a responsible and responsive local government, but also transparent provision of information about local revenues and expenditures (Bird, 2000a). Finally, the physical area within which benefits are received should be matched as closely as possible to the area within which those who pay for the services are located.8 When these requirements are adequately satisfied, the three groups relevant to local decision-making – those who decide what is to be done (accountability), those who benefit from what is done (beneficiaries) and those who pay for what is done – will correspond, thus ensuring the most economically efficient outcome.9 One implication of this approach is that localities should be constrained from ‘exporting’ tax burdens to those who do not benefit from local services (McLure, 1967; Bird 1993). For example, businesses should generally not be subject to higher effective property tax rates than residents, since such discriminatory taxation will undesirably bias business investment decisions and hence reduce national welfare.10 Another implication is that, as already mentioned, the most desirable form of local finance is to charge properly for services provided, both to current users (for example, for water and sewerage or waste removal) and to prospective future users (by borrowing to finance infrastructure and servicing the debt through taxes and charges on future users). In designing local revenue and expenditure systems close attention must also be paid to the institutional context (how local government accountability is established and monitored and how

M4439-BAHL_9781786435293_t.indd 169

01/03/2018 11:08

170

Fiscal decentralization and local finance in developing countries

intergovernmental transfers are designed and operated in the national political setting) along with such other factors as the area, population size and density, income level and administrative capacity of the locality. To apply a strict benefit approach in determining the appropriate level and structure of local revenue systems, three strong assumptions need to be satisfied: 1. The business of local governments should be solely to provide local services to residents and businesses (and usually also to some nonresident beneficiaries). All the additional activities that people seem sometimes to expect from their local governments are ignored, and it is assumed that who is to do what is clear to all. 2. Local governments are permitted to exercise their responsibilities freely. They have (at least potentially) access to sufficient resources and are not subject to detailed control of their expenditures by higher levels of government. 3. Local governments are not directly concerned with redistributive policy. Their actions may of course have distributive consequences, but presumably these consequences are taken fully into account when higher-level governments determine their own, over-riding redistributive policies. Although this last point is often contentious, economists have long argued that subnational governments should not focus unduly on vertical equity. The main reason is because, as Musgrave (1983) emphasized, since income redistribution is central to national fiscal policy, the primary level of government concerned with achieving distributional goals is inevitably the national government. Another reason is because pursuit of redistribution at the subnational level is unlikely to succeed. For example, a local government that taxes the rich and subsidizes the poor may drive away the former (and thus reduce its tax base) while attracting the latter (and thus increasing its expenditure needs). Many attempts to make local and regional government taxes more progressive often amount to little more than adopting some change that appears to do something to reduce the fiscal burden on those with low incomes and then declaring success. Examples are imposing progressive property taxes or establishing user charges that exempt or heavily subsidize certain classes of users. Such moves may seem attractive in principle. In practice, however, their main effects are often to complicate local fiscal systems by making administration more costly and difficult, with little if any effect on distribution, and possibly adverse long-term effects on investment and growth. Most taxes that can be effectively implemented at

M4439-BAHL_9781786435293_t.indd 170

01/03/2018 11:08



Financing local and regional government ­171

the local level are unlikely to have much redistributive impact. As a rule, they should be kept as simple as possible and viewed primarily as ways to raise revenue. Similarly, although a good argument can be made for limited user charge subsidies (‘lifeline’) with respect to such services as water and sewerage, all too often reducing user charge funding of public services tends to reduce rather than increase the access of the poor to such essential services by reducing the funds available to expand services and making it more attractive to expand services to the rich who pay rather than to the poor who do not (Bird and Miller 1989). Local and regional governments wishing to help their poorer residents should make more effort to spend well on properly designed and implemented public expenditures that will improve the quality of life for low-income families and worry less about financing their spending in a nominally progressive fashion. Richer countries with more administrative capacity have more options to attempt redistribution with regional and local taxes.11 Poor countries should generally avoid making taxes more difficult to administer by cluttering them up with reliefs intended to benefit particular groups (or with incentives aimed at inducing economic actors to behave in particular ways). Most such countries are short of revenue and have so much difficulty in securing more that they are ill-advised to waste much effort on the revenue side trying to achieve progressivity. This does not mean that they should feel free to impose blatantly regressive taxes and charges whose effects cannot easily be offset by compensating expenditure policies. In spite of the problems to which they may give rise, in some instances the lesser of two evils may therefore be to exempt a few key products from a general sales tax, or even to reduce a (otherwise meritorious) high excise tax on some product consumed widely by the poor.12 The best general rule, however, is to follow the KIS (keep it simple) principle of policy design when it comes to designing local taxes. Ideally, local governments should be fully accountable to local citizens for how they spend local resources. Like the other conditions listed earlier, however, this is seldom the case in low- and middle-income countries. Regional and local governments operate in many different institutional settings. Usually, they offer some (excludable) services that are consumed by specific persons, others that are consumed jointly by the community, and still others that spill over local boundaries. Sometimes they are owners of enterprises that compete for sales in the private sector. Frequently, they also act as agents delivering redistributive services financed by higher levels of government. Especially when there is some degree of democracy – and even sometimes when, as in China, there is not much – local governments are inevitably in the business of redistribution, even if their generally open local economies make these efforts largely ineffectual.

M4439-BAHL_9781786435293_t.indd 171

01/03/2018 11:08

172

Fiscal decentralization and local finance in developing countries

In all countries, central governments force or induce local governments to act in accordance with national policy objectives. In many, local governments have little discretion with respect to either the services they offer or how they pay for them, and many local services are in the end paid for by someone other than beneficiaries. For these and other reasons, the benefit approach to local government finance can never tell the full story. Nonetheless, this approach deserves more attention than it has received either in the literature or in practice because it provides a useful base case against which to judge the economic efficiency of local revenue systems. One aim of decentralization around the world has been to improve the efficiency, flexibility and responsiveness (and perhaps also the credibility) of government. An important lesson from studies evaluating the effects of fiscal decentralization on the provision of such services as health and education as well as its links with corruption, stability and growth is that what governments do and how well they do it are inseparably entangled with the question of how they are financed (Smoke, 2014). Strengthening the linkage between local expenditures and local revenues – what Breton (1996) called the ‘Wicksellian connection’ – cannot and should not be the only aim of those concerned with local government finance. But it does provide an important baseline in developing a decentralization program to achieve the objectives noted above, and should have a prominent place in the toolkit of local government reformers. How to Strengthen the Wicksellian Connection The Wicksellian approach to local finance sketched above, although it underlies the accepted model of intergovernmental tax assignment laid out in Musgrave (1983), has been all but forgotten in developing countries. From this perspective, the optimal way to design a local tax system is first to determine the desired size and nature of local expenditures and then to put in place a tax (and transfer) system that confronts local decisionmakers with incentives that will lead them to choose to finance precisely that package of expenditures. In practice, decisions on the two sides of the local budget are usually made independently, with both subnational government expenditures and revenues being heavily influenced by the priorities of the central authorities. One consequence is that accountability at the local level is often both confused and confusing. Another is that local governments are usually considerably less effective or efficient than they could and should be. In principle, there are three ways in which the Wicksellian connection between local services and revenues may be strengthened: (1) by changing the ‘package’ of local services; (2) by altering the ‘package’ of local

M4439-BAHL_9781786435293_t.indd 172

01/03/2018 11:08



Financing local and regional government ­173

revenues; and (3) by altering the way in which these two packages are tied together. For example, it is common to suggest that some local expenditure should be financed in whole or in part through user charges. However, there is less enthusiasm about the extent to which specific local revenues should be explicitly ‘earmarked’ to specific expenditures. Earmarking is almost as unpopular with experts in local finance as it is with experts in budgeting.13 Well-known economists like Nobel Prize winner William Vickrey (1963) and municipal finance expert Dick Netzer (1966) have argued that local property taxes may to some extent be thought of as a surrogate ‘user charge’ – a way in which residents (including businesses) who benefit from local services may pay for those services. Similar reasoning underlies the argument mentioned earlier for constraining local governments from imposing taxes whose burdens are exported to non-residents who do not benefit from the local services (Bird, 1993). Here, we combine these and other ideas, and sketch what a Wicksellian local finance system – one that takes seriously the desirability of tightly linking local taxes and local expenditures – might look like. The obvious starting point is the case for more and better use of local user charges. Financing local services through appropriate user fees provides not only essential funding but also, importantly, information on which services should be provided, in what quantity and quality, and to whom. Good user charges thus improve the efficiency with which scarce public resources are employed, giving people more of what they want (and are willing to pay for) instead of what someone else decides they should have. When people are not explicitly charged for consuming a service, the implied value of the last unit they use is effectively zero. When no charge is imposed for a service, more of it is likely to be (politically) demanded and (inefficiently) consumed than people would be willing to pay for if they had to face the full costs of providing the service. Underpricing – not charging people the full cost of providing a particular service – often results not only in over-consumption but also in inadequate maintenance of existing public capital and eventually still more ill-advised investment in new capital. For example, when subsidized roads become crowded, the political pressure to widen them becomes greater. Over-investment in underpriced facilities is sometimes condemned because corruption may arise in the contracting process. More importantly, however, even if every step in the process is carried out honestly, the end result is always that some scarce public resources are wasted building infrastructure that is not worth what it costs. Underpricing public services is thus the ‘black hole’ of local government finance in the sense that revenues go in but nothing of equal value to society comes out. Good user charges can avoid such waste. Although it is usually difficult to avoid completely the

M4439-BAHL_9781786435293_t.indd 173

01/03/2018 11:08

174

Fiscal decentralization and local finance in developing countries

adverse political economy factors that too often lead to establishing low and badly designed user charges, much more attention needs to be paid to this issue by those who want to see local communities get what they pay for and want.14 Beyond user charges, two additional basic principles for assigning revenues to local governments are suggested by the Wicksellian approach: ●●

●●

First, sufficient own-source revenues should be assigned to the richest local governments to enable them to finance from their own resources all locally provided services that primarily benefit residents.15 Second, to the extent possible, local government revenues should be collected mainly from those living within the boundaries of the local government area, preferably in relation to the perceived benefits they receive from local services. Ideally, revenues from other sources (including local business activities and non-residents) should similarly match the benefits they receive from local services (Oates, 1996).

When services like water and sewerage connections are provided to specific locations it may be better to pay for at least the costs of accessing such services through charges related to relevant characteristics of properties (such as size of lot, frontage or building height) or to property values. Other services (or components of services) such as arterial streets, utility lines and public transit, as well as major parks and recreation facilities, may also be area-specific in the sense of being most accessible to those nearby. Since the value people attach to such services should be reflected in property values to a considerable extent, a suitable form of financing may be a special assessment based on property values. Still other services may provide city-wide or even region-wide benefits: again, such benefits should affect property values, and an appropriate form of financing would again appear to be a property tax. Another suitable source of finance may in some instances be a development charge or other form of (pre- or post-) value capture system of the sort that is used successfully in some Latin American cities.16 Since the cost and benefits of providing local services may differ widely from business to business, property taxes are unlikely to be the best way to finance business-related services, not least because their employees (who enjoy lunch in the local park), their customers (who benefit from locally provided business inputs like streets) and their owners (who similarly benefit from cost-reducing local services) are not always residents of the locality. In addition to business-related services that may indirectly benefit non-residents, non-residents may also benefit from locally provided ­services

M4439-BAHL_9781786435293_t.indd 174

01/03/2018 11:08



Financing local and regional government ­175

when they visit a locality as commuters (working but not living there), as tourists (presumably enjoying locally provided amenities) or simply as visitors coming to shop, to dine or for some form of entertainment or recreation. While some of the cost of providing services to non-residents may be recouped through user charges and taxes on business, a case can also be made for imposing some additional local taxes or charges on nonresidents.17 However, as with all business taxes, from the perspective of efficiency it is also important to be sure that the taxes and charges imposed on such non-voting beneficiaries are not excessive. In rich countries and poor countries alike, local politicians and local voters generally find it considerably less painful to tax non-residents than themselves and their neighbors.18 A key question to be asked about all local revenues from the benefit perspective thus relates to the economically undesirable but politically attractive possibility of tax exporting – the shifting of tax burdens to non-residents such as: (1) commuters (non-resident labour); (2) tourists and other visitors (non-resident consumers); (3) non-resident owners of local businesses (external capital); and (4) non-resident consumers of city exports (e.g. financial services). Since non-residents may gain from the joy of living next to the amenities of a city – even if they don’t use them (‘option demand’) – a related issue is the extent to which tax exporting matches possible offsetting benefits from local services. For instance, local sales taxes fall on the purchases of visitors and locals alike; taxes on hotels and entertainment are often aimed mainly at visitors but may also impact residents, while a local payroll tax falls on non-resident commuters as well as residents. As Musgrave (1983) emphasized, the assignment of taxes should follow the assignment of expenditures in a decentralized system. This linkage is crucial in three respects. First, as discussed in the next section, it tells us how large the subnational government tax base must be to achieve the desired level of fiscal balance. Second, it also provides some useful guidance on how to divide the local revenue structure between taxes, transfers, user charges and borrowing. Third, it provides some guidance, as just discussed, to the appropriate composition of the subnational tax base. As Bahl and Linn (1983, 1992) discuss, the appropriate mix of revenue sources for any subnational government depends on the expenditure responsibilities assigned to that government. Consider the following four categories of spending: 1. For publicly provided goods and services, where the benefits accrue to individuals within a jurisdiction and where the exclusion principle can be applied in pricing, user charges are the most efficient financing instrument. This is a particularly relevant argument for public utilities

M4439-BAHL_9781786435293_t.indd 175

01/03/2018 11:08

176

Fiscal decentralization and local finance in developing countries

such as water supply, sewerage, power and telephones, but also for public transit and housing.19 These services may involve some external benefits, but most of the benefits are likely to be local (or regional) in nature and can therefore be handled by subsidies financed from other subnational revenue sources. 2. Other local government services, such as general local administration, traffic control, road maintenance, street lighting, security, primary schools, local clinics and parks and recreation are local public goods whose primary benefits accrue to the local population. The same may be said at the regional level of secondary schools, universities, mental hospitals, trunk roads, bridges and the like. Since the exclusion principle in pricing cannot usually be feasibly (or at least politically) applied to most such services, they are most appropriately financed by taxes whose burden is local (regional) so that “the electorate is confronted with the true opportunity cost involved” (Musgrave and Musgrave, 1976, p. 665). Those who bear the tax burden should experience the benefits of the expenditures financed by that tax. Following this rule not only supports the assignment of property taxes to local governments, as is common in many countries, but also the assignment of broad-based consumption and income taxes to metropolitan and regional governments (see Chapter 8). 3. For services in which substantial spillovers to neighboring jurisdictions commonly occur – such as health, higher education and certain types of infrastructure expenditures – provincial or national intergovernmental transfers should obviously contribute to financing (see Chapter 7). Full local financing would lead to under-provision of these services from a regional or national perspective, and full financing from transfers would not recognize local benefits. In principle, the portion of these services that provide local, regional or national benefits should determine the share that needs to be financed by local (regional) taxes, although the allocation of benefits are often difficult to determine, and few countries do so with any rigor. 4. Finally, borrowing is an appropriate arrangement for financing capital outlays that have a long service life, e.g., public utilities or mass transit (see Chapter 4). Again, efficiency considerations suggest that the debt should be serviced from local taxes and user charges if the infrastructure benefit zone is local, and from higher-level government subsidies to the extent the benefit zone is regional or national. Table 5.1 provides an illustrative summary of the appropriate financing of some major subnational government expenditure categories by types of current revenue.

M4439-BAHL_9781786435293_t.indd 176

01/03/2018 11:08



Financing local and regional government ­177

Table 5.1  Appropriate local revenue source by category of expenditure Service Water supply Sewerage Drainage Electricity Telephones Markets and abattoirs Housing Land development Streets Highways Public transit Refuse disposal Refuse collection Parks and recreation Fire protection Law enforcement General administration Education Health Welfare

Local taxes

User charges

Transfers**

Borrowing

S S P – – S S – P S S S P P P P P P S S

P* P* P* P P P P P* S* P* P P P – – – – S S –

– – – – – – S – – P – S – – – – – P P P

A A A A A (A) A A A A A A (A) (A) (A) – – (A) (A) –

Notes: P 5 primary source; S 5 secondary source. A 5 Borrowing appropriate for major capital expenditures; (A) 5 borrowing is appropriate for capital expenditures but likely to account for a small share of spending. * 5 Development charges (special assessments, valorization charges, etc.) are appropriate where benefits are spatially well defined within a jurisdiction. ** Transfers may be from a regional or central government. Source:  Adapted from Bahl and Linn (1983).

THE BOTTOM LINE: THE HARD BUDGET CONSTRAINT Both experience and theory suggest that if local governments face a ‘soft budget constraint’ (Kornai, 1986) – that is, their decisions are influenced by the expectation that any shortfalls in revenues will be offset by additional resources provided by others (or by successful subterfuge) – they are unlikely to operate efficiently.20 For decentralization to produce the

M4439-BAHL_9781786435293_t.indd 177

01/03/2018 11:08

178

Fiscal decentralization and local finance in developing countries

best results subnational governments must be required to balance their budgets in such a way that they face a hard budget constraint at the margin: that is, when they decide to spend an additional dollar on an activity they must simultaneously determine how they are going to raise that dollar (McLure, 2000). While this is simply another way of stating the importance of internalizing all benefits and costs to local decisionmakers if they are to make the right decisions, it is useful to consider the question a bit more formally in order to define the budget constraint and to set out the framework for the later discussion of the design and role of intergovernmental transfers. There are two basic points here. The first is that if local governments can shift some of the costs of their decisions to others through such means as exporting taxes to non-residents, receiving transfers or subsidies from other levels of government (including loans where repayment is forgiven) or by creating a fiscal crisis and being bailed out (Inman, 2003), they will have little incentive to spend within their means. Preventing such inefficiencies requires an appropriate legal framework including such rules as, e.g.: financing of current spending with taxes and charges; limiting central transfers to financing demonstrable spillovers and (perhaps) equalization grants; and monitoring local accounting and borrowing, preferably both by higher level governments and through relatively competitive markets (Bird and Tassonyi, 2003). Finally, if all else fails, appropriate provisions should be made to handle local bankruptcy. The second point is that local government leaders do not to like to tell their constituents that expenditure demands cannot be satisfied without a tax increase. Luckily – from their perspective, if not society’s – local decision-makers can often find ways to deliver services now while shifting the payment burden to future budgets and future leaders. ‘Fiscal mischief’ (as this practice may be labeled) obviously does not square with the ideal of a hard budget constraint. But it fits well with the high time discount rate that affects the decisions of most elected (and often appointed) political leaders. We return to this matter below. The concept of a hard budget constraint is somewhat difficult to grasp, in part because there are many different concepts of a deficit, all of which are to some extent relevant. As the formal discussion below shows, there is a current account budget deficit, a capital account budget deficit, a total budget deficit and, finally, a more inclusive concept that accounts for future claims on the resources of the subnational government. For each of these accounts there is a concept of balance, and all need to fit together to tell the complete story about the financial condition of the subnational government.21

M4439-BAHL_9781786435293_t.indd 178

01/03/2018 11:08



Financing local and regional government ­179

Defining a Hard Budget Constraint It is helpful to define the hard budget constraint for subnational governments more precisely. The basic condition is that the amount in any period that any local government can spend (G) is constrained by the amount of taxes (T) and other local revenue (OR) such as fees and charges that it plans to raise in that period, plus the amount it expects to receive in transfers (TR) plus any amount that it borrows (B), plus any planned draws from its accumulated (unrestricted) cash balances (SC): G 5 T + OR + TR + B + SC

(5.1)

If borrowing is only permitted to finance capital spending (as should be the case), then current spending (GC), including debt service, must be financed from current revenues (CR), which include current intergovernmental transfers (TRC), so that: GC ≤ CR 5 T + OR + TRC + SC

(5.2)

The budget balance rule commonly imposed on local governments requires that GC (current spending) cannot exceed CR (current revenues).22 But a local government can have current savings (SC 5 CR – GC) which may be used either to finance investment spending or be accumulated in the form of what may be thought of as a precautionary balance – in effect, a contingency fund to support future spending.23 There is a separate (annual) hard budget constraint for the capital account. Some transfers may be earmarked for investment purposes only (TRK 5 TR – TRC). The amount of local borrowing (B) required to finance the amount of investment spending in the current period (GK) is thus:

B 5 GK –TRK – SA

(5.3)

GK ≤ TRK + B + SA

(5.4)

or

where B is the annual amount of new borrowing that the local government is able to do for purposes of financing capital expenditures, and SA is the amount received from the sale or lease of assets.24 There are thus two hard budget constraints for subnational governments. The first is that recurrent expenditures must not exceed recurrent revenues (Equation 5.2). The second is that capital expenditures (expenditures for

M4439-BAHL_9781786435293_t.indd 179

01/03/2018 11:08

180

Fiscal decentralization and local finance in developing countries

the purchase of long-lived assets) must not exceed the revenue available from designated capital financing sources (Equation 5.4). To meet these two tests of a hard budget constraint, subnational governments are bound tightly by proper definitions of what is a recurrent expenditure and what is a capital expenditure. These two components of the hard budget constraint are shown as rows 3 and 6 in Table 5.2. It is often argued that local and regional governments financed largely from transfers inevitably face a soft budget constraint. This is not correct. A hard budget constraint can be imposed on subnational governments even if they are fully financed from intergovernmental transfers, provided the transfers are correctly designed (Bird and Smart, 2002). If, however, transfers are frequently subject to discretionary changes by the central government, or if they are specifically intended to finance deficits, spending is unlikely to be efficient owing to the uncertainty of subnational budgetary planning and the perverse incentives. In principle, even if transfers are very large, the outcome is likely to be considerably better if the transfer is an entitlement, i.e., a guaranteed transfer of a predetermined amount of revenue from central taxes. We discuss this more fully in Chapter 7, but we should note here (and many countries have demonstrated) that even guarantees are not ironclad when it comes to intergovernmental transfers. Local government deficits are not always locally caused, and they are not always locally resolved. With respect to the latter, some expenditures may be reassigned to higher-level governments, although this can be politically difficult and it reduces local control over spending. Alternatively, more taxing power could be assigned to the local government. But this too is seldom easy to accomplish and, even if it is, local governments may be unwilling to impose new taxes. Sometimes, especially within metropolitan areas, responsibility for a function may be shifted to a special district or a public entity that may be able to impose increased user charges. Finally, of course, higher-level governments may simply increase the level of intergovernmental transfers to cover local revenue shortfalls.25 Budgetary experts are not of one mind about this subject, but here we treat the capital account as a separate though linked entity. The linkages are important.26 As outlined above, the capital budget should also be subject to a hard budget constraint. If infrastructure is financed with borrowing it is efficient to repay the loan over the life of the asset, provided the loan and the asset life match. The test of fiscal health is then whether the surplus of revenues over other current expenditure commitments is sufficient to enable timely repayment of debt service as well as cover annual operation and maintenance (O&M) costs. Additions to capital revenues may also be financed by capital grants (grants earmarked for capital purposes) or from accumulated savings (cash balances and assets that may be converted to cash).

M4439-BAHL_9781786435293_t.indd 180

01/03/2018 11:08



Financing local and regional government ­181

Table 5.2  How to soften the budget constraint Budget line

Comments

1. Recurrent revenues 1a. Tax and non-tax revenues

Revenues received regularly each year Includes user charges and delinquent tax payments Intergovernmental transfers regularly available to local governments for current expenditure purposes, but not including year-end gap-filling grants Accumulated (unrestricted) balance from previous years All expenditures regularly made each year, excluding any capital expenditures Wages and salaries, plus normal operating and maintenance expenditures Includes transfer payments to individuals, transfers to other governments for current services, transfers to government enterprises Repayment of principal and interest due Regular contributions to pension funds, health care funds, etc. Row 1 minus Row 2 Long-lived capital assets acquired; excludes any current expenditures Revenues available to fund capital expenditures Conditional grants that must be spent for capital purposes Accumulated from current account surpluses Buildings, land, businesses, etc. Long-term debt increases Row 4 minus Row 5



1b. Recurrent transfers



1c. Cash balances

2. Current expenditures

2a. Operating expenditures



2b. Subsidies and other



2c. Debt service 2d. Contributions

3. Total current surplus/deficit 4. Capital expenditures 5. Capital budget: sources of funds

5a.  Capital grants from higherlevel governments 5b. Cash balances

5c. Sales or lease of assets 5d. Borrowing 6. Capital budget surplus/deficit 7.  How to get around the hard budget constraint 7a. Deferred payments



7b. Unfunded contributions

M4439-BAHL_9781786435293_t.indd 181

Push the payment of employees or suppliers to future fiscal years; fail to pay utility bills owed; fail to pay other governments for services provided Delinquent or inadequate contribution to pension fund, health care funds, etc. or ‘raids’ on balances in pension funds, etc.

01/03/2018 11:08

182

Fiscal decentralization and local finance in developing countries

Table 5.2  (continued) Budget line

7c. Reclassify expenditures



7d.  Bailout by higher-level government 7e. Debt rollovers 7f.  Deficit covered by short-term loans 7g. Backdoor approaches

8.

Exhibit: contingent liabilities 8a. Underfunded pensions, etc. 8b. Loan guarantees 8c.  Underfunded debt repayment schedules

Comments Shift recurrent expenditures to capital account Debt repayments forgiven or reduced Debt service paid by issuing new debt Bank overdraft to cover current deficit Create extra-budgetary accounts and/or levy unauthorized taxes or charges Other claims on budget revenues

Finally, we note that there often are local government liabilities that are not included with the annual current and capital budgets. One such liability is sometimes called ‘floating debt’ (unpaid bills from purchases in prior accounting periods).27 Others are contingent liabilities (for example, loan guarantees in support of state enterprises) and unfunded liabilities, such as inadequate (or unpaid) contributions to pensions or health insurance funds. All these items, although rarely mentioned in financial reports, are in principle as much claims against subnational government resources as any other debt. There might be some exceptions. When such liabilities are mandated by higher-level governments or are the result of some unforeseeable disaster that has devastated the local economy, central government rescue may be warranted. But subnational governments that take on such liabilities of their own free will should do so from a strong current account position, and preferably only after having set aside a solid precautionary fund. If they misjudge, then they should have to bear the consequences. Unfortunately, all too often it is the governments with the thinnest margins of safety that carry the largest contingent claims. It is often difficult to discover the real situation of local finances in many countries, but it is likely reasonable to underline that neither the size of the real deficit nor the extent to which it is due to imprudent local government policies can usually be determined simply by looking at the bottom line of the local government budget. Such budgets do not reveal such common problem areas as: overcommitting on employee salaries; d ­ eferring ­maintenance of capital assets; failing to raise taxes or user charges rates to

M4439-BAHL_9781786435293_t.indd 182

01/03/2018 11:08



Financing local and regional government ­183

at least the average level (for localities with similar characteristics); providing tax exemptions that are too liberal; and failing to collect delinquent taxes. They also do not help one understand the extent to which such problems may be beyond the control of the local government, for example, because of restraints and conditions imposed by higher-level governments or local calamities such as floods or storm damage. Fiscal Mischief As experience in many developing countries demonstrates, subnational governments often search for a way around central government restrictions either to do what they are supposed (or ordered) to do, or perhaps simply what they want to do. The result is often what we call ‘fiscal mischief.’ Although this label is pejorative, not all such mischief is bad policy, and not all of it is attributable to bad behavior by local governments.28 Still, ‘mischief’ is not a bad way to describe how local governments often manage to escape the nominally hard budget constraint they confront. For example: ●●

●●

●●

Local leaders may disobey hard budget constraint rules because following the rules complicates their political position. Raising local taxes (or user charges), or even enforcing them more rigorously, is never popular. The opposite can be true for local spending, especially in election years. Local voters may go along with the mischief because they do not want to raise local taxes either, because they believe they will be rescued ‘from above,’ or because they just do not know what is going on. Local governments often find themselves between a rock and a hard place. They may have very limited own revenue-raising powers, an expenditure budget stuffed full of items beyond their control (many of which were imposed by higher level governments), and intergovernmental transfers that are inadequate to cover the gap. Breaking the rules may be the only way to escape budget rigidity. Finally, because bad things happen even to good people, even the best-run local government may be caught short by a natural disaster, conflict, the fallout from poor central macroeconomic decisions, or an unexpected large cut in central transfers. In such cases, central government monitors of the hard budget constraint might be right to look the other way. Even when local governments disobey budget rules with less cause, higher-level governments sometimes let it go because it is expedient to avoid the political cost of cracking down in, say, an election year, or a recession.

M4439-BAHL_9781786435293_t.indd 183

01/03/2018 11:08

184

Fiscal decentralization and local finance in developing countries

Panel 7 of Table 5.2 lists several ways local governments may cope with deficit problems by softening the budget constraint. A current account deficit may be dealt with simply by securing a gap-filling transfer (row 7d), by borrowing (7f) or by reclassifying expenditure from current to capital in order to make it eligible for debt finance (7c). While always popular with local governments, the first of these solutions is particularly undesirable. Rao and Chelliah (1991) noted long ago that ‘fiscal dentistry’ (a gap-filling transfer) sends exactly the wrong signal to subnational g­ overnments – ‘don’t worry about budget balance because you will be bailed out.’ However, for the most part, India’s states (and those who lend to them) continue to count on central government generosity to resolve their longstanding deficits. The second and third solutions mentioned above violate the golden rule of borrowing only for investment purposes (see Chapter 4), and also increase debt service costs and hence the likelihood of further budget imbalance in subsequent years. Two bad practices are to defer payments to creditors (row 7a) or to underfund contributions to pension or health care accounts (7b). Even the best of these ways of financing a current account deficit is not sustainable, and none address the fundamental problem of an imbalance between current revenues and current spending. Fiscal discipline requires painful decisions to cut expenditures (and presumably service levels) or raise taxes. At the root of all these problems is the failure of the central government to address the fundamental fiscal imbalance inherent in its intergovernmental fiscal structure. The hard budget constraint rule for the capital budget may also be ­softened, often in similar ways. For example, the central government may simply forgive debts owed by subnational governments (row 7d), or such governments may get out of trouble by issuing new debt and rolling over the proceeds to repay old debt (7e). Again, neither of these approaches deals with the basic structural problems underlying the budget shortfall. All else failing, local governments can get out of their budgetary fix by levying taxes and charges (and perhaps also undertaking some related expenditures) that are outside their legal powers (row 7g). A clear example happened in China in the 1990s when local governments dealt with revenue shortfalls by imposing ad hoc extra budgetary levies. Although the central government eventually acted to abolish these levies (Bahl, 1999), the problem remained serious a decade later when some such levies had even become what were referred to as ‘extraextrabudgetary funds’ (Wong and Bird, 2008). A similar problem occurred in Indonesia following the bigbang decentralization, when local governments took liberties in levying ‘inappropriate’ new taxes (Sidik and Kadjatmiko, 2004).

M4439-BAHL_9781786435293_t.indd 184

01/03/2018 11:08



Financing local and regional government ­185

How Large Should the Margin of Safety Be? It is difficult to be definitive about how large a margin should be maintained to protect a subnational government from a current account budget deficit. Rules of thumb about this – for example, that local governments should carry balances of 5 percent of budget expenditures – are not very helpful because situations vary so greatly from one local government to another. For instance, more revenue discretion at the margin is required: if the transfer system is characterized by uncertainty; if the central government’s own revenues are unstable; if higher-level governments have a tradition of imposing unfunded mandates on lower levels (for example, requirements for increased public sector wages and so on); or if subnational government borrowing practice is not adequately controlled by fiscal rules. Each of these problems is found in some developing countries – and a few have all of them. In addition to adequate margins for local governments to control current account deficits, cut spending or raise taxes, another key condition necessary for fiscal discipline is a well-structured grant system in which the amount that each local government may expect in any given year is fixed before its spending budget is set. If grants are negotiable (or, as in India, subsidized public lending is readily available), local governments will be tempted to overspend in the expectation that their deficits will be covered. Alternatively, if annual grants are ad hoc (and hence perhaps unduly soft in principle), the downside is that they may be among the first items to be reduced when central revenues are tight, as happened in many transitional countries in Eastern and Central Europe in the 1990s. In these circumstances, even healthy local governments are well advised to allow for a larger discretionary margin (contingency fund) than they otherwise would. Even if both the conditions just mentioned – adequate budgetary flexibility and a sound transfer system (see Chapter 7) – are satisfied, local leaders may still not be held politically accountable for failing to balance the budget. Like governments at all levels, subnational governments usually try to shift the blame for unpleasant action to factors beyond their control, such as the actions (or inaction) of other governments. Nonetheless, with more discretionary subnational tax power, more explicitly visible subnational government taxes and a well-structured transfer system, residents will have more opportunity to figure out what is going on – and, provided the system provides some clear lines of accountability, more incentive to do something about it. All else being the same, such local governments will have less need for a larger rainy day fund.

M4439-BAHL_9781786435293_t.indd 185

01/03/2018 11:08

186

Fiscal decentralization and local finance in developing countries

Hard Budget Constraints and Public Policy Governments at all levels largely determine which taxes should be imposed in terms of a political rather than an economic calculus (Hettich and Winer, 1999). They are influenced less by effects on resource allocation than by perceptions about what tax changes may mean for their remaining in power. Since the political cost–benefit calculus almost invariably leads to the conclusion that it is relatively cheaper to beg (or demand) funding from others than to tax one’s constituents – which is one reason why local governments do not always use additional taxing power even when they have it – the problem reduces to the same issue discussed above with respect to the hard budget constraint: how can one best make ‘hardness’ credible in terms of creating incentives for local governments to behave properly without being unduly cruel to those facing difficult conditions? At the local level, the usual alternative to bankruptcy is to replace local government with an appointed financial regulator (as in the US cases of New York and Detroit, for example).29 In most low-income and many middle-income countries subnational governments have limited revenue discretion, usually only for such politically contentious levies as taxes on land or property and user charges. Because of the lack of comparable data on subnational government fiscal health, it is difficult to say much about the extent to which such limited revenue autonomy is a contributory factor to their running current account deficits. Of course, similar problems occur in rich countries also (Bahl, 2011a). In Japan in the early 2000s, for example, local governments were incurring a financing gap which was addressed by central government transfers and subsidies in one form or another. A so-called ‘trinity reform’ was introduced to replace much of the grant funding with local government taxation authority (Ikawa, 2008). This reform was only a partial success because it put the greatest pressure on the budgets of those local governments with weaker financial capacity. German state governments (Länder) regularly run current deficits and finance them by both shortand long-term borrowing (Rodden, 2003b). Owing to the equalizing nature of intergovernmental fiscal transfers, the poorer states have little incentive to restructure their fiscal system to find sustainable budget balance (Spahn and Föttinger, 1997). Similar problems could occur in Canada owing to its equalization system; however, to date they have largely been controlled by shifting most transfers to straight per capita transfers (Smart and Bird, 2010) and by forcing provincial borrowers to be subject to the requirements of an internationally competitive capital market (Bird and Tassonyi, 2003). Central governments in many high-income countries have been willing to assign important expenditure responsibilities and autonomy to

M4439-BAHL_9781786435293_t.indd 186

01/03/2018 11:08



Financing local and regional government ­187

provincial and local governments, to share central government revenues with them, to give them significant taxing powers, and to give them independence in borrowing to finance capital projects. In a few middle-income countries, such as India and Brazil, state governments (and to a lesser extent local governments) also have some significant taxing power. In most low- and middle-income countries, however, broad-based and revenueproductive taxes remain off limits for lower level governments. As a result, most regional and local governments in such countries face large vertical imbalances and are unable to finance from their own resources even those activities for which they are primarily responsible. If the assigned tax bases cannot yield adequate revenues at reasonable rates, the result may be both deficient public services and a haphazard gap-filling strategy that includes the imposition of a variety of undesirable and distortionary fees, levies and informal charges. A significant amount of revenue-raising discretion at the local level is necessary if local governments in developing countries are to bring discipline to their fiscal decision-making process. Inman (2009) suggests two promising measures are establishing a rainy day fund at the local level, and providing for central transfers to address emergency situations once the contingency fund is exhausted. We agree; but still there is the question of how the local fund will be capitalized if not with local taxes and charges, and of how local political leaders will be forced to take some responsibility for the deficits that they played a role in creating.

THE EVOLVING ECONOMIC THEORY OF TAX ASSIGNMENT The traditional starting point for thinking about normative tax assignment is Musgrave’s (1983) multi-level budget framework, which assigns the stabilization and distribution functions to the central government, leaving only part of the responsibility for the allocation function to subnational governments.30 This division of responsibility provides some, but not much, guidance about the placement of various instruments of taxation at the central, regional and local government levels.31 Progressive taxes with a distributional goal and taxes containing automatic stabilizers would be left with the central government. Subnational government taxes, which are seen essentially as charges for services rendered by those governments along the lines discussed earlier, are to be financed mainly by taxes levied on immobile bases. In reality, subnational governments are inevitably involved both in macroeconomic policy and in shaping the distribution of real incomes

M4439-BAHL_9781786435293_t.indd 187

01/03/2018 11:08

188

Fiscal decentralization and local finance in developing countries

(Bahl and Linn, 1992). As the rate of urbanization has increased in most countries in recent decades, such traditional local functions as utilities, sewerage and drainage, and local transportation have become increasingly important not only to local residents and businesses but also to national economic growth (Frank and Martinez-Vazquez, 2016; Glaeser and Joshi Ghani, 2015). The benefits and costs of local public services spill over existing jurisdictional boundaries, especially in countries in which subnational governments are responsible for such important components of investment in human capital as education and health. A broader assessment of tax assignment than that provided by the conventional model is necessary to accommodate these realities. The share of subnational taxes increased only by a modest amount in both developed and developing countries between the 1970s and the 2000s (Bahl and Wallace, 2005; Bahl, 2014). Some developing and middle-income countries increased their level of subnational government taxation, but most did not. Regional and local governments in developing countries raise about 2.4 percent of GDP in taxes, which is about one-third the rate in OECD countries. Some richer countries in this group allowed their subnational governments to move into the income tax and consumption tax fields (Bahl, 2011a). A few countries in Latin America (Argentina, Brazil, Colombia) also assigned some power to tax consumption to subnational governments, with the same result. But most developing countries held to the traditional pattern, and continued to expect subnational governments to get their part of the assigned job done with user charges, land and property taxes, and minor levies. However, the job assigned to these governments has changed dramatically with urbanization and globalization. Some years ago, Brennan and Buchanan (1980) emphasized the importance of competition among subnational governments as a way of controlling the size of governments. If subnational governments taxed mobile factors (consumption, income) they ran the risk of driving away jobs and capital. At the same time, however, the threat of losing a tax base would provide a strong incentive to deliver services in a more efficient way, with one result perhaps being lower tax rates and smaller governments. Subsequently, Oates (1996) suggested that if local taxes were viewed as benefit levies, then they should also be imposed on non-resident beneficiaries, which would generally require taxing mobile factors to some extent. If non-residents come to town to work or shop or operate a business, tax them! But subnational governments are now called on to compete in terms of both tax rates and public service levels. Particularly in the bigger cities, as we discuss in Chapter 8, the problem they face is not just to deliver basic services at low cost but also to deliver more, higher-quality and more

M4439-BAHL_9781786435293_t.indd 188

01/03/2018 11:08



Financing local and regional government ­189

expensive services. Cities compete not simply in terms of their tax rates but also their service package. Another caveat to the tax competition argument is that factor mobility is less of a constraint where the taxing region is larger. Again, larger local governments such as metropolitan regions and provinces may face different problems and have more choices than smaller urban local governments or rural localities. In addition, one must always remember that those who make political decisions are usually more concerned about how their decisions on taxes will affect their retention of power than about the effects on resource allocation (Hettich and Winer, 1999). Shifting tax burdens to non-residents is always and everywhere so attractive an option for local politicians that there is good reason to be very careful about giving them too free a hand to do so. The importance of this factor has begun to be realized to a limited extent in the so-called ‘second-generation’ approach to fiscal federalism (Weingast, 2009; Lockwood, 2006), which moves away from the assumption of a median voter supported by a benevolent dictator to self-interested politicians who pursue their own objectives. However, as Oates (2008, pp. 329–30) notes: While the form of the analysis has new elements, the nature of the problem remains essentially the same: the issue is one of a tradeoff between the capacity of a centralized solution to provide ‘coordination’ of local outputs (i.e., internalize spillover effects) and the ability of a decentralized system to tailor outcomes to the preferences . . . of the local jurisdiction.

It is thus not surprising that there is little difference between how the theoretical median voter of the traditional approach and the theoretical politician of the second-generation approach would see things. Votemaximizing politicians would like to tax mobile factors to export burdens to non-beneficiaries. But so would any median voter, to reduce the taxprice paid for local services. A self-interested politician would shy away from such unpopular taxes as the annual property tax and embrace those whose ill-effects are less transparent, such as taxes levied only on property sales or mainly on non-resident businesses. Again, however, so would most residents in any jurisdiction. Similarly, getting more intergovernmental transfers is always a vote winner: who does not like paying less in taxes? Increased factor mobility and to some extent the spread of democracy in developing countries have stimulated the demand for fiscal decentralization (Bahl, 2008; Stegarescu, 2009). These changes have also altered our view about how decentralization should be financed. On the expenditure side, the decentralization theorem is straightforward: push all expenditure responsibility down to the lowest level that is consistent with efficiency considerations. On the revenue side, however, it is more difficult to state a

M4439-BAHL_9781786435293_t.indd 189

01/03/2018 11:08

190

Fiscal decentralization and local finance in developing countries

corresponding rule. Perhaps the best we can do is to posit much the same rule but in a more qualified way: all taxes should be assigned to subnational governments to the extent required to finance the services assigned to subnational government unless macroeconomic, redistribution or efficiency considerations dictate otherwise. As we remarked earlier, in Chapter 2, it is important to measure what we can as well as we can. However, the fact that we may be able to measure something does not imply it is more important than something that we cannot measure as well or at all. The classical federalism principle of ‘every tub stands on its own bottom’ – as embodied, for example, in constitutions like India’s – says each level of government should have its own tax base. The economics literature on fiscal externalities sees potentially high welfare losses arising from tax-base sharing (Dahlby, 1996; Keen, 1998; Keen and Kotsogiannis, 2004). The efficiency losses considered in this analysis are those related to distortions arising from undue exporting of tax burdens to non-beneficiaries that lowers the tax price of services to residents of the taxing jurisdiction. But such losses may be offset in whole or part if the resulting increased autonomy of the local government in deciding on its budget expenditures and its increased tax discretion at the margin sufficiently improves trust (social capital) between government and society. Viewed from this perspective, perhaps a lesson emerging from even the limited extent to which the second-generation approach to tax assignment incorporates the political economy dimension, may be that more – perhaps most – tax bases should be shared to some extent between levels of government. Whether one agrees with this conclusion or not, because it is usually costly to undo well-entrenched although flawed tax assignments, it is important to lay out some basic principles in structuring subnational government taxes and to note the perils of ignoring these principles. First, however, we should make clear what we understand to be a decentralized or subnational tax. There are several ways that a ‘local tax’ might be defined: ●● ●● ●● ●● ●● ●● ●●

A tax that regional or local governments have the power to decide whether to impose or not. A tax for which they determine the base. A tax whose rate they establish. A tax for which they determine the liability of particular taxpayers. A tax that they collect and enforce. A tax whose revenue accrues to them. A tax for which certain combinations of the above conditions exist.

As Blöchliger and King (2006) discuss in detail, there are many possible ways to mix and match these characteristics. They distinguish 13 distinct

M4439-BAHL_9781786435293_t.indd 190

01/03/2018 11:08



Financing local and regional government ­191

combinations in the 30 OECD countries, taking into account only the degree of discretion over tax policy decisions, and not such questions as whether and how more than one level of government may control some aspects of administration. The extent to which a government controls its own revenue sources is critical. As Martinez-Vazquez et al. (2006b, p. 21) note: “If fiscal decentralization is to be a reality, subnational governments must control their own sources of revenue. Subnational governments that lack independent sources of revenue can never truly enjoy fiscal autonomy, because they may be – and probably are – under the financial thumb of the central government.” Simply having a constitutionally fixed share of central revenues may seem to guarantee autonomy in this sense. But spending autonomy financed from on high is much less likely to be spent as local citizens would wish than if they had to find the funds from their own pockets. History and theory suggest strongly that for local governments to be autonomous they must be responsible to local citizens for how they raise as well as spend revenues. Democratic local elections and transparent reporting to citizens about local spending patterns are two obvious ways to move towards this goal (Bird, 2000a). But even the most democratic local government is likely to spend more responsibly if it also bears some responsibility for raising revenue by imposing taxes on residents in a visible and accountable way.32 The most important factor ensuring that subnational governments are accountable to their citizens is probably to make them clearly and visibly responsible for determining tax rates. The tax rate is for most people the most visible and understandable characteristic of any tax (McLure, 2000). The more power regional and local governments have in terms of collecting revenue – choosing which taxes to impose, how the tax base is defined, and assessing and collecting the tax – the greater their fiscal autonomy. But without the ability to establish and alter tax rates, even if only within some limits, the transparency and accountability of the local revenue system is likely to fall short of what is needed to support the fundamental economic case for fiscal decentralization. Provided regional and local governments can meaningfully establish tax rates, as we discuss later they do not need to administer taxes themselves provided the (presumably higher-level) government has a vested interest in administering the tax well. If the local government sets a reasonable nominal tax rate and the higher-level government administers the tax badly, the result may be a low effective tax rate. Some subnational taxes should be administered locally because it is efficient to do so and because doing so makes the ultimate accountability clear. But fiscal and administrative decentralization, while related, are distinct: it is possible to have much of one with little or none of the other, as we discuss further below.

M4439-BAHL_9781786435293_t.indd 191

01/03/2018 11:08

192

Fiscal decentralization and local finance in developing countries

These matters are resolved very differently in different countries. In China, Russia and many of the former Soviet bloc countries a subnational government ‘tax’ is essentially a central government levy whose revenues are fully assigned to the lower-level governments. In contrast, in India and Brazil a subnational tax is one for which lower-level governments have some degree of discretion in setting the tax rate and/or base. We are more in tune with the latter position: a necessary condition for a real subnational tax is that the regional or local government can set the nominal tax rate. But four quite different approaches to subnational taxation exist around the world, each with different outcomes in terms of the autonomy given to state and local governments and the incentives for them to behave in one way or another:33 1. Most consistent with fiscal decentralization is the full assignment of taxing powers and tax administration responsibility to subnational governments, e.g., as is common especially at the regional level for some taxes in the US and Canada. 2. Subnational governments may ‘piggyback’ on a central government tax base. Under this regime, the subnational government sets the tax rate but the central government defines the base and administers the tax. Examples include the local income tax found in the Nordic countries (Lotz, 2012), the federal–provincial tax collection agreements found in Canada (Bird and Vaillancourt, 2006) and shared state-local sales taxes in the US. 3. Although both the tax rate and the tax base are determined by the central government, responsibility for assessment and collection may be assigned to lower-level governments, which in return receive a prescribed share of collections which may, or may not, be related to the amount they collect. This is roughly the system used for certain taxes in Germany and in many transition countries, and for the (former) business tax in China.34 In this case – though less visibly than in the first two approaches – the subnational government can, by altering its administrative effort, directly affect the level of revenue collected. 4. Under the fourth approach – often called ‘revenue sharing’ and common in many developing countries – the central government sets the tax rate and base, collects the tax and assigns a portion of collections to the subnational government.35 Because the recipient governments have no direct influence on the amount received, and hence no fiscal autonomy, such systems are better considered a special form of intergovernmental transfer rather than a subnational tax.36 As Bird (2010) discusses in detail, finding a good subnational government tax is essentially a process of eliminating the less good, namely

M4439-BAHL_9781786435293_t.indd 192

01/03/2018 11:08



Financing local and regional government ­193

those taxes that seem likely to violate efficiency, equity, administrative and macroeconomic goals by the widest margin in the jurisdiction in question. Although the government officials and politicians who make these choices likely attach more weight to revenue and political capital than to considerations of efficiency and equity, they too must strike a difficult balance between the weight they attach to different goals and their assessment of the connection between different taxes and the achievement of those goals. In the end, where experts and decision-makers stand on an issue is often determined largely by where they sit. Different countries make different decisions about local taxation, but there are some patterns. For example, developing countries rarely devolve payroll taxes to subnational governments, in part because central governments are concerned about protecting this base for themselves, notably for financing social insurance (Alm and Wallace, 2004; Bird and Smart, 2014). Excises are often off-limits because of revenue or administrative considerations, particularly the tax on motor fuels. Income taxes may be vetoed as interfering with central government income distribution ­objectives.37 But some of the old walls are coming down. Turnover taxes are now used in several developing countries because they are revenue productive and relatively invisible. Even the value-added tax (VAT), long considered to be both undesirable and infeasible at any but the national level, is now important at the subnational level in some countries (Bird, 2015a), even though other local sales taxes sometimes coexist rather awkwardly with the national VAT.38 Whatever the current tax assignment may be, considerable political fuss often arises when significant changes in tax assignment are contemplated. The primary goal for local and regional taxes is to raise revenues to meet assigned expenditure needs. If subnational governments have important expenditure responsibilities, a tax base cobbled together from the property tax and such minor taxes as those on entertainment, restaurants and signboards is unlikely to cover service delivery costs. Nor will reliance on such politically and administratively difficult taxes as the agricultural income tax in India or the property tax in most countries permit even the largest subnational governments to be close to self-sufficient in terms of financing. In Indonesia, for example, although 39 percent of all government expenditures are made by subnational governments, subnational government taxes finance less than 8 percent of these expenditures (Harjowiryono, 2011). In contrast, in some more developed decentralized countries broadbased taxes are often available to subnational governments. For example, subnational governments are empowered to tax income in the Nordic countries, Spain and Switzerland, and to impose general sales and income taxes in the United States and Canada. In other countries, however,

M4439-BAHL_9781786435293_t.indd 193

01/03/2018 11:08

194

Fiscal decentralization and local finance in developing countries

such as Australia and the Netherlands, subnational governments do not levy broad-based taxes. This is also the common situation in developing countries, with a few exceptions such as Brazil and India. Relative to GDP, subnational government taxes are five times greater in OECD countries than in developing countries (see Table 2.1).39 A common argument in the literature is that subnational government tax bases should be stable rather than cyclical, both because subnational governments provide essential public services and because they (usually) cannot budget for deficits since they are unable to borrow to smooth out large swings in the flow of revenues. Heavy reliance by subnational government taxes on minerals or agricultural products that sell in world markets is not a good idea for this reason.40 Some countries, like the US and Canada, do allow subnational governments to tax natural resources, but few if any developing countries do so. Moreover, in such countries, when subnational governments rely on transfers from a central government that itself is heavily dependent on natural resource revenues (Bolivia, for example), local budgets are often equally vulnerable. Taxes on profits, income and consumption are also subject to economic downturns, of course; but the general view is that it is easier to deal with swings in such revenues by increasing tax rates or cutting expenditures. The property tax, the mainstay of local government finances in most developing countries, is probably the least sensitive to economic downturns, in part because weak assessment practices seldom capture property value changes quickly.41 In some ways, these arguments about the stabilizing impact of different taxes seem a bit odd. Even though income-elastic tax bases may decline during economic downturns, long-run income elasticity is essential at the subnational level, especially for countries that are urbanized and growing. The demand for local public services will expand with development. A tax system that requires regular appeals to support new discretionary tax increases every time they are needed is unlikely to prove successful for long. If a cyclical downturn leads to a revenue decline that is severe enough to threaten essential services, some form of higher-level intervention may be needed – but is seldom likely to occur in hard-pressed developing countries. Reacting to political pressure against local tax increases by preventing subnational governments from accessing broad tax bases is not the way to go, however tempting though it may be for higher-level governments to impose restraints on others rather than themselves. More desirable alternative approaches might include: providing a more diversified tax base; establishing a budget stabilization fund (a rainy day fund) at the subnational government level; setting up a countercyclical grant program implemented by the central government and triggered by a defined decline

M4439-BAHL_9781786435293_t.indd 194

01/03/2018 11:08



Financing local and regional government ­195

in the economy; or even simply a less formal agreement that the central government will step up in the event of a recession.42 The larger subnational governments are often those that most need a broad-based and elastic revenue structure, both because they have the greatest revenue-raising potential and are usually assigned (or deliver) more important expenditure responsibilities. As we discuss further in Chapter 8, such governments are also those that in most countries should come closest to being self-financing (Bahl and Linn, 2015). Smaller, rural local governments may more easily be able to live with the usual limited revenue sources at their disposal – the property tax, user charges, fees, licenses and intergovernmental transfers.

POTENTIAL NEW SUBNATIONAL TAXES The new frontier of practice in fiscal decentralization is to explore and develop more and better sources of own revenue for subnational governments. In OECD countries, the subnational government share of taxes is nearly twice that in developing countries, and as a share of GDP it is nearly three times higher (see Table 2.1). Broad-based taxes have been devolved to regional and local governments in some OECD countries. In low- and middle-income countries, however, intergovernmental fiscal systems have remained heavily centralized on the tax side. Even in the relatively few countries in which regional and local taxes produce significant revenue – as in Brazil (state VATs and local turnover taxes) and Argentina and Colombia (turnover taxes at provincial and local levels, respectively) – these levies have been widely criticized, in part because of structural problems and in part because of doubts about how the proceeds are used. In other countries, subnational taxes that were revenue productive (though badly structured) have been abolished – as in South Africa (a payroll and turnover tax called the Regional Services Council (RSC) levy), India (octroi) and East Africa (graduated personal tax) – without being replaced by other local revenue sources. Finally, even though the property tax seems to be every scholar’s favorite local government tax – as we discuss in detail in the next chapter – almost no developing country collects much from this source, even when taxes on property transfers and higher effective taxation of business property are included. Regional and local governments in low-income countries remain dependent on intergovernmental transfers. One result is that few of the promised improvements in governance, accountability and efficiency potentially achievable by decentralizing expenditures have been realized in most such countries.43 Some might think it is not so bad that subnational g­ overnments

M4439-BAHL_9781786435293_t.indd 195

01/03/2018 11:08

196

Fiscal decentralization and local finance in developing countries

do not have significant taxing powers, perhaps because they believe that most such taxes are likely to be inefficient or perhaps because they think that a more modern local tax system will naturally emerge as GDP rises and urbanization proceeds. Support for the second of these arguments is undermined by the evidence from such developed countries as Italy and Greece which suggests that there is nothing natural about good local taxes unless someone consciously gets them accepted and implemented (Slack and Bird, 2014). Equally, the first argument ignores the evidence from countries such as Colombia and Argentina that substantial improvements are possible in both the design and implementation of local taxes if both the national and local levels have the political will to put better taxes in place and to provide the support needed to ensure that they are effectively implemented.44 To see what is required for successful tax decentralization we might revisit the first principles for ‘getting it right’ that are usually put forward as best international practice. We consider both the gains that might be expected from assigning more taxes to subnational governments and the dangers of doing so, before addressing the fundamental question of whether the implementation of a broader subnational government tax system is worth the transaction and political costs incurred. Our review of the practice of subnational government taxation in developing and transition countries suggests that many countries, rich and poor alike, have gotten some things right and others wrong. We conclude with some ideas about how to get it more right than wrong by getting around the impasse that seems to have so far held back subnational government taxes in most developing countries. Because the ‘right’ subnational government tax structure differs from country to country, getting it right may sometimes mean that some of it may be wrong in the sense of violating some of the commonly proposed rules for best practice. Decisions to decentralize taxing power to regional and local governments imply costs and benefits, as do decisions about which taxes to decentralize. Some costs and benefits are economic. Others relate to bureaucratic politics, technical issues of tax administration, social equity, a paternalistic central attitude toward local government, and the degree to which the central government sees itself as the sole guardian of tax policy. Even if the rules of good subnational government taxation are followed, and even if the finance-follows-function mandate is obeyed, there are benefits and costs that need to be carefully weighed in each specific context when it comes to choosing good taxes. The core benefit from more decentralized revenue assignment is the welfare gain that results from financing decentralized services with decentralized taxes owing to the increased accountability of provincial and local

M4439-BAHL_9781786435293_t.indd 196

01/03/2018 11:08



Financing local and regional government ­197

government officials for the quality of services that are delivered to those who elected them and pay the taxes that have been imposed. Other benefits that may emerge from a more decentralized revenue structure are that: (a) it may capture some gains from diversifying the administration of the tax system; (b) it may improve fiscal discipline by enabling the imposition of a more effective hard budget constraint on subnational governments; and (c) it may enhance revenue mobilization. We discuss each of these in turn. Increasing Accountability Locally imposed taxes make locally elected officials more accountable to their voting population for the public services that they deliver than if the services were financed by intergovernmental transfers (Olson, 1969). At the margin, residents are likely to see transfers as an increased share of the national revenue pie gained at the expense of another subnational jurisdiction – a process they often see as imposing no additional cost to them, and hence view as a windfall gain.45 An increased local tax, on the other hand, requires local voters to pay more for the public goods they receive, so accepting such a tax implies that they are willing to pay for more local public services. Since local politicians may be called to account if they do not deliver the goods, they will pay more attention to the quality of services in a decentralized regime. Ideally, the result is that better public services are provided for which people are more willing (or less resistant) to pay, which in turn will drive down the political and administrative cost of revenue mobilization. For this process to work, subnational governments must have the authority to impose higher (or lower) levels of taxation, as well as the incentive and the capacity to do so. This in turn requires both that they have the power to control their spending at the margin and that local residents have some power to control and influence their governments (usually through elections but also through voicing public complaints without fear of reprisal). Moreover, sufficiently accurate information must be available so that not only subnational governments but also their constituents know what they are doing and can evaluate their fiscal decisions (Bird, 2000a). As always, the real world seldom hits these targets. The benefits of accountability (local voters get what they want from their elected officials) are not easy to achieve in many developing and transition countries. The electoral process might not be open and contested (China, Vietnam, Sudan) or elections may have been suspended (Nepal). In other countries, voters are less than fully mobile and have not learned how to use the vote to hold their officials accountable. In many, the local government revenue structure is so dominated by intergovernmental transfers that any benefits from increased subnational government taxes are unnoticeable. Finally, since the process

M4439-BAHL_9781786435293_t.indd 197

01/03/2018 11:08

198

Fiscal decentralization and local finance in developing countries

of deciding to increase spending (and taxes) is never all that transparent, voters may not know who should be held accountable in any case.46 Accountability does not automatically result from decentralization. Several components of the intergovernmental policy framework must be in place to get the right result, including such features as the following: ●●

●●

●●

●●

●●

●●

People need to feel some pain from taxation if accountability is to work. The minor levies and nuisance taxes found at the local level in many developing countries will not do the trick. Neither will taxes that are not visible to taxpayers or taxes whose burden is exported. As we argue in Chapter 8, although few politicians are likely to have the courage (or the self-interest) to do so, it may sometimes be desirable to put pressure on the larger, richer districts that should be essentially self-financing, perhaps by cutting central budgetary support to such districts. As we discuss further below, decentralizing tax administration responsibilities is not essential; but when it exists it may help taxpayers to distinguish which taxes belong to the subnational government and where accountability should rest (Mikesell, 2007). Ideally, all tax proposals should begin by preparing, releasing and discussing in public a note that describes the expected impact of the proposal, including the distribution of its burdens and benefits. Few developing countries do this. The process of budget determination should be made more visible by publicizing budget proposals as widely as possible. Efforts should be made to educate the media on how to report revenue proposals, and to tie these proposals to identifiable elected officials. Few politicians are likely to be eager supporters of an initiative to this effect. People could be more engaged in the budgetary process if local governments publicized and held local meetings, and perhaps if some financial incentives were given to local governments to make efforts along these lines.47 The awareness of future voters may be broadened by introducing discussions of some of these issues in the school system. In the case of rural local governments in India, it has been shown that there is a link between the level of literacy and the level of local government taxes (Bahl et al., 2010a).48

Increasing Revenues An objective in many decentralization programs has been to increase subnational government revenues in order to relieve the central ­government

M4439-BAHL_9781786435293_t.indd 198

01/03/2018 11:08



Financing local and regional government ­199

of the need to support local government expenditures. Why might tax decentralization enhance overall revenue mobilization, i.e., increase the tax-to-GDP ratio? Three reasons might be hypothesized. The first is that there may be greater willingness to pay for services delivered under a decentralized fiscal system. People ought to be willing to pay more to get services that they want. The second is that the decentralization of taxing powers may lead to the implementation of new taxes, and therefore to an overall revenue increase. In many countries, provincial and local governments have broadened their tax net with a variety of special tax instruments and administrative measures such as levies on the sales of assets, licenses to operate, betterment charges, and various forms of property and land taxation (Bird and Wallace, 2004; Smolka, 2013). Without the local discretion that comes with fiscal decentralization, some of these new revenue sources may not have emerged. Third, and perhaps most importantly, increased revenue may result from the base-broadening that might result from dividing tax bases according to comparative advantages in assessment and collection (Bahl and Linn, 1992; Bahl, 2011). There is an especially strong administrative case for the property tax to be a local government levy (see Chapter 6). The information advantages of subnational governments seem most relevant in developing countries (Bahl and Bird, 2008), where subnational governments may reach some who are hard to tax under centralized regimes. For instance, state and local governments oversee a variety of licensing and regulatory activities, and they track property ownership and land-based transactions. They can identify local businesses and gain some knowledge about their assets and scale of operation. Because the potential revenue gain is much more important for them in relative terms, subnational governments have more incentive to tax such activities effectively than do national governments. Detailed local knowledge of the tax base may lead to the capture of some who presently do not fully comply with national taxes or evade taxes altogether, such as the self-employed and small businesses, who often under-declare taxable income and consumption (Alm et al., 2004). Local governments have a comparative advantage with respect to small taxpayers who usually are less easily reached by central tax systems. Tax authorities in many developing countries bemoan the fact that the selfemployed and the informal sector mostly escape taxation. Most revenues from major central government taxes can be traced to a relatively small number of taxpayers. The individual income tax is paid mostly by withholding tax on incomes in the formal sector of the economy (Bird and Zolt, 2005), and the VAT is collected mainly from larger firms (Bird and Gendron, 2007).49 These failures at capturing the low end of the tax base are institutionalized in many developing countries which set relatively high

M4439-BAHL_9781786435293_t.indd 199

01/03/2018 11:08

200

Fiscal decentralization and local finance in developing countries

thresholds for the broad-based taxes. For example, the VAT threshold is often set relatively high in terms of gross sales, precisely to exclude small taxpayers that are not cost effective to reach (Keen and Mintz, 2004). Similarly, corporate income taxes usually cover only large firms, and the exemption level for the individual income tax is often set well above the average income level, in part to exempt subsistence incomes but in part also in recognition of the difficulty in collecting tax from the informal sector (Bird and Zolt, 2005). In contrast, local governments often impose various types of taxes and fees on small businesses excluded from the ambit of central taxes. The amounts of revenue raised through such levies are usually not large. Nonetheless, particularly in larger cities, these revenues are often both important and elastic, and more use can and should be made of better-designed local business taxes (Bird, 2006a). There also are reasons why tax decentralization might have a dampening effect on the overall rate of revenue mobilization. Local people may not buy into the idea that higher taxes will result in better services.50 They may think new revenues are more likely to result in over-bloated payrolls stuffed full of politicians’ relatives, or be spent to provide perquisites for local officials or to satisfy the whims of bureaucrats and politicians rather than the needs of ordinary citizens. Because better-off (and usually more influential) people often have good substitutes for such public services as schools, security and refuse collection, they are perhaps especially unlikely to support increasing local taxes.51 Increased subnational government taxing power may also encroach on the taxing space of higher-level governments, increasing opposition to higher central government tax rates.52 The empirical evidence on whether increased subnational government taxation crowds out central revenues and reduces overall revenue mobilization is thin. Lotz (2006) observed that it is unclear whether decentralized taxation power increased the overall level of taxation in OECD countries. The Nordic countries have both high levels of tax assignment to subnational government and high overall levels of taxation; on the other hand, Austria and Switzerland also have a decentralized tax system but much lower levels of taxation. Bahl and Cyan (2011) in an econometric analysis of ‘crowding out’ for 58 developed and developing countries in the 1990s found evidence that increased local taxes reduced levels of central tax revenue for higher-income countries with more open economies, but that this finding did not hold for developing countries. The crowding-out threshold (subnational government taxes as a percent of GDP) was estimated to be about 8 percent; that is, in countries below that level, an increase in subnational government tax revenues is likely to add to overall revenue mobilization. Summing up, this study found little evidence of a displacement effect.

M4439-BAHL_9781786435293_t.indd 200

01/03/2018 11:08



Financing local and regional government ­201

Increasing Tax Competition An obvious danger in decentralizing tax power is that some subnational governments will use taxation as a measure to gain advantage over other subnational governments, which in turn may result in efficiency losses and harm economic growth. One reason is simply because, as always, politics and self-interest get in the way of good economics. Politicians at all levels quickly understand that tax exporting (and the protection of local industries) is popular with local residents, and that most people are sufficiently baffled by the tax system to provide political cover for this financing strategy. The correspondence (matching) principle discussed earlier would limit subnational governments to choosing taxes whose burden rests on residents or on non-residents who benefit from services provided by that subnational government. Land, labor and capital employed in a locality but owned by non-resident beneficiaries should be taxed. Non-residents who commute into an area to work or to shop in local markets should pay taxes just like locals since they are also using city services, though likely to a lesser extent. Such reasoning supports city and county government income and retail sales taxes in the US. Several Eastern European countries similarly impose local payroll taxes based on the place of work (Golovanova and Kurlyandskaya, 2011), as does Mexico (Bahl, 2011). But this argument does not imply that subnational governments should be able to tax mobile factors as they wish. If a locality can export some of the burden of its taxes to (non-benefiting) residents in other jurisdictions, its residents will face lower tax-prices for local government spending, and are therefore likely to spend more than they would otherwise do. Politicians and residents of the taxing jurisdiction may both like this outcome, but it is not efficient from a social viewpoint. Most countries therefore put limits on subnational taxing power. One way to limit the degree of tax exporting is to deny the right to levy discriminatory taxes on non-residents. This rule applies to US state and local governments, but does not prevent significant tax exporting (Phares, 2005). Another restraint on tax exporting is the threat of retaliation from other regional and local governments. A third, and common, approach is to prohibit subnational governments from adopting taxes where there is a significant potential for exporting the tax burden. And a final limitation found in some countries is to establish a maximum rate to limit the amount of potential exporting, often coupled with a minimum rate to limit the potential to ‘steal’ tax base from other jurisdictions. Nevertheless, origin or source-based taxes that permit the export of burdens to non-residents, including some who do not benefit from local

M4439-BAHL_9781786435293_t.indd 201

01/03/2018 11:08

202

Fiscal decentralization and local finance in developing countries

public services provided by the taxing region, are common in developing countries. For example, subnational origin-based turnover taxes are levied in both Argentina and Colombia; and, as mentioned above, subnational payroll taxes are levied in Mexico and several countries in Eastern Europe. Other examples are natural resource levies, pre-retail stage sales taxes and, to some extent, non-residential real property taxes. Regions and localities may also be willing to create some perverse incentives that harm economic growth. In this regard, the ‘fiscal war’ among state governments in Brazil has attracted much attention (Afonso and Barroso, 2007). De Mello (2012) found evidence that when Brazilian states began offering VAT rate discounts to attract firms – making up for the revenue loss by imposing higher rates on essential local consumption (e.g. fuel or utility bills) and less mobile production activity – other states reacted strongly to protect their competitive position, to some extent resolving the problem.53 Perhaps in part because the extent and nature of the resulting distortions are less clear, similar reactions to the tax exporting inherent in such cascading gross receipts taxes as those levied by Argentine provinces and Colombian cities have not occurred, and such taxes continue to retard economic growth.54 Some have argued that one problem in letting subnational governments decide how much to tax is that they are unlikely to utilize fully the revenue sources open to them, which may lead to a deterioration of the level and quality of local public services provided – a version of the infamous ‘race to the bottom.’ Others suggest that this is unlikely to be a grave problem (Bird and Slack, 2008). If the service in question is one of national importance or one in which there is a strong national interest in maintaining standards, it should presumably be funded and monitored by the central government. If it is not a matter of national interest, why should the central government be concerned if one region chooses to have a larger government than another? Central governments are usually keen to control corporate taxes for revenue, stabilization and regulatory purposes. Such taxes are difficult for small governments to administer effectively, let alone efficiently. Taxes on international trade, like those on international investment flows, are also both difficult for regional and local governments to implement properly and highly distorting in efficiency terms. On the other hand, as North American experience shows, regional and even local sales taxes are possible, though since even national jurisdictions have so far been unable to tax some cross-border business activities fairly, efficiently and effectively, subnational jurisdictions are even less likely to be able to do so (Bird, 2015a). Although it is easier for subnational governments to tax personal incomes (as in Switzerland, the US and the Nordic countries) or to impose

M4439-BAHL_9781786435293_t.indd 202

01/03/2018 11:08



Financing local and regional government ­203

payroll taxes (as in Australia and Mexico) than to tax corporations, central governments may again be reluctant to permit them to do so – for example, because maintaining a degree of visible progressivity in direct taxation may be considered necessary for political stability (Bird and Zolt, 2015a). The regulatory role of taxation may also play an important part in shaping a continued central role in excises on such products as alcohol, tobacco and petroleum, even though such taxes have sometimes been suggested as more suitable for subnational governments (McLure, 1997) and are extensively used by such governments in some countries, as we note below. The policy constraints imposed on subnational taxation may sometimes lead local governments to engage in a desperate search for revenues by piling on all sorts of specific local levies (entertainment taxes, advertising taxes, business taxes and so on) which are usually costly to collect, arbitrary in their administration and economically distorting. At one point, for example, China revised its revenue-sharing regime and then implicitly allowed local governments to create an array of ad hoc, piecemeal and sometimes clearly illicit ways of filling the revenue hole that had been created (Bahl, 1999; Wong and Bird, 2008; Bird et al., 2011).55 The outcome of pushing growing cities to rely on such a hodgepodge of revenue sources is unlikely to be equitable or efficient, let alone to encourage responsible accountability. One reaction is to impose still stricter constraints, as when Russia reduced the list of local taxes from 22 to two in 2005 (Golovanova and Kurlyandskaya, 2011). It is easy to go too far in this direction. Making it difficult for local governments to tax local businesses may reduce administrative costs, compliance costs and distortion costs. But if the result is not only to weaken the link between local government and local businesses but also to reduce the incentive for local governments to favor investment and growth, the game may not be worth the candle. Increased Inequality? Another much-cited risk is that increased subnational government taxation will result in greater regional inequity in public service levels. The wealthier subnational governments will have larger tax bases, more administrative capacity and stronger demand for more locally financed services. In fact, in most countries where tax decentralization has been embraced, higher-income places do better. Such disparities are not inconsistent with what one finds in many countries – that higher ‘effort’ (actual as a share of potential) is often found in poorer than in richer areas.56 Our view is that higher-income subnational governments should be encouraged to raise more revenues even if the result is to make fiscal disparities greater, in part because the migration of factors (including

M4439-BAHL_9781786435293_t.indd 203

01/03/2018 11:08

204

Fiscal decentralization and local finance in developing countries

labor) to regions of higher productivity is an inevitable and desirable component of economic growth.57 However, the obvious downside of increased regional disparity as an apparently necessary accompaniment to economic growth is that political unrest may ensue when wealthier places gain most. The correct response is not to discourage the rich from trying harder in terms of taxing themselves more heavily, but rather to design an intergovernmental transfer system to make sure that poorer regions and localities do not fall behind too far. The key to unlocking this puzzle is to get the transfer system right in terms of equalization, as we discuss in Chapter 7.

WHAT ARE THE BEST TAXES FOR SUBNATIONAL GOVERNMENTS? Even the best local or regional tax administration can only succeed if it has the right taxes to administer. We consider here which major subnational taxes might contribute, say, an additional 1–2 percent of GDP in revenues over the next decade. Experience in countries such as Argentina, Brazil, Colombia, South Africa and China (among others) shows that it can be done, but that doing so usually requires throwing away at least part of the conventional guidelines to good subnational taxation. Developing countries that have established significant subnational revenue systems have done so by focusing on the simple aim of mobilizing revenues that can be implemented satisfactorily in a politically acceptable way. They have not worried that much about either economic efficiency or equity (interpersonal or interjurisdictional). Revenue is what it is about. It follows that the place to look is to where the potential tax base is largest, namely, metropolitan areas and the richer regions. Following the money means taking an asymmetric approach to tax decentralization. As we argue in Chapter 8, let the richest places exploit their tax bases, and leave the rest more dependent on grants until they get far enough ahead economically to have the capacity to take advantage of decentralized taxing power. Leaving taxes on land and property to the next chapter, four areas – payroll taxes, selective consumption/sales taxes, gross receipts taxes and general business taxes – may offer potential ways the better-off regions in developing countries can raise more revenues fairly quickly. None is ideal from any economic perspective, but this is where the money is – and where it may perhaps be collected without undue administrative or political strain.58

M4439-BAHL_9781786435293_t.indd 204

01/03/2018 11:08



Financing local and regional government ­205

Payroll Taxes A flat rate tax on payrolls with no personal exemptions may be thought of as a rough sort of benefit tax to finance services provided by the subnational jurisdictions in which the taxed activity takes place (McLure, 1998). Such a tax imposed on the employing firm is not difficult to administer, at least at a regional or metropolitan level that encompasses most of the relevant labor market. It can generate significant revenue from a tax base that will expand with urbanization and growing formal sector employment. It is roughly consistent with the correspondence principle because whether employers or employees bear the ultimate burden, much of the tax is paid by those who benefit from local services. No tax is perfect, of course. Local or regional government payroll taxes do not reach the hard to tax informal sector, and have the significant disadvantage of discouraging employment in the modern sector (Bird and Smart, 2014; Alm and Wallace, 2004). When jurisdictions are small, too much tax may be imposed on nonresident workers relative to the benefits they receive, so the tax price paid by residents is too low. One of the few developing countries in which subnational governments can tax payrolls is Mexico, where state governments have the power to levy and administer a wage tax. However, none has used this power very intensively. Rates vary little among the states, with a top nominal rate of about 2.6 percent, and collections are only about 0.27 percent of GDP (Revilla, 2012). Some other countries permit subnational governments to levy various types of rudimentary income taxes or poll taxes, often based on presumptive assessments and invariably with small revenue yields: examples include the community tax in the Philippines and the professions tax in India and Pakistan (Bahl, 2011a). In Rwanda, a graduated rate tax on rental income, imposed by national law, accounts for about 18 percent of local government revenue (Cyan et al., 2013). In 2003, Uganda abolished its graduated personal tax, long an important source of local government revenue. In the same year, Tanzania eliminated its ‘development levy’ – essentially a poll tax – which had been the major source of rural local government revenue. South Africa, which had permitted regional governments to impose a tax based on both payrolls and turnover, abolished it in 2003. In none of these cases were local governments given any compensating taxation power. Although many countries are reluctant to consider allowing subnational taxation of payrolls because they have generally earmarked such taxes for financing social insurance spending (Bird and Smart, 2014; Alm and Wallace, 2004), further exploration of the potential of such taxes for regional and metropolitan governments seems warranted.

M4439-BAHL_9781786435293_t.indd 205

01/03/2018 11:08

206

Fiscal decentralization and local finance in developing countries

Selective Consumption/Sales Taxes Selective taxes on consumption play an important role at the subnational level in some countries. In Colombia, for example, excise taxes on beer and tobacco, together with the profits of public liquor producers, provide important revenues for departmental (regional) governments, with most of the proceeds being earmarked for health-related spending. In addition, these governments – like local and regional governments in many other countries – also receive revenue from a variety of other consumption taxes, especially those on motor vehicle ownership, licensing and use. Selective local taxes on entertainment, hotels, billboards and a variety of other goods and services are commonplace in developing countries. To the extent such taxes can be administered effectively on final consumers they are unlikely to have any serious effects on economic efficiency, or to be too regressive. Selective sales taxes on electricity and telecommunications can be more revenue productive and constitute an important component of the municipal tax structure (e.g., in South Africa), although often such taxes are largely paid by non-residential users and may hence discourage growth. However, few selective consumption taxes produce much revenue, and most are costly to administer and to comply with. It may often be difficult to tax sales on a destination basis, a problem that in India led to the long-standing use of the ‘municipal import tax’ (octroi), a tax that was administered mainly by erecting customs barriers on roads leading into urban areas, and which until very recently provided substantial revenue in Mumbai (Pethe, 2013). When there are many small shops and informal traders dealing in taxed goods, collecting at the time of final sale is administratively costly and often impracticable. Colombia has, over the years, developed a system under which producers allocate alcohol and tobacco taxes to regional governments according to shipping invoices, although there remain substantial evasion problems owing to smuggling and sales in the informal sector (Bird, 1984). Brazil and Pakistan levy more general taxes on the sale of services. In Pakistan, provincial governments levy a sales tax on selected services, collecting 0.5 percent of GDP in 2014 (Ghaus-Pasha and Pasha, 2015). Brazilian cities, especially the larger ones, raise 0.7 percent of GDP, about one-third of their own-source revenues, from a sales tax on all services except communications and interstate and intercity public transportation, which are taxed by the state VAT. Municipalities can set a rate generally between 2 and 5 percent of turnover, although more presumptive methods of assessment are used in some cases (Afonso and Araújo, 2006; Rezende and Garson, 2006; Wetzel, 2013). There is an especially strong case for taxing motor vehicles at the local

M4439-BAHL_9781786435293_t.indd 206

01/03/2018 11:08



Financing local and regional government ­207

and regional level (Bahl and Linn, 1992). They are easy to tax, doing so is relatively progressive, driving generates serious externalities (congestion, pollution, accidents) and the tax base usually expands quickly with growth. Many of the associated costs (both externalities and road construction and maintenance) are inherently localized and likely to differ substantially from place to place. Charging properly for road use is arguably one of the most important things governments can do to ensure the sound and sustainable development of urban areas. As we discuss further in Chapter 8, fuel taxes, vehicle registration and licenses, parking taxes and tolls may all have a role to play in this respect (Bird, 2005a). In fact, however, relatively little subnational revenue is collected from this rich and expanding tax base in most developing countries. Gross Receipts Taxes Gross receipts taxes are imposed, as the name implies, on the total value of sales (turnover). They are thus more broad-based than property or payroll taxes and can produce considerable revenue at low rates. They are also less transparent, and seem generally to be less objectionable than most other (technically superior) forms of taxation, both to the firms that pay them – perhaps because of the low rate and the simple reporting required – and to their customers, who likely bear most of the burden but are unlikely to know that the tax is even imposed. Such taxes are widely used in various forms by subnational governments, and have sometimes been very successful in terms of producing significant revenue (Bahl, 2011a; Martinez-Vazquez, 2013a). Such taxes impose efficiency costs on the economy. They tend to ­‘cascade’ through various stages of production and distribution, and hence end up imposing a heavier burden on final consumers than the revenue they generate for the treasury, distorting resource allocation, encouraging vertical integration and discouraging growth. When imposed at the subnational level, turnover taxes reduce accountability to the extent the tax burden is exported to non-beneficiaries of services provided by the taxing jurisdiction. Such taxes may be tricky to administer below the central level because of the ‘headquarters problem’ that arises when a firm pays its taxes on all its sales wherever its headquarters is located. The gross receipts tax definitely is not a member of the good tax family. But it does produce a lot of money with relatively little effort, and may enable subnational governments to raise significant revenue and achieve greater fiscal autonomy. In Argentina, the most important own-revenue source of the provinces is the turnover tax, both the rate and base of which they control. The tax rate varies widely by type of product, with rates set higher for transactions

M4439-BAHL_9781786435293_t.indd 207

01/03/2018 11:08

208

Fiscal decentralization and local finance in developing countries

closer to the retail level to reduce distortions. This tax raises about 3 percent of GDP. Half the revenue is allocated to the location of sales and half to the location of employment and other non-material costs (Braun and Webb, 2012). A gross sales tax (the industry and commerce tax) is also levied by municipal governments in Colombia within rate limits set by the central government which range from 0.2 to 1 percent on estimated gross sales and vary by sector and by class of municipality. Although the effective rates are often eroded by preferences and administrative practices, they are highest in the capital, Bogotá. This tax provides about 45 percent of total subnational revenues, with almost all this amount collected in the largest cities (Bird, 2012a). A local tax on gross sales in the Philippines is levied as a license, and accounts for a significant percent of local government revenues. The tax base and rate limits are defined by the central government. A similar gross receipts tax is the largest own-revenue source for municipalities in Tanzania. The rate is set by the central government, but collection is by the local governments. Calculation of the base is enabled by using company tax returns provided by the Tanzanian Revenue Authority. In Rwanda, local governments tax businesses with a trading license on gross receipts. Many other countries have variants of such business taxes, often based on presumptive assessment systems that may bear very little relation to reality (Bird and Wallace, 2004).59 Local Business Taxes The line between gross receipts taxes and local business taxes is a fine one, frequently blurred in practice. How local governments tax business varies substantially from country to country. Some rely mainly on property taxes; some on a range of specific charges and fees; some on some type of turnover tax; and some on some type of levy differentiated by type, size and location of the business, and often, like the patente in some francophone African countries, fixed in amount for a period of years.60 Such taxes are often convenient fiscal handles for local government: they can generate significant revenue and seldom lead to severe political reactions. No matter how often visiting economists note that such taxes are economically undesirable, countries are going to continue to use them. Significant revenue raised at low political cost trumps the advice of tax policy experts every time. Interestingly, one of the few systematic reviews of local business taxation in a developing country, Kenya, concluded that while it would be better in principle to impose a flat-rate tax on turnover (or, preferably, net income or sales), the new business license system should instead follow the

M4439-BAHL_9781786435293_t.indd 208

01/03/2018 11:08



Financing local and regional government ­209

fixed lump-sum tax model, varying by business type and size. This was in part for administrative reasons and in part to avoid overlapping central income and sales taxes (Devas and Kelly, 2001). Since such local business taxes produce between 5 and 30 percent of urban local government revenues in anglophone African countries (Fjeldstad et al., 2014), they need closer study than they seem to have received. In any case, this approach is unlikely to prove sustainable in the long run because economies grow and urban areas become larger and more complex, and require considerably expanded funding of urban local services. A more promising path towards a better – not perfect by any means, but better – local business tax than that now found in most developing countries has been gradually developing in countries such as Italy, Japan and, most recently, France in the form of a local tax on value-added which is quite distinct from, though administratively related to, the VATs imposed at the national level. What such a tax may look like is sketched in brief in Box 5.1. Although considered briefly in South Africa (Hunter van Ryneveld, 2008) and set out in some detail in Bird (2014, 2015a), the possible viability and utility of this form of local business taxation has not yet been field tested in the context of a developing country. It may deserve closer consideration, especially in metropolitan areas like those we discuss in Chapter 8.

TAX DECENTRALIZATION AND DECENTRALIZED ADMINISTRATION One last consideration that many consider important with respect to assigning taxes is whether the government to which they are assigned can feasibly administer them. We review this question briefly in this section.61 Of course, administration is not the only consideration that matters. The property tax, for instance, is in some ways surprisingly difficult to administer well at the local level (Slack and Bird, 2014; Bahl, 2009); but administrative feasibility is not the only relevant factor in determining tax assignment. Chapter 6 makes the case for maintaining and improving the role of the property tax as a mainstay of local government finance in developing countries. Where subnational government taxes should be administered remains an open question. In developing countries, Dillinger (1991) suggested that the choice was between incompetence (because local administrations are likely to be less capable) and indifference (because central administrations are not likely to pay much attention to local concerns). This view is too simple but not irrelevant. When administration has been turned over to

M4439-BAHL_9781786435293_t.indd 209

01/03/2018 11:08

210

Fiscal decentralization and local finance in developing countries

BOX 5.1  A BETTER WAY TO TAX LOCAL BUSINESSES? Most local business taxes are imposed on certain inputs (labor, capital, land) or gross receipts. Taxing inputs or gross sales distorts production decisions.1 Most countries have now substantially reduced the efficiency costs of their national tax systems by taxing business value-added instead of turnover. Businesses add value by combining labor and capital with other purchased inputs. The value added by labor is the cost of labor (wages and salaries), while the value added by capital is the cost of capital (both debt and equity). Value-added taxes (VATs) are usually imposed in a way – the invoice-credit system – that permits businesses subject to VAT to deduct the VAT they pay on inputs (including purchases of capital goods) from the VAT due on their sales. A possible way that local governments can adapt this approach to fit their circumstances may be by imposing what is sometimes called a ‘business value tax’ or BVT.2 A BVT is imposed on essentially the same base as VAT, and can therefore be based on the annual value-added tax base reported for VAT purposes. However, as the table below shows, it is quite different in several important ways. Essentially, the BVT taxes business income, not consumption, and does so based on the location of the seller (origin) rather than the location of the buyer (destination). Moreover, the base of the BVT may be calculated simply from annual accounts as sales minus the purchase of inputs, including capital purchases (less depreciation allowances), from other businesses.3 If all capital purchases were deductible, the base would be equivalent to the tax base of a conventional VAT.4 BVT vs. VAT Base Principle Application Timing

VAT Consumption Destination Invoice-credit Monthly

BVT Income Origin Accounts Annually

However, since the BVT taxes profits as well as wages, it is a tax on income and not, like VAT, a tax on consumption. Moreover, since it is imposed on an origin rather than a destination basis, it is in effect a tax on production – that is, on both investment and consumption – and not a tax on consumption. Also unlike VAT, BVT taxes exports (sales outside the jurisdiction) but not imports (purchases from other jurisdictions). BVT is thus clearly a tax on business activities and not a form of sales tax.5 Finally, BVT is administered differently, being imposed on an annual accounts basis – but on gross receipts rather than net income as required for income tax purposes. This base is roughly the same as the total of the monthly (or quarterly) invoicing of sales transactions required for VAT, though without requiring similar information on purchases. Introducing an income-based BVT as a local business tax would improve local business tax systems in most countries in at least three ways. First, such a tax would be more neutral (less distorting) than most other forms of local business taxes, such as differentially high property taxes, which discriminate against

M4439-BAHL_9781786435293_t.indd 210

01/03/2018 11:08



Financing local and regional government ­211

i­nvestment. Second, a BVT would be less susceptible to base erosion because with a larger tax base, the rate required to produce a given level of revenue would be lower. Finally, to the extent that the principal economic rationale for taxing business rests on benefit grounds, a BVT is in effect a generalized user charge, and may thus not only be an equitable way to finance local expenditures but also one that should induce more efficiency and effectiveness in local budgeting and spending. Local (or regional) powers to impose a local BVT should clearly be regulated to ensure, for example, that the tax base is defined uniformly in all localities. Moreover, although localities should be able to determine their own tax rate for accountability, it would seem sensible to allow such discretion only within a range of rates, with a maximum to restrict tax exporting and a minimum to restrain tax competition. Importantly, there should also be a uniform formula for allocating business value added among localities. Businesses that operate in only one locality would be taxed solely by that jurisdiction. However, when businesses operate in more than one jurisdiction, the tax base would, like that of any income-based tax, need to be allocated according to formula weights based on such factors as payroll, sales (on an origin basis) and capital. Notes: 1. Taxing net profits may not impose similar efficiency losses (if profits are correctly defined, which they seldom are) but can seldom be done effectively at the subnational government level. 2. Various versions of this approach in different countries are described in Bird (2014). 3. This tax base may also be calculated in two other ways. The first is called the addition method, by adding wages to the base of a business income tax (such as corporate income tax) as usually calculated. The second is to approximate the same result by imposing separate taxes on wages (a payroll tax) and an equivalent tax on net capital income. 4. This summary oversimplifies matters in several respects: for a more detailed comparison of business income taxes, conventional VATs and BVT, see Bird and McKenzie (2001). 5. Taxing investment and exports, as BVT does, may not seem to be a particularly good idea. Although few seem bothered by the fact that all income taxes do the same thing, the potential distorting effects provide a strong reason for keeping the tax rate low.

local governments, the results have not always been good.62 High administrative costs and substantial leakages in the base may sometimes reflect the badly designed taxes that local governments are supposed to administer. But they may also reflect weak administrative capacity, incompetence and corruption at the local level. Even when tax policies are well designed and administrations at both levels of government are well run, as in Canada, a single administration may be more cost-effective than multiple administrations.63 The introduction of modern tax technology has in many ways made centralization even less costly, as evidenced by the almost worldwide tendency of central administrations to reduce the number of local tax offices (OECD, 2015). On the other hand, there are strong arguments in the other direction when countries decentralize because, as emphasized

M4439-BAHL_9781786435293_t.indd 211

01/03/2018 11:08

212

Fiscal decentralization and local finance in developing countries

earlier, decentralized governments need substantial fiscal autonomy if they are to function effectively and efficiently. This suggests that, whenever feasible, regional and local taxes should be administered by the governments with the most revenue at stake. Every country must weigh the conflicting arguments for centralizing and decentralizing tax administration, with how the balance is struck with respect to any tax depending on a variety of institutional and empirical factors that can be dealt with only in the specific context concerned.64 Because there is a fixed ‘set-up’ cost in collecting any tax, decentralizing administration tends to increase collection and compliance costs. Evasion and avoidance may also increase with decentralization for taxes where the base is mobile or straddles more than one jurisdiction. The economies of scale and scope associated with the information systems on which modern tax systems are increasingly dependent (Bird and Zolt, 2008) may be achieved through more effective and coordinated use of specialized staff in a more centralized administrative structure. Compliance costs are reduced when taxpayers submit fewer returns, interact with fewer officials and offices, and deal with a more uniform and harmonized administrative structure.65 Administrative costs are reduced when it is simpler to communicate laws and regulations to officials and establish exchanges of information and coordination of action without complex interagency or interjurisdictional negotiations. A more uniform and better-coordinated administrative system should provide more equal procedural treatment to all taxpayers and be able to cope more effectively with evasion, and especially complications arising from cross-border transactions. It may also be less susceptible to political interference and corruption. On the other hand, subnational governments are often correct to worry that the central government will be less enthusiastic about collecting their taxes than collecting its own. Central governments are unlikely to have much incentive to do a good job. After all, it’s not their money, and national political and other concerns seldom turn on the needs and concerns of lower levels of government. In the terms mentioned earlier, it may perhaps prove easier to remedy local incompetence than central indifference. Bringing administration closer to the people it is supposed to serve may also yield improved accountability and efficiency gains, and perhaps (as discussed earlier) even increased revenues because people can more easily identify how fairly taxes are being administered and what the money is being spent on. Better local information and knowledge, more flexibility in organizing and staffing to fit local conditions, greater scope for innovation when there are several tax agencies rather than one monopoly, and above all the greater incentive for local decision-makers to collect and spend local revenues effectively may all make increased local control over

M4439-BAHL_9781786435293_t.indd 212

01/03/2018 11:08



Financing local and regional government ­213

tax administration a vital component of the fiscal autonomy required to achieve the theoretical benefits of fiscal decentralization. A standard complaint about developing countries is that local administrative capacity is inadequate to do the job properly. While this may sometimes be not only true but also inescapable – for instance, small local governments are unlikely ever to be able to operate a standard creditinvoice VAT efficiently66 – there are many ways around this problem. As with fiscal decentralization in general, there is undoubtedly a learning curve, so it may take time (perhaps quite a lot of time) for regional and local governments and their citizens to learn how to run things effectively, let alone efficiently – especially in countries in which subnational officials have little experience of such matters. But they can and do learn; and since better local tax administration will build additional capacity in financial administration, local financial management and expenditure outcomes may also improve as a result. Additional costs associated with decentralized administration may thus be offset to some extent by benefits in terms of improved efficiency, equity, acceptability and accountability. For example, although residential property taxes are not only unpopular but also relatively costly to administer well, the higher costs may be fully justified both by the benefit tax aspect of such taxes and by the resulting increase in government accountability (Bird and Slack, 2014). In contrast, although taxes imposed on local businesses that export most output to other jurisdictions (and therefore some tax burden) are usually popular with locals and may be cheap to collect in terms of administrative cost per dollar of net revenue collected, they not only reduce accountability but also economic efficiency; therefore, limits on the freedom of local governments to tax such activities may be needed in the interests of national growth. Central governments often fear that tax decentralization may encourage competition between governments for tax revenues. This argument assumes the amount of ‘tax room’ is fixed. However, it may be increased if, for example, local administration improves accountability by making it clearer to taxpayers what their taxes are buying and they become more willing to pay (Mikesell, 2007). Surveys in Colombia, for example, suggest that citizens in all economic groups felt they were getting more out of paying taxes to local than to national governments (Acosta and Bird, 2005). As discussed earlier, if local administrators do a better job of identifying and assessing the tax base, overall revenue mobilization may increase. In Armenia, for example, the delegation of property tax collection responsibility to the local government level in 2003–2005 reportedly led to a 38 percent increase in collected tax revenue. On the other hand, the centralization of sales tax administration in Kyrgyzstan in 2009 resulted in a

M4439-BAHL_9781786435293_t.indd 213

01/03/2018 11:08

214

Fiscal decentralization and local finance in developing countries

decrease in the amount of collected tax (Golovanova and Kurlyandskaya, 2011). In Peru, those municipalities that created autonomous local tax offices on average raised their own-source revenues by 81 percent from 1997 to 2008, compared to an increase of only 61 percent in those that did not. The locally run offices were found to have improved in terms of less political interference, better client focus and more trust, less corruption, more innovation and better cooperation with other tax administrations (Fretes Ciblis and Ter-Minassian, 2015). Decentralized administration may also permit some new forms of taxation to be implemented. Sometimes regional and local governments have broadened the tax net through a variety of special tax instruments and administrative measures such as levies on the sales of assets, licenses to operate, betterment charges and various forms of property and land taxation (Bahl and Bird, 2008). As mentioned earlier, Bahl and Linn (1992) suggest that dividing tax bases according to comparative advantages in assessment and collection may broaden the tax base that can be effectively reached, especially in the so-called ‘hard to tax’ sector (Alm et al., 2004).67 In developing countries regional and local governments usually oversee a variety of licensing and regulatory activities, and may be better able than the central government to track property ownership and land-based transactions as well as to identify local businesses and gain some knowledge about their assets and scale of operation. Because the potential revenue gains are much more important for local governments, they have more incentive to attempt to capture small businesses and the self-employed who may not fully comply with national taxes or evade taxes altogether. However, even when there is a good case for local administration, some tasks such as valuing complicated properties or identifying the appropriate local tax base may be complex or require coordinated action between different local, state, and national agencies and departments. Even when the property tax is a local tax, the central (or regional) government often plays a substantial role in such tasks as setting valuation standards, training valuers and monitoring the quality of local assessments, especially when, as is common, intergovernmental transfers are based to some extent on estimates of the potential local tax base.68 Sometimes, though, central valuation agencies have shown little willingness to support local needs in improving and maintaining the local tax base unless there is some direct gain for them in doing so. The increasing ‘informatization’ of the world and the greatly expanded reliance on information technology (IT) to deal with routine administrative processes has been a two-edged sword when it comes to tax ­decentralization. On one hand, central tax administration can capture economies of scale through, for example, centralized data processing and

M4439-BAHL_9781786435293_t.indd 214

01/03/2018 11:08



Financing local and regional government ­215

record-keeping, uniform approaches to assessment and audit, the development of centralized training programs and so on (Vehorn and Ahmad, 1997). On the other hand, IT also makes it possible to achieve more uniform service levels more efficiently and fairly in a more decentralized fashion without requiring every locality to have highly specialized skills – in effect, enabling a country to multiply the skills available locally as well as monitoring outcomes more effectively (Bird and Zolt. 2008). Such a system may, for instance, be one way to check the common concern about excessive corruption at the local government level.69 Sometimes, both technical and political factors are binding constraints at the local level in developing countries. In the state of Lagos, Nigeria, for example, a study concluded that the degree of both corruption and evasion was so high that the best solution was a two-level approach of improving substantially the state’s tax administration (e.g. better training and more and better use of IT) while simultaneously devolving some administrative activities and outsourcing collection, and even some core administrative activities, to private agents (Enahoro and Olabisi, 2012). There is no costless way to address all the constraints binding local tax administrative capacity, but trade-offs are always possible. A common approach is to restrict state and local governments only to certain taxes. Mexico and Australia allow states to impose payroll taxes; Argentina and Canada let them impose certain types of sales taxes; many countries allow state and/or local governments to impose taxes on the ownership or use of motor vehicle licenses; and many permit state and local governments to impose some form of business licenses as well as property taxes. Too often, however, such levies are poorly designed (most presumptive levies), economically inefficient (property transfer taxes) or (like many property taxes) simply badly administered, with low coverage rates, arbitrary assessment and large delinquent lists. A different approach to reducing the costs of collecting local revenue is to change the structure of local taxes. One example we discuss further in Chapter 6 is using area-based assessments for property taxation instead of more sophisticated valuation approaches based on comparative sales values (Slack, 2006). Another is to impose a business license based on the estimated volume of sales rather than a sales tax based on actual sales records.70 Such shortcuts may make a tax easier to administer at the expense of making it less effective in other ways – for example, in the property tax case just mentioned by moving it from being a tax on property value, and hence reducing its potential role as a surrogate form of benefit taxation (Bird and Slack, 2014). A better approach may sometimes be for regional and local governments to piggyback on the tax base of higher-level governments, as Canadian

M4439-BAHL_9781786435293_t.indd 215

01/03/2018 11:08

216

Fiscal decentralization and local finance in developing countries

provinces do with respect to most of their taxes, thus in principle allowing them to be fully politically accountable without having to take on the task of tax administration. Accountability may be adversely affected as taxpayers no longer view the tax as local because it is centrally administered and collected.71 If the central government decides to alter the tax base – for example, to favor a particular industry or sector – local taxes are similarly affected (Martinez-Vazquez and Timofeev, 2010). Although this effect may be offset, the tax base is still essentially set at the central level and the lines of accountability are somewhat confused.72 Reforming taxes and tax administration is never easy but it can be done, as many countries have demonstrated.73 However, it takes time, patience and consistent support – none of which are readily available in many parts of the world. In South Africa, for example, a major source of revenue at the time the post-apartheid government took power was a local tax (the Regional Services Council levy) that was so poorly structured that it was doomed to fail at some point. It did, with the result that it was then abandoned completely, removing a significant revenue source from local governments – although it could have been saved by some restructuring (Bahl and Solomon, 2003). Despite numerous attempts, it has not proved possible to reassemble this Humpty-Dumpty once it was pushed over the wall, and South Africa’s burgeoning cities have been deprived of a broad-based, non-property tax revenue source to cope with their pressing spending needs (Steytler, 2013). Impatience can be fatal to the success of any aspect of fiscal decentralization including tax administration. Central politicians and officials, like local voters, seem often to expect too much to work too well too soon – and then to react too adversely when their unrealistic expectations are not met. Efficient administration does not always require that subnational taxes should be centralized. For example, a credit-invoice VAT might be ruled out as a subnational government tax in most low- and middle-income countries because a well-administered central government VAT is not in place or because the taxation of inter-local transactions is still well beyond the reach of most subnational governments. The same argument might be made in the case of a broad-based retail sales tax. However, as some countries have shown, ways can be found around many of the conventionally cited problems with subnational administration even of VAT, at least for larger and better-run subnational jurisdictions (Bird, 2015a). Even a less than perfectly administered subnational tax may be better than such alternatives as depending on even more distorting local taxes or on transfers or centralizing service delivery. In developing countries, broad-based taxes are generally the exclusive domain of the central governments, which can capture such economies of

M4439-BAHL_9781786435293_t.indd 216

01/03/2018 11:08



Financing local and regional government ­217

scale as centralized IT and record-keeping, uniform approaches to assessment and audit, and the development of centralized training programs (Vehorn and Ahmad, 1997). When taxpayers such as large companies operate on a country-wide basis, they can be more effectively taxed on a national basis. Central officials are unlikely to be willing or able to devote much time or effort to solving the problems of subnational tax offices that may need access to central records to do their jobs. Moreover, they may think that corruption is especially likely in local tax administrations where officials often know and may even live alongside significant taxpayers.74 Centralized administration carries some other advantages. For example, the distributional objective of income tax, though usually limited in developing countries (Bird and Zolt, 2005), is likely best served by central design and administration of the tax. Central governments also like to be able to regulate and influence businesses through company tax policy, while taxes on international trade are of course inherently centrally administered. Finally, since fairness in taxation requires uniform implementation of the tax code, a single (national) tax administration is again likely to be best. Central administrations do not do any of these things very well, but most local administrations are unlikely to do better. Often local taxes consist of an array of presumptive levies on retail shops, motor vehicles, factories, registered professionals and real property. Such taxes are often badly administered, with a low rate of coverage, arbitrary assessment and large delinquent lists. When combined – as is often the case with such specific levies such as entertainment taxes, advertising taxes, etc. – the result is an arbitrary, costly and not very effective local tax administration which imposes high compliance costs and may well drive business away (International Finance Corporation, 2011). Nonetheless, as argued earlier in this chapter, without a real local tax base, the linkage between subnational governments and taxpayers, and hence accountability, is substantially weakened. Countries have taken very different approaches to resolving this problem. For example, Russia reduced the list of local taxes from 22 in 2004 to 2 in 2005 (Golovanova and Kurlyandskaya, 2011), and China acted to reduce the excessive imposition of ‘extra-budgetary’ and other levies on rural taxpayers by a major ‘taxfor-fee’ reform in the early 2000s that essentially eliminated a wide range of sources of local finance (Bird et al., 2011). More recently, the Chinese have eliminated the business tax, the largest source of local government revenue (Bahl et al., 2014). Indian states similarly eliminated the octroi, a sort of internal customs duty that had been a major source of subnational revenue in many areas. Similar pressures are found even in developed European countries: Italy, for example, has several times moved to eliminate not

M4439-BAHL_9781786435293_t.indd 217

01/03/2018 11:08

218

Fiscal decentralization and local finance in developing countries

only a generally well-designed regional business tax (IRAP) but also local residential property taxes. Unfortunately, in none of these cases did central governments provide (or promise to provide) a better tax to replace those they abolished or wanted to abolish. A better way to deal with problematic subnational government taxes is to improve them. Ironically, one of the main examples along these lines was an earlier reform in Italy that introduced the IRAP – a tax subsequently emulated to a considerable extent in later reforms in Japan and France and seriously considered at one point in South Africa (Bird, 2014). Another example is the move in some countries to a simplified form of property tax using area-based assessments for property taxation instead of more sophisticated valuation approaches based on comparative values, as in Patna and Bangalore in India (Rao, 2008; Rao and Bird, 2011). Another is a business license based on the estimated volume of sales rather than a sales tax, whether classified by line of activity as in Kenya (Devas and Kelly, 2001) or as a gross receipts tax as in Colombia’s industry and commerce tax (Bird, 2012a). Of course, all these shortcuts are inferior in some ways to better-designed taxes. For example, if one objective of the property tax is to tax property value – whether as a (poor) proxy for a wealth tax or as a (perhaps somewhat better) benefit tax – area-based assessments clearly do not do the job, as we discuss further in Chapter 6.75 Perhaps the best way for subnational governments to reap the benefits of access to broad-based taxes without incurring the costs of administration and the dangers of fragmented administration is to piggyback on the tax base of a higher-level government. Provided they can impose their own tax rate, and are politically accountable for doing so, this seems the ideal solution, and is indeed the way such broad-based subnational taxes as income tax in Sweden and VAT in Canada are implemented. Of course, no solution is without its own problems. Taxpayers may not distinguish the central from the subnational tax, although they are perhaps more likely to do so when rates vary from region to region than when a uniform national ‘subnational’ levy is imposed, as is commonly done in tax-sharing arrangements.76 Moreover, so long as the central government sets the tax base, some of the cost of any tax preferences it grants may be passed on to the subnational governments, although this concern may be offset (as in Canada) by explicit compensatory provisions in the relevant intergovernmental agreement.77 Experts have frequently suggested some version of the piggyback approach in developing countries, but as yet no one has followed this path, perhaps because it implicitly assumes that subnational governments are in some sense more ‘equal’ to national governments than is true in most countries. Another approach to improving subnational tax administration stressed in almost every advisory report (but almost never implemented) is for the

M4439-BAHL_9781786435293_t.indd 218

01/03/2018 11:08



Financing local and regional government ­219

central government to invest heavily in improving subnational administrative infrastructure. All too often central governments refuse to assign significant revenues to local governments because the latter’s administration is too weak to do the job; they then ensure that this assumption remains correct by providing inadequate assistance for appropriate training programs and IT improvements, as well as failing to pass tougher enforcement laws to support collection efforts. Only experience and experimentation can give subnational governments the on-the-job experience necessary to improve their tax administration capacity. But everyone needs a leg up to begin with, and many localities do not have a leg to stand on. Rome was not built in a day, and its tax administration is no doubt still far from perfect. Still, Rome has long been one of the world’s great cities; millions have managed to live there not too badly, and much has been done to improve matters over time. Even in very poor countries, it is sometimes possible to improve local administration substantially in a relatively short period of time if the right people are willing to do the right things. In Sierra Leone, for example, a careful analysis of why decentralization worked well in some municipalities and not so in others points out the great importance of local characteristics such as history, social settings, and the will and capacity of specific political champions of reform (Jibao and Prichard 2015).

CONCLUSION Subnational governments in developing countries are beset with two problems on the tax side of their budgets. The first is that they have insufficient revenues to finance the services needed to upgrade the quality of life of their citizens and to keep their economic bases competitive. The second is the inadequate control they have over the revenue tools in their power. The two issues are related. On revenue adequacy, though there have been a few success stories, there has scarcely been any movement in the average share of revenue mobilization by subnational governments in the last two decades. One reason has been the disinterest and unwillingness of local political leaders to impose taxes on their constituents. Another is the lagging growth in the local formal sector. But perhaps the main reason has been the fear on the part of higher-level governments that letting subnational governments tax more would compromise their own revenue collections, which in most countries are concentrated in the largest urban areas. Instead of d ­ ecentralizing taxing power, many countries have instead continued to reform their system of intergovernmental transfers. As we discuss in Chapter 7, they

M4439-BAHL_9781786435293_t.indd 219

01/03/2018 11:08

220

Fiscal decentralization and local finance in developing countries

are right in seeing that there is a strong connection between transfers and subnational taxes, and that both need reform. But the problems of one will not be resolved by tinkering with the other. Urbanization may perhaps break the logjam on regional and local government revenue mobilization. Financing the expenditure demands in growing urban areas is usually more than central government revenues (which themselves have not risen significantly in the last two decades) can handle.78 Since much new revenue may end up financing urban services in any case, national governments may begin to see more clearly the wisdom of letting at least the larger urban areas look after themselves, a case we consider in more detail in Chapter 8. There are good reasons why central governments in developing countries are unwilling to give up much control over revenues. In addition to wanting more revenue for their own purposes, they are responsible for stabilization policy and have the last (and by far the most important) say on redistribution, so the more freedom they have on the revenue side the better. However, unless regional and local governments are given some degree of freedom with respect to revenues, including the freedom to make mistakes (for which they are accountable to their citizens), the development of responsible and responsive subnational governments is likely to remain an unattainable mirage. When government is as centralized as it is in most developing countries, the process of decentralizing taxing power is almost certain to be fraught with mistakes and excesses, especially in the early years of implementation – not least because subnational government officials have little experience of making independent tax policy decisions, and the learning curve may be gradual. Practice may not make perfect, but it helps a lot! However, regional and local governments seldom get passes for bad behavior from either their residents or the central government. Regardless of how badly central politicians and officials run things themselves – or the extent to which the problems underlying local failures often reflect deficient central policies – they often find it advantageous to attack mistakes made by subnational governments, or even the smallest whiff of corruption at lower levels. A common reaction is to slap heavier controls on local governments and to take away even the little ‘tax freedom’ that they may have, as illustrated by such cases as the abolition of the RSC levy in South Africa and the octroi in India, as well as the restrictions imposed on local property taxes in many countries. People are likely to be equally impatient when they feel that local taxes are unfairly administered or they are not getting adequate services for their money. How to get both central governments and local citizens to endure the inevitable period of trial and error of the learning curve needed for countries to get much benefit from

M4439-BAHL_9781786435293_t.indd 220

01/03/2018 11:08



Financing local and regional government ­221

decentralization is a problem few have solved. Paradoxically – or ironically, if one prefers – decentralization intended in large part to increase accountability may even prove to be particularly difficult to achieve in countries where governments are already more democratically accountable to their usually impatient citizens. As many now in power like to say, autocrats can act much more quickly and decisively than can democrats, precisely because they can do what they want rather than whatever it may be that most citizens want. Unfortunately, it is hard to cite many instances in which a developing country has stuck with decentralization long enough to have learned these lessons. The main lesson emerging from the long history of existing, relatively successful decentralized developed countries is that each case is very different and has developed in its own way and at its own pace. Moreover, as recent experience in countries such as the UK, Belgium, Spain, Canada and the US demonstrates, the process of institutional evolution is one that is unlikely ever to end because the reasons for and against decentralization, the interests supporting each side and the conditions in which the discussion takes place from country to country are continually evolving (Vaillancourt and Bird, 2016). The issues discussed in this chapter are thus unlikely ever to be fully settled to the satisfaction of all. But that, perhaps, is exactly as it should be.

NOTES  1. There are always exceptions to every generalization. For example, in the late 1980s some Argentine provinces paid some bills by issuing payment orders on public financial institutions under their control. While such ‘money’ was subject to Gresham’s law (bad money is driven out by good) and was palmed off whenever possible as worthless change to visiting foreigners who paid real money for services, it was practices like these that led to much of the early concern about the possibly destabilizing effects of decentralization expressed by such authors as Tanzi (1996), Prud’homme (1995) and Dillinger et al. (2003). A balanced appraisal of the impact of decentralization on stabilization policy may be found in Boadway and Shah (2009).   2. However, loans from national governments to local and regional governments – like foreign loans to national governments – are sometimes either never repaid or have such ‘soft’ terms that they are hard to distinguish from grants. We do not explore this issue in detail here: while we recognize that in practice it is not always easy to distinguish loans from unrequited transfers, when we use the term ‘loan’ we assume the amount is to be repaid at (more or less) market rates, in contrast to a grant or transfer that does not have to be repaid.   3. Elsewhere in this book, we spell out other ways in which local revenues may sometimes be usefully linked to specific local expenditures (for example, as with respect to infrastructure spending in Chapter 4).   4. Much of the argument in this and the next section is spelled out in more detail in Bird and Slack (2014). As Bird (2000) stresses, regional (state, provincial) governments do not usually fit neatly within the benefit box. Sometimes, as in federal countries, this middle

M4439-BAHL_9781786435293_t.indd 221

01/03/2018 11:08

222

 5.  6.   7.

  8.

  9.

10. 11.

12. 13. 14. 15.

16. 17. 18.

Fiscal decentralization and local finance in developing countries level of government has some relatively independent (sovereign) power bestowed constitutionally and may behave in some ways like a mini-central government. Sometimes, regional governments are essentially financed by central transfers and operate as little more than regional branches. In either case, regions often act as providers of services (and perhaps finance) to a group of local governments. We return to many of these points later. For a good discussion of how the benefit model fits the theory of tax assignment, see Martinez-Vazquez (2013a). For a good discussion of the Tiebout model and the research it has generated, see Fischel (2006). Obvious distributional questions may be raised about this argument, but for the moment we leave them aside. In chapter 7, we discuss the extent to which this goal can be achieved at the (economically relevant) margin by the appropriate design of local revenue and intergovernmental transfer systems even when a substantial fraction of local revenues in most localities are provided by transfers (Bird and Smart, 2002). When there are spillovers of benefits from one jurisdiction to another – for example, in the country mentioned earlier in the text, the reform of the school system resulted in many students living in one locality and attending school in another – arrangements must be made for appropriately matching financing with benefits. Set out explicitly in Olson (1969), this ‘fiscal equivalence’ approach is also central to the arguments of Bird et al. (2003). Note also that it holds important implications for the appropriate institutional setting within which intergovernmental transfers are designed and implemented, as discussed in Chapter 7. Higher taxes on business may be justified in some cases as an economically efficient way to tax the ‘rents’ that firms may reap from agglomeration economies, particularly in metropolitan areas. Bahl et al. (2002) studied the redistribution choices of state and local governments in the United States, and concluded that states view revenue and expenditure distribution policies as complementary and pursue distributional policies with both. Most state governments exempt food from their retail sales tax, largely to target low-income families for relief, and 34 states have progressive personal income tax rate structures (Fletcher and Murray 2008). In Spain, regional governments may adjust their individual income tax rate schedule provided that the schedule is progressive and has the same number of brackets as the central income tax (López-Laborda et al., 2006). For examples, see the detailed discussion of the incidence of the (then proposed) general sales tax in Jamaica in Bird and Miller (1989a) and Bahl (1991), and the well-known discussion of excise taxes by McLure and Thirsk (1978). The pros and cons of earmarking and the many varieties found in practice are outlined in Bird and Jun (2007). The difficult political economy setting usually confronting those advocating user charges is discussed in more detail (though in the context of a developed country) in Bird (2017). This argument (set out initially in Bird, 1984) assumes that local governments provide only local services and impose charges where possible on service beneficiaries (whether residents or not), and that central transfers are so designed that they compensate properly for spillovers and ensure the desired degree of equalization and redistribution. Each of these points is discussed in more detail elsewhere in this book. See Smolka (2013). Colombian cities have, with varying degrees of expertise and success, made extensive use of such a system to finance certain urban public works for almost a century (Rhoads and Bird, 1967). Oates (1996) sets out the benefit case for imposing taxes on non-resident beneficiaries; see also Martinez-Vazquez (2013) as well as Box 5.1. As one example, the revealed preference of politicians, both local and provincial, to tax non-residential property much more heavily than residential property is documented in detail for a Canadian province in Bird et al. (2012). If local politicians are influenced

M4439-BAHL_9781786435293_t.indd 222

01/03/2018 11:08



19.

20. 21. 22.

23.

24. 25.

26.

27.

28.

29.

Financing local and regional government ­223 by money rather than votes, they are probably less likely to impose differentially heavy taxes on business. Arguably, much the same can be said with respect to pricing the use of congested roads and streets, something that technology makes increasingly easy – though few countries or cities have yet had the courage to emulate Singapore’s early leadership in this respect. We return to this issue in Chapter 8. For useful reviews and discussion of this topic, see Fischer and Easterly (1990), Rodden et al. (2003), Vigneault (2005), and Liu and Waibel (2010). For a broader discussion of how one may assess the fiscal health of subnational governments, see Bird and Slack (2015) and references cited there. There is a gray area here because some transfers in some countries are distributed on an ad hoc basis, so that the annual entitlement of the local government is determined each fiscal year. In other cases they are explicitly deficit grants. As we discuss in Chapter 7, when transfers are adjusted to fill a current expenditure gap that would otherwise arise, they soften the budget constraint. In the presentation here, we simplify by assuming that all transfers are ‘predetermined’ for each period by a higher-level government – i.e., they are unaffected by the current behavior of the local government. Much local government financing is on a cash rather than accrual basis, so SC is often directed to a ‘reserve fund’ where it can later be used to cover future current expenditure needs or invested, although we do not go further into such details here. The considerable virtues of the accrual approach are laid out in Cavanagh et al. (2016); but they note, as we did in Chapter 3, that an essential prerequisite is a solid cash accounting system – something not yet achieved by many regional and local governments in developing countries. To simplify, we assume that revenues from the sale or lease of assets are earmarked for capital spending. If the problem is simply a cash flow problem – for example, the timing of tax receipts may not match the fiscal year or the project spending may be behind schedule – the situation may sometimes be handled by providing access to a short-term loan (less than one year). Alternatively, if the deficit results from a natural disaster or other development completely beyond local control, a special relief transfer may be given. Earlier, we mentioned the importance of the linkage between local budgets and intergovernmental transfers, an issue discussed further in Chapter 7. It is also important to consider the variety of ways that the level and structure of local budgets are constrained in different countries (e.g. Dafflon, 2002) as well as the many ways in which subnational borrowing is often constrained. We discuss this issue only very briefly in this book: for more thorough treatments, see e.g. Poterba (1996), Rattsø (2002), and Canuto and Liu (2013). Some years ago one of us spent several days going through the records of the public works department of a Latin American government to find that the accrued liability from unpaid prior accounts (mainly bills from private contractors) exceeded by a substantial margin the total budgeted capital expenditure for the current year. In China, for example, local governments faced huge infrastructure needs to accommodate the rapid urbanization of the past two decades. Since they had no borrowing powers and no recourse to raising own-source revenues, they invented a system of using the proceeds from the sale of government-owned agricultural land to collateralize loans issued by local investment companies. The plan was seriously flawed and not sustainable, but much of the needed infrastructure was built remarkably quickly (World Bank, 2012; World Bank and Development Research Council, 2013, 2014; Bahl et al., 2014). For discussion of this and alternative institutional solutions in the context of Argentina some years ago, see World Bank (1996a, vol. 2). Bednar (1999) provides a useful general discussion of the difficult problem of encouraging better outcomes in a heterogeneous intergovernmental environment. The same source also discusses how to avoid (in effect) ‘bankruptcy’ at the regional level in a federal setting; see also the case studies in Rodden et al. (2003).

M4439-BAHL_9781786435293_t.indd 223

01/03/2018 11:08

224

Fiscal decentralization and local finance in developing countries

30. For a discussion of Musgrave’s major contributions to the tax assignment question, see Bird (2009). 31. Perhaps paradoxically, most discussions of fiscal federalism are more suited to a unitary than to a federal or highly decentralized country (Bird and Chen, 1998) and give little guidance about how taxes should be divided between the regional (provincial, state) and local government levels. 32. This familiar proposition has become one of the main tenets of ‘second-generation’ theories of fiscal federalism (Weingast, 2009). 33. For a related but distinct characterization of different models of local government finance, see Bird (2011a). 34. This ‘dual subordination’ system is discussed for the Russian case in Martinez-Vazquez et al. (2008), and for China in Bahl and Wallich (1992) and Bahl (1999); the German system is described in Ulbricht (2008), and its effects analyzed in Baretti et al. (2002). 35. The allocation between subnational governments may be made on either a derivation (source) basis or on the basis of a (more or less) equalizing formula, or in an ad hoc way as discussed in Chapter 7. 36. Of course, subnational governments may still be able through their policies to promote the growth of the local tax base, and thus the level of revenue that they receive. As Qian and Weingast (1997) suggest, this result may perhaps be a good thing from the perspective of encouraging economic growth. However, as with other ways in which subnational governments can affect outcomes, it may also lead to ‘gaming’ the system, particularly if the intergovernmental grant system is not well designed. Chinese local governments, for example, used a ‘backdoor’ approach to increase the share of tax collections, and directed these to local extra budgetary accounts (Bahl, 1999; Wong et al., 1995). 37. As with virtually every generalization in this field, however, there are exceptions: states in Mexico and Australia levy payroll taxes; excise taxes are the main regional government tax base in Colombia; and income taxes are a major subnational revenue source in many OECD countries. 38. For example, China eliminated its local business tax, in large part because enterprises in the VAT system were not receiving full credit for taxes paid on their input purchases. But other countries cannot get past the local autonomy issue; e.g., the turnover tax and the VAT still coexist in Argentina. 39. As discussed in chapter 2, the available data (e.g. IMF) exaggerate the actual taxing autonomy of subnational governments in both OECD countries (Stegarescu, 2005) and developing countries (Ebel and Yilmaz, 2003). 40. Musgrave (1983, p. 11) argued that, in principle, “tax bases which are distributed highly unequally among sub-jurisdictions should be used centrally”; but clearly the weight attached to this argument in any country depends heavily on the historical and institutional setting. 41. For a detailed consideration of the stabilizing effect of the property tax in the UK, see Muellbauer (2005). See also the discussion in Chapter 6 below about the experience with the property tax in the aftermath of the 2009 recession. 42. In Mexico, for instance, subnational governments covered their projected deficits during the 2009 recession with a combination of short-term borrowing that was collateralized with future flows into their revenue stabilization fund and other short-term borrowing that was not registered at the Ministry of Finance (Revilla, 2012). 43. See the case studies in Martinez-Vazquez and Vaillancourt (2008) as well as Chapter 2 above. 44. As Jibao and Prichard (2015) show for Sierra Leone, it is not sufficient for national governments to give a green light to local tax reform: local governments need to heed the signal and act on it effectively for local tax reform to be successful. For this reason, the effectiveness of any local tax reform inevitably varies across jurisdictions, as we discuss later (especially in Chapter 8). 45. For a nice example of this argument, see Winer (1983).

M4439-BAHL_9781786435293_t.indd 224

01/03/2018 11:08



Financing local and regional government ­225

46. To take an example from perhaps the most democratic tax system in the world, consider the careful discussion by Dafflon and Daguet (2012) of the precision and care with which Swiss legislators and courts formulate and interpret the laws establishing taxes and user charges and the detailed accounting and legal requirements imposed to ensure that they are properly implemented and fully reported to the public. 47. Many aspects of a more open budget process are discussed in Khagram et al. (2013); see also www.internationalbudget.org/opening-budgets/open-budget-initiative/. 48. On the other hand, there is little evidence that the major early efforts to increase ‘tax consciousness’ through educational programs in the Philippines (as noted in Romualdez et al., 1973) had any discernible effect on levels of public understanding and discussion. 49. Of course, the economic incidence of these taxes may spread more widely across the economy. 50. They may have the same view about higher central government taxes but feel that central government decisions are too far removed for them to have an influence. 51. In one small West African country with exceptionally poor roads, for example, the small, well-off business class had no problems getting about because they had expensive and comfortable four-wheel drive vehicles. Similarly, they did not care about poor local schools because their children went to private schools. 52. Higher joint tax rates and dual administration may also result in higher compliance costs. For example, Plamondon and Zussman (1998) estimate that a single administration of the Canadian federal and provincial business taxes would reduce compliance costs by 1.3 percent of collections. 53. As Bednar (2009) stresses, the possibility of such retaliation is in some ways the ultimate safeguard against interjurisdictional competition, especially when in a federal country such competition amounts to shirking constitutionally assigned responsibilities. 54. As many analysts and reports have noted: see for example Piffano (2005) on Argentina and Bird (2012a) on Colombia, as well as a recent expert report to the Colombian Ministry of Finance (Comisión, 2015). 55. Wong (1998) reported that McDonald’s restaurants in Beijing paid 31 fees on average, of which only 14 were legal. But abolishing these fees was not easy because many financed essential services, including the wages and salaries of local government employees. 56. For such an analysis for Indian states, see Garg et al. (2017). 57. Recent extensive empirical investigations of this and other effects of fiscal decentralization on interjurisdictional (as well as interpersonal) equality in OECD countries illustrate both the complexity of the issues and the tentative nature of the conclusions that emerge even when data problems are much less severe and countries much more similar than those in the developing world. For a useful example of this work, see Bartolini et al. (2016). 58. For detailed reviews of subnational government taxes from a more conventional economic perspective, see e.g. Bird (2011) and Bahl (2011a). 59. For a more favorable appraisal of such taxes, see Engelschalk (2004). 60. For a more detailed discussion of local business taxation around the world, see Bird (2006a). 61. For an extended discussion, see Bird (2015), and Martinez-Vazquez and Timofeev (2010). 62. In Indonesia, for example, local administrative costs eat up over 50 percent of the revenue collected (Lewis, 2006); in Argentina, administrative costs at the provincial level vary from 1.1 percent (less than the 2 percent at the national level) to over 10 percent (Fretes Cibils and Ter-Minassian, 2015). 63. As noted earlier, unified administration of federal and provincial corporate income taxes was estimated to reduce compliance costs by 1.3 percent of the amount collected (Plamondon and Zussman, 1998). A decade after this study, when Canada’s largest province (Ontario) shifted administration of first its corporate income taxes and later its sales taxes to the federal tax office, substantial reductions in administrative costs were cited as a key reason for doing so.

M4439-BAHL_9781786435293_t.indd 225

01/03/2018 11:08

226

Fiscal decentralization and local finance in developing countries

64. Local experts, who know the terrain, may seem best equipped to decide such matters but are often too attached to the status quo and too encumbered by politics. Outside ‘fly-in’ experts too often come equipped with knowledge of their favorite country’s system and are tempted to replicate it rather than learn enough to design for local success. 65. The compliance costs shown in the well-known World Bank (annual) ‘Paying Taxes’ publication, for example, are substantially influenced by the number of administrative offices involved in tax collection. 66. Although, as the province of Quebec in Canada has demonstrated, some regional governments can do so. Indeed, as we discuss above and as some countries (Japan, Italy, France) have shown, even subnational governments can use a different type of VAT effectively (Bird, 2015a). 67. Segmentation of tax bases in terms of both structure and administration is also suggested in a more recent review of how to improve local government taxation in Africa (Fjeldstad et al., 2014). 68. See, for example, the discussions of valuation in Bird and Slack (2004) and of transfers in Bird and Smart (2002). 69. As noted in Box 3.4 earlier, however, clever people can sometimes defeat clever use of IT. 70. This is how the industry and commerce tax operates in Colombian municipalities (Vázquez-Caro and Ospina, 2006). For a discussion of the single business permit in Kenya, see Devas and Kelly (2001). 71. This shortcoming might be addressed by providing regular updates on the success of the piggyback rate in revenue mobilization, and by linking the rate decisions to local and regional-level councils. 72. In Canada, for example, the central government is required to adjust intergovernmental transfers if central tax changes have a marked effect on provincial revenues. To a limited extent, provinces may also impose credits and surtaxes to offset federal base changes. 73. See, for country examples, the cases in Thirsk (1997), as well as Slack and Bird (2014) on local tax reforms. 74. Once one of us was visiting a local tax office in a developing country with a senior official from the central revenue department when the local official flatly refused to let his superior see any files of local taxpayers: “They are my taxpayers,” he said. The central official did not argue, explaining later that the local official in question was well connected and could not be touched. Such situations can and do set the stage for corruption. 75. Of course, it is not difficult to see how one can impose area-based assessments initially and then gradually move to what is in effect a value-based tax, as spelled out for example in Slack et al. (1998). 76. Another commonly expressed concern about piggybacking, that the central government may not be as aggressive in enforcing local as central taxes, is unlikely to be of any importance since the two are collected simultaneously and in the same way. Of course, if fiscal times are tight, there is always the possibility that the subnational share may not be paid on time. 77. While we do not discuss this point further here, the result of providing a transfer to offset the effect on local budgets of a central change in local taxes is equivalent in a sense to what in the US context is often called a ‘funded mandate’ – that is, a central law that requires subnational governments to do something while reimbursing the costs. Most such laws, of course, as Inman and Rubinfeld (1997) stress, are unfunded in the US, as elsewhere. 78. See, for example, the rising costs of infrastructure discussed in Chapter 4.

M4439-BAHL_9781786435293_t.indd 226

01/03/2018 11:08

6.  Taxing land and property The classical economists of the early 19th century, beginning with David Ricardo, recognized that in theory, the land value tax was almost the perfect tax. (Netzer, 1998, p. x)

Most proposals to reform local government finance in developing countries include upgrading the property tax. The tax seems well suited to local governments, it is generally assigned to them and it is almost always poorly structured and administered. External donors have invested significant resources in strengthening the capacity of local governments in low- and middle-income countries to levy the property tax, and new technology has made it cheaper to administer effectively. Yet property tax revenues continue to be less than 1 percent of GDP and less than 4 percent of tax revenues in most developing countries. Despite the mountain of literature on this subject (some of it by us), property tax is still for the most part a non-starter.1 Certainly, there is no groundswell of popular support for more reliance on property tax to finance local government services. Indeed, taxpayers, officials and politicians seem to be united in not liking this tax, for a variety of reasons: ●●

●●

●●

The property tax is highly visible, so most taxpayers know what they pay. Since people usually feel that they pay more in taxes and charges than they get back in services, and local governments do not want to raise expectations about public service levels, neither side is likely to object if tax liabilities are made less transparent. Academics may like transparency in the tax system, but it is not clear that anyone else does. The property tax is inherently presumptive. Both the notional definition of the tax base and the judgmental nature of the ­assessment – how much would your house sell for if you sold it, or what is the normal rent that might be paid for the flat that you own? – are objectionable to most taxpayers. Most people think they have a better sense of the value of their property than some government fellow with a clipboard, a calculator and a book of reference values. Moreover, since the usual value-based property tax is to some extent a tax on unrealized increases in wealth, taxpayers seldom agree that their capacity to pay has increased as much as their tax bill. 227

M4439-BAHL_9781786435293_t.indd 227

01/03/2018 11:08

228

Fiscal decentralization and local finance in developing countries

There are significant political obstacles to ‘improved’ property taxation. Many countries have had past experience with property tax reforms that were introduced with great expectations but yielded few positive results. The list of mistakes is long: badly structured reforms that failed; wellstructured reforms that failed; well-intentioned but often unwise investments to improve valuation systems that ignored other key elements of the tax; and conceptually attractive ideas that came to nothing for political reasons. Perhaps a bigger reason is that it is difficult to believe that the answer to local and regional government financing problems in any country lies solely in taxing land and property more effectively. In particular, as we have argued extensively elsewhere and discuss further in Chapter 8, big cities (and regional governments with major roles in delivering services like education and health) are most unlikely to be able to perform their assigned tasks adequately without access to a wider portfolio of taxes (Bahl, 2011, 2017; Bird and Slack, 2013). Nonetheless, the property tax has many desirable features as a subnational tax: it approximates a benefit levy for some local services; it is generally not regressive; it is less distorting than income or consumption taxes; it has significant revenue potential; and it can be administered effectively. Moreover, in most countries it already exists as a local tax: taxpayers may not see it as a friendly devil, but it is a familiar one. We argue in this chapter that the key to better use of the property tax in developing countries is to ensure that it is structured to fit the circumstances of the country, and then to establish a plan to shape its evolution over time into an efficient and effective local tax. The property tax may not be the whole answer to local fiscal problems, but it is an important part of the answer. We begin by discussing some reasons why the conventional wisdom about property taxation seldom travels well to developing countries. Next, we explore the variation in the revenue performance of the property tax to see if there are patterns that might explain why some countries do better than others. We then review the key elements of the tax – base, rates and ­administration – before turning to some other ways of taxing real property, such as the (more popular but much less desirable) property transfer tax and alternative ways of capturing the increases in land values that generally accompany increased urbanization. We conclude with a sketch of how the property tax in developing countries can be moved closer to realizing its revenue potential.

M4439-BAHL_9781786435293_t.indd 228

01/03/2018 11:08



Taxing land and property ­229

WHY CONVENTIONAL WISDOM DOES NOT WORK Apart from the politics, one important reason why taxing land and real property has not been the winner that conventional economic wisdom suggests it should be is because its advantages have often been overstated. Economists have, for example, long argued that the distorting economic effects of property taxes are lower than most other taxes, essentially because the tax base is immobile and hence relatively inelastic (unresponsive) to tax. There is good empirical evidence supporting this statement in some high-income countries (Johansson et al., 2008). However, conditions are different in developing countries in several important respects (Bird and Slack, 2004). For example, much of the property tax base in most developing and transition countries consists of commercial and industrial structures which are usually taxed much more heavily than the residential sector. Since many such activities are mobile, increasing effective property tax rates may drive investment to jurisdictions with lower taxes or to sectors other than real estate – although, given the low effective rates of tax in most such countries (discussed below), such effects are unlikely to be very marked. As we noted in Chapter 5, to some extent taxes on land and property may be considered to be ‘benefit taxes’ on properties that benefit from public services because the value of such services is usually at least in part capitalized into increased property values (Vickrey, 1963). In developing countries, however, not only are such services often less beneficial but they are also less likely to be reflected in higher (taxable) land values, largely because assessments are usually dated and not reflective of market values. Despite the usually low level of administration, and contrary to what many believe, the burden of the property tax is basically progressively distributed because the average national tax on property falls on owners of capital,2 and the rich have more capital than the rest of us.3 However, the fact that most owner-occupied properties are exempted or subject to very low rates, combined with the fact that property values are generally underassessed, substantially weakens the progressivity of the tax in most developing countries. Moreover, those owning property of the same value (or paying the same rent) seldom pay the same tax. Because most taxes are riddled with exemptions and preferential treatments – ironically, often in the name of progressivity – and assessment is often based on physical area, building material and age of the asset rather than on value, the property tax in most developing countries is seldom horizontally equitable. Finally, even a well-designed, well-administered property tax may not have great revenue potential. Gravelle and Wallace (2009) estimate that the net stock of residential fixed assets in the US is equivalent to about

M4439-BAHL_9781786435293_t.indd 229

01/03/2018 11:08

230

Fiscal decentralization and local finance in developing countries

1.3 times the level of GDP. We do not have good, comparable data for developing countries, but it seems likely that this potential tax base is smaller in lower-income countries.4 Moreover, real property wealth is harder to tax in low-income countries for many reasons: ‘produced capital,’ which includes structures and urban land, accounts for only 16 percent of all wealth (World Bank, 2006); the quality of information on property values is weak; governments have few competent appraisers; land titling is often in disarray; and collection rates are low. The potential base for any property tax – and hence its yield – is unlikely to be large. The conditions for real property taxation in developing countries are thus considerably less favorable than they are in high-income countries, many of which have themselves not been all that successful in imposing significant taxes on land and property (Slack and Bird, 2014). Property markets are not well developed, there is a paucity of reliable evidence on transaction values, administrative capacity is limited and no one wants to pay taxes for services that are often inaccessible to many and usually poor for all. If one adds to the mix numerous legal exemptions, as many countries do, the result – low tax yields – is almost inevitable. As countries develop, property markets will develop, urban land will become more valuable and good property tax practices may gradually be implemented. Until then, however, the property tax in most developing countries is likely to remain stuck at its present low-level equilibrium. This chapter shows how such countries may be able to move towards a more efficient and productive property tax.

REVENUE PERFORMANCE In most low-income countries property taxes are such an inconsequential source of revenue that, if eliminated, they would hardly be missed – at least at the national level. IMF data on the revenue yield of property tax in developing countries – the best comparable data available – show that the average is only about 0.6 percent of GDP (Table 6.1). The true average is likely even lower because some countries with very low property tax revenues are excluded from the IMF figures. The data in Table 6.2 compare the ratio of property tax revenues to GDP for 20 countries, including many not included in the IMF data, with most of these being below the 0.5 percent level. Property taxes are clearly not important in most developing countries. Since on average the ratio of total taxes to GDP in developing countries is only about 16 percent (Bahl, 2014), taxes on property provide about 3 percent of total tax revenue in these countries. By contrast, in OECD countries, where the average total tax ratio is about

M4439-BAHL_9781786435293_t.indd 230

01/03/2018 11:08



Taxing land and property ­231

Table 6.1  Property tax as share of GDP (%)

OECD countries   (number of countries) Developing countries   (number of countries) Transition countries   (number of countries) All countries   (number of countries)

1970s

1980s

1990s

2000s

1.24 16 0.42 20 0.34 1 0.77 37

1.31 18 0.36 27 0.59 4 0.73 49

1.44 16 0.42 23 0.54 20 0.75 59

2.12 18 0.60 29 0.68 18 1.04 65

Note:  The average for the 2000s is for the years 2000 and 2001. Source:  Bahl and Martinez-Vazquez (2008) based on data from IMF GFS (various issues).

34 percent, taxes on property provide 6 percent of the total and are thus approximately four times as important as a share of GDP (OECD, 2015a). But while taxes on land and property are relatively unimportant nationally, they are often considerably more important at the local level. For example, Thailand’s house and buildings tax accounts for about 80 percent of local government own revenue (Varanyuwatana, 2004), and the property tax is more than 20 percent of local government revenues in Kenya and Senegal (Franzsen and McCluskey, 2017), 36 percent in Chile and 40 percent in Poland. In the 36 largest cities in India, the property tax accounts for 28 percent of own-source revenue (Mathur et al., 2009). De Cesare (2012) reports a survey of 64 municipalities in Latin America that shows the property tax to account for an average of 24 percent of local government tax revenue. Using panel data for 70 developed and developing countries and treating the property tax share as endogenous in the model, Bahl and Martinez-Vazquez (2008) find that higher levels of fiscal decentralization drive up the use of the property tax. This suggests that one important reason why property tax revenues are less important in low- and middle-income countries is that such countries tend to have more centralized fiscal structures. If so, perhaps all that anyone who likes the property tax may have to do is simply wait. As countries become more developed, local governments generally play a larger role in service delivery, and may also rely more on property taxes for finance. The property tax may perhaps be seen as one important way in which the growing urban middle class, which usually drives the expansion of local public services, contributes to financing those services. If so, countries where there is not much of a middle class – for example, much of sub-Saharan Africa – may not see much growth in

M4439-BAHL_9781786435293_t.indd 231

01/03/2018 11:08

232

Fiscal decentralization and local finance in developing countries

Table 6.2  Property taxation in selected countries Country (year)

Property tax as % of GDP

Argentina (2007)

0.32

Bolivia 2007)

0.62

Brazil (2007)

0.43

Colombia (2007)

0.54

Mexico (2007)

0.09

Chile (2010)

0.80

Indonesia (2009)

n.a.

Jordan (2010)

0.46

Panama (2010)

0.43

Ecuador (2008)

0.14

M4439-BAHL_9781786435293_t.indd 232

Comments

Source

Progressive rates, collection by Local Governments Progressive rates and valuation set by central government Rates set by local governments within a range Large local governments set rates within a range and do valuation Rate and base determination varies by state

Fretes Ciblis and Ter-Minassian (2015) Fretes Ciblis and Ter-Minassian (2015)

Central government tax policy and administration Local governments set rates and determine values Center sets rate and base, local governments collect Progressive rates, collection by local governments –

Fretes Ciblis and Ter-Minassian (2015) Fretes Ciblis and Ter-Minassian (2015) Revilla (2012); Fretes Ciblis and Ter-Minassian (2015) De Cesare (2012)

Haldenwang et al. (2015) Bahl (2010)

Bahl and Garzon (2010)

De Cesare (2012)

01/03/2018 11:08



Taxing land and property ­233

Table 6.2  (continued) Country (year)

Property tax as % of GDP

Guatemala (2008)

0.16

Paraguay (2008)

0.39

Peru (2008)

0.16

Uruguay (2008) Pakistan (2008) India (2007)

0.71 0.10 0.20

Liberia (2011)

0.17

Uganda (2012)

0.15

Mozambique (2006)

0.32

Egypt (2010)

0.72

Comments

Source

Tax rates are centrally set, valuation and collections are municipal Rates are set centrally, collection is local Tax rates and bases are set centrally – – Local governments set rates within bounds Centrally set rates and collections Centrally set rates, local collections Centrally set rates, local admin. Central rates and administration

De Cesare (2012)

De Cesare (2012)

De Cesare (2012) De Cesare (2012) Bahl et al. (2015) Mathur et al. (2009) Franzsen and McCluskey (2017) Franzsen and McCluskey (2017) Franzsen and McCluskey (2017) Franzsen and McCluskey (2017)

property tax revenues until growth and urbanization increase to levels similar to, say, many countries in Latin America in which the middle class has become a more decisive factor in political and fiscal decisions.5 But even with urbanization and economic growth, the property tax may not become a dominant revenue source because it may be crowded out by other sources of revenue, especially in larger urban areas. Argentina, Brazil and Colombia are cases in point. An important barrier to more use of property taxes is that the tax is

M4439-BAHL_9781786435293_t.indd 233

01/03/2018 11:08

234

Fiscal decentralization and local finance in developing countries

not cheap to administer. Few countries have the requisites for a good ­administration: a full and up-to-date survey of all land (urban and rural); records of title that would enable determining tax liability; reliable data on the sales price of properties; and good valuation expertise. Bahl et al. (2011) argue that given the current yield of the property tax in many countries, it is difficult to justify the expensive outlays needed to bring all these things up to par. As a result, when developing countries do try to reform the tax, they usually make only marginal upgrades to its administration rather than undertaking a full scale upgrading of the tax register and a comprehensive revaluation. Unfortunately, the result is that the revenue impact of reform also tends to be marginal, especially when, as is usually the case, the basic design of the tax remains riddled with exemptions, reliefs and preferential treatments. A further problem is that, in most countries, decisions on both policy and administration are made by central governments that generally care little about increasing taxes that largely go to urban local governments. In Indonesia, for example, the property tax and property transfer tax were shared responsibilities between the central and subnational governments until 2009, when they were devolved. When policy and administration were centralized, there was little or no growth in property tax revenue. A key element in the reform was to empower local governments to adapt their property tax structures and property tax administration systems to the local environment. Although the early results in the post-reform period were promising in terms of revenue mobilization, good implementation takes time and most local governments were slow to move their property tax towards its full potential, so it is still too early to evaluate the long-term effects of this change (Kelly, 2015; Haldenwang et al., 2015). Another explanation for the slow growth in property tax revenue may simply be that local governments are able to find less politically painful ways to secure finance. Property taxes are hard to increase without rate increases or costly revaluations that are highly visible to businesses and middle-class homeowners. If subnational governments have access to less visible and more elastic sales taxes, they are more likely to turn to this revenue source. In Argentina, indirect taxes (primarily the turnover tax) account for about twothirds of subnational government tax revenue, while property tax accounts for only about 12 percent. In Brazil, local sales taxes raise twice as much revenue as local property taxes. Colombia’s larger cities raise more from the industry and commerce tax (gross receipts tax) than from the property tax. In Mumbai (India), where the property tax is 24 percent of local government revenue, the octroi (a form of sales tax) until recently provided 44 percent of revenue and its revenue elasticity was significantly larger (Pethe, 2013). More importantly, however, because most people appear to believe that

M4439-BAHL_9781786435293_t.indd 234

01/03/2018 11:08



Taxing land and property ­235

grants from a higher-level government are free (to them), residents and local politicians almost always prefer intergovernmental transfers to the property tax. In Mexico, for example, intergovernmental transfers have crowded out nearly all subnational government taxes. An econometric analysis in Argentina concluded that increased transfers reduced the share of property taxation in the subnational government revenue structure, suggesting that transfers substitute for property taxation (Artana et al., 2015). Interestingly, however, in a similar study for Colombia, Sánchez Torres et al. (2015) found that increased grants appeared to stimulate local tax efforts, so to some extent the jury is still out on this issue.

THE BASE OF THE PROPERTY TAX A good property tax in any country rests on three pillars: the choice of an appropriate tax base, a sound rate structure and a good administration. In most developing countries, the administrative problems are both obvious and numerous. However, before dealing with administrative issues, it is critical to be sure that both the base and rate of the tax are right. Two basic rules should be kept in mind. First, a simple tax structure makes it much easier to administer the tax properly; a complex system with many exemptions and differentiated rates makes it difficult. Second, the more closely the tax fits the context in which it operates – the land tenure arrangements, the tax ‘culture’ and the resources available to administer it (human or otherwise) – the more successful it will be. As always, there is no one size fits all solution. The legal structure and implementation of property taxation vary widely from country to country. The tax may be levied by local governments (Indonesia), regional governments (India) or essentially by the central government (Chile). Different countries tax different bases, use different rate structures and provide different types of exemptions. The main thing they have in common is that they do not raise much money from the property tax. Four distinct approaches are employed to determine the base of the property tax in developing countries: (a) annual rental value, (b) capital value of land and improvements, (c) capital or rental value of land only, or (d) physical area. Sometimes a mixture of these bases is used. In a survey of 122 countries McCluskey et al. (2010) report that 52 countries use some form of capital value, 37 use annual rental value and 16 have some form of unimproved value base (site value or land value), with many variations within each of these general categories. For instance, four countries tax only improvements (mainly structures), 44 use an area basis for determining the taxable amount (measured in capital or annual rental terms) and several allow subnational governments to choose their base. Moreover, the

M4439-BAHL_9781786435293_t.indd 235

01/03/2018 11:08

236

Fiscal decentralization and local finance in developing countries

legal tax base may be assessed in different ways in different countries: some rental value countries use capital value assessment methods, some capital value countries use area-based assessments and so on. The four main assessment methods are discussed in Box 6.1. Which Tax Base Is Best? There is no easy or generally agreed answer to this question for developing countries. If ‘best’ is defined as the most common practice in low- and middle-income countries, the answer would seem to be a capital value system where both land and improvements are valued and taxed. Capital value features continue to creep into the other systems, either in structure or assessment practices. Both the UK and South Africa switched to a capital value system in recent years. On the other hand, theory suggests that taxing land value alone is the best approach. The argument is straightforward. If the property tax is levied only on the land, the owner will have an incentive to invest in improvements that maximize the return on the property. In contrast, with a capital value system that covers both land and improvements, investment in improvements is discouraged because they increase taxes. Advocates argue that taxing land alone will encourage more efficient land use, thus reducing public expenditures and increasing economic development, and hence the potential tax base. Moreover, a land value tax will be more progressive in its distribution of burdens and less costly to administer because improvements are excluded from the tax base.6 However, this argument assumes that the land value tax is levied at a sufficiently high rate to induce a major improvement in land use patterns. The empirical evidence on this is not very helpful. The imposition of a higher rate on land than on improvements in some cities in the US state of Pennsylvania does not seem to have had much effect on land use choices (Oates and Schwab, 1997). One review of studies on this subject in the US suggests that “moving toward land value taxation may increase building activity, or at least does not decrease it” (Anderson, 2009, p. 118). While there is little evidence of the impact of land value taxation in low- and middle-income countries, it is clear from a wealth of anecdotal evidence and observation that landowners take account of property tax burdens in making investment decisions. As Bird and Slack (2004) note, the original arguments for site (land) value taxation (George [1879] 1958) were made in a context in which cities were growing rapidly in the United States. Land that was worthless one day was worth a fortune the next, owing largely to the rapid influx of population. Proposing to tax the ‘unearned increment’ seemed to be both sensible and fair. Perhaps the same might be said now in developing countries undergoing rapid urbanization.

M4439-BAHL_9781786435293_t.indd 236

01/03/2018 11:08



Taxing land and property ­237

BOX 6.1  DETERMINING THE TAX BASE Rental Value Many countries – Côte d’Ivoire, India, Malaysia and Trinidad are examples – base their property tax on the annual or ‘rental’ value of properties. In principle, the tax base is the rent that can be reasonably expected in a fair market transaction between owner and renter, i.e., it is a notional rather than an actual amount. Whether the base that is taxed approximates market rents depends on the method of assessment used. Hong Kong’s tax does because it uses both data on actual rents paid and income statements from landlords to develop estimates of the tax base (Brown, 2013; Pang, 2006). In most countries with this system, however, the tax base bears little relationship to market rents. Some countries require submission of rent contracts. Some use evidence pulled together by experts, usually private valuers or real estate sector professionals, which may sometimes be close to the mark. Others rely on declarations of rents paid, which, like most taxes on honesty, do not work very well. In addition to the general problem of insufficient or unreliable ­evidence, the rental value approach is hard to apply to industrial or other properties where there is no active rental market. Most countries use capital value assessment methods, such as depreciated replacement cost, for factories and larger c­ommercial ­establishments, and then some capitalization rate – for example, 10 percent in Malaysia but 3–8 percent in Senegal – to convert these capital value estimates to a rental value equivalent. Another common problem is dealing with vacant land, unoccupied properties and premises subject to rent control. Often a different regime is applied to unoccupied properties and vacant land: for example, the Central African Republic taxes vacant properties at twice the rate of developed properties; in contrast, Jordan taxes vacant land at a differentially lower tax rate. In India, where the courts ruled that the controlled rent was the proper basis for taxation for many years, the revenue potential and the equity of the property tax were significantly dampened (Nath, 1987). Finally, after nearly a century of protracted legal battles, a few local governments, beginning with the city of Patna but soon followed by such other large cities as Ahmedabad, Bangalore and Delhi, found their way around the rent control ordinance by shifting to a modified area-based system (Rao, 2008). However, other local governments such as Mumbai continue to employ rental value taxes, and hence suffer from the rent control ordinance (Pethe, 2013). Almost everywhere the rental value approach is followed, the result is substantial understatement of the tax base. Although few developing countries carry out sales-assessment ratio studies to gauge the quality of their assessment, Mathur et al. (2009) used survey data on market values to show assessment ratios of 20 percent in Kolkata and 30 percent in Lucknow and Nagpur. Other estimates suggest assessment ratios of 50–70 percent in Jordan (Bahl, 2010) and only 20–30 percent in Pakistan (Bahl et al., 2015a).

M4439-BAHL_9781786435293_t.indd 237

01/03/2018 11:08

238

Fiscal decentralization and local finance in developing countries

Capital Value The common form of property tax in OECD and Latin American countries (also used widely in Southeast Asia) is the capital value approach based on the market value of property, i.e., the amount the land and any improvements would sell for in an open market. The legal taxable base is sometimes the full market value and sometimes a percentage of that value. This approach escapes some of the problems of using the rental value base. For instance, there is no conceptual problem in defining the tax base for vacant land or rent-controlled properties. More importantly, although assessments are still notional (presumptive), a formal record of real estate transfer prices is generally required to register a change in ownership, so there is potentially a better information base for this approach. Moreover, not only is the concept of taxing property value more easily understood than the concept of taxing rental value, it may also perhaps be fairer and it will certainly lead to the imposition of substantially lower statutory tax rates – always a politically popular characteristic. It is thus not surprising that the capital value approach is becoming more widespread (McCluskey et al., 2010). Of course, this approach also has its problems. In most countries, there is little good comparative data on the sale prices of properties. The sale values declared for registration purposes are usually understated to avoid taxes on property transfer and are rarely effectively monitored. Since industrial properties are infrequently sold, they are usually valued on an estimate of the depreciated cost of replacing the facility at the same location, which in practice often means little more than applying an arbitrarily chosen depreciation rate to some available book value figure. Similarly, commercial properties are often valued by capitalizing their estimated annual rental value, often with little solid evidence supporting either the annual rent assumed or the capitalization rate chosen. Imperfect though it usually is, the administrative cost of the capital value system can be high. Even if the data on comparative sales are good, as in the US, the cost of revaluation in the early 2000s was estimated at $20 per parcel (Dornfest, 2010). Though costs in developing countries are obviously lower – UN Habitat (2011) estimates US$0.82 per parcel in Hargeisa, Somalia – they are still high relative to per capita collections because ownership of properties is often uncertain, the basic data needed for valuation are not usually available and trained staff are in short supply. Land Value A land (site) value tax is based on the market value of land, inclusive of the value added from clearing, grading, installation of utilities, etc.1 This approach is used in only a few countries – Australia, New Zealand, Denmark, Estonia, Jamaica and Kenya (Franzsen, 2009). However, some of the key features of a land value tax are found in many more countries. For example, the separate valuation of land and improvements in Brazil and the imposition of higher tax rates on land than on improvements in Namibia capture some of the advantages of a land value tax, as does the separate rate regime with higher effective tax rates on vacant and underdeveloped properties found in many countries. Excluding structures from the tax base lowers administrative costs. Since the physical characteristics of land seldom change while structures are always

M4439-BAHL_9781786435293_t.indd 238

01/03/2018 11:08



Taxing land and property ­239

depreciating or being improved or replaced, the costs of maintaining the tax roll is also less costly. On the other hand, in built-up urban areas – where most of the property tax base is inevitably found – it can be difficult to find adequate reliable data on transaction values in land. Although there are several accepted techniques to ‘back out’ a land value from data on sales of improved properties (Gloudemans, 2000; Bell et al., 2009), Franzsen (2009) suggests that these methods are compromised when appraisers must rely on sales from adjacent neighborhoods to estimate land values because there are not enough vacant land sales in the jurisdiction. Because the site value approach has a smaller tax base it requires a higher tax rate to yield any given amount of revenue, a feature that never makes people happy. The land value approach also places a lighter burden on properties with visible, high-valued structures, which again makes other taxpayers unhappy. Policy-makers may agree that a broader tax base would be better. In Kenya, for example, where the land value tax has been in place since 1923, some have argued for a switch to a capital value system because “site value taxation does not provide a broad enough base to raise the revenue required to improve service levels and infrastructure” (Franzsen and McCluskey, 2008, p. 377). However, site value rating has been under attack in Kenya for so long that it may turn out to be one of those good ideas that will not go away easily (Smoke, 1994). Area The idea behind an area-based system is simply to tax each property at a specific rate per area unit of land and per area unit of structures.2 This approach has been around for a long time: (McCluskey et al., 2010) identified 44 countries where it is used, including several transitional countries in Eastern and Central Europe. It is also often used to some extent in rural areas in many other countries (Franzsen and McCluskey, 2017). The area approach is more easily understood and administered than value-based systems. It makes it possible to impose a property tax in countries where there is only a fledgling property market or no property (land) market at all, and gets around the problem of the shortage of valuers because it requires only measurement rather than valuation. Area-based systems range from a pure form where the tax is imposed on the physical area of land and buildings irrespective of value, to a hybrid where the specific tax rate varies according to property characteristics such as location or access to services (Bell et al., 2009). The main requirement with a pure area tax is that all the taxable area has been accurately reported and located, and properly classified (residential, commercial and so on). It is thus much like a specific rate excise tax. More commonly, however, the area approach is adjusted to reflect some indicators of value, for example, applying different rates to zones where different levels of public amenity are available (Holland, 1979).3 In Chile and Jordan, for example, the value per square meter for each zone is based on municipal land value maps that take account of the use and location of a property (Bell et al., 2009). The administrative ease of an area-based property tax comes at a cost. Revenues can grow only if tax rates are increased or the zone to which properties are assigned is changed, as is done for example in Bangalore, India (Rao, 2008). Since establishing tax zones requires considerable judgment, each usually

M4439-BAHL_9781786435293_t.indd 239

01/03/2018 11:08

240

Fiscal decentralization and local finance in developing countries

­­ contains a wide mix of property types and qualities, with some variations in access to amenities, which makes the tax less horizontally equitable.4 Notes: 1. The problems of differentiating improvements to the natural land (e.g., drainage and grading) from physical structures are discussed by Oldman and Teachout (1979). 2. A capital value tax may be administered in a similar fashion, as noted above, with each parcel valued at a specific rate (per area unit of land and per area unit of structures) but the tax rate is then applied to the total value. 3. Technically, this is like the ‘banded’ property tax as in the UK (Slack, 2004; McCluskey et al., 2017) except that the ‘bands’ are zones rather than value ranges. 4. The same problem of course arises under any system where rates vary by zones: in a capital value system where values are ‘normalized’ on (in effect) a zone basis by assuming unsold properties have increased in value at rates similar to the average within the comparable zone. When beachfront properties go up by 100 percent, for example, if the assessment process includes some non-beachfront properties in the same zone it is likely to overestimate the value increase in the latter.

Area-based systems may also deserve a better hearing than they have received in much of the professional literature. A version of the areabased approach has long existed in many countries where a common approach to valuation is to determine a land value map for the jurisdiction, to estimate land values as best as possible for each neighborhood area, and then to value buildings according to construction cost and physical characteristics. The most important advantage of adopting an area-based approach is undoubtedly administrative: it enables the imposition of a property tax when there are few valuers and when there is no reliable direct evidence on the market value of properties. For example, most transitional countries of Eastern and Southern Europe had no developed property markets or value information; but they did have registers containing information on the areas of both land parcels and structures, so it made sense for them to use these data as the initial base for a property tax. In effect, this approach converts the property tax into a type of specific excise tax, with some adjustment in the tax rate for location and perhaps construction quality. On the other hand, the area system is both fundamentally unfair in equity terms and unlikely to support the revenue growth needed to support the increasing cost of local government services, so it is hard to see this as anything but a transitional approach. The best system for any country depends on the present state of property markets; on the skill set of the property tax administration; on what the country most wants to accomplish with its property tax; and, as always, on political and other relevant circumstances. All we can do here is suggest how such a choice might be reasonably made. To illustrate:

M4439-BAHL_9781786435293_t.indd 240

01/03/2018 11:08



Taxing land and property ­241 ●●

●●

●●

●●

●●

●●

●● ●●

If the primary goal is revenue, the base should be as broad as possible. In principle, any of the bases discussed in Box 6.1 could do the job, provided the relevant authorities were willing to set the tax rate at a high enough level, to hold exemptions to the minimum possible and to revalue property regularly. If equity issues dominate, and equity or fairness is defined in terms of property value, then any value-based system could work if properly designed and administered. Of course, the more tax assessment is based on such essentially presumptive methods as notional assessment or mass appraisal, the less fair the system will be. To some extent, the choice will be guided by such ‘givens’ as land tenure arrangements and the availability of good information about real estate prices (McCluskey et al., 2010). If most property is held in leasehold form, then the focus might be on annual rental value; if tenure is dominated by owner-occupancy, there is a stronger argument for an (improved) capital value base. In countries that are still mostly rural, but with some large cities, a combination of a value-based system for urban localities and an area-based system for rural localities may work best. If there is no private ownership and no property market, an areabased system – perhaps a hybrid version with taxes assigned according to location and amenities available – may be the only feasible choice. If the property tax is considered as an extension of the income tax – for example, as a backup to offset the common under-declaration of rental income – a rental value system might be the best fit, although the serious problem of imputing rents to owner-occupiers is unlikely to be resolved. On the other hand, if the property tax is seen as mainly a wealth tax, a capital value (or land value) system is called for. If a major problem is the underuse or inappropriate use of land, elements of a land value tax might be built in, for instance, through separately assessing land and buildings with differentially higher tax rates for land or by a separate tax regime for vacant or underutilized properties.

To some extent countries can even ‘mix and match’ different approaches for different purposes, provided, of course, the needed administrative and political support to do so is in place.

M4439-BAHL_9781786435293_t.indd 241

01/03/2018 11:08

242

Fiscal decentralization and local finance in developing countries

Exclusions, Exemptions and Reliefs Whatever base is chosen, an additional critical issue is how broadly it should be applied. All too often the property tax base is narrowed significantly by exclusions, legal exemptions and preferential treatment: sometimes to achieve a social or economic goal such as encouraging investment in real estate; sometimes for equity reasons (such as exempting low-income housing); and sometimes simply to curry political favor. Whatever the reason, the result is likely to be a less fair property tax and is certain to be one that yields less revenue. The policy question, of course, is whether the benefits gained are worth the costs. Some types of property tax relief are seldom controversial. Nearly all countries, for example, exempt certain properties – such as those freed of tax by international convention (e.g., foreign embassies) or those whose use is considered meritorious (e.g., schools and churches).7 Another common exemption found in most countries is for government-owned properties and those occupied by nonprofit enterprises. In Brazil, the Constitution forbids the taxation of government-owned properties even if used for non-government purposes. This issue is sometimes contentious because such exemptions deprive local governments of the right to charge for some of the land use within their boundaries even though these activities make considerable use of local public services, thus requiring others to pay higher taxes (Bahl and Linn, 1992). In India, for instance, a working group was established by the government in 1996 to study the issue. The state governments wanted to tax government properties and the central government opposed this, so no agreement could be reached (Mathur et al., 2009). Although the central government is required to pay a service charge in lieu of property tax in Tanzania, it is delinquent on these payments (McCluskey and Franzsen, 2017). Some other reliefs are purportedly intended to protect low-income families. Although not all countries provide special relief for this purpose, one common approach is to exclude properties valued below some threshold level – a move that may be particularly just when the poor for the most part live in areas that are not well serviced. Informal settlements of various types are a related problem in the urban areas of many developing countries (see Box 6.2). Much housing may have low assessed values, and the cost of collection may even exceed the revenue gain. However, the threshold approach also provides relief to the non-poor and may be costly in revenue terms. In Chile, for instance, over two-thirds of properties registered in the fiscal cadaster (or cadestre) are exempt (Irarrazaval, 2004), and in São Paulo (Brazil), 40 percent of all properties are exempt (de Cesare, 2012). If general relief is the goal, a lower statutory rate might be the

M4439-BAHL_9781786435293_t.indd 242

01/03/2018 11:08



Taxing land and property ­243

BOX 6.2  TAXING INFORMAL SETTLEMENTS Many low-income countries face problems in determining the appropriate property tax treatment of informal settlements. As Smolka and De Cesare (2006, p. 14) put it: tenure rights in informal settlements are often obscure or even unknown; buildings are constructed gradually over time, self-construction is common, and the whole unit may never be finished; property value depends on vague or intangible factors such as the security provided by community organizations; the occupant or even the legal owner may be too poor to pay taxes; administrative costs of collection are higher than in the formal areas, whereas assessed values are often much lower; and there is hardly any public investment in infrastructure and services.

For many countries, this problem is too big to be neglected. Billions of people around the world live in such informal settlements as slums, pirate subdivisions where sales occur without clear title, illegally built housing and tribal land. Some popularly held notions about informal settlements are more myth than reality. It is not true that such informal settlements are populated only by people who are unwilling and unable to pay property tax and who work, if at all, in the informal sector, or that almost all property exchanges in such areas occur through nonmarket transactions. Moreover, alternative approaches such as taxing occupiers and using local information sources to enumerate parcels and to identify trading values are possible in some instances. Generalizations are difficult, and the difficulty of levying property taxes in informal settlements varies substantially from region to region, as the following examples suggest: ●

In the large urban area of Mumbai, over half the population lives in slums where conditions are deplorable because of a lack of access to basic amenities: only 78 percent of slum dwellers use tap water; 37 percent use communal toilet facilities, with 24 percent walking 0.2–0.5 km to latrine facilities; and only 84 percent of slums have approach roads that accommodate motor vehicles. Nonetheless, many living in the slums have incomes above the poverty line, and have demonstrated both demands for better public service and some willingness to pay. Their problem is less poverty as such than ‘shelter poverty’ (Rao, 2009). ● In Africa, the problems with informal settlements are often complicated by tribal ownership of the land. When land use rights are assigned by traditional leaders, neither ownership nor value is clear, so that implementing a property tax is seldom easy both for administrative reasons and because it may be viewed as a challenge to the role of the traditional leaders (Solomon et al., 2002; Franzsen and McCluskey, 2017).

M4439-BAHL_9781786435293_t.indd 243

01/03/2018 11:08

244

Fiscal decentralization and local finance in developing countries

better strategy. Other attempts to provide property tax relief – usually only to homeowners who, not coincidentally, are perhaps most likely to vote in local elections – are even less well targeted. Providing preferences for the elderly, for instance, is again likely to benefit many who are not poor. Sometimes, larger families are favored, whether rich or poor. In Serbia owner-occupiers receive a 40 percent reduction in taxes, which is increased 10 percent for every member of the household up to a limit of 70 percent (Begović, 2004). Exemptions are often introduced for what may be called ‘social engineering’ reasons. One of the most common (and costly) preferential treatments is that for owner-occupiers. Perhaps in some places it may at times make sense to assume that homeowners are the foundation of the nation, but the general reason for this preference seems to be political. However, owner-occupier reliefs not only provide larger benefits to rich and middle-class families but they also increase administrative costs because every parcel needs to be classified in terms of tenure. Such relief is often provided by applying differential nominal tax rates or differential assessment ratios. In Pakistan, where owner-occupied properties with areas less than 1,360 square feet are exempt from property tax, 66 percent of owner-occupied properties fall into this class (Bahl et al. 2015a); an earlier study estimated that bringing owneroccupied property fully into the tax base would triple the level of property tax revenues (Bahl et al., 2008). An owner-occupier pays only one-tenth the rate of that paid by a renter in Pakistan. Slovenia also exempts owneroccupied housing if it does not exceed a certain floor area. Still other concessions are intended to provide incentives for certain activities. Some countries exempt new construction (say, for 10 to 20 years) to stimulate investment activity. Countries using rental value systems sometimes provide standard deductions for maintenance. For example, Jordan gives a standard deduction of 20 percent of assessed value for all properties, which reduces revenues by 40 percent (Bahl, 2010). Since such deductions are rarely if ever tied to actual maintenance expenditures, they amount to a general rate reduction. Although the revenue loss due to exemptions can be quite large, few estimates are available because governments seldom maintain good valuation records with respect to exempt properties. In India, for instance, though Mathur et al. (2009) reported that about 10 percent of property in the 36 largest cities were exempt, no reliable estimate of the revenue forgone could be made. In Panama, where at least one-third of the tax base is exempt (and only about 13 percent of structures are taxed), the revenue cost is likely even higher owing to the concentration of value in higher-rate brackets (Bahl and Garzon, 2010). As with all taxes, the best way to deal with the erosion of the property

M4439-BAHL_9781786435293_t.indd 244

01/03/2018 11:08



Taxing land and property ­245

tax base through such reliefs is to try to keep them to a minimum and to establish a systematic approach to review and monitor them. Ideally, one might first place a moratorium on new preferential treatments, and then establish a sunset (termination) date or at least a schedule for reviewing existing preferential treatments, as well as a requirement for a formal cost–benefit analysis to be prepared and approved before adopting any new incentives or concessions. A second key feature is to require that all property, including exempt property, should be regularly valued and that the cost of all concessions should be publicly reported. Obviously, there would be some administrative costs incurred, but this procedure would make the costs of concessions clear to all and open to debate. Third, it is a good idea, if it can be done, to charge a payment in lieu of property tax for government properties as well as for most non-profit uses of property, and to show these payments as explicit budgetary costs.8 If there is a constitutional provision forbidding the taxation of government property, then a service charge should be imposed. Such a practice is followed in some countries (for example, India and Kenya), while Malawi and Namibia reportedly tax government properties at a discounted rate. In Mumbai, for example, the payment in lieu is set at 80 percent of the tax that would have been paid if the property was privately owned (Tata Institute and University of Mumbai, 2001).9

SETTING PROPERTY TAX RATES In discussing rates, it is important to distinguish between the nominal (legal or statutory) tax rate and the effective tax rate. The nominal tax rate is simply the ratio of tax liability to taxable assessed value (TL/TAV), where TL 5 tax liability and TAV 5 taxable assessed value. But the effective tax rate – the ratio of tax liability to the market value of the property (TL/ MV) – is the one that really matters. The effective tax rate may be expressed as the product of the nominal tax rate (as just defined), the exemption rate (the ratio of taxable to total assessed value) and the assessment ratio (the ratio of assessed to market value):10 TL/MV 5 (TL/TAV) (TAV/AV) (AV/MV). What this simple identity shows is that the effective tax rate changes when either the nominal tax rate (TL/TAV), the exemption rate (TAV/AV) or the assessment ratio (AV/MV) is changed. It is critical to understand this relationship between the three components of the structure of the property tax because they are not independent. Some countries have chosen a progressive nominal tax rate structure, but then largely offset it

M4439-BAHL_9781786435293_t.indd 245

01/03/2018 11:08

246

Fiscal decentralization and local finance in developing countries

by exempting or systematically underassessing higher-valued properties. Others have obscured the full extent to which business property is taxed by hitting it with both a higher nominal rate and a higher assessment ratio than other property. Such policy trickery complicates the system, increases administration costs and of course makes it more difficult for taxpayers to understand what is going on – which may perhaps be the desired outcome. A wide variety of legal tax rate structures (on the different bases discussed above) exist in developing countries. Rates may be flat or progressive; they may be different on land and structures and on different classes of property (often lowest on agricultural land and sometimes especially high on vacant land or land thought to be underutilized). Usually, rates are applied to each landholding separately, but occasionally they apply to the total value of real property owned by a taxpayer (though seldom very successfully in practice). Some countries have very complex rate structures. In Panama, for example, there were three separate rate structures: (a) a progressive rate structure with slabs and marginal rates; (b) an alternative progressive rate structure with a lower top marginal rate for those who have obtained a certified private sector appraisal; and (c) a special rate regime for multi-unit buildings. Equally, there is considerable variation in the responsibility for determining the rate structure: sometimes it is done nationally and uniformly; sometimes regional and/or local governments set rates; sometimes governments are permitted only to choose a rate within a prescribed range (Colombia); and sometimes the rate is set by the central government (Chile). To some extent, one may perhaps read the degree of autonomy that local governments are given in setting rates as an indicator of how committed the higher-level government is to fiscal decentralization: how far do they trust local governments? Interestingly, almost no one seems to focus on the economically relevant effective rate. The first step to getting the property tax right in any country is to make the difference between the effective tax rate and the statutory tax rate as small as possible – for example, by establishing a base, limiting exemptions and assessing properly so that the role of the nominal rate is primarily to determine the amount of revenue to be raised by the taxing government. In practice, in most countries the rate and base regimes are usually jointly determined by the central government in the presence of severe administrative constraints, with the main concerns of policy-makers being more to avoid creating problems for themselves or making their constituents unhappy than to getting the local policy mix right. One hardly ever finds the sort of simple, flat-rate structure that most experts suggest. Discriminatory or progressive rates are less effective and less desirable than, say, a modest low-income exemption and a good assessment base, but they look good and seem to be more saleable politically. Similarly,

M4439-BAHL_9781786435293_t.indd 246

01/03/2018 11:08



Taxing land and property ­247

layering separately earmarked property tax rates for street lighting, parks or whatever – attractive though this approach may sometimes be in making higher taxes a bit more acceptable – is seldom a good idea and tends to complicate both tax administration and budgetary efficiency.

ADMINISTRATIVE ASPECTS No matter how well structured a property tax might be, in the end how well it works in practice depends largely on how well it is administered. Poor administration of the property tax is a major reason why it yields so little revenue and often scores so poorly in terms of fairness and equity. This is not a secret. Indeed, attempts to reform the property tax usually make better administration their top priority. The key pillars in any administrative reform are simple: update the valuation roll, discover untaxed properties and improve the collection rate. But it is not so easy to get the right results since doing so requires capable professionals, good management practices, good information and strong political commitment. Few administrative reforms have resulted in wholly successful revenue outcomes. Most taxpayers see administrative reform as code for higher taxes, and understandably react strongly when such reform is coupled, as it often is, with the elimination of such beloved (bad) practices as the undervaluation of property and exemptions for owner-occupiers, as well as tighter enforcement. Sometimes their reaction results in further ‘reforms’ that make the tax even less fair and less productive in revenue terms.11 Politicians who need electoral support or seek more popularity with articulate propertyowners are likely to respond to adverse popular reaction by measures that as a rule tend to move back towards the pre-reform system.12 Getting property tax administration right requires identification of properties, an efficient registration system, good valuation, a modern information system to track performance, and a working collection and enforcement system – all of which depend in part on substantial cooperation between different departments and agencies: the administrative barriers to reform are hard to overcome in countries in which officials are often unprepared to do the technical work and find it difficult to cooperate with each other. Technically, the costs of property tax reform – though still high relative to the revenue likely to be collected at present effective rates – are probably less now than in the past, owing to the widespread availability of such new technology as geographic information systems (GIS) (De Cesare, 2012). But new human skills are needed to gather and use new information properly, and local and regional governments must be able to absorb and utilize both different kinds of staff and different tools than in the past. A precondition

M4439-BAHL_9781786435293_t.indd 247

01/03/2018 11:08

248

Fiscal decentralization and local finance in developing countries

for successful reform that has often been unduly neglected is the need to alter the legal system to accommodate the effective use of new approaches.13 Another problem often encountered is that all four pillars of reform – property identification, current information system, valuation and ­collection – need to be in place for success. If one of these pillars is weak the system may collapse; or at least, to be a bit less dramatic, it may fail to achieve any sustainable long-term effect on revenue. A well-known example is an early reform effort in the Philippines (Dillinger, 1988). The project was successful in producing tax maps and updated property assessments. But little revenue ensued because little was done to strengthen collection, so that although the assessed tax base increased by 37.5 percent and the assessed tax liability by 13.6 percent, the actual increase in tax revenue was only 1.1 percent. Important pieces of the system are often left out in many attempts to reform property tax administration. Few governments are willing to take on the bureaucratic (and political) efforts required to enforce the cooperation needed when, for example, one ministry is responsible for finding and registering properties and property transfers and another is responsible for maintaining the tax roll. Enforcing the rules – for instance, by seizing delinquent properties when taxes are not paid – appears to be very hard to do in most countries. The development of a system to record accurate sales values of property – whether through structural reform or administrative effort – also seems to be ‘mission impossible’ and is ignored in most developing countries. Little can be done to reform property tax administration in these circumstances. Identifying the Tax Base Of course, one could start from the other end and simply do a better job in collecting the taxes assessed. But one cannot build a sustainable property tax administration solely from the ‘sharp end’ of the administrative process – collection. For sustainable reform, it is critical that the location, physical description use and ownership of the tax base are properly identified and recorded in the tax roll (property register, cadaster) (Almy, 2004; de Cesare, 2012). Most countries have some form of title registry, and of course almost everywhere there are now aerial or satellite photographs as well as relevant information in other public data bases.14 Often, as Kelly (2014) notes, information about properties and taxpayers is collected in partnership with various agencies as well as private firms – for example, by cross-referencing existing cadastral information with new requests for building permits and subdivision, business licenses and public utility connections. In some cases, self-declaration of property characteristics has also proven to be helpful (Box 6.3).

M4439-BAHL_9781786435293_t.indd 248

01/03/2018 11:08



Taxing land and property ­249

BOX 6.3  SELF-ASSESSMENT Some countries have turned to self-declaration (often called self-assessment) to get around the limited availability of data on property ownership, characteristics and value. The strategy is a simple one: require all owners to declare information about their property, and use this information (either directly or indirectly) in the valuation process. Sometimes, as in India, taxpayers must declare the physical characteristics of their property; sometimes, as in Colombia, they are also asked to estimate its value, perhaps within a specified range. Taking this path has some advantages. It makes use of the owner’s knowledge of their property and gives them some involvement in determining their tax liability, thus removing some of the mystery about how the property tax works. It also flips the normal property tax process on its head, reversing the sequence in which the government sets the taxable base and waits for the owner to object: with selfdeclaration, often the owner can set the value and wait for the government to object. This approach may provide essential information about the stock of real estate, and even some idea of how people value it. In some cases (as in Bangalore, India and Bogotá, Colombia), it has proved to be quite successful in building up the tax roll. When valuation is not possible but a tax may be assessed, it might sometimes be the only feasible way to fill in the blanks. Bangalore adopted an area-based system featuring self-assessment, though a better label might perhaps be something like self-calculation of tax liability based on a formula supplied by the local government and owner declaration of the physical characteristics (Rao, 2008). Essentially, the government supplied the valuation rates for each location and selected the physical attributes of the property to be used in the formula, and the taxpayer did the arithmetic. There was significant, voluntary participation by taxpayers and revenue collections tripled (though the effective tax rate was still quite low). Official valuation was provided for in the event of suspicious declarations. The program was well implemented and featured a strong taxpayer awareness program. However, as we discuss later in this chapter, with respect to taxes on property transfers, taxes on honesty rarely work in the long run. Self-declaration can bring people into the property tax system and provide some basic information that can greatly help the development of a full fiscal cadaster. But it is likely to do so only when the authorities can and do carry out independent valuations that are sufficiently credible to dissuade owners from under-declaration. Of course, the higher the tax rate, the more stringent such monitoring will need to be. Even if a voluntary system may occasionally be used in the early stages of a property tax reform, it is at best only a transitory approach that should soon be replaced by a more regular official valuation regime.

But even with the use of GIS systems, tax mapping and other types of data bases (for example, on roads), building up and maintaining reliable information on the tax base is not a simple matter in countries in which there is often substantial urban migration, much informal building, the land titling system is in disarray and recorded property boundaries are

M4439-BAHL_9781786435293_t.indd 249

01/03/2018 11:08

250

Fiscal decentralization and local finance in developing countries

often incorrect. It is not surprising that cadasters are far from complete in most low- and middle-income countries. Reportedly, for example, one in every four properties in Peru is not included in the potential tax base, although some (less than 40 percent) of those omitted are slum properties that would likely not bear much (if any) tax in any case. In Delhi, only 38 percent of all properties are on the tax register (Mathur et al., 2009). In Chile, half of recent new building is not included. Franzsen and McCluskey (2017) report many examples of thin coverage of the property tax base in Africa. For instance, in Maputo (Mozambique), only about 7 percent of properties are registered; in Cameroon, only 6 percent of identified plots are included; and in Addis Ababa (Ethiopia), 45 percent of all structures are not on the tax roll. As discussed in Chapter 5, there is no necessary link between who gets the revenue from the property tax and who administers the tax. Many think that the task of establishing and maintaining a cadaster should be assigned to a central (or perhaps regional) authority rather than to (often small) local governments. In Lithuania, for example, the task is assigned to a national agency, the State Land Cadastre (Aleksiene and Bagdonavicius, 2008). Many countries, both developing and developed, similarly centralize the valuation process to an agency that is better able to gather the needed combination of expertise and technology to do the job right. Others assign oversight of how well local governments perform the task (e.g. through sales-assessment ratio studies) to a higher-level government. Of course, centralization does not guarantee better administration. Panama, for example assigns registration to notaries public and the Public Register, valuation to the cadaster office and taxpayer identification to three different central government agencies. The results have been poor, largely because of the lack of communication between these offices (Bahl and Garzon, 2010). Similar poor outcomes are not an uncommon issue in other developing countries. Such problems can be fixed: in Jamaica, for instance, the National Land Agency brought together the core land information functions of government under one agency (Gainer, 2017). The greater familiarity of local governments with local conditions, and their interest in getting the most revenue they can, suggests that they should also play an important part in determining the local tax base. Some Latin American countries (e.g., Guatemala and Mexico) have shifted more administrative responsibility to local governments (de Cesare, 2004). As mentioned earlier, Indonesia has decentralized property tax administration. Often, large cities have special authority in such matters. In Uruguay, only the capital city (Montevideo) looks after its own cadaster and valuation, as Bogotá has long done in Colombia, where several other large cities and one region (Antioquia) also have their own cadastral services.

M4439-BAHL_9781786435293_t.indd 250

01/03/2018 11:08



Taxing land and property ­251

In Brazil, urban local governments are fully responsible for property tax administration, but the central government plays this role in rural areas. As usual, there are no hard and fast rules telling us that central, local or shared administration is ‘best practice’ for any particular locality, region or country at any particular time. Maintaining the Tax Roll Once the tax base is identified, it – the ‘tax roll’ – must be maintained and kept up to date. Routine matters like this are not just boring details: if they are not done right, the property tax will not work well. The record (the master file) should include a physical description of each property as well as ownership information. To do so requires tracking all improvements to properties as well as changes in ownership and any sub-division of properties. Moreover, all the annual tax information for every property – ­assessments, exemptions, payments and delinquency – needs to be on record and accessible. It is a lot of work and often a painstaking task to do all this, but once developed a current recordkeeping system is extremely useful not only in implementing the tax properly but also in evaluating and reporting on its performance. Putting this information system in place is made unnecessarily difficult in many developing countries by the complicated structure of the property tax. For example, if owner-occupied properties are given preferential treatment, occupancy as well as ownership status must be tracked. Different rules and rates for different land uses mean that not only must that land use be tracked but so must changes in the intensity of mixed (residential-commercial) properties. Few developing countries have adequate staff, proper information systems or sufficient interdepartmental coordination to do this. Not only is the result that it is next to impossible to know who is being taxed how much or to evaluate the likely effects of policy and administrative changes, but it is also often difficult to administer the tax properly. De Cesare (2004) reports that in Guayaquil, Ecuador, simply integrating the public registry of properties and the cadaster online increased the number of properties recorded from 165,000 in 1993 to 418,474 in 2001. Many countries do not even have unique parcel identification numbers, and their tax rolls are usually out of date and far from inclusive (UN Habitat, 2011). For example, in Jamaica the stamp duty office, which recorded values at transfer, and the land valuation department used different identification numbers (Sjoquist, 2005).15 In some Balkan countries, such as Croatia, the property registry office did not inform local governments when properties changed hands, making it very difficult to bill properly for taxes due.

M4439-BAHL_9781786435293_t.indd 251

01/03/2018 11:08

252

Fiscal decentralization and local finance in developing countries

Valuation The property tax differs from taxes on income and consumption because it is usually based on a notional estimate of value rather than on an actual flow. The tax base is supposedly the sales price of a property if that property had sold (in an arm’s length transaction) in the tax year. The only way we know to reduce the inherently subjective nature of assigning this value (the assessment process) is to make the process of determining the value on which taxes levied as professional, open and objective as possible, and to monitor its outcomes. The best way to achieve this aim is by keeping assessed values as close as possible to the actual market values of comparable properties and by taxing comparable properties at roughly the same rate.16 In attempting to achieve this goal, three valuation methods are commonly employed for value-based property taxes: comparative sales, depreciated replacement cost and an income measure (UN Habitat, 2011; Franzsen and McCluskey, 2017). Few low- and middle-income countries have the resources to take into account the characteristics of every property. Since they seldom have good data in any case, it is not surprising that the methods used in practice vary considerably from country to country. In the few cases where data are available – as in Hong Kong, for instance, where landlords and occupiers are required to submit financial statements and evidence on rental agreements (Brown, 2013; Pang, 2006) – market rental values might be imputed to all properties in a neighborhood. Since such data are seldom available, valuers in most countries must make do with what data they can gather from surveys, third parties like banks and real estate companies and self-declarations, and then perhaps make adjustments for such factors as floor area. About the best that can be expected from this approach is for variations in average rents paid in different areas to be roughly reflected in the assessment roll. Matters are little better in capital value systems. The approach taken is often different for the valuation of land and the valuation of buildings. For example, in Porto Alegre, Brazil, ‘market value’ is simply the sum of land value and building costs (de Cesare and Ruddock, 1999), with land values (per unit of local area) being established by reference to the reported value of recent sales of vacant land in zones considered to be relatively homogeneous, as adjusted by the physical characteristics of the site and the availability of infrastructure services. Building values are based on the estimated average cost per area unit of construction of various types of buildings, adjusted by a depreciation rate. Similar systems have been applied for decades in other countries in Latin America (e.g. Mexico and Colombia). In Botswana, where land and improvements are valued separately, the

M4439-BAHL_9781786435293_t.indd 252

01/03/2018 11:08



Taxing land and property ­253

appraisal work is contracted to private sector valuers. In contrast, in South Africa, sales data (encompassing both land and improvements) from the property transfer tax are sometimes used for property tax valuation purposes, even when suspect (Bowman, 2002) – though in other cases the declared transaction values supplied by the stamp duty office are supplemented by expert judgment and other evidence of land values (e.g., bank mortgage information and real estate listings). Some think the answer to valuation for tax purposes lies in the magic of Computer Assisted Mass Valuation (CAMA). The idea behind CAMA is to develop an empirical model relating location and the principal characteristics of individual properties to sales prices, and then use the results to estimate the notional price for properties in similar locations and with similar characteristics that were not sold (Ward, 2002; Eckert, 2008). This approach is the main way in which most residential and small commercial properties are now assessed in Canada and the United States, and is widely used in other OECD countries and in a few developing countries, including South Africa and Ghana.17 CAMA appears to bypass much of the expensive legwork involved in developing a cadaster and assessing all properties, so at first glance it may seem like an ideal approach for countries with few resources to develop a good property tax system. However, since developing an appropriate model requires a substantial amount of reliable data on sales prices of real property it is not an obvious solution for countries in which the basic problem is that they have no such data and little capacity to fill the gap. The reality is that even the few developing countries that report salesassessment ratios – the usual ‘standard’ for appraising assessment p ­ ractices – report poor performance. De Cesare (2012) reports assessment rates at 50 percent or lower in Latin America; Mathur et al. (2009) report that the assessment ratio for India’s 36 largest cities ranges between 20 percent and 40 percent; and Kelly (2000) reports that the ratio varied between 20 and 80 percent for Kenyan municipalities. One important reason why underassessment is so widespread is because revaluations seldom occur. Legally, many countries require revaluation of all properties every three–five years, with new tax rolls to be implemented as soon as revaluation is complete. But hardly anyone actually does such things. In 2010, for instance, the general tax roll in Panama was 40 years out of date. Unsurprisingly, in countries that have had prolonged conflicts (such as Sierra Leone and Liberia), scheduled revaluations had also never taken place. Even when countries do revalue (as Ghana did in 2007, after a 20-year delay), no government could possibly sell the public on the huge tax increase that the estimated 457 percent increase in property values would have entailed – assuming, of course, unchanged tax rates (Franzsen and McCluskey, 2017).18 As the discussion in Box 6.4 underlines, a big revaluation is a very big deal.

M4439-BAHL_9781786435293_t.indd 253

01/03/2018 11:08

254

Fiscal decentralization and local finance in developing countries

BOX 6.4  THE SPECIAL PROBLEM OF REVALUATION The most costly and difficult aspect of establishing and maintaining a property tax is the need for periodic revaluation. With levies like value-added tax (VAT) or income tax, once an initial taxpayer register is set up, all that must be done is to add new taxpayers to it and to track and audit information supplied by taxpayers (or third parties such as employers) about their tax base. With the property tax, however, the tax base is generally established by an appraisal process that must be redone periodically (say, every three or five years) – a process that usually produces large, one-time increases instead of the relatively smooth increases (or decreases) that characterize the income or VAT base. Preparing a new list of values is a costly affair, and implementing the resulting new tax roll is almost always contentious in any country. As Prud’homme (2006, pp. 97–8) noted about France, “the few courageous Ministers of Finance who have attempted to undertake reassessments paid a high political price for doing so.” The more out of date the tax roll, the greater the problem of revaluation. Taxpayers never react well to potentially big changes in their tax liability – changes that they often assume will be as big as the change in the size of the tax base. Politicians, who usually compensate for the expected taxpayer reaction by lowering effective rates one way or another, do not like to use up their political capital for a tax change that in the end may yield less revenue than trouble. Kelly (2000) notes that a revaluation in Nairobi, Kenya (which taxes on land values only) in the early 1980s increased residential values by 600 percent and commercial values by 250 percent, but reduced industrial land values by 225 percent. It is difficult to imagine an elected political leader in any country who could ‘sell’ the resulting shift in tax burdens if effective rates were unchanged! Similarly, a revaluation in Cape Town, South Africa, in the 1990s led to a significant increase in assessed values and to a shift in burdens from commercial to residential taxpayers (Franzsen, 1998); and a 2007 revaluation in Ghana showed a 500 percent increase in property values compared to 1988 (Franzsen and McCluskey, 2017). With changes like these it is not surprising that governments are often reluctant to implement new valuation rolls. Politicians frequently try to soften the problem by reducing the tax rate or capping the increase in taxable assessed values. Such practices have long been common in the US and are also found in places like Buenos Aires, Bogotá and Egypt, where legal limits have been imposed on property tax increases. To reduce the pain of the big hit with long-delayed revaluation, some countries (such as Colombia) index assessments between revaluations; but this approach, while softening the magnitude of the blow, nonetheless leads to inequities over time when property values grow (as they usually do) at different rates in different sectors and in different neighborhoods. The failure to revalue imposes a significant revenue cost. Had Punjab province, Pakistan, brought in its newly completed valuation roll in 2006, property tax revenues would have doubled (Bahl et al., 2011). In an econometric study of Colombia, Sánchez Torres et al. (2015) find that each additional year since the last cadastral update resulted in an undervaluation of properties of 7 percent in urban areas and of 4.3 percent in rural areas. Since on average the last urban cadastral update in 2009 (the year of the study data) was around five years, the compound loss in the property tax base of Colombian municipalities owing to the failure to update was

M4439-BAHL_9781786435293_t.indd 254

01/03/2018 11:08



Taxing land and property ­255

over 40 percent. On the other hand, Belo Horizonte (Brazil) successfully introduced a new valuation roll in 2009 that increased assessments and revenues by 20 percent by rolling back the statutory rate, increasing the tax threshold, spreading the increased tax over two years and offering a 7 percent discount for early payment (Domingos, 2011). Packages like this need careful design and ­implementation – and phasing out over time – if they are not to end up worsening inequities and losing rather than increasing revenue.

Five questions should be asked about any valuation system: 1. Is there is clarity about whether the legal base is the full market value or some fractional assessment? 2. Is it clear how assessments are to be carried out? Although the estimation approach described in any assessment manual is almost certain to look arbitrary to some taxpayers, clarity in the law should make the process much more transparent, and hence make life simpler (if not necessarily happier) not only for assessors (public or private) and courts but also perhaps even for taxpayers. 3. Are there enough qualified appraisers (in public and private sectors) to make the system work? For example, McCluskey et al. (2017) report the deadly combination in some African countries of a limited number of professional valuers and a tax law that requires professional valuation.19 Supplementary questions here are whether there is a good training and certification system for appraisers, and how large is the wage gap between public and private appraisers. 4. Is there a system in place to monitor the fairness and accuracy of the valuation process? It may not be feasible to carry out sales-ratio studies like those mentioned earlier; but at the very least there should be some panel of credible experts charged with making regular published assessments of the quality – the average degree of underassessment – of official assessments. 5. Finally, does the country impose a high tax rate on property transfers? This is important because if so there is a strong incentive to understate values which may undermine the prospects for being able to significantly improve annual property taxes. We discuss this issue further below. Collection Unless property taxes are collected, the coverage of the base, the state of record-keeping and the accuracy of valuation do not make much

M4439-BAHL_9781786435293_t.indd 255

01/03/2018 11:08

256

Fiscal decentralization and local finance in developing countries

­ ifference. Some experts have therefore argued that property tax reform d should begin at the sharp end, with collection. This makes short-term sense. If more revenue is the goal, collect the taxes that have already been assessed before doing anything else! However, if the system is badly designed and operated, collecting more from those trapped within the system while continuing to give a free pass to those who are not presently on the roll or are grossly undervalued may make things worse in the long run although in the short run the government has more revenue than it would otherwise have. Moreover, because even a poorly run tax is likely to be at least partly capitalized, to some extent the market may have already corrected some of the initial inequities. There is clearly much room for improvement in property tax collection in developing countries. Kelly (2014) estimates that the collection rate in most countries is between 30 and 60 percent.20 Numerous case studies support this conclusion. For example, the collection rate in the Philippines is 50 percent (Guevara, 2004), in Kenya 60 percent (Kelly, 2000), in Montenegro 43 percent, in Macedonia 15 percent (USAID, 2006) and in India 37 percent (Mathur et al., 2009); however it is higher in Croatia, at 70 percent (USAID, 2006) and Colombia (75 percent), rising to as much as 90 percent in Bogotá (de Cesare, 2004). Collection rates are low largely because there is too little enforcement and too little penalty for non-payment. Local governments are sometimes blocked by higher-level authorities from enforcing taxes on the politically powerful who often own valuable property.21 Owners of higher-valued properties often appeal: they usually do not have to pay until the appeal is settled, and the courts are so slow that many years pass before a decision is reached. Even if the decision favors the government, the usually inadequate interest charges applied to such delayed payments make the real tax paid often less than what would have been paid in the first place, especially if, as is likely there has been some inflation during the period between the initial assessment of liability and the time payment is made.22 Compliance costs may also be a factor. Sometimes taxpayers are unreasonably required to pay on a designated day at a distant collection point, or expect to have to pay additional bribes to tax collectors. Taxpayers may hold back on the usually small but visible direct property tax because they think they get little or nothing from government for their taxes, and know they will not suffer much for not paying. Foreclosure and sale at auction are rarely used. In Venezuela, the law prohibits the eviction of those who do not pay taxes. Tax clearance certificates are effective only when land on which taxes are due is sold. The names of tax delinquents are seldom made public, perhaps in part because tax evasion may be seen at times as

M4439-BAHL_9781786435293_t.indd 256

01/03/2018 11:08



Taxing land and property ­257

more a badge of honor – a symbol of brave resistance to the exploiter, perhaps – than of shame. While it can be difficult to turn all this around, it can be done – if governments are willing to do so. The seizure of delinquent property would likely be effective, but it appears to be a step too far in most countries and it is expensive to implement. Other draconian measures have been used with some success, for example, shutting off the electricity supply if property taxes are not paid (as in El Salvador and South Africa) or blocking the bank accounts of property tax delinquents (as has been discussed in Kenya). Other ‘hard’ approaches considered in some countries include turning delinquent accounts over to private collectors on a commission basis and sequestering funds from bank accounts. Softer approaches include making the property tax friendlier through public relations campaigns and, more importantly, implementing structural and administrative changes that simplify the tax, make its operations more transparent and make it easier to comply with. For example, simply moving collection points to banks, post offices and the internet has been found to increase compliance (Kelly, 2014).23 Possibly, one might even provide an incentive for taxpayers in the form of a discount for early payment.24 Another incentive might be given to local governments by increasing transfers when property tax collections are increased, although this raises the difficult question of how to measure improved revenue performance at the subnational level (Box 6.5).

PROPERTY TAXES IN RURAL AREAS To provide services to rural people and to build up the institutional social capital needed for development in general, local communities need to be encouraged and enabled to do what they can to tax themselves rather than depending on the (always unreliable) kindness of strangers in central government or in the legislature. For rural communities to provide meaningful services to rural people and to be maintained as viable places to live and work, they need to be as financially self-sufficient as possible. In many countries, the only way to make any progress towards this goal is by taxing rural land. Rural property taxes usually produce so little revenue that they may get lost in the rounding error at the national level. But property taxes can matter a great deal to rural communities. The rural property tax is often one of the main connections between local government and local citizens, and hence one of the few instruments through which local officials can be made directly accountable for the quantity and quality of services provided. Whether they like it or not, rural local populations may be brought closer to government by paying taxes.25 If fiscal decentralization

M4439-BAHL_9781786435293_t.indd 257

01/03/2018 11:08

258

Fiscal decentralization and local finance in developing countries

BOX 6.5  ASSESSING PROPERTY TAX PERFORMANCE Provided the administrative structure of a property tax is set up and maintained so that it provides the necessary information, much can be learned about critical policy and administrative issues. For example, questions like the following should be asked to assess the performance of any property tax, though few countries do so in any systematic way: ● ● ● ● ● ●

What is the pattern of collections by value class and by residential/non-residential properties and so on? What are the characteristics of delinquent taxpayers, and what are their ownership patterns, i.e., residential vs. non-residential, value class, etc.? What is the linkage between assessed value, public service levels and sales prices of properties in various neighborhoods of the local government? What is the revenue cost of various exemptions and preferential treatments? What are the characteristics of those who benefit from these preferences? How much should be invested in improving valuation vs. collections vs. cadaster? What would be the potential impacts of various policy actions – e.g., raising the tax threshold, increasing the tax rate, removing an exemption, etc. – on revenue and on the distributive and other effects of the tax?

Too often the only indicator used to assess the performance of the property tax is its revenue yield, and the only indicator used to assess property tax administration is the cost–revenue ratio, where costs are defined as the budgetary outlays on staff, information technology, vehicles and so on. Ideally, costs should also include investments in maintaining and upgrading the underlying infrastructure, such as certifying valuers and paying for the appeals process, although it is not always clear that this is done. In OECD countries, annual administrative costs are usually in the range of 2–5 percent of revenues, with ratios over 10 percent indicating serious problems (Almy, 2004). While there is little reported on this matter in developing countries, Chile – well known as perhaps the best tax administration in Latin America, and where the property tax is centrally administered – reported that the property tax cost–revenue ratio in 1991 was in the OECD ballpark, at 2.2 percent (Irarrazaval, 2004). Even if one has such data, however, it is not clear what such numbers tell us about administrative efficiency or the appropriate role of property taxation. A higher cost–revenue ratio may simply show that tax revenues are low or that the tax base is more complicated. In theory, one could learn more from a hypothetical calculation of the cost of collecting a given amount of revenue based on targeted norms for assessment efficiency, collection rate and so on, much as one might assess the general performance of a tax administration.* The point we emphasize here, however, is simply that even the best-administered property tax – especially a value-based tax – is unlikely to be a low-cost way to collect revenue owing to the high cost of property identification and valuation. Property tax revenues may be more economically ‘efficient’ in terms of their economic effects than income or consumption tax revenues, and they may provide substantial incentives for more

M4439-BAHL_9781786435293_t.indd 258

01/03/2018 11:08



Taxing land and property ­259

accountability in local governance. But, relative to the revenue they produce, they are not cheap to collect. Note:  * For a recent example of such an analysis, see Dabla-Norris et al. (2017). A review of the earlier literature (and a proposal for a different approach) may be found in Vázquez-Caro and Bird (2011).

is to benefit the rural sector, providing rural local governments with some source of own revenue is an important part of the equation. The economic geography of rural local governments is diverse. Many are small, remote and poor. However, others may have large populations and some are located on the urban fringe (Bird and Slack, 2008). Countries vary widely in terms of how the rural public sector is organized and how much discretion it has with respect to property taxes. The property tax is about the only source of own revenues for India’s village governments (gram panchayat), which provide essential services to about 70 percent of the national population (Bahl et al., 2010a). In Brazil, the rural property tax is separate from the urban property tax and is administered by the central government. Indonesia exempts most improvements in rural areas from the tax base. Tanzania does not tax rural property at all. Property taxes in the rural sector are usually simple and often levied on an area basis (or even a house basis) rather than on value. The yields are low because the tax base is small, the tax roll usually does not cover all properties and collection rates are low. But the other side of this coin is that relatively small improvements in identifying the tax roll and in valuation can sometimes add large proportionate increases to own-source revenues. In India, for example, rural local governments raise only about 6 percent of own-source revenues, with the property tax being the most important source (Bahl et al., 2010a). Tax rate ceilings are set by the state governments, and assessment practices are ad hoc. Rao et al. (2004) estimate that collection rates are so low that in some cases the cost–revenue ratio was perhaps 50 percent – half the revenue collected. Nonetheless, in an econometric analysis of over 3000 rural local governments in West Bengal state, Bahl et al. (2010) found that higher literacy rates were associated with a higher level of local government revenue mobilization, perhaps showing that when people demanded more services even this weak tax base was important. A key question in developing the property tax base for rural local governments is how agricultural land is taxed. Often there is a benefit justification for taxing agriculture because it benefits from some locally provided services (schools, water, roads, livestock services and the like).

M4439-BAHL_9781786435293_t.indd 259

01/03/2018 11:08

260

Fiscal decentralization and local finance in developing countries

But increasing the taxation of what Rajaraman (2004) has termed the ‘hardest to tax’ sector inevitably raises serious political and administrative problems. The simplest approach is to tax land area at a uniform rate, without regard to differential income-generation capacity apart from such broad characteristics as, say, irrigated vs. non-irrigated land or mountainous vs. level land (Bird and Slack, 2008). Land taxes assessed on such adjusted land area bases are relatively simple in terms of structure and administration. The information requirements are minimal: the area of the property, its location, its classification and the name of someone to whom the bill can be sent. Since the tax is lump sum and not based on output, it may even have the economically desirable impact of inducing owners to improve the use of their land. A quite different approach, discussed in Rajaraman (2004) and developed more fully in Bahl et al. (2011, 2015a), is a presumptive tax based on land area and crop. Such a levy would capture the significant differential net returns from different crops (and arguably the differential productivity of the land). All farms above a certain acreage could be subject to this tax. Assessment would seem manageable, with the important proviso being that a full survey of agricultural land is in place and up to date. Some elasticity could be built into such a regime by indexing the specific rate to crop prices. Much ingenuity has been devoted to working out systems to tax agricultural land in developing countries that seem to be both feasible and likely to produce more progressive, elastic and economically efficient results than a simple area-based tax. Such systems have been implemented in a few countries (e.g. Uruguay). However, none has yet proved to be particularly successful.26 Farmland on the urban fringe often faces pressure for urban development. Rapid development on the urban fringe may lead to sprawl and the high costs associated with providing infrastructure and services to new developments on the outskirts of cities. Or, as has often been true in metropolitan areas from Mexico City to Cape Town, it may result in sprawling slums. In the United States and elsewhere, farms in urban fringe areas are often taxed at their value in current use – defined as their selling price if used as a farm – rather than their market value (which supposedly reflects the highest and best use of the land). The usual argument for such favorable treatment is to protect farm families. In fact, however, it is so difficult to differentiate between family farms, hobby farms, corporate farms and land being prepared for subdivision that the main beneficiaries are often land developers. There is no good evidence of any desirable outcomes from decades of such tax relief in the US (Youngman, 2005). In developing countries, many attempts have been made over the years to design refined versions of rural land taxation that single out land owned by certain groups

M4439-BAHL_9781786435293_t.indd 260

01/03/2018 11:08



Taxing land and property ­261

(foreigners, corporations, owners of multiple properties) or land that is underused in some way, but none has been very successful in attaining such non-fiscal objectives (Bird, 1974). Moreover, such efforts usually absorb administrative effort that could have been more productively employed in establishing a better basic revenue instrument for rural local governments. The problem in most countries is not so much how to design a rural property tax (including a tax on agricultural land) – mainly, keep it simple and devote adequate resources to establishing the tax roll – but rather how to ‘sell’ such a tax to those who matter, namely, those who must pay it. Potential taxpayers must be convinced that they will get something for their money. In countries in which the main experience rural people have with government is coping with bureaucratic hassles and arbitrary impositions, the central government is unlikely to have much success with such marketing efforts. The only real chance for effective local property taxes in rural areas in most developing countries is probably to emphasize their local character and the direct and visible benefits that can accrue to local communities when and if they collectively decide upon, and pay for, local public works and services. Although it is difficult to establish an effective land tax system in poor rural areas at the local level, most countries could do better than they have. Unfortunately, all too often countries introduce special treatment for agricultural land that can weaken both the general property tax (by providing an avenue for speculation in urban fringe lands, as discussed above) and agricultural development more generally (by locking in inefficient land use). There is also more to rural property taxation than just taxing agricultural land. Often, much of the potential rural property tax base consists of residential and non-residential buildings and land devoted to non-agricultural pursuits, giving rise to many of the other issues discussed earlier in this chapter. In rural areas – in which politics is usually more personal – coupling property tax reform with some decentralization of spending power would seem in principle simpler than in more complex urban areas. It may perhaps also have a better chance of success than a ‘stand-alone’ reform of either side of the rural public budget.

TAXING PROPERTY TRANSFERS Taxes on property transfers sometimes take the form of a stamp duty on the transfer document; others are imposed as a separate property transfer tax, usually at the national or, in some countries, the regional level, but sometimes with the revenue flowing to local governments. The liability for payment may rest with the buyer or seller, or it may be split between

M4439-BAHL_9781786435293_t.indd 261

01/03/2018 11:08

262

Fiscal decentralization and local finance in developing countries

them. Transfer taxes have long existed in many countries, sometimes at high rates – for example, 15 percent in Senegal and 8 percent in Namibia (Norregaard, 2013) – though they have long been roundly criticized by economists.27 But their staying power is great, for several reasons. First, they are easy to administer. Most buyers and sellers want a legal record of ownership, and therefore voluntarily comply and provide some sales data to the agency that records such transfers. Second, they produce a fair amount of revenue – in some countries more than the property tax – and cost little to collect in most cases. However, the apparently low cost is deceptive. If governments tried to ensure that the values reported are accurate, costs would be much higher, which is no doubt one reason they seldom check – in Jordan, perhaps 10 percent of transactions are checked (Bahl, 2010). Consequently, reported values are probably greatly understated, thus, as noted earlier, making it more difficult and costly to establish the data base needed to apply a good, value-based annual property tax. Finally, there are many fewer taxpayers involved in any given year than for a general property tax, which both reduces costs and the likelihood of widespread opposition. Indeed, if most who buy high-value properties are thought to be rich (or foreign), there is often widespread popular support for transfer taxes. If, as is likely, not only is property ownership concentrated in the higherincome classes and turnover is greater for higher-income properties, the incidence of such taxes may indeed be progressive.28 To the extent the transfer tax reaches wealth held in the form of real property, it captures part of the potential tax base that otherwise largely escapes taxation in most countries. On the other hand, because the tax is based on the gross transfer value (like the property tax itself), it may act as a deterrent to investment in property improvements. Nonetheless, even buyers and sellers may like transfer taxes if they think it cools down an overheated real property market, which is one reason such taxes have sometimes been imposed. Although there is little evidence of their effectiveness in achieving this goal, introducing such taxes – as several cities in Canada did in 2016 and 2017, for example – can play a useful political role by allowing politicians to demonstrate that they are responding to popular demands to curb speculation in housing without doing anything that significant groups of potential voters will think hurts them. People are unlikely to realize that imposing an additional cost on real estate transactions will tend not only to reduce prices but also the volume of transactions, thus hampering the development of the real estate market. In developing countries, an additional undesired result will likely be to shift more transactions to the informal ‘off-the-books’ market as well as to increase the incentives to underreport the value of formal transactions.

M4439-BAHL_9781786435293_t.indd 262

01/03/2018 11:08



Taxing land and property ­263

This last point brings us back to the major problem with property transfer taxes in developing countries: taxpayers in effect establish the tax base by declaring the sales price. Few if any developing countries routinely check these declared values for accuracy, and the evidence is that underdeclaration is commonplace. The incentive to cheat is great: the transfer tax is often levied at a high nominal rate, and underreporting appears to be detected only very rarely. The result is not only that revenue is lost but also that the recorded sales data – the necessary base for any objective assessment of the annual property tax – are suspect. Although little attention appears to have been paid to this problem in most countries, high transfer tax rates lead to under-declaration, which fatally compromises the use of techniques like CAMA. The result is that appraisal is dependent on thirdparty data (for example, from banks and brokers) and subjective estimates. This outcome is especially pernicious when, as is often the case, the annual property tax is the main own-source revenue of local governments and the  property transfer tax, which usually wrecks the prospect of a good local property tax, is imposed by higher-level governments. Several directions for reforming property transfer taxes come to mind. One approach is simply to abolish the property transfer tax, making up any revenue loss at the local level by an increased annual property tax and at higher levels by increased income and consumption taxes. A less drastic approach is to lower the transfer tax rate significantly and to make much more effort to monitor declared values: for example, by requiring certified appraisals at the expense of buyers/sellers; by upgrading valuation staff at the government level; and by imposing significant penalties for under-declaration. A quite different approach is to replace the property transfer tax with a capital gains tax on real property. In principle, such a tax embodies a ‘selfchecking’ feature because buyers and sellers have opposing interests that could lead to a more accurate statement of sales values and hence provide better sales data as needed to strengthen valuation for the annual property tax.29 Of course, this alternative would be difficult; and those developing countries that have in the past attempted to tax capital gains on real property have not managed to collect much revenue, and have sometimes given up the effort.30 Assessing capital gains would obviously be a notional exercise, at least initially; but it would perhaps be no more administratively difficult to tax capital gains on real property sales than to properly monitor the self-declared sales values that drive property transfer taxes. A final alternative that deserves more thought relates to the treatment of property transfers under the value-added taxes (VATs) that now exists in most countries. Despite the many ways in which these VATs differ from an ‘ideal’ VAT (James, 2015), one way in which they almost all differ from each

M4439-BAHL_9781786435293_t.indd 263

01/03/2018 11:08

264

Fiscal decentralization and local finance in developing countries

other is in how they treat transfers of real property (Schenk et al., 2015). As yet, no one appears to have figured out (either in theory or practice) whether and how such sales should be taxed under a VAT. One interesting proposal is that sales of real property should, like any other sale, be subjected to VAT. Cnossen (1996, p. 245) characterizes the ‘anachronistic’ taxes on property transfers as no more than a poor “proxy for the VAT that should have been levied on the increase in the value of immovable property realized in the sale . . . [which] represents the capitalized value of the increase in the value of the . . . services of the immovable property that belongs in the VAT base.” He suggests that they should simply be abolished and replaced by VAT.31 Like the capital gains alternative, however, the VAT option leaves two problems unresolved. First, since the replacement tax in both cases would be a national tax, local governments, if they had previously received the revenue from the transfer tax, would either need to raise some other tax (e.g. the property tax) or receive increased transfers to compensate. Second, while the information base on property sales might be improved by either the capital gains or the VAT approach, for local governments to receive any direct benefit they would obviously need full and immediate access to the improved data base – a requirement that experience with intergovernmental data transfer in many countries suggests that few countries will be able or perhaps even willing to provide. The best approach in most countries might be to follow the gradualist path of lowering the rates of transfer taxes, enforcing them more rigorously and ensuring better use of the sales data in improving the administration of the annual property tax.

VALUE CAPTURE Urbanization increases the demand for land and for serviced residential and non-residential properties. Changes in real estate values are driven in part by the relaxation of government constraints on urban development (e.g., by zoning changes that allow development on the urban fringe) and in part by government expenditures (e.g. on infrastructure investment). In principle, such value increases are reflected to some extent in the property tax base. In practice, however, lags in revaluation and the generally low effective rates of the property tax in developing countries mean that the amount of such value increments captured by the property tax is likely to be low. There has been considerable discussion of, and some experience with, using other fiscal instruments to capture a portion of the land value increments attributable to government actions, often with the funds being

M4439-BAHL_9781786435293_t.indd 264

01/03/2018 11:08



Taxing land and property ­265

used to support the financing of public investments and public services.32 With high rates of urbanization expected to continue over the next two decades, significant increases in public investments and property values are expected to continue. There is a strong case for the public sector to receive its fair share of the increment in land values attributable to government action. Those who benefit from the action of the collectivity should not be able to retain all the benefits for their own use. If an investment of, say, US$10 million in a new road increases property values in the affected areas by $20 million, it seems only right that at least the cost of the project should be borne by those who benefit. Since such land value increments are ‘unearned’ because the benefiting property owners did nothing to generate them, it may even be considered fair to claim a larger public share. This approach seems especially advantageous since the tax base is (by definition) a ‘rent’ so that taxing it would, unlike most taxes, create no economic distortion. Government actions that create new value for private owners without any effort from them – whether building a road or changing land use regulations or zoning (e.g. altering permissible land uses) – provide a potentially important, equitable and efficient revenue base. Some have argued that the use of value capture tools may permit local governments to shape urban development in better ways. While further exploration of this possibility takes us well beyond our remit in this book, we underline the important fact, sometimes underplayed in these discussions, that there are significant costs in applying these tools, most of which are intended to generate revenues to finance specific public works projects. Designing and implementing projects, estimating potential land value increases, identifying potential beneficiaries and negative externalities, and getting all the relevant political actors on board are all complex, costly and time-consuming tasks. It is much easier to think of such ideas than to turn them into reality, particularly given the limited analytical and financial resources usually available at the subnational level in developing countries. Nonetheless, it can be done and some large-scale projects in several countries have been successfully financed this way (Smolka, 2013). Land value capture is hardly a newly discovered approach. The idea – originally proposed by George (1879 [1958]) – was a basis for the municipal land value increment tax in the early 1900s in Germany (Backhaus, 2000) and was introduced in Colombia as the valorization tax (valorización) in the 1930s, where it began to be used more extensively to finance local public works in the 1960s (Rhoads and Bird, 1967). In the same period, a similar levy called ‘land adjustment’ began to be widely used in Korea (Doebele, 1979). These and other features of the main approaches to value

M4439-BAHL_9781786435293_t.indd 265

01/03/2018 11:08

266

Fiscal decentralization and local finance in developing countries

capture are discussed in Box 6.6. All have two key ingredients. First, a local government determines that it has a marketable service product to sell – e.g., some combination of improved public services, land, development rights, building permits, increased floor area ratios (FAR) or zoning changes.33 Second, some private party – such as a developer or property owner who expects to benefit if government carries out the activity on offer – either chooses to or is required to pay for some or all the expected benefit. Payment is usually made by (directly or indirectly) channeling a share of the expected increase in land values to the local government. The precise arrangements are often determined in part by the nature of the project: for example, whether it is a road improvement, a large-scale urban redevelopment project, an increase in building heights or the extension of public services to the urban fringe (Box 6.6). Bureaucratic and electoral politics play an important role in determining the effectiveness of land value capture projects. Officials may downplay good economics to enhance the possibility of selecting their favored projects, usually those that fall within their ministry or department. In China, for instance, provincial and local government officials were arguably overenthusiastic about land leasing because of the direct (pocketbook) and indirect (political gains from expanded growth) benefits they received as a result. Elsewhere, elected officials may have been tempted by the lure of apparent shortcuts enabling them to move ahead with project implementation. With so much money (and/or influence) at play, and with the rules shaping value capture programs being more ad hoc than formal, the scope for corruption may sometimes be all too great, especially when it is often so unclear who really benefits (and how much) and who really pays (and how much) under many value capture schemes. The extent to which value capture and beneficiary financing overlap is a subject ripe for research. It is important to remember that capturing land value increments through special instruments is not that different in many ways from the annual property tax. Both tax increases in land values, albeit usually at different times, and both use the funds to support the provision of local public services – although, in a less closely related way, much the same can be said of property transfer taxes. All three instruments are related and should be more closely coordinated in their design and administration than is now the case in most countries. For example, the base of the property tax and the property transfer tax should ideally be the same – and would be if valuations were based on market value and kept up to date. Such a valuation base would also provide a level starting point for estimating potential land value increases in addition to providing a data base that could be used to improve such estimates substantially,

M4439-BAHL_9781786435293_t.indd 266

01/03/2018 11:08



Taxing land and property ­267

BOX 6.6  INSTRUMENTS USED FOR VALUE CAPTURE Betterment Levies Probably the best-known fiscal instrument used for value capture is the betterment levy – a tax or charge levied to recover the costs of a public investment from specific beneficiaries (Doebele, 1998). The tax rate may be set to recover costs, or it might be set as some portion of the expected increase in land values due to the project; and the liability is usually distributed among beneficiaries according to such indirect measures of benefits as the size of their property and its distance from the project. It may also be adjusted to take account of such other factors as land use and ability to pay. Such taxes and charges have been around for a long time, for example, in the form of special assessments to recover the cost of sewer and drainage projects, and other public works in North America. In Brazil’s 20 largest municipalities betterment levies finance over 10 percent of total municipal investment (Vetter and Vetter, 2011). Similarly, some Colombian cities collect significant revenues from valorization charges. In Medellín, for example, valorization accounted for 45 percent of local government revenues in the early 1980s, and in Bogotá, 24 percent of revenues in the 1990s. Local governments have some discretion as to how they levy the charge: some recover investment cost; some recover value increase; and different formulas are used to allocate the levy among those determined to be within the scope of the project. Rhoads and Bird (1967) suggest that a principal reason why this system was more successful in Colombia than in such countries as Ecuador and Mexico was because of the greater attention paid both to careful engineering and cost and benefit analysis before determining the tax rate, and more effort to secure the consent of taxpayers before allocating tax liability. In general, collections and the level of satisfaction with the process seem good (Borrero Ochoa et al., 2011). Although the results of this approach are generally equitable and efficient, the process of assigning payment shares to beneficiaries, though generally sensible and in some cases inventive, is inevitably subjective. A problem with relying heavily on this means for financing public investment is that one outcome may be to push public investment too much toward higher-income neighborhoods where larger land value increases are likely to occur and where there is likely to be both more willingness and ability to pay. Exactions and Charges for Building Rights Changes in government regulations such as zoning and allowable floor-area ratios (FAR) may increase land values, but they may also impose new costs on society such as increased congestion and the displacement of low-income families. Exactions, which often take the form of a payment in kind, are a way to capture some of these benefits for public use and to compensate for some of the costs. Exactions are often designed in such ad hoc ways as requiring developers to provide open public space, to pay compensation for the costs imposed on public services by increased congestion, and to provide low-income housing. In Rio de Janeiro, for example, the developer of a downtown office building was required to

M4439-BAHL_9781786435293_t.indd 267

01/03/2018 11:08

268

Fiscal decentralization and local finance in developing countries

renovate nearby historic buildings; in other Brazilian cities developers were required to extend sewerage trunk lines (Smolka, 2013). In some cases, if FAR is increased or zoning variances granted, a process may be set in motion to recover some of the expected land value benefits – e.g., up to 50 percent of the land value increase under Colombia’s plusvalia system. Brazil imposes a direct charge for changes in FAR, based on the expected increase in land values. Over a five-year period in the late 1980s, payments for additional building rights were US$650 million in São Paulo (Sandroni, 2010). An alternative approach, common in some areas in North America, is to impose ‘up front’ development charges on new developments intended to cover the costs of expanding public infrastructure – sewers, water, roads, schools, etc. – to accommodate the expected population increase in the area. In principle, such charges could not only provide substantial revenue but could also be an important urban planning tool. Well-run cities like those in Ontario, Canada, collect substantial funds from development charges. However, even in these cases, the opposition of powerful political groups (and sometimes even corruption) often results in such charges being applied on a simple unit-cost basis that fails to accurately reflect the important cost differences associated with different projects (Slack and Tassonyi, 2017). Land Sales and Leases Leasing or selling public land is another area where the issue of land value increments often arises. The most striking recent example is undoubtedly the leasing of urban land by local governments in China. This policy enabled the financing of the large amount of infrastructure needed to absorb nearly half a billion migrants into Chinese cities. By 2013, lease revenues accounted for about one-third of subnational government revenues (inclusive of intergovernmental transfers) and 7 percent of GDP. On the other hand, it also resulted in dispossessing farmers from urban fringe land with little compensation, the channeling of a substantial amount of money into well-connected private hands, and over-extended local borrowing and a subnational debt crisis (Bahl et al., 2014; World Bank and DRC, 2013; World Bank, 2012). When land is owned by the government and land use rights are leased, local governments have in their hands a powerful tool for both revenue mobilization and for shaping the urban form. This has long been recognized and can be well used, as the example of Canberra, Australia – all of which is built on leased public land – makes clear. Although local property taxes have now largely replaced lease rentals as a source of income, sales of land-use rights continue to be an important source of revenue in Canberra (Sansom, 2009). As Rao and Bird (2014) recently noted in India, the existence of substantial public lands in large cities (e.g. defense establishments) might similarly be used to provide revenue to finance the needed expansion of urban infrastructure. Land Adjustment Some countries have financed expansion into the urban fringe or large-scale redevelopments with changes in land use by land adjustment schemes (Hong and

M4439-BAHL_9781786435293_t.indd 268

01/03/2018 11:08



Taxing land and property ­269

Needham, 2007). Under such schemes, landowners transfer a portion of their (unserviced) land to the government (or to a designated entity) in return for serviced public provision in the project area. The idea is that although owners will end up with land elsewhere that may not be the same size, they will nonetheless also end up with more valuable land. The government then uses some of the land it gets in exchange to finance such public services as roads or parks and sells the rest to recover its costs. Land adjustment, though its main purpose is usually to assemble larger parcels of land, has also been used for cost recovery in Germany and Japan, and more recently in Korea (Doebele, 1979) and Israel (Alterman, 2007) as well as Brazil. This approach is particularly useful in land value capture when there are numerous owners/occupiers, and even where the existing plots are irregularly shaped. However, it is seldom a simple or quick process. Although it avoids expropriation of land, it requires the consent of many landowners, which may entail significant transactions costs (Hong, 2007). CEPACS Often, the major problem in value capture is to estimate the increase in land values. Neither rules of thumb nor econometric analysis may provide a satisfactory (acceptable) answer to this problem. An innovative Brazilian program of bonds called Certificates of Additional Construction Potential (or CEPACS, from the Portuguese name) takes a different approach by letting the market estimate the increase in an auction of development rights. CEPACS, which can be used only in an urban area that a city government has designated for public investments, are denominated in area units, classified by type of development right and tradeable. They are then auctioned by the Federal Bank of Brazil, with the number on offer being controlled by the municipality. CEPACs have been most successful in São Paulo for urban development projects, though other applications have included road projects and revitalization projects. The city issuing the bonds obviously benefits from the revenue, while developers benefit not only because of the expected future increase in value but also because they can decide on the timing of their investments according to market conditions. However, as Sandroni (2010) notes, this approach is not likely to work so well in many other countries. Certificate holders may bear high risks and need to have considerable resources and financial expertise, especially since it may take a long time for a secondary market for trading certificates to develop. Equally importantly, a system like this – essentially a way to sell the rights for future development – makes sense only when urban property markets are buoyant.

for example by permitting more precise estimates of the likely effects on values of, say, improved road access. A unified tax roll would also, of course, substantially reduce administrative costs and should help make land and property taxation a significantly more robust and elastic source of local revenue.34

M4439-BAHL_9781786435293_t.indd 269

01/03/2018 11:08

270

Fiscal decentralization and local finance in developing countries

HOW TO REFORM THE PROPERTY TAX The property tax has not lived up to its revenue potential in developing countries. In some ways this is a self-fulfilling prophecy because most governments neglect its administration and often erode its base with preferential treatments. Taxpayers do not think the tax is fair, and few politicians are interested in being its champion. The usual recommendations for reform – revalue accurately and collect better – are sound but are no more likely to be accepted in the future than in the past. The underlying problem remains the absence of strong enough incentives to stimulate governments to make their property tax a more relevant part of their revenue mobilization system. Nonetheless, the science of better property taxation is fairly clear, so we conclude by suggesting a few ways to move forward with property tax reform. To begin at the end, since revenues are the main interest most governments have in the property tax, the first question to ask is how any reform will affect revenues. To do so, we begin by setting out the linkages between the components of the tax and its revenue yield in the form of a simple identity: TC/Y 5 (TC/TL) (TL/TAV) (TAV/AV) (AV/MV) (MV/Y) where TC 5 property tax collections, TL 5 property tax liability, TAV 5 taxable assessed value, AV 5 total assessed value (including exempt properties), MV 5 market value and Y 5 GDP. As this equation shows, the property tax ratio – its share of GDP – results from the combination of five other ratios: ●● ●● ●● ●● ●●

the relationship between real property wealth and GDP (MV/Y); the assessment ratio (AV/MV); the exemption ratio (TAV/AV); the statutory tax rate (TL/TAV); and the collection rate (TC/TL).

Taking the relative importance of real property wealth as given for any country at a particular time, the property tax ratio is determined by the other four terms in the equation, each of which is in principle within the control of the government.35 Property tax reform can then focus on any or all of these items – assessment, exemption, rate or collection. As an illustration of the use of this identity, consider the case of a hypothetical ‘average’ country shown in Table 6.3. The first numerical column in Table 6.3 is the baseline calculation,

M4439-BAHL_9781786435293_t.indd 270

01/03/2018 11:08



Taxing land and property ­271

Table 6.3  Simulation of impact of alternative reforms (%)

Revenue  as % of GDP (TC/Y) Collection  rate (TC/TL) Statutory  rate (TL/TAV) Exemption  rate (TAV/AV) Assessment  Ratio (AV/MV) Real  property wealth ratio (MV/Y)

Baseline case

Collection-led reform

Valuation-led Comprehensive reform reform

0.006

0.009

0.009

0.018

0.5

0.8

0.5

0.8

0.0214

0.0214

0.0214

0.0214

0.7

0.7

0.7

0.9

0.4

0.4

0.6

0.6

2.0

2.0

2.0

2.0

Note:  See text for details and designations.

with assumed values for the policy and administrative determinants that are roughly in line with (more or less) international averages in low- and middle-income countries based on information from case studies and data sources. We assume an assessment ratio of 40 percent, a collection rate of 50 percent and an exemption rate of 30 percent. We further assume that the level of revenue mobilization is the average for low- and middle-income countries (0.6 percent, as calculated from IMF data). The most difficult parameter to calculate is the ratio of real property wealth to GDP: here we use the average ratio of total wealth to GDP for a sample of countries, and the ratio of reproducible capital and urban land to total wealth (both as reported in World Bank, 2006). Finally, we calculate the statutory tax rate implied by the average reported revenue ratio.36 The other columns of Table 6.3 report the results of three simple simulations. Columns 2 and 3 compare the revenue impacts under the two most frequently advocated approaches to property tax reform – s­ tarting

M4439-BAHL_9781786435293_t.indd 271

01/03/2018 11:08

272

Fiscal decentralization and local finance in developing countries

from the end (collection) or the beginning (assessment) of the range of tools available to government. In principle, either a collection-led or a valuation-led reform could produce the same revenue outcome. For example, in this case a 50 percent increase in revenue (with the same tax rate) could be achieved either by increasing the collection rate from 50 to 80 percent or increasing the assessment ratio from 40 to 60 percent. The collection-led approach would almost certainly be less costly to implement and would increase revenues more quickly. But it would do nothing to alter the basic problems of significant and usually very uneven assessment. On the other hand, the more expensive valuation-led approach would rectify these problems but would do nothing to catch evaders and non-payers. The best way is to follow an approach that collects some revenue up front from these groups while also rectifying the more fundamental assessment problems. The comprehensive approach illustrated in the last column of Table 6.3 goes further, and tackles not only both ends (collection and valuation) but also the middle (base erosion) by lowering of the exemption rate to 10 percent. In this illustration, such an approach could, without increasing any legal rates, triple revenues to nearly 2 percent of GDP. The moral of the story told in Table 6.3 is as obvious as it is well known. The best solution is to improve everything – base coverage, assessment accuracy and collection. The result will be an economically efficient, fair and productive property tax. Leave out any component –back (valuation), middle (base erosion) or front (collection) – and none of these aims will be satisfied. As we have emphasized throughout, there is no ‘one size fits all’ solution in property tax reform. There are also no easy shortcuts. We conclude with two sets of guidelines for reform. The first is a list of key points that would-be property tax reformers in any country should bear in mind: ●●

●●

To begin, have as clear an idea as possible of exactly what you want to achieve and why you want to achieve it. Is the concern primarily revenue mobilization? To tax property wealth more effectively? To stimulate more intensive land use? To establish a sound fiscal basis for decentralization? Clear thinking about the primary objectives of reform and a thorough analysis of precisely how the existing property tax system is out of line – economically, legally, ­administratively – with these objectives is the place to begin developing a sound reform strategy.37 Make sure that the legal underpinnings of the property tax – from the constitution to the implementing regulations – are clear to all

M4439-BAHL_9781786435293_t.indd 272

01/03/2018 11:08



Taxing land and property ­273

●●

●●

●●

●●

●●

and consistent with the desired definition and coverage of the tax base, and the tax rate structure. Make sure that the division of property tax administration between different levels is equally clear and consistent with policy intentions. As discussed earlier, this may, for example, suggest that certain administrative activities should be more centralized when local governments are weaker and that the optimal division of responsibilities should be different in large cities and poor rural areas. Pay special attention to the need to improve the necessary information infrastructure and to ensure that the necessary cross-agency information flows take place. For example, accurate records on property transfers and new buildings are critical to effective property tax administration. Make sure the structure of the tax is consistent with the desired objectives, which will in most instances require broadening the tax base, removing preferential treatments, and simplification. At a minimum, all exemptions should be reviewed periodically, the tax expenditure implied should be annually calculated and reported, and a sunset period should be set to review and reconsider every exemption. Since graduated property tax rate structures and classification systems complicate the administration of the tax and sometimes introduce unwanted distortions, they should be avoided. A better and fairer approach is a flat rate tax applied to all types of property, perhaps exempting most low-income housing.38 Avoid ‘fiscal engineering’ – attempting to achieve this or that specific goal by tinkering with tax rates or tax exemptions: this game is almost never worth the candle. The system of property and land taxes in many countries is a kind of schedular system, with urban property subject to one rate schedule, agricultural use to another and sales of property to another. Moving this system towards one that taxes a more uniform base and a single rate structure will usually reduce administrative costs, increase revenues and have desirable economic effects. Finally, tougher enforcement and a more realistic set of penalties are usually needed, as well as some legal and policy changes to take such actions – for example, making it possible to seize delinquent properties, to intercept bank accounts to collect delinquent taxes or to increase the financial penalties for late and missing payments.

There is little new in any of these guidelines. Many have said most of these things for decades in many countries. But experience has unequivocally demonstrated that, as with most tax reform issues, the key points are not

M4439-BAHL_9781786435293_t.indd 273

01/03/2018 11:08

274

Fiscal decentralization and local finance in developing countries

economic (why are we doing this?) or technical (how should we do this?), but political: how might we persuade people (and politicians) that they should do this? It is more difficult to set out a similar set of guidelines to help would-be reformers on this front, but some useful ideas do emerge from the literature. For instance, Table 6.4 lists some ideas put forth in a recent paper by Slack and Bird (2015) on how to reform property taxes in Canada. Most of the problems listed in the first column of Table 6.4 seem equally important in developing countries, and most of the possible solutions to these problems listed in the other two columns may also be relevant at least in some instances. Equally, some of the points set out below with respect to improving the ‘political economy’ environment for property tax reform in developing countries may prove equally relevant in the many developed countries (from the UK to Greece) that are currently wrestling with how best to reform their property tax systems (Slack and Bird, 2014). This last comment is applicable to a key ‘political’ point: namely, that to be successful any significant property tax reform needs a political champion at the center – someone willing and able to play the role of building sufficient support for reform and then carrying it through. Candidates for this task seldom leap to mind, but perhaps those who are strong supporters of fiscal decentralization for whatever reason may find it appropriate to sell property tax reform as an essential strengthening plank for local government finance. In addition, if decentralization is the goal then, as argued earlier (and developed further in Chapter 7), it is essential to think of both local taxation and intergovernmental transfers if one is to get either one right. One way to tie the tax transfer knot more securely may be to condition the growth of transfers to local governments to some degree on how effectively they tap their potential property tax base, a point we discuss further in the next chapter. Finally, since the largest potential tax base, the greatest administrative capacity and the biggest local financing needs are usually found in the biggest cities, another key to success may be to direct support (and incentives) initially to those cities where local champions are keen to increase revenues and willing to implement reforms. Subsequently, other localities that are willing to make serious efforts along these lines may similarly be given both more authority and more access to increased financial resources.39 For such an asymmetric approach to reform to be feasible the constitutional and legal structure, as well as political sensibilities to differential treatment, need to be carefully taken into account. As a rule, such an approach makes sense only when the central (or higher-level) government is willing to step in and play a more direct role in administration in the localities that have insufficient resources – and perhaps also, though this is always more arguable, insufficient will – to do the job adequately without special support.

M4439-BAHL_9781786435293_t.indd 274

01/03/2018 11:08



Taxing land and property ­275

Table 6.4  Problems blocking property tax reform, with possible solutions Issues and problems

Promising approaches

Problematic approaches

Salience: property tax is more visible than other taxes

Couple tax reform  with improvements in local services Phase-in gradually Withhold tax at  source and other payment options

Assessment limits Property tax  capping

Liquidity constraints: imperfect association between taxpayers’ incomes and property taxes, especially for seniors

Means testing,  Tax deferrals for seniors More payment options Phase-in

Assessment limits Property tax capping,  exemptions based on age

Perceived regressivity: taxes higher as a percent of income for low-income taxpayers

Property tax credits Tax deferrals Bundle with other tax reforms Package with  expenditure changes Low-income housing  exemptions

Classified tax rates Progressive tax rates Assessment limits Property tax  capping, owner-occupied exemptions

Volatility: potentially large swings in taxes for some taxpayers

More frequent reassessment Index base Taxpayer education  and communication in understandable form Phase-in

Assessment limits Property tax  capping, frequent statutory rate changes

Presumptive tax: tax base inherently arbitrary

Taxpayer education Accessible appeal process Phase-in

Self-assessment Classified property   tax rates Assessment limits Property tax capping,   preferential treatment

Inelasticity (a problem for local governments, not for taxpayers): taxes do not increase with growth

Frequent reassessments Index base, eliminate  tax preferences, keep property tax roll up-to-date, modernize collection procedures and introduce more stringent penalties

Frequent statutory   rate increases Tax amnesties

Source:  Based on Slack and Bird (2015).

M4439-BAHL_9781786435293_t.indd 275

01/03/2018 11:08

276

Fiscal decentralization and local finance in developing countries

NOTES   1. For reviews of property taxation in developing and transition countries see e.g. Bahl (1979), Youngman and Malme (1994), Bird and Slack (2004), Bahl et al. (2008, 2010), UN Habitat (2011), Norregaard (2013) and Kelly (2014). Recent country or regional studies include de Cesare (2012), Bahl et al. (2014), Fretes Cibils and Ter-Minassian (2015), and Franzsen and McCluskey (2017).   2. For a discussion of the equity and efficiency dimensions of the property tax in the US, see Zodrow (2006). For a discussion of the incidence of the property tax in developing countries, see Bahl and Linn (1992).   3. In Rio de Janeiro, for instance, the 15 percent of households earning more than 10 minimum wages held 35 percent of the aggregate residential wealth of all homeowners, while the 25 percent of households earning less than two minimum salaries held only about 15 percent of the property wealth (Vetter et al., 2014).   4. For example, comparing data from a study of the total wealth of countries in 2000 (World Bank, 2006) with estimates of GDP, the ratio of total wealth to GDP was 10.2 in the US, 7.2 in Brazil and 4.5 in India.   5. This is a variant of the fiscal contracting argument made for some Latin American countries in Bird and Zolt (2015a).   6. For a comprehensive discussion of the impact of a land value tax, see Dye and England (2009).   7. Ideally, such exemption should apply only when property is used for the stated main purpose of the entity: for example, churches may not be taxed but the exemption would not be extended to all lands owned by the church.   8. For a discussion of the practice of payments in lieu of taxes in the US, see Kenyon and Langley (2016).   9. In Canada, it is the federal government, and not the taxing authority, that determines the values and rates to be used in the payment in lieu calculation with respect to federal property. 10. To make this illustration simpler, we assume that the base of the property tax is the market value of all real property. We also assume that all sectors (commercial, industrial, residential, etc.) are treated the same. These assumptions are easily removed, but of course the presentation of the model would then be more complicated. 11. A classic example was the well-known Proposition 13 introduced in the US state of California which capped the growth in the property tax base at the rate of inflation, provided for new valuation only in the case of new construction or resale of property, and froze the statutory tax rate. For discussion of the property tax limitation movement in the US, see Youngman (2016). 12. For a detailed account of the virtual reversal of many of the key reforms during the decade following a major administrative reform in Ontario, Canada, see Bird et al. (2012). 13. See e.g. the discussion in Bird and Zolt (2008) of some of the problems encountered in the past in various countries when adopting new levels of tax technology. See also McCluskey et al. (2017) for a discussion of the problems with attracting and holding qualified valuers in the government sector. 14. For example, some years ago simply spending an hour going through readily available aerial photographs of a large city in Argentina revealed that about 45 percent of the built-up urban area was not recorded in the cadaster. 15. As Gainer (2017) discusses, however, this problem should now have been resolved by unifying these (and other) departments. 16. There are always complications. For example, the legal taxable base may be a specified fraction of market value or it may be fixed between revaluation periods. The issue of what properties are ‘comparable’ is also not always easy to resolve. 17. For a detailed analysis of the strengths and weaknesses of this approach in one Canadian province, see Bird et al. (2012).

M4439-BAHL_9781786435293_t.indd 276

01/03/2018 11:08



Taxing land and property ­277

18. As Slack and Bird (2014) discuss, the reluctance to revalue and the resistance delayed revaluation generates are equally apparent in OECD countries such as the UK and Italy. 19. For example, there are fewer than 100 professionals in such countries as Namibia, Malawi, Uganda and Swaziland. 20. The collection rate is measured as the amount of tax collected from current year liability expressed as a percentage of the tax liability for the same year. Payments for arrears should be excluded from this calculation, although it is not clear that this has been done in all the studies mentioned. 21. As an example, in one small country all efforts to reform the property tax in the capital city were blocked by a large local landowner who was extremely closely connected to the governing party. 22. In fact, there may be no net revenue for the government: an unpublished background study underlying Musgrave (1981, pp. 349–50) found that (in inflation-adjusted terms) the administrative costs of dealing with tax appeals substantially exceeded the additional revenues obtained. 23. One can go too far with friendliness, however. Mauritius, for example, allows taxpayers to write off debts older than five years, a provision that would seem to reduce rather than increase the incentive to pay. 24. Many developing countries find it hard to charge interest on deferred taxes. Perhaps they should consider increasing the tax rate and then providing a discount (back to the original rate, assuming the increased rate just allows for an appropriate public sector interest rate as the cost of delay) for earlier payment that would reward early payers, penalize late payers and ensure that both government and taxpayer took the ‘time value of money’ properly into account. 25. Colombia, for example, considered a major reform of rural property taxes as part of its attempt to reincorporate localities in the countryside, long dominated by various guerrilla and anti-guerrilla forces, into the governance system (Garzón and Vázquez-Caro 2004). 26. For a book-length treatment of this issue, now out of date but unfortunately still all too relevant, see Bird (1974). 27. For example, Bahl (2004, 2009) and Alm et al.(2004). For a brief review of earlier experience, see Bird (1967). 28. To the extent taxes on land are capitalized into land values, they are borne by all owners of land. If landownership is concentrated in the higher-income brackets, the distribution of any tax on property is then progressive. 29. Sellers would, as with the property transfer tax, of course prefer to report lower prices. However, buyers who contemplate selling at some future date – as most perhaps do – would prefer to report a higher price (which would lower their tax liability on future sales). Both sides could collude and split the ‘tax saving’ difference, but any compromise would still likely be higher than the present ‘seller only’ reported price. 30. For an early examination of this issue, see Amatong (1968). 31. For further discussion, see e.g. Bird and Gendron (2007), who suggest something like the Cnossen proposal – but only for non-residential property. 32. Land value recapture instruments are most developed in Latin America, where workable approaches have been developed (e.g. in Brazil and Colombia), and much recent discussion has been fostered by the Lincoln Institute of Land Policy. For thoughtful reviews, see Smolka (2013) and Hong and Needham (2007). 33. The floor area ratio is the ratio of floor area to the net surface of the undeveloped land (where net surface is defined as excluding rights of way and environmental set-asides). 34. Bahl et al. (2011) propose such a unified roll in Pakistan. 35. This approach is set out in Bahl and Linn (1992). The same identity has of course been used by many others to discuss the revenue responsiveness of property tax revenues to changes in rates, base and administration. 36. As shown in Tables 6.1 and 6.2, the effective tax rate is of course much lower than the legal rate. In fact, since only 0.8 percent of GDP is assessed, only 0.56 is taxable and

M4439-BAHL_9781786435293_t.indd 277

01/03/2018 11:08

278

Fiscal decentralization and local finance in developing countries

only half the taxes assessed on this amount (0.28) are collected at the legal rate, the effective tax rate on property wealth on average in column 1 of Table 6.3 is only about 0.6 percent. 37. For examples of this approach see, e.g., the studies of property tax reform in Macedonia and Montenegro in USAID (2006), Jamaica in Sjoquist (2005) and Pakistan in Bahl et al. (2015a). Of course, as the remainder of the text list suggests, much more than good initial design is required for successful reform. For reviews of reform efforts in a variety of countries, see Bird and Slack (2004), Bahl (2009), Slack and Bird (2014), and Franzsen and McCluskey (2017). 38. Income distribution concerns should not be a major issue in designing property taxes in developing countries. Landownership is usually concentrated in the top brackets, most low-income residents are likely to be outside the tax net, and the effective rates of even a reformed tax are in any case usually low. 39. As mentioned earlier, a good example of the importance of such a local champion may be found in the Sierra Leone study by Jibao and Prichard (2015). As this study shows, the best leadership is not always found in the biggest cities.

M4439-BAHL_9781786435293_t.indd 278

01/03/2018 11:08

7.  Intergovernmental transfers While the economic objectives of the transfer system are clear, the transfer system actually represents a political compromise. . . . Nevertheless, it is important to keep in mind the economic objectives because that serves as the benchmark for reforming the transfer system. (Rao and Singh, 2006, p. 222)

Intergovernmental transfers are the fiscal cornerstone of subnational governments in developing countries. Transfers are a compromise between centralization and decentralization. They allow central governments to control most revenue-raising and distribution decisions while decentralizing some expenditure responsibility to subnational governments. But countries compromise to different degrees and in different ways. China retains all legislative revenue-raising powers at the center but assigns about 85 percent of expenditure responsibilities to subnational governments. Canadian subnational governments account for about 75 percent of total government spending, but they finance over 85 percent of these expenditures from their own revenue, with the balance coming from (mostly unconditional) transfers. Subnational governments in Colombia account for about a third of total government spending, but depend on transfers (most of which are conditional) for about half their revenues. In all cases, the results now observable are the outcome of substantial shifts over time, reflecting considerable technical and political investments. We begin this chapter by noting the importance of fiscal transfers in developing countries. We then discuss the various rationales and objectives that underlie such transfers, and explore the considerable variations found in practice around the world. Next, we turn to the impacts of fiscal transfers, the evidence about whether these impacts appear to match the goals their designers had in mind, and the sorts of reforms that appear to offer some promise of improving outcomes. Finally, we take up the issue of monitoring and evaluation. Three points emerge clearly from this discussion. The first and most important is that the right transfer system for any country depends on both its policy objectives and its circumstances. As usual, there is no magic solution applicable everywhere. Second, it is critical to view intergovernmental transfers as a system and to assess its impacts not only in total but also in terms of its components. Third, as with most policy 279

M4439-BAHL_9781786435293_t.indd 279

01/03/2018 11:08

280

Fiscal decentralization and local finance in developing countries

areas, both art ­(politics) and science (economics) are required to design an effective transfer system. Despite the rhetoric that so often surrounds transfers, it is not all politics. Clear economic and management principles are needed to design and implement a good intergovernmental transfer system.

THE IMPORTANCE OF TRANSFERS As Mathur (2012) notes, reliance on intergovernmental transfers by local governments in low- and middle-income countries has grown significantly in recent years. Subnational expenditures have increased with population and income growth in many countries, but few have loosened their tight control over local government taxing and user charge powers. Mostly, intergovernmental grants have expanded primarily to fill the gap in local budgets. Although transfer systems almost everywhere remain a work in progress, and some major issues remain unresolved, we have learned much over the years about how developing countries approach the job of structuring transfers, what they seem to be trying to do with their transfer systems, and to some extent how much success they have had – although the kind of empirical work needed to really answer that question is, as usual, severely hampered by data problems (as shown in Box 7.1). As Bahl and Wallace (2007) show, the ‘vertical share’ of subnational governments – defined as intergovernmental transfers (Tr) divided by total central government taxes (Tx) – may be decomposed as follows: Tr/Tx 5 [Tr/SE * SE/Y] / Tx/Y where Tr/SE 5 intergovernmental transfers as a percent of subnational government expenditures (the transfer dependency effect); SE/Y 5 subnational government expenditures as a percent of GDP (the decentralization effect); and Tx/Y 5 central government tax revenues as a percent of GDP (the revenue mobilization effect).1 These three components of the vertical share for a sample of low- and middle-income countries are shown in Table 7.1.2 These data show that, although subnational government expenditures are financed more by transfers (40 percent) in developing than in high-income countries (38 percent), the vertical share of transfers in total central tax revenues is substantially lower in developing countries (13 percent) than in richer countries (19 percent). The lower level of taxation in developing countries constrains their ability to give more transfers to subnational governments,

M4439-BAHL_9781786435293_t.indd 280

01/03/2018 11:08



Intergovernmental transfers ­281

BOX 7.1 ASSESSING THE SUCCESS OF INTERGOVERNMENTAL TRANSFERS Intergovernmental transfers go by many different names: grants, shared taxes, revenue sharing, entitlement programs, subsidies and subventions. Two programs with the same name may be quite different, while two programs with very different names may be almost identical. What matters are the details of how programs are designed and work, not what they are called. From our perspective, the essential feature of an intergovernmental transfer is that a higher-level government redirects some of its revenues to the budget of a lower-level government.1 One may be tempted to argue that another feature of a transfer (as distinct from a local tax) is that the receiving government cannot influence the amount of the revenue transferred. However, this is not always the case. Sometimes subnational governments can influence the amount they receive by ‘buying into’ a cost reimbursement (matching) grant at a higher rate or by doing better than others in a competition for shares of a fixed amount of some transfer program. However, provided local governments do not have discretion to affect the rate or base of the revenue source that funds an intergovernmental transfer, we classify it as such rather than as either a local revenue or a national direct expenditure. A clear definitional rule is important both to make meaningful comparisons over time and space and to draw on experience elsewhere for inspiration and guidance on policy analysis and design. To understand just how any transfer design functions in any setting one needs to understand in detail how the transfer is distributed, under what conditions it can be spent and how the behavior of recipients is monitored. Even when definitions are clear, evaluating success in one setting is a surprisingly difficult and complex task, and extending that success to a different setting is even more complex (Deaton and Cartwright, 2016). The analysis of grant impacts is constrained by the absence of sufficiently detailed, comparable data on specific types of transfer.2 For example, there is no data set for developing countries that will allow comparison of the importance of equalization grants in transfer systems or the degree or nature of conditionality of transfer payments. We often cannot even tell if a transfer system is composed of many small grants or a single large transfer. Without plunging into the institutional intricacies of the country or countries concerned, one can learn very little about the practice of transfers. Notes: 1. Some transfers (e.g. subsidized loans) may not show up in full or at all in the budget of the recipient government. Nonetheless, they add to the total resources available to spend – to some extent at least, as discussed further below – at the discretion of that government. 2. IMF (2014a, pars. 5.100–105) GFS defines general government grants (line 133) as transfers “that do not meet the definition of a tax, subsidy, or social contribution.” GFS data are ‘net’ transfers (that is, deducting any transfers that flow in the opposite direction). A ‘transfer’ is a payment for which the donor government does not receive any ‘good, service or asset’ in return. However, as we discuss later, this is not an accurate way to describe many intergovernmental transfers, since they are often explicitly intended to affect the behavior of the recipient government so that it will, in effect, provide a ‘service’ for the donor government in the form of doing what the donor wants rather than what it might want to do if left on its own. Moreover, the only sub-classification made in GFS data is between ‘capital’ and ‘current’ grants, with anything a government chooses to label a ‘capital’ transfer being classified as such. All this is understandable, given resource limitations and the desire for international comparability, but using such data to draw conclusions about what is really going on may be misleading.

M4439-BAHL_9781786435293_t.indd 281

01/03/2018 11:08

282

Fiscal decentralization and local finance in developing countries

Table 7.1  The importance of intergovernmental transfers Transfers as % of central taxes High-income  countries Developing  countries Transition  countries

Transfers Subnational Transfers as % of expenditures as % of GDP as % of subnational GDP expenditures

19.0 (22) 13.3 (72) 14.2 (22)

38.1 (25) 40.1 (18) 29.4 (23)

5.5 (25) 2.2 (73) 2.9 (22)

15.4 (21) 6.4 (32) 9.5 (22)

Central government taxes as % of GDP 31.3 (23) 17.5 (80) 23.9 (20)

Note:  Numbers in parentheses indicate number of countries included in each sample. Source:  Bahl and Wallace (2007).

and the lower level of expenditure decentralization reduces the demand for transfers. An econometric analysis of these data produced results consistent with these hypotheses: higher levels of expenditure decentralization exerted a strong positive effect on the claim of transfers on central government taxes, and a higher rate of central government revenue mobilization (with revenue mobilization treated as endogenous) had a negative marginal effect (Bahl and Wallace, 2007). The dictum that ‘money sticks where it hits’ appears to be applicable: when central governments in developing countries get more revenue, at the margin they are unlikely to give more of it to subnational governments.

THE RATIONALES FOR INTERGOVERNMENTAL TRANSFERS Perhaps the real question is: why do central governments transfer any of the taxes they collect to subnational governments? Five principal reasons may be identified: 1. To close the ‘vertical gap’ – the difference between the revenues that can be raised by local governments (if they make a normal effort) and the expenditures required to cover assigned local government responsibilities (at some normal level). 2. To reduce unwanted disparities in tax burdens and in the level of services provided by subnational governments.

M4439-BAHL_9781786435293_t.indd 282

01/03/2018 11:08



Intergovernmental transfers ­283

3. To internalize the external benefits and costs of local government expenditure programs. 4. To compensate for assigning taxing powers to the level of government with the lowest marginal cost of raising funds. 5. To respond to political concerns. Addressing so many different objectives with one policy instrument is difficult, which is one reason why there are so many types of transfer and so much complexity in the world of transfers. The same factor probably explains in part why attempts to derive rigorous explanations of the impact of the grants are difficult and seldom satisfactory. We may sometimes be able to get it right for an individual grant but are unlikely to be able to do so for the system as a whole. Transfer design does not lend itself to such simple maxims as ‘broaden the base and lower the rate’ in tax design or ‘finance follows function’ in the assignment of taxes and expenditures. Vertical Balance: Closing the Gap Vertical balance refers to the extent to which subnational governments finance their budgets from revenues that they raise from their own sources. One of the most vexing problems in intergovernmental finance is finding the right balance. The bottom line on this is that the best balance will vary from country to country, depending on a whole host of institutional, economic and political characteristics of the country. One way to approach the job of finding the right vertical balance is to start with a situation in which all local government expenditures are financed with local own-source revenues – no intergovernmental transfers – as the counterfactual. Then the rationale for departing from this baseline, and its impacts, can be explored in a systematic way. The main reason more transfers are usually required is because as economies develop, subnational expenditures tend to increase more quickly than subnational taxing powers are expanded. Some level of vertical imbalance persists in all countries for three additional reasons. First, central governments may simply be unwilling to devolve more revenue authority to regional and local governments (China, Egypt). Second, even if they do so, some regional and local governments invariably still end up with too little fiscal capacity to finance even whatever minimum level of public services the central government considers essential – and others may not be willing to do so. Third, since some expenditures for which subnational governments are responsible generate external benefits, some central fiscal assistance may be needed to achieve the proper level of expenditure on such services. No matter how one gets to a decision about the right vertical balance

M4439-BAHL_9781786435293_t.indd 283

01/03/2018 11:08

DFj=

bjTi Ti 1 CT

DFj= 284

bjTi Ti 1 CT

Fiscal decentralization local finance in developing countries 0IE and0IE 0G 0C

= · · 0FD 0G 0C 0FD for subnational governments, doing so is unavoidably a subjective exercise 3 Even when we try for a that may easily turn into a political adventure. 0IE 0IE 0G 0C = · · more formal approach to defining 0FD fiscal 0G balance, 0C 0FDthis subjectivity gets in the way, For instance, the vertical gap (or vertical fiscal imbalance) can be defined as the difference between the amount 0EG T(a) 0EG 0Csubnational governments c =ifa they bexert a a b‘normal’ revenue effort can raise from own-revenue sources 0FD y 0C 0FD and (b) the amount they must spend to provide a ‘minimum’ level of the Tcassigned government services that have0EG been 0C Assuming the two 0EG to them. =any a particular ba b terms are clearly defined, the0FD gap for y 0C 0FD level of subnational government is then: GAP 5 a ( Ei 2 Ri ) (7.1) &



&

i

~ the revenue raised from own sources at normal effort by local where 5R GAP 5 a ( Ei 2 Ri ) i ~ 5 the amount of i expenditure needed to provide a government i and E i minimum level of assigned services in local government i. The targeted a (GAP) vertical share (VS) – the share VS 5of central taxes allocated to subnational CR transfers – is then: &

&

xa1i(GAP) x2i G 1 g (7.2) G Ri VS 5 g5 1 2 CR x a ax x1i gap that the x2i central government where a is the percent of the & financing Gand 1g G total amount of g1 2 Ri Ri 5 system, intends to cover with the transfer CR is the ax ax revenues raised by the central government from current sources.4 & Few countries make such calculations, of course, or even attempt to Ri by such key terms as normal tax effort formalize exactly what they mean or minimum service provision. In many countries, most decisions on transfers seem often to rely on little more than a few crude indicators combined with someone’s intuition about expenditure needs and a close reading of the political tea leaves to determine what is feasible and who needs to be made most happy by the result. Perhaps little more is needed. Expenditure ‘need’ is inherently a subjective concept; and since the public generally has little idea of the real tax price of local public goods in any case, the demand for local government services is always likely to outweigh the capacity (or willingness) at any level to finance them. There will thus always be some gap to be filled – and it is usually, though seldom completely, filled by central transfers. Still, our view is that it is better to be as systematic as possible in such things, and to edge away from intuition and toward transparency wherever possible. Working with a conceptual framework such as described in equations (7.1) and (7.2) might help us move in this direction.



M4439-BAHL_9781786435293_t.indd 284

01/03/2018 11:08



Intergovernmental transfers ­285

The key policy question in any country is not whether there should be transfers – there will be – but how large should they be. Economics fails us here, and we are left to work with sensible conjectures. One possible rule of thumb is that when there are no benefit spillovers, the richest subnational governments should be able to cover the cost of providing all assigned (local benefit) services at prescribed minimum levels and with (again) normal tax effort (Bird, 1993). This can work if expenditure assignments are mostly for local benefit services and if taxing powers are a reasonable match for expenditure assignments, as for example is the case in South African metropolitan cities. Of course, local fiscal capacity in most other localities would be inadequate to achieve this result, leaving a gap to be filled by transfers. But transfers are not the only way to close fiscal gaps, vertical or horizontal. Alternatively, the central government may reduce the expenditure responsibilities of lower-level governments and/or give them more revenueraising powers. In most countries, subnational governments appear to prefer transfers to any of these options. Central governments often agree. Most low-income countries assign relatively few expenditures to subnational governments and are reluctant to give them sufficient taxing authority to pay for them on their own. Regional and local government taxes account, on average, for less than 3 percent of GDP and less than 12 percent of total tax revenues in less-developed countries (see Table 2.1). In higher-income countries, on the other hand, subnational governments seem on a path of financing more of their spending from their own revenues. Although the share of transfers in subnational government expenditures varies greatly among OECD countries, from 26 percent in Korea to about 1 percent in New Zealand in recent years, this share has tended to decrease in most countries (Blöchliger and Vammalle, 2010). Equalization Almost all developing countries are characterized by large inter-regional variations in the strength of their local economies. In a study of East Asia, for example, Hofman and Guerra (2007) point out the wide differences in per capita GDP between the richest and poorest provinces: Indonesia (17 times), China (11 times) and Vietnam (9.5 times). De Silva et al. (2009) report the difference in Russia to be even greater, with the richest being 69 times better off than the poorest.5 With urbanization, these income ­differences may grow even larger as the metropolitan areas continue to capture the benefits of agglomeration while the poorer places fall further behind (see Chapter 8). We argued in earlier chapters that effective fiscal decentralization

M4439-BAHL_9781786435293_t.indd 285

01/03/2018 11:08

286

Fiscal decentralization and local finance in developing countries

requires some significant devolution of revenue-raising power. But giving more independent revenue raising powers to subnational governments, other things being equal, tends to widen fiscal disparities.6 Stronger economies go hand in hand with stronger tax bases, and often with better ability to administer taxes. The same may be true on the expenditure side of the budget, that is, richer urban areas have more ability to absorb devolved local functions and do a good job with service delivery. It is thus not surprising that many countries have compensatory measures intended to address fiscal disparities. Political leaders everywhere at least wave their arms at the issue of large fiscal disparities among regional and local governments, even if they are not very successful in doing much about the problem. In OECD countries, for example, there is a strong relationship between the degree of tax decentralization and the strength of equalization transfers, with “a 10 percent increase in the sub-central tax share . . . associated with a 15 percent increase of equalizing grants” (Blöchliger and Vammalle, 2010, pp. 176–7). However, as we discuss below, it is not clear that the same can be said in less-developed countries. At first glance, it seems simple enough to design an equalizing transfer system: measure the extent of fiscal disparities; decide how much of the inter-regional gap should be eliminated; and then adopt a grant formula that will produce the desired degree of equalization. In fact, the design issues are not so simple, and few developing countries have successfully implemented an equalizing grant system. To do so requires not only establishing the level of the (appropriately defined) fiscal gap to be filled in each jurisdiction but also finding a distribution formula that will get the job done, both tricky tasks that we discuss in more detail later. In addition, it must be decided whether the grant will be unconditional or earmarked for specific purposes; whether the transfer (or at least the task of monitoring any conditionality) will be the responsibility of line ministries or a central agency; and whether any conditions should focus on short-run programs (for example, upgrading specific public services in poor localities) or longer-run targets (e.g. upgrading the infrastructure for major public utilities). Moreover, even developed federal countries with long-standing and extensive equalization programs find it necessary to revise and alter them from time to time. None of this can be done once and thereafter left alone: an equalization system needs continual monitoring and modification as circumstances change.7 Finally, we take the view that, contrary to practice in some countries, equalization grants should be unconditional. There is a place for conditional grants in the intergovernmental transfer system, but that place is not embedded in a general equalization transfer, where the primary objective is usually to establish as level a playing field as possible for all

M4439-BAHL_9781786435293_t.indd 286

01/03/2018 11:08



Intergovernmental transfers ­287

local ­governments (rich and poor) by enabling them to provide a minimum package of local public services with normal tax effort.8 Externalities Another rationale for intergovernmental transfers is to stimulate increased provision of local public services when some benefits ‘spill over’ to residents of other jurisdictions. The design of the transfer should provide an incentive for local decision-makers to take externalities into account. In addition, the grant should be conditional, i.e., earmarked for the activity in question. The size of the required transfer depends on four things: (a) the degree of external benefits involved; (b) the marginal cost of providing the good; (c) what the granting government can afford; and (d) the extent to which the local government will respond to the grant incentive (that is, the price elasticity of demand for the good). Presumably the donor government knows what it can afford (c) and it can perhaps make a reasonable estimate of the cost (b). It can also assess the political gains from providing central financing for the service in question. However, even in the datarich environment of industrialized countries the other two key elements mentioned – the measurement of externalities (a) and the price elasticity (d) – are seldom based on much more than guesswork. Almost never does any government investigate ex post whether the effects of such conditional grants are more or less than they ‘should’ be in any economic sense. For these reasons, despite the prominence of externality-compensating grants in the literature, imposing conditionality on more general equalization grants makes little sense (Smart and Bird, 2010; Bahl, 2010a). Relatively few local expenditure responsibilities in most developing countries generate large spillovers of benefits to other areas on which local governments may otherwise underspend significantly. There is almost no evidence that the matching rate (the share financed by the transfer) implicit in the design of most spillover transfers reflects either the size of the spillover generated or the likely demand responsiveness of recipient governments. If a country decides that a general transfer is a good idea, whether for gap-filling or equalization, it would be wise to ignore the temptation to clutter up the transfer design by tying its use to specific expenditures or imposing other ‘targeting’ conditions. A final factor to be kept in mind with respect to the effectiveness of conditional grants is the fungibility of money. A simple example makes the point. State A receives a $100 conditional grant for primary education. If the state responds by spending the full $100 for primary education, all else being the same, the conditional grant is fully effective: it generates a $100 net increase in expenditures. However, at the same time as it increases

M4439-BAHL_9781786435293_t.indd 287

01/03/2018 11:08

288

Fiscal decentralization and local finance in developing countries

expenditure on education, State A may reduce taxes by $100: did the grant finance $100 additional spending on education or a $100 tax cut? Alternatively, State A may have always intended to budget another $100 for primary education, and the receipt of the grant allows it instead to buy some new chairs for the governor’s office or to channel an extra $100 in resources to the water supply agency. To the extent budget resources are fungible, conditional grants can become unconditional and desired stimulative impacts – even if they appear in the numbers – may not be real. Reducing Costs The simple fact that broad-based taxes are generally best administered at the central level often means some type of transfer program is required to flow some central revenues to lower-tier governments (Martinez-Vazquez and Timofeev, 2010; Mikesell, 2007). Administrative and collection costs per unit of revenue collected are especially likely to be high in smaller and more rural areas (Bahl, 2013). As discussed in earlier chapters, some taxes and charges – such as the property tax and local business fees – may often be better administered by regional and local governments. However, the major tax sources – income and consumption – are usually imposed at the central level; so if the subnational sector has been allocated more functions than it can finance from the package of revenues within its own control, it sometimes makes good sense for the central government to collect taxes and then to transfer some of the revenue to subnational governments. As we discuss later, how this is done may have different implications for the effectiveness of fiscal decentralization. An additional reason for preferring central administration that one often hears from central officials is that regional and local officials and politicians are less capable (and more liable to corruption) than they themselves are. However, experience in a variety of countries casts doubt on the general validity of such arguments, and sometimes corruption may even be less at the local than at the national level. As we discussed in Chapter 2, the empirical evidence on such matters is, to say the least, inconclusive. Political Justifications Political factors shape intergovernmental transfer policy in many ways, but probably more in the direction of increasing the vertical share of transfers than reducing it. At one level, for example, central politicians and officials may be unwilling to accept the loss of control, influence, power and possibly even position that may result if subnational governments gain more autonomy in either raising revenue or spending it. Even if political impera-

M4439-BAHL_9781786435293_t.indd 288

01/03/2018 11:08



Intergovernmental transfers ­289

tives require giving something to lower-level governments, the center often continues to hold the reins as much as possible, for example, by denying significant taxing power to regional and local governments, by attaching various conditions to any transfer payments and by imposing expenditure mandates. One reason for devolving expenditure responsibilities to subnational governments – common, for example, in the transitional countries of eastern and central Europe in the 1990s (Bird et al., 1995) – was simply to shift some of what would otherwise have been the central budget deficit down to subnational governments by assigning them expenditure responsibilities without providing any funding. Even when some transfer funding is provided, it is often subject to annual discretionary changes by central government. Sometimes the distribution of transfers to individual local governments may reflect transparent political objectives, as in India and the Philippines where national legislators have been allowed to allocate some amount to local governments at their discretion. Even when there is not such an overt and obvious ‘political’ component, political leaders, factions and groups commonly influence the flow of transfers to different regions and localities in a variety of ways. In some instances, subnational governments have given up their taxing powers in return for a claim on the federal tax base. Two prominent examples are Argentina and Mexico, where the states surrendered their taxing power in exchange for a guaranteed share of central tax collections.9 Most other cases in which the central government replaced subnational taxes by a compensatory transfer have been less voluntary on the part of local governments. In South Africa, for example a combination payroll and turnover tax levied by local governments – admittedly a poorly structured tax (Bahl and Solomon, 2003) – was replaced by central transfers. In Bangladesh, India and Pakistan, a local ‘import duty’ (octroi) was similarly abandoned in favor of a compensating grant to local governments. Tanzania and Kenya abolished their local government personal taxes in favor of a central grant. Such changes may well have eliminated distortionary local taxes. But they were also driven to some extent by the fact that central politicians could gain favor with voters without having to deal directly with the revenue loss implied: they had no problem in abolishing someone else’s tax. Nor is it a great surprise to learn that the promised full compensation with transfers did not usually occur. In all the cases just mentioned, for example, the supposedly compensating transfers turned out to be less income-elastic than the taxes they replaced, and the local governments ended up as losers. As Mathur (2012) noted in South Africa, for example, local governments had to draw on other transfers to make up for the revenue losses from the abolished tax.

M4439-BAHL_9781786435293_t.indd 289

01/03/2018 11:08

290

Fiscal decentralization and local finance in developing countries

Broader political trends may also impact transfer policy. In Russia during the Yeltsin years, political factors led to more power being given to the regional governments; with Putin, the pendulum swung back to centralization. In Indonesia, after the fall of Suharto in 1999, the new political leadership embraced decentralization with some fervor and follow-through. Something similar happened in South Africa immediately after the end of apartheid. In other countries, as discussed in Chapter 2, regional conflicts and secession movements and other factors have similarly shaped both the level and the design of transfer systems over time.

THE ARCHITECTURE OF TRANSFER SYSTEMS Every intergovernmental transfer has two dimensions: the vertical share, the pool of revenues to be distributed to local governments; and horizontal sharing, the formula or other way in which the distributable pool is allocated to the recipient units. Both dimensions are integral to the design of any transfer. A taxonomic approach to transfer design developed by Bahl and Linn (1992) allows us to classify the main types of transfers with a simple two-way classification, vertical sharing in the columns and horizontal sharing in the rows, as set out in Table 7.2. Although there is much more to transfer design than a simple 2 × 4 matrix can encompass, this framework is a good starting point to discuss how different combinations of vertical and horizontal sharing choices may result in very different impacts. Table 7.2  The architecture of intergovernmental transfers Allocating the divisible pool among eligible units (horizontal sharing)

Specified share of national or state government tax

Annual budgetary decision

Origin of collection   (derivation) Formula Cost reimbursement  (matching) Ad hoc

A

Not applicable

B C

E F

D

G

Note:   For discussion of type A–G transfers, see text. Source:  Adapted from Bahl and Linn (1992).

M4439-BAHL_9781786435293_t.indd 290

01/03/2018 11:08



Intergovernmental transfers ­291

Determining the Vertical Share In principle, the share of central revenues transferred to subnational governments should be determined by the goals the central government wishes to achieve.10 In practice, however, political compromises, attitudes toward decentralization, history, institutional factors and data availability may all play important roles in shaping the architecture of transfer programs. The two main approaches taken in developing countries to determining the vertical share are tax sharing and discretionary budget allocations. Tax sharing Under the tax sharing approach, the central government allocates a share of national collections of some tax (or all taxes) to the regional governments and/or local governments.11 At one extreme, countries may share collections from all (or nearly all) taxes with subnational governments. The cornerstone of Indonesian decentralization in 2001 was to share 26 percent of all ‘domestic’ revenues with local governments. The Philippines allocates 40 percent of internal taxes (excluding import duties) based on the third preceding year, to local governments. In India, all proceeds from union (central government) taxes are assigned to the divisible pool: in 2016, the share assigned to the states was 42 percent. In Pakistan, the provincial government share is 57.5 percent of central taxes. Mexico allocates 20 percent of central taxes collections to a general revenue sharing pool for distribution to state governments. Another approach is to share revenues from only certain taxes. China, for example, designates 60 percent of income taxes and 50 percent of the VAT for provincial governments. Peru shares 10.5 percent of national VAT collections with municipal governments. Russia shares 73 percent of the enterprise income tax and all of the personal income tax on a derivation basis (De Silva et al., 2009). Latvia earmarks 75 percent of revenues from personal income tax for local governments, while Hungary and Poland share up to 40 percent of personal income tax revenues with local governments. Revenue sharing between Brazilian central and state governments is funded by 21.5 percent of the revenue from federal income tax and VAT. The tax-by-tax approach is also used in some industrialized countries. Japan’s ‘local allocation tax’ is really a transfer funded by 32 percent of central personal income tax and liquor tax revenues, 35.8 percent of ­company income tax revenues, 29.5 percent of consumption tax revenues and 25 percent of tobacco tax revenues. Australia allocates all the proceeds of its GST (VAT) to the states. Germany has different sharing rates for most major taxes. Sometimes, as in most of the cases mentioned, a relatively large share

M4439-BAHL_9781786435293_t.indd 291

01/03/2018 11:08

292

Fiscal decentralization and local finance in developing countries

of central revenue is transferred; in other cases, only a small share may be reallocated. For example, only the property tax is shared with local governments in Jamaica, and only 5 percent of Kenya’s income tax is used to fund its Local Authority Transfer Fund. Subnational governments usually prefer shared taxes because this arrangement gives them a claim to an income-elastic source of revenue and provides a degree of security that is unattainable when transfers are subject to discretionary changes. Money received without either having to persuade one’s constituents to tax themselves or having to crawl to the capital to beg from the central government is not easily turned down by any politician. Provided the money flows without conditions and can be spent as local governments wish – as is the case, for example, in Indonesia, the Philippines and India – there is little downside from the view of subnational governments. In some cases, however, shared taxes are earmarked for specific purposes. In Peru, for instance, although a specified share of natural resource taxes is shared with subnational governments, the money comes with the condition that it must be spent on infrastructure (CanavireBacarreza et al., 2015). Tax sharing has much going for it: it clearly addresses the revenue– expenditure gap, is transparent, often income-elastic and frequently gives recipient governments a fair amount of autonomy in spending the money. Because it is an entitlement, it may be considered part of the glue joining different regions into a nation. Not all is roses, however. First, sharing taxes clearly limits the fiscal flexibility of the central government. The center is obligated to transfer a fixed share in good times and bad. Of course, there are many ways to define just what is shared, so central governments have in practice often proved quite capable of dodging this bullet when times are bad, keeping to the word but not the spirit of the law. Viewed from below, a concern is that tax sharing makes subnational revenues sensitive to central government tax policy changes. Shifting VAT from a production to a consumption base in China, for example, had an important impact on the revenues of subnational governments. With sharing, central tax incentives may in part be at the expense of subnational governments. When sharing is limited to certain taxes, central governments may choose to use their political capital and fiscal efforts more effectively in raising non-shared taxes rather than those where some revenues flow to others. Or central governments may simply decide to reduce the legal sharing rate to deal with a deficit or to fund something else, as happened in China in its latest rounds of revenue sharing reform. Even if central governments do not make discretionary changes in tax structure or sharing rules, to the extent the shared tax base is influenced by business cycles, the pain is shared by subnational governments through the shared tax system.

M4439-BAHL_9781786435293_t.indd 292

01/03/2018 11:08



Intergovernmental transfers ­293

The global economic downtown in 2008–2009 significantly compromised the revenue position of many Eastern European local governments (for example, Bucharest, Budapest and Zagreb) that rely heavily on shared personal income taxes (Bahl, 2011; Golovanova and Kurlyandskaya, 2011). Questions may also be raised about whether the sharing rate is too high (i.e., too much goes to subnational governments) or too low (too little is going to subnational governments). If it is too high the result may be to dampen the enthusiasm of the central government to collect taxes that largely go to others. Too high a sharing rate may also both discourage local governments from mobilizing own revenues and leave the central government with too little to cover its needs. China dealt with this perceived problem by reducing the sharing rate. Pakistan, which has also felt the pinch, is trying to regain balance with more expenditure decentralization. On the other hand, if the sharing rate is too low, local governments may argue that they do not have sufficient resources to cover the expenditure gap. Even if no one can measure this rather subjective concept in a way that cannot be questioned, it is not difficult to use this argument in a political context. Under such pressure, local governments that act as collectors of national taxes may sometimes reduce the transfer of revenue to the central government by paying some of their own bills directly from the national taxes they collect (Bahl, 1999; Wong, 2013). A seemingly small issue that can be surprisingly troublesome is whether the subnational government share is measured against budgeted or actual central government revenue. If the base is actual revenue, there is a timing problem since actual collections are often not known for a year or even more, thus introducing a degree of uncertainty in subnational government budgeting. The Philippines attempts to deal with this problem by sharing 40 percent of actual total central revenues collected three years earlier – e.g., 2017 revenue sharing is based on 2014 tax collections. The opposite approach is taken in Mexico, where the mandated transfers are paid to subnational governments daily, based on actual collections. If the sharing base is budgeted collections, both the central and subnational governments assume some risk, especially in countries that are revenue dependent on products sold in world markets. In the Indonesian decentralization of 2001 the central government disbursed transfers to the regions based on budgeted central revenues. When the rupiah depreciated against the US dollar, actual revenues (and expenditures) were driven up, and the central government saved an amount equivalent to about 0.5 percent of GDP (Fengler and Hofman, 2009). Does the central government see revenue sharing as an inviolate contract, or is it likely to reduce transfers if it thinks it necessary to preserve its own programs? In the Philippines, where the constitution and the

M4439-BAHL_9781786435293_t.indd 293

01/03/2018 11:08

294

Fiscal decentralization and local finance in developing countries

local government code would seem to allow the government to reduce the general grant allocation in the event of an unmanageable public sector deficit, governments in the past did so (Capuno et al., 2001). However, in 1999, when the President attempted to withhold an additional 10 percent of the allocations, he was overruled by the Supreme Court (Diokno, 2009); the actual allocations ranged from 87 to 98 percent of the required amount in 1998–2001 (Manasan, 2009). Other countries have faced similar issues: for example, Yatta and Vaillancourt (2011) report underfunding of entitlements in Ghana, Côte d’Ivoire, Gabon and Cameroon. Finally, as mentioned earlier, central governments are not above gaming the system to avoid sharing the full entitlement with subnational governments, for example, by relabeling some taxes or otherwise excluding them from the sharing pool. In Argentina, a significant new tax on financial transactions was introduced in the 2000s but was not included in the revenue sharing pool. In 2003, the Netherlands changed the definition of the tax sharing base so that revenue devolution would be reduced (OECD, 2006). Sometimes central governments earmark certain revenues for specific central expenditures largely to reduce the size of the general revenue sharing pool. Or they can simply change the law, as Russia did in 2002 when it switched the base of its equalization fund from a fixed percent sharing of all federal taxes to a fixed real amount (Kurlyandskaya, 2005). Colombia did much the same thing, though it provided for some growth in the real amount of the grant fund (World Bank, 2009a; Sánchez Torres et al., 2015). Discretionary budget allocations The alternative approach to fixing the size of the vertical share is to make it part of the annual budget process. The central government (or the Congress) determines that some amount will be distributed to regional and/or local governments for some year (or period of years). Compared to the tax sharing approach, this gives the central government more certainty with respect to the flow of transfers that it must budget for, but leaves the subnational governments in a considerably less secure position. The budget allocation approach has some advantages over the tax sharing approach. It gives the central government some breathing room to introduce a program that might not become a long-term mainstay of local government budgets; i.e., because its effectiveness may not be all that certain, it might be a candidate for privatization, or it might involve funding for a project or program with a finite life. By contrast, funding such activities from tax sharing may lead local governments to treat the transfer as an entitlement. Other advantages depend on where one sits in the intergovernmental

M4439-BAHL_9781786435293_t.indd 294

01/03/2018 11:08



Intergovernmental transfers ­295

finance system. Programs tend to be owned, usually by a line ministry, and therefore usually have champions. The main defense for local governments against cutbacks is the combination of their political clout and the support of the various line ministries concerned with the delivery of different services (e.g. education, health). Because the political fortunes of the various ministries and programs change, the resulting vulnerability of subnational government budgets can be a major drawback of the discretionary budget approach. A capital project might be cancelled, leaving construction of a rapid transit system only partly finished (as in Jakarta and Bangkok) or the country might decide to de-emphasize line ministry grants (as in India).12 Some smaller programs may not have a strong constituency and can get lost in the crowd when it comes to budget cutting. In 2004, for example, Tanzania’s conditional grants were contained in 21 different budget votes (Boex and Martinez-Vazquez, 2006). In Australia, about 40 percent of transfers are made in 90 conditional grants for both current and capital purposes (Hull and Searle, 2007). About 50 percent of all grants in Mexico are from eight conditional grant programs (Revilla, 2012). In a time of budget crisis, programs that are funded annually are often the first to be cut. On balance, there are probably more discretionary transfers than those financed by tax sharing. The discretionary approach is more suitable for smaller programs and for tackling specific needs, and has less an air of permanency than tax sharing. Most countries use both approaches, although country practice is quite varied both in terms of the extent to which vertical sharing is discretionary and the extent to which conditions on how it may be spent are imposed. Tax sharing is the approach most often used for general-purpose aid to subnational governments; but the discretionary approach is also sometimes used, with general revenue sharing programs for subnational governments as a separate annual budget line. In Colombia’s revenue sharing program, the ‘pool’ is a fixed, indexed amount where the annual budget line is different every year. In Canada’s federal–provincial equalization system, the size of the distributable pool allocated by the federal budget, which is determined based on recent trends in some average ‘basket’ of provincial revenues, is also different every year.13 Many countries that follow the discretionary budget approach to ­vertical sharing use conditional grants to some extent. Some are quite large, as in the case of India’s largest transfer program for urban public sector assistance (Mohanty, 2014; Ahluwalia et al., 2014). South Africa’s major transfer to local governments (‘the equitable share’) was also structured this way. More often, however, the conditional approach is used mainly for such purposes as special initiatives for disaster relief or for some other specific purpose deemed worthy by the central government.

M4439-BAHL_9781786435293_t.indd 295

01/03/2018 11:08

296

Fiscal decentralization and local finance in developing countries

The major drawbacks to the budgetary allocation approach are that: it is not very transparent; it is subject to frequent one-sided ‘re-contracting’ which invites political manipulation; it results in considerable budgetary uncertainty for subnational governments; and it perhaps encourages the central government to view this use of resources as being of lower priority than competing expenditure requests. Despite the argument made above for keeping grants intended primarily for equalization unconditional, the budgetary discretion approach may perhaps encourage the imposition of higher levels of conditionality in order to exert stronger central control over local use of the funds. On the other hand, the fact that the transfer is discretionary may encourage local politicians and officials to seek relief from the central level instead of trying first to deal with their own problems by better managing their spending and perhaps raising more local revenue. Local politicians often welcome the opportunity to blame local problems on someone higher up and somewhere else. Horizontal Sharing The distribution of the revenue sharing pool among eligible subnational governments usually follows one of four methods: a derivation approach, a formula approach, a cost reimbursement approach (including matching grants) and an ad hoc (discretionary) approach. Since the derivation approach applies only when the size of the pool is determined by tax sharing, this gives seven distinct ways to design a transfer (see Table 7.2). Examples of all seven are found in the practice in developing countries. The derivation approach The derivation approach (Type A in Table 7.2) allocates a shared national tax to subnational governments in accordance with the collection of that tax within the geographic boundaries of the relevant subnational government. For example, 50 percent of VAT collections in China are allocated to provincial governments largely according to VAT collections within the province.14 The issues of the appropriate taxes to share and sharing rates are summarized in Box 7.2. Derivation-based sharing may sometimes be justified by the lower marginal cost of funds under central tax administration. The central government likes it because it retains full control over its tax base and rates. Subnational governments like having access to more productive and elastic revenue sources. Since richer and often politically stronger regions will presumably get more from this system than poorer ones, they will find this approach especially attractive. Poorer regions are less likely to be happy. The political calculus will differ from country to country and from time to

M4439-BAHL_9781786435293_t.indd 296

01/03/2018 11:08



Intergovernmental transfers ­297

BOX 7.2  WHAT TAXES TO SHARE?1 Two key decisions in structuring a tax-sharing system are (a) which taxes should be shared, and (b) what the sharing rate(s) should be. Often these decisions appear to be made, as it were, backwards along the following lines: 1. Current economic and political conditions require that $X needs to be transferred to subnational governments to close all or some proportion of their fiscal gap. 2. The current collections of Tax A, or 50 percent of Tax B, or 10 percent of all taxes yields roughly the amount required. 3. End of story. Much the same ‘number-matching’ approach is sometimes used to determine not only the size of the transfer but also its allocation.2 Although the amount needed to meet immediate needs is an important consideration, more thought should be given to both the taxes and the amounts shared, as a few examples may suggest. If regional and local governments are to hitch their revenue hopes to a share of national taxes, they are well advised to seek a share of the whole pie rather than any particular piece of it, in part because every piece has its own problems and dangers. From a subnational perspective, most taxes give rise to various problems in terms of administration, volatility, allocation or susceptibility to central discretionary action. On the other hand, from a central perspective, if some but not all revenues are subject to some degree of sharing, national tax policy decisions are inevitably distorted: what government is willing to take political heat for increasing taxes that mainly raise revenues for other governments? Sharing VAT is tempting: this tax produces a lot of revenue and tends to grow with the economy, so even a small cut of the take may be lucrative. However, the tax is more complex than many think, and dividing up the pie on a regionally sensible ‘derivation’ basis is by no means simple. It can be done without severe damage to the tax or the economy only, as in Canada, by severing the amount allocated to any region from the amount of collections actually made in that region and tying it to some economically logical base such as the share of taxable final sales in the region.3 Similarly, a sensible way to deal with exports and imports appears to be that employed in Canada and for the most part within the European Union (EU), although the system is considerably more porous in the EU owing to the lack of an overlapping union-wide tax and the totally unintegrated tax administration (Bird, 2015a). Other alternatives that have been employed elsewhere are flawed.4 If they cannot get a piece of the value-added tax (VAT), most lower-level governments would probably like to have a piece of the corporate income tax (CIT), since everyone likes to tax someone else and no one is ever sure who ends up bearing the burden of CIT. Unfortunately, as McLure (1998) shows, both in principle and in practice CIT is a poor candidate for either local taxation or derivation-based revenue sharing.5 Although whatever is done along these lines is more likely to be driven by subjective and political concerns than economic reasoning, those who advocate for access to this tax base should be warned that the CIT is both the most cyclical of all taxes, and the one most vulnerable to policy twitches at the central

M4439-BAHL_9781786435293_t.indd 297

01/03/2018 11:08

298

Fiscal decentralization and local finance in developing countries

level. Regional and local governments are generally ill-advised to tie the financing of the essential services that they provide to such an unstable revenue source. Another chunk of national revenues usually comes from a personal income tax (PIT), although such taxes are significantly less important in developing than developed countries (Bird and Zolt, 2005). In developing countries, most PIT usually comes from wages in the formal sector. Still, it is more logical as a basis for a derivation-based transfer than VAT or CIT because the share collected locally is presumably borne mainly by local residents. Most revenues will also be collected in richer areas, which is one reason China reduced the provincial share of PIT from 100 percent to 40 percent in 2010. Similar problems arise with payroll taxes, which, like PIT, may also lead to allocation problems when many employees reside in other localities, as in the metropolitan area including Mexico City (Bahl, 2011). Finally, some countries share certain excise taxes with subnational governments on a derivation basis. Mexico earmarks a percent of federal taxes on tobacco, beer and alcohol for transfer to subnational governments, as well as a vehicle use tax and a tax on the registration of new vehicles. In Indonesia, local governments receive a fixed share of revenues from four provincially collected taxes – motor vehicle tax, vehicle transfer tax, fuel excise tax and groundwater extraction and use tax. As experience shows – see the case of Colombia discussed in Bird (1984) – it can be difficult to determine the destination of goods like alcohol, tobacco and fuel that are almost invariably taxed at the production level. Either a cumbersome and perhaps expensive tracking system must be created or an approximate allocation formula applied. A presumptive system can often be found that does the job adequately, although the very arbitrariness that makes such schemes workable means that the results are always arguable. Notes: 1. See also the discussion in Chapter 5 of most of the taxes mentioned here. 2. Argentina offers an example. The allocation formula established in 1988 was said to have as its only virtue that it “crystallized the outcome of the political negotiations in the inflationary 1980s” (World Bank, 1996a, II-33). The current coparticipación formula, as tax sharing is called in Argentina, still bears traces of that compromise. 3. The Canadian system, which has the additional important virtue from the perspective of decentralization of permitting provinces to set their own VAT rates, thus making it a ‘local tax’ rather than just a transfer, is set out in detail in Bird and Gendron (2010). 4. Calculating regional shares on the (origin) production basis, for example, as in Brazil and perhaps (though this is not yet quite clear) in India, perpetuates many of the economic distortions associated with pre-VAT forms of sales tax. Changing the base of the tax to something much closer to a retail sales tax would (as often suggested in the EU) again make it more economically distorting as well as potentially susceptible to being used for protective purposes. Allocating by imposing some sort of split on ‘headquarters’ collections, as in China, is subjective and arbitrary. 5. Even Canada only managed to make subnational corporation income taxes workable by applying an essentially arbitrary formula for base-sharing country-wide – after spending several decades arguing about the matter (Smith 1998). As Bird and Wilson (2016) show, there is still good reason to worry about the results of the present system.

M4439-BAHL_9781786435293_t.indd 298

01/03/2018 11:08



Intergovernmental transfers ­299

time. As we note below, the issue has been a particularly contentious one when it comes to sharing natural resource revenues. This approach has been widely used in the transitional countries emerging from the former Soviet bloc, and is also found in many other low- and middle-income countries. Although such shared revenues are sometimes labeled as ‘local taxes’ they are in fact really transfers because subnational governments have no control over the legal tax rate or the legal tax base, and their entitlement to any share of the revenue is determined by central legislation. Even if, as in Brazil and Argentina, tax sharing is constitutionally based, control over rates and bases remains entirely under central control. Interestingly, an important variant of the derivation approach, tax base sharing, under which subnational governments are free (sometimes within limits) to set their own tax rate and sometimes also (invariably within limits) even to determine some aspects of the tax base, with the central government collecting these taxes along with its own, appears almost never to be used in developing countries.15 A common concern with the derivation approach is that poorer localities get smaller per capita transfers than richer regions with stronger tax bases and better collection rates. Regional disparities in tax revenues tend to be widened with derivation-based tax sharing. As Zhang and MartinezVazquez (2006) note, nine of China’s 28 provinces collect 70 percent of the derivation-based income taxes. São Paulo, Brazil’s most developed state, accounts for 22 percent of the national population but 43 percent of all tax collections. For this reason, some countries have created additional equalizing transfers, sometimes indirectly financed by reducing the share of the divisible pool allocated according to derivation. Another important point relates to accountability. If derivation-based tax sharing provides the bulk of subnational revenues, regional and local governments are not forced to make tough taxing decisions. Since people are likely to understand that both the taxes they pay and how much flows to subnational governments is determined at the center, they have less incentive to hold local elected officials accountable for the quality of local services. When most taxes are hidden in sales prices anyway, even the idea of a tax price for public services may be forgotten. However, when derivation-based transfers are unconditional (Martinez-Vazquez and Timofeev, 2010) people know local officials have considerable freedom when it comes to spending the money. Even though most are likely to be less concerned about how ‘other people’s money’ is spent than their own, they may thus hold local officials accountable to some extent for the ­quality of local services. When there is some link between what local governments do and the amount of taxes collected in their jurisdiction that they receive, they

M4439-BAHL_9781786435293_t.indd 299

01/03/2018 11:08

300

Fiscal decentralization and local finance in developing countries

presumably have some incentive to try harder. In some countries, local governments have some leverage over local tax effort. In Russia, for example, tax administration is centrally controlled; but in the past, regional and local governments were responsible for some of the compensation of the tax administration – for example, housing and fringe benefits (MartinezVazquez et al., 2008). Local governments may also sometimes have close ties with local enterprises and some influence on their tax compliance. In China, provincial governments had responsibility for assessment and collection of business taxes and retained all revenues collected (Bahl et al., 2014). Moreover, the closeness of Chinese subnational governments to local enterprises gives them some influence over tax compliance, and appointed local officials are directly rewarded for their success in economic development and revenue mobilization. As Qian and Weingast (1997) argued, they had strong incentives to encourage the development of a broader tax base. On the other hand, local discretion may sometimes lead to granting special tax favors to enterprises, and hence reduced collections (Bahl et al., 2014).16 The sword may cut both ways with respect to certainty also. On one hand, derivation-based sharing makes local budgeting and fiscal planning simpler and more certain because the entitlements are transparent. On the other hand, it can be hard to predict tax collections and they may sometimes be volatile, as many Eastern European local governments found with respect to shared income taxes in the late 2000s (Bahl, 2011). Moreover, derivation-based sharing leaves some subnational governments more susceptible than others to changes in central government tax policy. When China shifted from a production- to a consumption-based VAT the average provincial revenue loss (with 100 percent collection efficiency) was estimated to be about 30 percent, but with big differences in the revenue share accruing to producer vs. consumer provinces (Ahmad et al., 2004). Central preferences given to industries concentrated in certain regions will also have differential regional effects. An advantage of the derivation approach is that its administration is relatively simple. This was particularly true under the old socialist system, where the taxes paid by enterprises were simply remitted to the bank accounts of the sharing governments as specified by the financial plan. Matters are more difficult when taxes are collected from people as well as enterprises by a central administration and then remitted to the accounts of the designated recipients. The major problem in many cases is how to divide the revenues when companies operate in different regions but pay taxes only where their headquarters are located. China has a proration formula which (with some ad hoc twists) is used to divide VAT revenues by factors intended to reflect where the transac-

M4439-BAHL_9781786435293_t.indd 300

01/03/2018 11:08



Intergovernmental transfers ­301

tions occurred. Similar formula allocations are used in other countries where VAT is regionally shared, including those in which subnational governments have some control over the VAT rate (Bird, 2015a), and are also used to allocate the income tax when there are separate subnational taxes.17 Central governments under budgetary pressure have sometimes reduced the sharing pool by declaring some taxes to be excluded or by earmarking some taxes for financing designated central government expenditures. When local governments act as collectors of central taxes, then they too may play such games – for example, understating collections and channeling them to extra-budgetary local accounts. Some years ago, a study reported that “there is evidence in Russia that local collectors would fill the coffers of the local governments first and then remit to the center” (Martinez-Vazquez et al., 2008, p. 206). The extensive use of extrabudgetary revenues at the local level in China has also frequently been noted (Bahl, 1999; Wong and Bird, 2008). Formula grants Another popular approach to the allocation of intergovernmental transfers among subnational governments is to use a formula (Types B and E in Table 7.2). The Type B version is particularly in step with the objectives of fiscal decentralization because it can give the subnational government both an entitlement to a share of central government taxes and a guaranteed share of the revenue sharing pool. The formula approach offers an alternative to the inherently regressive derivation approach. The differences in the distribution of transfers received depends on the formula used, but can be dramatic. For example, in China the simple correlation among provinces between per capita derivation-based shared taxes and per capita GDP was +0.91 in 2009; but for per capita equalization grants distributed on a formula basis, it was −0.41 (Bahl et al., 2014). Formula grants are often used to distribute grants among local governments. However, it is not always clear whether the primary goal is to reduce fiscal disparities or to backfill the revenue gaps that occur because higherlevel governments will not authorize local government taxing powers. The distinguishing feature of a formula grant is that usually relatively objective quantitative criteria are used to allocate the pool of revenues among the eligible subnational government units. An advantage of this approach is transparency: everyone knows the exact criteria by which distributions among local governments are made. Transparency is facilitated when the formula is relatively simple and everyone (legislators, bureaucrats and at least some voters) understands both what it means and how changes in its different components will

M4439-BAHL_9781786435293_t.indd 301

01/03/2018 11:08

0EG Tc 0EG 0C =a ba b 0FD y 0C 0FD

GAP 5 a ( Ei 2 Ri ) affect the final allocation. Once ithe distributable pool (G) is established, allocating it among the recipient governments is in principle a simple matter of arithmetic. Assume there are two formula components, X1 and X2 (for example, per capita income and population), and that they receive weights of g1 and g2, respectively, in the formula, where g1 + g2 5 1. This a (GAP) formula will then allocate VS to 5 jurisdiction i an amount (Ri) determined as CR follows: 302

Fiscal decentralization and local finance in developing countries & &



Ri 5 g1

x1i

ax

G 1 g2

x2i

G a x (7.3)

& If one knows the right numbers for X1 and X2 for region i and the weights Ri attached to each in the formula – say, for example, that the two factors are weighted equally – then one knows exactly the amount the region should receive. The grant process is completely transparent.18 Transparency does not mean that intergovernmental transfers are completely objective. The donor government can alter how much any region receives by adjusting its choice of factors or the weights attached to them in the formula. Presumably, these elements are chosen because of the objectives of the grant program. The usual objective is to close the financing gap between the resources available to different subnational governments and the costs of providing some basic level of services in a way that takes account of the differences between regions in both needs and resources. Sometimes the formula is weighted more toward expenditure needs, sometimes it is focused more heavily on fiscal capacity, and sometimes on the gap between expenditure needs and taxable capacity. Sometimes the objective is sought through a single formula, and sometimes different formulas are used for different grants, perhaps reflecting the different emphasis the central government puts on different local activities and on the (often conflicting) needs to respond to the political imperatives that inevitably shape intergovernmental negotiations in all countries. Ideally, a formal national deliberation – for example, in the context of a legislative debate, an official study or an intergovernmental or expert body specifically established for the purpose (such as a grants commission) – would decide on such matters. All too often, however, discussion does not advance much beyond a few ­politicians asking: what options are politically acceptable? and a few officials then asking: what data do we have?19 Grant design is an art, a science and a political exercise. Most designers seem to focus on indicators that reflect expenditure needs; others give heavier weight to fiscal capacity (presumably countries that have already devolved some taxing power); and some look for more balance between expenditure needs and fiscal capacity indicators.

M4439-BAHL_9781786435293_t.indd 302

01/03/2018 11:08



Intergovernmental transfers ­303

The expenditure needs focus is most common. An important task for those who design (or reform) grants is to find good proxy measures for the cost of providing a basic level of services, for example, land areas to be serviced, population or the proportion of people associated with higher service costs.20 Sometimes data are not available at all, and so only a rough categorization is used; for example, only a crude label may be used such as the ‘category’ of a municipality (for instance, a capital city vs. a rural municipality) or a general indicator of the level of development (for example, a ‘backward’ region). Many such indicators may be found in various combinations in such formulae around the world, as Box 7.3 illustrates. Those who design formulae naturally focus on conditions in their own country. For example, Indonesia determines needs by a composite index based on relative population, relative area, relative construction price index, the inverse of the Human Development Index (HDI) and the inverse of relative nominal per capita GDP;21 it then multiplies the result by per capita aggregate spending for the past year to estimate the value of the expenditure need component. The expenditure needs approach is related to the input approach that was widespread in the former Soviet bloc, where ‘expenditure norms’ were used to assess minimum expenditure needs in terms of the costs of physical levels of services, for example, hospital beds required, necessary square footage of classrooms and so on This approach is intuitively appealing but too complicated to make operational. To convert physical norms (somehow BOX 7.3  SOME INDICATORS OF EXPENDITURE NEED ● ●

● ●

●

●

Population, i.e., a straight per capita distribution (e.g. Pakistan, Bolivia). Often, a basic ‘floor’ per capita amount of grant is provided for all subnational units. Physical factors that may lead to greater costs of service provision such as land area, population density, urbanization, remote locations (e.g. Ethiopia, Indonesia). Equal absolute amounts to every local government to reflect the fixed cost element of local governance (e.g., Philippines). Measures to reflect the concentration of people costly to serve, for example, the proportion of families living below the poverty line, numbers of people on pensions or school-aged children, the Human Development Index (HDI) or other indices, as available. Indicators of infrastructure needs, such as miles of paved highways, percent of households with access to adequate water supply, infrastructure needed to support economic development, etc. (e.g. Colombia, South Africa). Category of municipality, e.g. capital city status to reflect the special public servicing costs in provincial capitals, or categorization by size of population and/or budget (e.g. Brazil, Colombia).

M4439-BAHL_9781786435293_t.indd 303

01/03/2018 11:08

304

Fiscal decentralization and local finance in developing countries

set) to relevant current monetary equivalents for a wide variety of expenditure functions (and to keep them up to date) would have been a complex and costly task.22 As pointed out by Martinez-Vazquez and Thirsk (2011) for Ukraine, budget allocations were in practice negotiated by the regional government authorities and the Ministry of Finance in Kiev. In the past, both China and Russia used simple inflation adjustments to update baseyear estimates of expenditure requirements. Adjusted or not, if a ‘norm’ is based on existing infrastructure (e.g. costs per hospital bed) the result may be to give more to those that already have more.23 Another problem with this approach is that experts or politicians seeking votes may for their own reasons set the norms at unaffordable levels. Although this can readily be offset by altering the level of the grant to keep it affordable, failing to finance the announced ‘minimum’ is unlikely to be greeted with applause. Those who make the promises have already often received their rewards, while those who fail to deliver the undeliverable get the blame. Many countries take expenditure needs into account mainly in combination with some measure of fiscal capacity.24 Such grants provide larger transfers to jurisdictions with lower capacity to raise taxes. Some may base this approach partly on the assumption that weaker fiscal capacity implies greater needs, although this is not necessarily the case. Most, however, attempt to balance expenditure needs and fiscal capacity. For example, Japan, Korea and Australia define needs by formulas including physical indicators of desired levels of service and then calculate the amount of the grant by comparing the result with a ‘normal’ level of revenue mobilization based on the size of the tax base (Kim, 2008; Ikawa, 2008). The most obvious proxy for fiscal capacity is a broad measure of the potential tax base such as regional or local GDP, where such figures are available. However, many developing countries do not have good measures of either subnational output or income, especially at the local (as distinguished from the regional) level. A well-known alternative approach is to use some measure of the potential tax yield if all subnational government tax bases were subjected to ‘normal’ rates. This so-called representative tax system (RTS) approach is used to assess tax capacity and, often, tax effort in countries such as Australia and Canada but has seldom been used in developing countries. An important advantage of this approach is that it provides an incentive for local governments to impose taxes at least at average tax rates. If they impose a lower than average rate, the grant they receive does not increase, so the amount available for them to spend is less than it would otherwise be. On the other hand, if they impose an aboveaverage tax rate, the grant they receive is not reduced, so they have more to spend. To put this another way, a ‘tax’ of 100 percent is imposed on the amount any government does not collect when their tax rates are lower

M4439-BAHL_9781786435293_t.indd 304

01/03/2018 11:08



Intergovernmental transfers ­305

than average, and a ‘tax’ of zero on any amount they collect by imposing above-average rates. When a grant is structured in this way, there is no need to include any special incentive to increase tax effort; the necessary incentive is inherent in the design of the formula.25 Few developing countries have adequate data to estimate the potential local tax bases essential to this approach, however, so most have taken a simpler approach. For example, India’s Finance Commission has used the ratio of tax revenue to state GDP as a measure of tax effort.26 Indonesia includes 93 percent of own-source revenues, 100 percent of non-resource tax revenue sharing, and 63 percent of resource-revenue sharing in its estimate of fiscal capacity. Some countries establish a ‘hold-harmless’ arrangement to provide a buffer against large public service disruptions that might result from formula changes. When Indonesia adopted a new transfer formula in 2001, it protected local governments from annual revenue losses of more than 10 percent during the transition period. Mexico also protected state governments from revenue loss below the 2007 level when it revised its grant formula in 2007. The formula approach is appealing mainly because of its transparency and objectivity. However, for recipient governments to have confidence in the grant system they must have confidence in the data on which it is based. A common problem, for example, is when a new census shows that some areas – usually large cities – have grown rapidly. There is often strong resistance from less-urbanized (and often poorer) areas to using the new population data in the grant formula. Since political systems seldom adjust political representation quickly, the voices of such protesters are often heard more clearly in the seats of power than the needs of the many. In the early 2000s, for example, the Indian Finance Commission still used population data for 1971 to preserve a larger grant share for slower-growing states (Rao, 2009). Grant designers are often caught between wanting to build a formula that captures the objectives of the program and using available data that are generally thought to be accurate. Frequently, there is little overlap between these two requirements. Well-intentioned social engineers often want to insert additional variables into the formula to achieve some particular effect. The effects are often negligible because the weight of the variable is so small, or it cannot be measured in a way needed to achieve the desired result, or its measurement is so tenuous that it introduces additional volatility as well as complications that reduce transparency even further. Well-done simulations during the design phase may sometimes help avoid this mistake.27 A good formula must be both acceptable politically and technically feasible. Ideally, it should also be as transparent and understandable as

M4439-BAHL_9781786435293_t.indd 305

01/03/2018 11:08

306

Fiscal decentralization and local finance in developing countries

possible in order to keep the lines of accountability – what government is responsible for what decisions – as clear as possible. One way to achieve these aims is to avoid having many different grant programs with different formulae and to keep the formulae as simple as possible. Many countries have listened to such advice. Indonesia’s general revenue grant (DAU) uses a fiscal gap approach with a formula that has four factors. India distributes most intergovernmental transfers through Finance Commission grants with five factors. On the other hand, Brazil’s transfers have different formulae and extensive earmarking; Mexico’s unconditional transfers are distributed under eight different heads, each with a different formula; and Ethiopia’s includes 14 indicators of expenditure needs and fiscal capacity. Even with more complex systems, however, it can be relatively simple to run a formula grant system: all that is required is to determine the annual vertical share, update the database, recalculate the entitlements and distribute them appropriately.28 Cost reimbursement (matching) grants Another way to allocate intergovernmental transfers is through cost reimbursement grants (Types C and F in Table 7.2). Such transfers usually have one or more of three features. The first feature is that the higher-level government specifies the functions (or objects) on which the money must be spent – that is, the funds are earmarked for a specific outlay. Although the local tax price to deliver the service is lowered by the grant, there is nonetheless a social cost in the sense that local spending decisions directed from above are likely to differ from how local people would prefer to spend the money. For example, a grant to reimburse all or a given share of teachers’ salaries will reduce the relative local costs of paying teachers. Although receipt of the grant may free up some local funds for other purposes, such transfers make it relatively cheaper to hire teachers than, say, to improve water quality – even if the latter is what locals may consider more valuable.29 Conditional grants, whether matching or not, are ways in which central governments can and do affect local spending decisions. The second relatively common feature is that the grant covers a certain percentage of the cost of the service or project, with the local government being forced to ‘match’ this share by covering the rest of the cost from its own sources. Sometimes, especially with infrastructure finance, the match may be less formal, e.g. a requirement that the local government will cover operation and maintenance costs. A third feature of such grants is that all sorts of other conditions – standards of performance, construction standards, employee qualifications, specification of which expenditures on a particular function are eligible, etc. – are often applied, thus constraining even further the degree of free-

M4439-BAHL_9781786435293_t.indd 306

01/03/2018 11:08



Intergovernmental transfers ­307

dom within which recipients may make decisions. The constraints imposed by such conditions, which are often difficult to satisfy and cumbersome to enforce, may sometimes be inappropriate given local conditions and needs.30 They usually reduce the net local benefits from receipt of a grant. Of course, since money is fungible the effectiveness of all such conditionality depends on the extent to which recipient governments are able to shift funds by using grants to replace locally funded spending on a function and then reduce local taxes or, more likely, to fund some other activity. Practice with respect to cost reimbursement grants is varied, both in terms of the objectives sought and the structure of the grants. Sometimes these are important continuing programs; sometimes they are one-off capital grants; often they are something in between. Such grants may be meant to lower the tax price of a particular type of spending. In Mexico an earmarked grant is given to cover the wage bill for teachers and medical professionals who were transferred to the states in the 1990s. Another approach is to compensate for an activity that the center fears the subnational government cannot afford, or for which the central government feels some degree of responsibility. China gives conditional grants to support the compulsory rural education program and to compensate rural local governments for the elimination of the agricultural tax. The cost reimbursement approach may also be used to support programs of the line ministries. India’s rural development schemes were allocated among states and required a matching contribution. Ad hoc transfers Finally, some intergovernmental transfers are distributed on an ad hoc basis every year; i.e., the grant amount is distributed on a discretionary basis by the legislature or the central government (Types D and G in Table  7.2). Such transfers are often all about politics: who’s in and who’s out with respect to the central government. A more economically rational version of ad hoc distribution would have subnational governments submit requests for funding from a special pool of funds, with the higher-level government choosing those projects to be funded. This approach is sometimes used with respect to funding infrastructure projects. If the winners are decided on strict cost–benefit terms, the outcome can be a more productive use of funds. However, this process may exacerbate regional disparities since richer areas are more likely to have the skills to develop and present good project proposals. Alternatively, projects may be allocated to attract votes in the next election or to penalize regions controlled by the opposition. Bad or good, some such discretionary grant programs exist in every country: whether in the form of emergency year-end bailouts for regions in trouble or special support to border regions (as in Mexico, for example); or

M4439-BAHL_9781786435293_t.indd 307

01/03/2018 11:08

308

Fiscal decentralization and local finance in developing countries

for disaster relief from flooding or an earthquake; or to offset the costs of, for example, environmental degradation in mining states.31 Some countries, like India, give legislators an allocation to be spent at their own discretion. Colombia also did this prior to its 1991 constitutional reform. While such programs can be criticized for such obvious defects as their lack of transparency and susceptibility to corruption, they may also sometimes be a necessary price to pay to gain and retain sufficient legislative support for a more structured and formal program of fiscal decentralization. Indeed, the price paid may be cheaper than the cost associated with the distorted transfer design that might otherwise have been required to meet political demands. However, by most standards of good intergovernmental finance policy, ad hoc grants fail. They are seldom transparent, may fluctuate significantly from year to year, and are unlikely to have clearly stated objectives such as revenue mobilization or equalization. Such transfers nonetheless exist almost everywhere because they have substantial advantages from the perspective of the higher-level government. The amount and allocation of these grants can be controlled in a flexible way reflecting budgetary circumstances, political necessity and perhaps changing policy priorities, and they are controllable and flexible enough to reflect the changing priorities of the center. They may also be used to permit a country to move through a transition period from one grant system to another without disrupting service delivery. Finally, and perhaps most importantly, ad hoc grants always seem to have powerful champions.

REVENUE SHARING FROM NATURAL RESOURCES In many countries, revenues from the natural resource sector are subject to a separate regime for revenue sharing. One reason is because of the special features of this sector: returns are driven by international markets, the assets are exhaustible and often the amounts involved are very large. This component of the intergovernmental transfer regime is often contentious. Once a region realizes that it has been blessed with a natural wealth, it wants to receive the benefits from its endowment – and is upset if it does not. If it does receive the benefits, other regions usually resent it. Either way, the issue of who gets how much is divisive, and has proved especially contentious in countries where natural resource revenue sharing and ethnic tensions overlap. Rebellion, terrorism, civil war and separation have been the result in some cases; so it is not surprising that many have explored the extent to which decentralization and regional autonomy may dampen – or possibly inflame – the otherwise divisive effects of sudden resource wealth.

M4439-BAHL_9781786435293_t.indd 308

01/03/2018 11:08



Intergovernmental transfers ­309

We do not enter this difficult territory here, but it is not surprising that – to repeat a note we have sounded often in this book – every case is different. The line that must be tread in distributing resource revenues is a fine one. To illustrate: one recent study concludes that “countries plagued with terrorism (but not larger scale insurgencies) need to carefully balance the degree of centralization and regional autonomy they give to the regions containing these resources. Countries affected by larger scale insurgencies should give regional groups a share in political power short of regional autonomy” (Dreher and Kreibaum, 2015, p. 16). Redefining the problem so that the right answer for any country turns on the point at which ‘terrorism’ becomes ‘larger-scale insurgency’ does not make the life of those charged with allocating resource revenues any easier. Similarly, a recent empirical study by Bhattacharyya et al. (2016) concluded that increased resource rents had little effect on subnational revenues and reduced the subnational share of total government expenditure for the sample studied (over 90 countries). However, the effects were strikingly different in ‘permanently democratized’ countries32 where both the subnational tax share and the share of transfers in subnational revenue were significantly increased. Although this study does not explicitly examine the effect of fiscal decentralization on the whole, its findings suggest that democracy, decentralization and more decentralized distribution of resource revenues tend to go together. This does not imply, however, that decentralizing resource revenues is a good idea. As the abstract of a recent case study of two Nigerian states says with commendable understatement, when there “are no effective mechanisms for ensuring subnational fiscal discipline and political accountability . . . political decentralization . . . may not necessarily result in improved natural resource governance” (Ushie, 2012). Summing up, in the frustrating phrase that policy-makers so often hear from sensible policy advisers, ‘it all depends’ – in this case on factors that are often difficult to unearth, let alone to understand how they interact in the specific situation at hand. Given such problems, it is unsurprising that when it comes to the interaction of rents and transfer design the result is usually an uneasy compromise that no party really thinks to be all that fair. All we can do here is discuss briefly three aspects of the question of sharing natural resource revenues. The first is the basic argument for and against giving subnational governments access to a share of the revenues coming from the extraction of natural resources; the second is how such revenues might be distributed; and the third is to briefly describe the wide variations found among low-income countries. Should natural resource revenues be shared in a separate transfer regime?33 To those who live in a region rich with natural resources, a share

M4439-BAHL_9781786435293_t.indd 309

01/03/2018 11:08

310

Fiscal decentralization and local finance in developing countries

of the return might be seen as an entitlement. McLure (1994, p. 199) puts it nicely: “Subnational governments have argued strongly that they may have the right to tax natural resources located within their boundaries, to convert resource wealth (their ‘heritage’) into financial capital . . . to turn oil in the ground to money in the bank.” There is also a cost reimbursement argument. Natural resource extraction is a dirty business that can pollute the environment and wear out the infrastructure, and it might attract a migrant population that disrupts the local social order. Finally, it might be better to provide a formal share of natural resource revenues to the subnational sector than to risk their imposing costly informal taxes and charges on mining companies or to institute local blockades and other disruptive tactics such as destroying pipelines, as has happened in some countries. But a good case can also be made against sharing revenue from resources. First, natural resource revenues are inherently unstable because they are tied to commodity prices that are determined in a world market. Subnational governments – usually bound to hard budget constraints and responsible for providing essential functions such as education and health – find it difficult to respond to these fluctuations, so heavy reliance on resource revenues is risky. Second, because natural resources are unequally distributed, channeling resource revenues to the location where the resources are extracted may lead to large interjurisdictional disparities and increased discontent in other regions, an outcome observed in countries from Papua New Guinea to Peru. Third, when a lot of money flows quickly into an area, it is often not spent sensibly, and often leads to such undesirable outcomes as rapid expansions in the number of ‘political’ public employees, ill-conceived public projects like ‘bridges to nowhere’ and opulent presidential residences, and often massive corruption and perhaps even armed conflicts between competing ‘robber barons.’ A significant amount of money received quickly and easily may not be spent any more wisely by resource-rich regions than by lottery winners: it might be used for investments in the non-tradeable sector rather than on developing a new export sector; it might be squandered on ill-conceived projects; it might simply be stolen; or it might be used to finance civil wars.34 Many countries have introduced some special regime for natural resource revenue sharing. But different countries have structured the arrangements differently with respect to the determination of the size of the vertical share for subnational governments, which governments will share in the distribution, and how this special regime is linked to any general revenuesharing program. A few examples illustrate the different approaches taken: ●●

In Indonesia, subnational governments receive 15 percent of oil revenue (3 percent for provinces, 6 percent for originating districts

M4439-BAHL_9781786435293_t.indd 310

01/03/2018 11:08



Intergovernmental transfers ­311

●●

●●

and 6 percent distributed equally for other districts within the same province). In the case of gas, the subnational share is 30 percent (6 percent for provinces, 12 percent for originating districts and 12 percent distributed equally to other districts within the same province). Subnational governments in the resource-rich provinces of Aceh, Papua and West Papua get an additional share of 55 percent for oil and 40 percent for gas. However, the general revenue-sharing program (DAU) is adjusted to offset some of this additional share by including the fiscal capacity of the natural resource sector in the calculation of their entitlement (Shah, 2012; Lamont et al., 2012). The vertical shares of subnational governments in Peru vary according to whether the revenue comes from the mining tax or from royalties. These revenues are distributed among regions on a derivation basis but among local governments by a prescribed formula, with the (highly concentrated) revenues earmarked primarily for infrastructure. Ten of the 24 departments in Peru receive over 90 percent of the transferred funds (Canavire-Bacarreza et al., 2011). In per capita terms, the maximum amount received from resource revenue sharing was 25 times greater than the maximum received from the general transfer to local governments. Unsurprisingly, some local governments were unable to efficiently absorb the huge inflow of revenues received during the boom. A different approach is followed in Bolivia, where revenues from gas and oil are the principal sources for the general revenue-sharing program for municipalities and prefectures. The direct tax on hydrocarbons is earmarked for municipalities and is distributed among them primarily on basis of population. Two other components of hydrocarbon revenues are distributed to regions (prefectures), one on a derivation basis to producing regions only and the other on a formula basis to all regions (Brosio and Jimenez, 2012).

Evaluating such approaches to natural resource revenue sharing is invariably a complex task, in part because the discussion usually gets bogged down in the specifics of how transfers should be evaluated: Are disparities too great? Should there be requirements about how the funds are spent? Do they encourage or discourage local fiscal autonomy? Complexity also arises because it is seldom possible to understand any of these issues unless one probes deeply into many other aspects of the relevant local situation.35 We do not go further into these important but invariably ‘place-specific’ issues here. Instead, we end this brief discussion by stressing again that the key question is whether natural resource wealth belongs to the entire nation or to the producing region.

M4439-BAHL_9781786435293_t.indd 311

01/03/2018 11:08

312

Fiscal decentralization and local finance in developing countries

If the resource is considered ‘national’ in character, then the revenues, like other national revenues, may be included in any general revenue-­sharing pool and allocated among subnational governments in some acceptable way. Ideally, of course, the central government should also take seriously its long-term responsibility to use the revenues to build up the economic base sufficiently to replace the exhaustible resource when the day inevitably comes. If instead the producing regions are the owners of the resource, although they should not then be expected to share their revenues, any general revenue-sharing system should of course take the resource revenue capacity of these regions into account. In this case, the producing regions become those responsible for spending the revenues sensibly in anticipation of the day when the money stops flowing. In most countries, however, what usually happens is that some middle ground is chosen, with two or more levels of government squabbling about who gets how much – and all too often no one taking seriously the long-term development implications of depending heavily on exhaustible resource revenues.

UNCONDITIONAL VS. CONDITIONAL TRANSFERS VS. VERTICAL PROGRAMS Most countries have both conditional and unconditional (untied) transfers, but as usual there is wide variation. In Mexico, for instance, the split is about even between conditional and unconditional grants. In Indonesia, most local government finances come from untied grants. In Brazil about 60 percent of transfers are unconditional grants, while in Colombia only about 10 percent are unconditional. Among other things, such differences reflect preferences for central control vs. local autonomy and views about the capacity of regional and local governments to put transfers to good use. To the extent conditional grants go to public functions, where there are significant externalities they may be efficiency enhancing, although, as noted earlier, this potential is more talked about than realized. A more important rationale in practice is that conditional grants are a way in which the central government can impose national uniformity and minimum standards in the delivery of some services, requiring, for example, road construction to certain standards or certain textbooks for schools. An alternative way to ensure minimum standards and national uniformity is to impose a mandate to the effect that subnational governments must provide services in a certain way at a certain level, without providing any commensurate funding. Such unfunded mandates would of course impose a considerably heavier, and generally undesirable, fiscal pressure on subnational governments.

M4439-BAHL_9781786435293_t.indd 312

01/03/2018 11:08



Intergovernmental transfers ­313

Yet another approach is simply to create a direct national expenditure program. Central governments can generally deliver whatever programs they want in all or parts of the country (unless barred constitutionally from doing so). If they want something done at the regional or local level, they do not need to establish a conditional grant to finance the regional or local government to do it; they can do it themselves – and of course fund it themselves. ‘Vertical programs’ – defined here as central (or perhaps regional) programs that are delivered in local areas by the central (regional) government – have no direct impact on either side of local budgets but are seldom well coordinated with whatever the subnational government may be doing with respect to the same activities. A common rationale for centrally financed and controlled programs is that the national government can do a better job than local governments. Another rationale, as with conditional transfers, may be that a national program is a particularly effective way to ensure that services characterized by significant externalities are adequately funded and delivered to meet minimum standards. Since both the fiscal base and probably delivery capacity tend to be weaker in smaller localities, the role played by vertical programs may be especially important in such areas (see Box 7.4). This approach has some obvious problems, however. While competition among service providers is sometimes as useful in the public as in the private sector, if the result is simply to duplicate local activities the main effect may be to increase costs. In Argentina, for instance, health services are delivered by five different government agencies. Another problem in some countries is that such programs are sometimes used by central line ministries as ways of keeping control over expenditure areas that have supposedly been devolved to local levels – the ‘empire’ fighting back, as it were. Yet another is that such central programs are typically not reported to the local government jurisdictions in which they occur, which makes it difficult to evaluate both the level of services provided and the degree of ‘expenditure need’ if these are matters of policy concern, as is often the case. In addition, of course, central programs are likely to be less responsive to local needs. When conditional grants are used, the conditions may sometimes be very loose, as when so-called ‘block’ grants simply require that the money must be spent on, say, primary education or health services. Such grants give subnational governments more discretion than if they were simply spending agents of the center, but allow the center to retain some control over how funds are used. The extent to which control can be effectively exercised depends on how closely the use of the funds can be monitored. A tighter degree of control is common with earmarked grants such as those sometimes created to help build the capacity of subnational g­ overnments

M4439-BAHL_9781786435293_t.indd 313

01/03/2018 11:08

314

Fiscal decentralization and local finance in developing countries

BOX 7.4  RURAL LOCAL GOVERNMENTS In some countries, urban and rural local governments are treated differently both in terms of expenditure assignments and financing options, and in terms of intergovernmental transfers. Almost everywhere, rural local governments are usually financed primarily by transfers because their revenue-raising capacity is small and their ability to spend the money efficiently is thought to be equally small. One obvious approach is to address the basic public service needs of rural people through direct national programs. However, it is generally neither possible nor desirable to leave rural local governments completely out of the picture, not only because rural people migrate – so their problems and concerns have a broader impact – but also because in many countries much of the population still lives in rural areas. In India, about 70 percent of the population lives in 250,000 mostly rural jurisdictions. Providing better public services – more of what people want – may slow the flow of migrants to cities. Perhaps more importantly, involving local residents more directly in governance is a significant nation-building activity. Moving local governments, some of which may become much larger over time, further up the learning curve is good when it comes to service delivery and revenue mobilization. It also, not unimportantly, ensures that rural residents contribute to at least some extent in financing as well as shaping the kind and level of services they receive. In practice, governments in developing countries take many different approaches to transferring funds to rural local governments. Some treat them just like urban local governments, as is common in Latin America as well as in Indonesia. Some federal countries, like Brazil and Nigeria, mandate that a certain proportion of federal-state transfers be passed on to local governments. Some countries have direct central transfers to municipal governments, as in the Philippines. In others, such as China and Argentina, it is left to the regional (provincial) government to decide on the distribution among local governments. In India, every state does it differently. The state of West Bengal, for example, has a separate sharing pool for urban and rural local governments, and distributes this amount among the respective local governments by different formulas. How states share revenues with their local governments is supposedly set by periodic State Finance Commissions which, like the Central Finance Commission, sit every 3–5 years (see Box 7.5). A recent study of West Bengal, where the population is 72 percent rural, found that only 17 percent of government expenditures are made by rural local governments, which derived 94 percent of their revenues from intergovernmental transfers (Bahl et al., 2010a). While every situation is different, some general ideas about structuring transfers to rural local governments are suggested by both experience and theory. A first point is that the differences between large urban and small rural local governments are often so great in terms both of their revenue base and their capacity to deliver services that it is usually a mistake to try to apply the same transfers at both ends of the subnational spectrum. A degree of asymmetry in transfer design is thus not a flaw but a necessity if the objective is to produce more uniform results. For example, national programs can be used to deliver certain services directly in rural areas, or small rural areas may be encouraged and supported in contracting certain activities to larger local governments or subnational governments higher up

M4439-BAHL_9781786435293_t.indd 314

01/03/2018 11:08



Intergovernmental transfers ­315

in the hierarchy or perhaps to other public agencies or private suppliers, as some smaller Colombian municipalities did when water services became a local function. The other side of the coin is that larger (regional and urban) governments should be given both more access to productive own-source revenues and greater expenditure responsibility, as we discuss further in Chapter 8. A second key point is to keep any allocation formula simple, both because spatially appropriate data are often unavailable and to make it as transparent to all as possible. For example, a simple formula based on population and land area may do the job. However, if it is worth having at all, even the smallest local government needs to have a budget sufficiently large to enable it to be responsible for some services that matter.

in areas new to them (Blöchliger and Vammalle, 2009) or intended to recognize special needs (for example, grants for rural education) or to compensate for the abolition of local taxes (e.g., the agricultural taxes in China). In addition, since there is almost always a fair degree of paternalism (‘father knows best’) in central–local relations, conditional grants may also be viewed as ways to save locals from such bad decisions (from a central perspective) as building a new sports field rather than the road the central authorities think they should really have. Central governments are often happier to bestow conditional rather than unconditional transfers on regional and local governments precisely because they restrict the recipients’ budgetary discretion. But the fact that conditional grants override the preferences of subnational governments – they get the road, not the sports field – is also their major flaw. There are all too many examples of central grants that, like foreign aid, sometimes saddle localities with maintaining public facilities that they did not want in the first place. This is perhaps one reason central governments commonly complain that local governments do not adequately maintain the infrastructure that has been so generously bestowed upon them: if they did not want it, why should they maintain it? It is harder to understand why central governments often have large unconditional grant programs. What do central politicians get out of giving money away without strings, and letting regional and local politicians claim credit for centrally financed expenditures? Here, the argument usually comes down to the extent to which the central government is serious about fiscal decentralization. When significant expenditure responsibility and (usually) less revenue-raising power is given to subnational governments, making up the resulting fiscal gap by providing transfers that leave major spending discretion to the recipient governments is perhaps logically consistent. But it is still not that easy to understand.36 If the system is well designed, the central government will, as we discuss below,

M4439-BAHL_9781786435293_t.indd 315

01/03/2018 11:08

316

Fiscal decentralization and local finance in developing countries

monitor outcomes and insist on good reporting and independent audits but otherwise leave spending decisions up to the locals. This ideal is seldom fully attained, however, and even unconditional grants are often ‘tied’ at least indirectly by expenditure mandates – for example, a requirement to pay centrally negotiated employee salaries (as with teachers in Colombia) or to spend at least some fixed proportion of its budget for education (as in Brazil). When it comes to conditionality, many stakeholders with different views need to be taken into account: politicians and officials at all levels, central vs. line agencies, employee groups, suppliers; and, not least those for whom it is presumably all about – citizens, both nationally and locally.37 It is not surprising that most countries end up with some mix of conditional and unconditional transfers. Ideally, if seldom in practice, the conditional grants should be limited to functions where external benefits can be identified and their size approximated. If an appropriate conditional grant cannot be developed and financed, the worst possible substitute – though one often found, perhaps because it appears to be costless to central governments – is an unfunded mandate or the equivalent of a budget allocation rule (x percent of the local budget must be spent on service A). Fiscal planners are caught on the horns of an efficiency dilemma when it comes to structuring the grant system (Bahl, 2010a). They can emphasize conditional grants and hope that they guess right on the externalities, or they can give unconditional grants to subnational governments and then deliver services with big external benefits through vertical programs. The latter approach may be indicated, for example, when external benefits are most significant – as, for instance, in the case of higher education.

POLITICS AT PLAY Transfer systems differ in different countries not only because of their different economic circumstances but also because of the important, usually critical, role politics plays in deciding policy. This variation may to some extent reflect changes in the (implicit) ‘fiscal contract’ between different groups and interests over time. Or it may occur simply because the problems facing different countries change and policy-makers search out solutions that may involve intergovernmental transfers. The structure of the transfer system may also reflect the natural tendency of decisionmakers at both the political and administrative levels in both central and local governments to push their agendas to the boundaries allowed by their political competitors and by their constituents.38 The self-interest of leaders may be to enhance their prospect of re-election or promotion,

M4439-BAHL_9781786435293_t.indd 316

01/03/2018 11:08



Intergovernmental transfers ­317

or simply to direct more money to their own pockets (or those of their families and supporters). Or it may be to achieve some greater and perhaps more socially valued objective such as ‘nation-building’ or stronger local autonomy (‘home rule’).39 Central governments often want to exert significant control over the level of spending, revenue-raising and borrowing by subnational governments. They always like some flexibility in determining the share of central revenues that flows to subnational governments. As a rule, they think that what they (the center) need is more important than what regional or local governments think they may need. After all, it is the central government that is responsible for achieving such broad goals as putting basic national infrastructure in place and maintaining economic stability. The central Ministry of Finance (MOF) is usually unlikely to be fond of tax-sharing regimes that entitle subnational governments to a large proportion of central revenues. Since MOFs (and most central line ministries for that matter) also think that most subnational governments are not able to deliver the public services for which they are responsible in a fiscally disciplined way, they are particularly unwilling to send them money when the central fiscal position itself is weak, as it often is in developing countries. To maintain more central ownership of its revenues and budgetary flexibility, as well as to protect against service delivery failure, most MOFs probably prefer the discretionary budget approach to determining how much of the pie (the vertical share) should go to subnational governments. Given the usual view that finance is first among ministries, it is a bit surprising that tax sharing is so widespread among developing countries. In some countries, even such politically centralized countries as China (Bahl and Martinez-Vazquez, 2006), the explanation seems to be that regional demands for more funds are politically hard to resist, and tax sharing at least avoids the need to devolve significant taxing powers to subnational governments. Sometimes this trade-off is explicit. In Mexico, for instance, state governments surrendered their taxing powers in 1980 in return for a guaranteed participation in the revenues from the federal tax system. An additional attraction for the central government, when tax sharing is not enshrined in the constitution, is that it can always change the rates, thus retaining a substantial degree of flexibility in budgetary policy. Finance ministries are more concerned to ensure that recipient governments face a hard budget constraint (see Chapter 5) than that they spend the money in any particular way. But other central ministries and many legislators may have different concerns. Transport ministries prefer conditional grants earmarked for roads, education ministries prefer grants earmarked for education and so on, both because such grants usually mean that more is spent on things they are concerned with and because this

M4439-BAHL_9781786435293_t.indd 317

01/03/2018 11:08

318

Fiscal decentralization and local finance in developing countries

enables them to maintain some degree of control over the way in which these services are delivered. Ministries of local government, like politicians in general, may instead favor ad hoc arrangements that give them more discretionary power to respond to and influence their regional and local constituents. Central politicians with regional power bases want to favor their regions, particularly when they can do so without giving credit to potential (local) rivals who may control regional or local governments. Looking at the design of a transfer system from the opposite ­direction – bottom up – almost everything is reversed. Lower-level governments typically favor transfer systems that have three characteristics: (a) a guaranteed, adequate revenue flow; (b) expenditure discretion; and (c) enough budgetary certainty to plan efficiently. Typically, they can agree on a tax-sharing approach to vertical sharing if it guarantees enough revenue flow through the intergovernmental transfer system, and probably also on a formula-based system for horizontal allocations. Many different interests and views may of course be found among subnational governments: urban vs. rural, rich vs. poor, large vs. small or remotely located vs. centrally located. Indeed, one reason subnational governments tend to have relatively little influence on the design of the intergovernmental transfer system is that they do not agree among themselves about what is best. Divided, they may not be conquered, but they are unlikely to dominate negotiations. For example, larger and richer places may prefer derivation-based tax sharing (Type A in Table 7.2), which bestows money to spend without requiring local politicians to get the support of local constituents. Less-wealthy subnational governments, while also likely to favor vertical shares determined by tax sharing, are more likely to prefer horizontal sharing formulae that recognize their expenditure needs (Type B or even Type C grants).

WHAT DO TRANSFERS REALLY ACCOMPLISH? To determine whether fiscal decentralization is a success or not, one needs to know whether intergovernmental transfers, a key component of fiscal decentralization, have achieved the goals that were set for them. Transfers fill much of the financing gap during the inevitable (and perhaps lengthy) period before local government revenue generation becomes significant. What we do not know is whether they have encouraged or discouraged subnational government revenue mobilization, and whether they have softened the unwanted fiscal disparities among subnational governments. Here, we consider the evidence on these two questions.

M4439-BAHL_9781786435293_t.indd 318

01/03/2018 11:08



Intergovernmental transfers ­319

Transfers and Revenue Mobilization Central governments frequently complain about the weak revenue mobilization efforts of subnational governments. Such complaints are sometimes self-serving justifications for the provision of poor public services, and are intended primarily to deflate calls for more local taxing power or for higher levels of intergovernmental transfers. Sometimes, however, there really is something to complain about. The problem emerges when subnational governments do not have sufficient resources to do what they are supposed to do.40 One reason may be that they do not have sufficient taxing power. Another may be that intergovernmental transfers are too small or too constraining. But central governments are right to suspect that a third reason may be that subnational governments do not adequately exploit the taxing power they have. If this last is the problem, another obvious question is whether and how their fiscal performance may be improved by changes in the design of transfers. Any intergovernmental transfer provides a local government with additional income. The local government may react by: (a) spending the full amount of the grant without reducing tax effort; (b) spending the full amount of the grant but reducing tax effort (sometimes called ‘fiscal ­laziness’); or (c) spending more than the full amount of the grant by increasing its own tax effort (the so called ‘high-powered money’ effect). What the recipient government does depends both on the structure of the grant itself (e.g., is it matching?) and on the relevant income and price elasticities of demand. The experience in developing countries may be used to support the hypothesis that any of these three outcomes may dominate.41 Large automatic transfers such as those from a large vertical share of central government taxes may so dwarf the revenue-raising potential of subnational governments that they are discouraged from making more effort to raise their own taxes. Why incur the anger of local taxpayers over an increase in drivers’ license fees when a grant of 100 times that amount is coming? In Pakistan, for example, where the provincial government entitlement is 57.5 percent of total central government taxes, provincial government tax effort has for decades been less than 1 percent of GDP (Bahl et al., 2015a; Ghaus-Pasha and Pasha, 2015). Manasan (2009) argues that the large tax transfers to municipalities in the Philippines have similarly discouraged local fiscal effort. Brodjonegoro and MartinezVazquez (2002) and Shah (2012) argue that there is a built-in disincentive in the Indonesian transfer system. Because increased tax revenue in local areas results in reduced grant revenues, local governments have an incentive to tax less. Alm and Boex (2008) find the same result for Nigeria; Mathur (2012) reports similar results for the Philippines, Vietnam, Nepal,

M4439-BAHL_9781786435293_t.indd 319

01/03/2018 11:08

320

Fiscal decentralization and local finance in developing countries

Korea and Hungary, as do Alexeev and Kurlyandskaya (2003) for Russia. Lower state government tax effort is also correlated with higher levels of unconditional grants in Brazil (World Bank, 2008). Other studies find that derivation-based natural resource revenue sharing has discouraged subnational government tax effort (Martinez-Vazquez, 2015; Brosio, 2015). However, some studies have found that transfers are ‘high powered’ and resulted in increased spending from local sources. For example, newly constructed roads financed through conditional grants are usually maintained with local resources and may induce the construction of more feeder roads, the provision of improved traffic controls and the stimulation of demand for better local services in the affected area. Unconditional grants may also be stimulative. Artana et al. (2015) found evidence that general transfers had a stimulative effect on property tax efforts in Argentina, as did Sánchez Torres et al. (2015) for Colombia. In Bangladesh, increasing grant levels were also found to be positively associated with own-source revenues (Lewis and Searle, 2011). Derivation-based distributions of transfers can be stimulative because of the built-in revenue mobilization effect (“the more you raise, the more you get”). In China, where the growth in derivation transfers was very high, although there was no formal subnational taxing power local governments could influence the aggressiveness of the tax administration and appointed local officials were directly rewarded for their success with revenue mobilization and economic development. Both local governments and the central government thus benefited as the tax-GDP ratio grew from only 10 percent in 1996 to 20 percent in 2013 (Bahl et al., 2014). Conclusions about the stimulative or substitutive effects of intergovernmental transfers on local revenue mobilization are thus mixed.42 Arguably, one problem with such studies is that they tend to be focused on system effects: that is, measuring the response of all transfers, some of which may be may be stimulative and others substitutive, depending on the specific design of the grants and the price and income elasticities of demand for the object of the grant. A more disaggregated analysis, especially for conditional grants, may give some insight into the determinants of revenue-mobilization effects. As Rao and Chelliah (1991) put it, designing grants to fill ­expenditure– revenue gaps is just ‘fiscal dentistry.’ For example, a well-designed unconditional grant will not provide an incentive to reduce tax effort as a simple gap-filling grant does. Many grant formulas do include measures of fiscal capacity and often tax effort to avoid this problem. Often the ratio of tax revenues to GDP (if subnational GDP data are available) is employed, though such measures are always questionable for a variety of reasons (Bird, 1976a). Perhaps the best approach is to estimate the potential tax

M4439-BAHL_9781786435293_t.indd 320

01/03/2018 11:08



Intergovernmental transfers ­321

base and to use this measure of taxable capacity directly in the grant formula (Bahl, 1972). Measuring tax effort by the revenues raised is not a good idea because it penalizes subnational governments that raise higher levels of taxes, and thus provides an incentive to dampen revenue mobilization. Apart from the basic data problem, another difficulty with such measures is that it may prove very difficult (perhaps impossible) to explain to either politicians or voters what is being done in a sufficiently persuasive way to make the resulting allocation of funds among localities acceptable to all. Many countries have attempted to go further and to introduce various elements into grant formulas intended to provide a direct incentive to extra tax effort. Few (if any) such attempts have been very successful, however, in part perhaps because such incentives are often so small that they are unlikely to stimulate much (if any) response from recipient governments. In the Indian Finance Commission formula, for example, at one point only a 7.5 percent weight was placed on a tax effort component, compared to the 50 percent weight placed on per capita income. The result of such incentives is usually more to complicate the whole exercise than to achieve anything useful. It is also usually a bad idea to reward those who increase their taxes most, as some countries have done, because this usually just gives more to those to whom the changing economy has already, by expanding the tax base, given most. Mexico, for example, gives a 30 percent weight to the amount by which state and local government taxes increase relative to the national average, and a 10 percent weight to the level of state and local government taxes relative to the nationwide level (weighted by population). India’s 12th Finance Commission (2005–2010) gave a 7.5 percent weight to tax effort and a 7.5 percent weight to revenue improvement in the formula for distributing transfers among the states. In contrast, the 14th Finance Commission (2015–2020) concluded (correctly, we think) that the best approach was to give no such specific incentive for revenue mobilization. Recently, many have advocated more reliance on ‘results-based’ and ‘performance grants’ – grants that require specified tests of performance to be met as a requirement for receiving the full amount – as potentially effective ways to improve public service delivery systems and public service outcomes.43 Most performance grants are relatively small capital grants, commonly requiring such conditions as an approved development plan, core staff positions and structure in place, final accounts produced on time, cash books and bank reconciliations up to date, no unsettled audit queries, and procurement entity in place (UNCDF, 2013; Lewis and Smoke, 2012). But implementing performance grants may be problematic. UNCDF (2013) reports a reasonable administrative cost (1–3 percent of the grant

M4439-BAHL_9781786435293_t.indd 321

01/03/2018 11:08

322

Fiscal decentralization and local finance in developing countries

amount), but local compliance costs may be considerably larger and may divert effort from other local government activities. Paying for the performance of what one is supposed to do in any case (e.g. prepare a budget or audit accounts) raises the question of whether improved behavior will continue when payment stops. Still, this approach may be particularly useful in rural areas where local governments are generally lowest on the learning curve in terms of good public administration practice. Although similar results-based grants are to some extent employed more broadly in a few countries (Boadway and Shah, 2009; Shah, 2010), it is not clear whether tying grants more closely to results in terms of public service outputs in developing countries is widely feasible. Do Equalizing Grants Equalize? The main objective of most equalization grants is to reduce regional disparities in public service levels. The importance attached to this objective is different in different countries (Beramendi, 2012), so it is perhaps not surprising that the evidence suggests that most low- and middle-income countries do not do a very good job of equalizing fiscal disparities. For countries concerned about inter-regional fiscal disparities being too large, it is not difficult to design a good equalization grant. The process is simple: (a) decide what is to be equalized; (b) measure the present level of fiscal disparities; (c) determine how much of the inter-regional gap is to be eliminated; (d) develop a formula to produce the desired equalization; and, finally, (e) put in place an effective monitoring program to guide the continuous fine-tuning of the program in an evolving economy. But, as so often is the case, something that is easy to do in principle turns out to be much more difficult when it comes to the details. The simplest basic equalization formula is: Gi 5 ai(Ei * – Ri*) 5 f(Xi) where Gi refers to the amount of grant that will go to the ith region; Ei* is the targeted level of expenditure for the ith region; Ri* is the level of revenues that can be raised at a normal effort by the subnational government in the ith region; ai is the proportion of this gap that the transfer is intended to cover in the ith region; and Xi are the proxy variables needed to make the formula produce the desired distribution of Gi. The sum of the Gi across provinces (states) is the total required grant pool, or vertical share for the equalization grant. The first step in the process – (a) – is obvious. The objective is to produce a grant distribution that will enable all localities (or regions) to offer at

M4439-BAHL_9781786435293_t.indd 322

01/03/2018 11:08



Intergovernmental transfers ­323

least a standard (average or minimum) level of local public services at a normal (usually average) level of tax effort. But it is often difficult to assess how successfully this aim is achieved, in part because it is seldom clear exactly what the equalization target is – the fiscal capacity to provide such services, the quantity of services provided or the quality of those services. The second step, the measurement of fiscal disparities (b), is also easy in terms of fiscal outcomes such as normal revenues and normal expenditures (though not easy if measured as public sector output). As mentioned earlier, the revenue component of the gap is often calculated as the revenue that the region could raise from the tax bases at its disposal if it applied the average national rate of regional taxation. The expenditure component may be based on some ‘ideal’ (normative) level, but is more likely also to be tied to some average national level of regional expenditures. In some countries, additional weight is attached to per capita income levels – or perhaps, as in Colombia, to some index of ‘unsatisfied basic needs’ – essentially to direct more money to the poorest communities. But few developing countries specify the minimum level of services that they intend to support. The third step, deciding how much of the disparity to eliminate (c), is usually determined by the size of the revenue pool designated to finance this transfer and the importance the government attaches to equalization. In our simple model, we can calculate the necessary a by adding up the (E* – R*) for all subnational governments and dividing by the amount that the government feels is affordable. ‘Affordability’ almost never produces a revenue sharing pool that fully equalizes fiscal disparities; nor should it. Even if a country followed the model of such developed countries as Australia and set out a precise and quantifiable target for equalization (Commonwealth Grants Commission, 2015), it is unlikely to aim at achieving full equalization in the sense that the post-transfer levels of per capita expenditures, revenues or service levels are the same for all subnational governments. Full equalization would have undesired effects on subnational government tax effort, interstate migration and perhaps political support, as well as probably overwhelming the central budget.44 In principle, the goal of equalization is to provide enough in intergovernmental transfers to allow every subnational government to provide a minimum level of public expenditures at a normal tax effort. However, few (if any) political leaders start the process with this goal explicitly in mind; nor do they usually monitor how successfully it has been achieved.45 Ignorance and non-transparency may not be bliss; but they do make it easier for politicians to declare victory when it comes to equalization – as well as harder for analysts to figure out what is really going on. Politicians who increase revenue mobilization are perhaps more likely to be rewarded than those who succeed in reducing regional inequality.

M4439-BAHL_9781786435293_t.indd 323

01/03/2018 11:08

324

Fiscal decentralization and local finance in developing countries

However, the fact is that there is no simple formula that will produce the desired equalization, i.e., ensure that every local government receives precisely the desired ai(E*−R*). To do so would require accurate estimates of ‘standard’ expenditures and ‘normal’ revenues for each potential recipient government every year, and no one can or does this – though Australia comes closer than most. In practice, countries commonly use intuition and expert knowledge about patterns of expenditure needs and taxable capacity, and hope this does the trick.46 How countries design such grants is driven largely by the availability of data, as suggested by the prevalence in formula grants of factors such as population, land area and even equal shares to all recipients (sometimes justified in terms of the estimated minimal cost of maintaining a basic local government framework). Few countries have much information on such matters as the concentration of poverty at the local level, or even the average level of personal income or GDP at the local level. Perhaps for this reason grants in countries such as the Philippines and Uganda have no equalization component. Population is sometimes treated as a rough indicator of expenditure need and land area as a cost-increasing factor that should be recognized. Some formulas may even include a nominal basic per capita amount to recognize the fixed costs of government. Per capita GDP may be thought of as related to revenue capacity and its inverse used as a proxy for low revenue capacity. Many countries add different ingredients to the formula to suit local conditions and tastes as best they can, and sometimes alter the ingredients to ensure that the way portions are doled out do not vary too much.47 Analysts charged with identifying whether transfer systems are equalizing have taken different approaches. Mostly, these studies are aimed at intuitive goals that inter-regional fiscal equity might suggest. Comparing revenue disparities among subnational governments before and after transfers, Hofman and Guerra (2007) found that transfers reduced revenue disparities in China, Indonesia, the Philippines and Vietnam, although fiscal disparities remained large even after transfers. A later study by Hofman et al. (2006) found even weaker revenue equalization in Indonesia, as did Shah (2015). Rao (2009) found evidence that intergovernmental transfers in India narrowed the gap in per capita expenditures of high- and low-income states but that, once again, per capita expenditure disparities remained high. Kurlyandskaya (2005) found that the equalization transfer in Russia reduced the gap in revenue sufficiency by 25 percent. Finally, in Vietnam, Rao (2003) estimated a higher cross-section income elasticity of per capita revenue (1.87) than per capita expenditure (0.27), a result which is again consistent with some equalization through transfers between rich and poor provinces.

M4439-BAHL_9781786435293_t.indd 324

01/03/2018 11:08



Intergovernmental transfers ­325

In contrast, studies looking for a statistically significant (negative) relationship between per capita transfers received and per capita regional income levels have been less successful.48 Manasan (2009) used correlations with income to show a pattern of counter-equalization among Philippine provinces. Capuno et al. (2001), whose results were similar, emphasized the absence of an equalization factor in the Philippine formula. Alm and Boex (2008) found a counter-equalizing result in Nigeria. The combined impact of unconditional and conditional grants was not redistributive in Mexico: “the higher the per capita income of a state, the more transfers it receives” (Ahmad et al., 2007, p. 5). Mathur (2012) reports studies showing similar counter-equalizing patterns in Ethiopia, Senegal and South Africa. Transfers in Brazil fall well short of equalization (Ter-Minassian, 2015). The provincial equitable shares grant in South Africa (the largest local government general grant program) is not distributed to favor poor provinces (Alm and Martinez-Vazquez, 2009). In China, since per capita total transfers are positively correlated with per capita GDP across provinces, not much equalization results (Bahl 1999; Bahl et al., 2014). Similar negative news comes from Colombia (Chaparro et al., 2005) and India (Mathur, 2012). This outcome is not surprising. Those who do not directly gain from equalization transfers are usually the richest and most politically important regions, so central governments often react by creating other transfers that go to better-off regions in sufficient quantities to more than offset any equalization grant. In Russia, for example, the equalization transfer accounted for only 33 percent of total transfers, and in China for only 18 percent. While the proposition is not easily tested, it seems likely that the more a country’s national objective is economic growth, the less weight is put on equalization. In addition, when the structure of national transfers is altered, the reform package often includes ‘hold-harmless’ provisions intended to ensure that no one loses much (or anything), thus muting (perhaps for many years) any impact on regional disparities (Ahmad et al., 2007). Because those who gain the most from equalization often have weak absorptive and delivery capacity, their reaction to increases in transfers is sometimes to displace local revenues, as discussed earlier. Finally, some supposed equalizing transfers are just poorly designed, for example, with only a small weight on the equalization component or the inclusion of offsetting components in the formula or in another transfer.

MONITORING AND EVALUATION A truth universally accepted by all who have analyzed intergovernmental transfer programs in developing countries is that the monitoring and

M4439-BAHL_9781786435293_t.indd 325

01/03/2018 11:08

326

Fiscal decentralization and local finance in developing countries

evaluation of transfers needs to be strengthened. Like all government programs, transfers should be periodically audited and, when flaws are found, reformed as necessary. Everyone knows this. Unfortunately, no one does much about it, and few intergovernmental transfers have sunset provisions that mandate such examinations. All countries with any degree of fiscal decentralization – which means just about every country in the world – should regularly and as rigorously as possible seek the answers to such questions about intergovernmental transfers as the following: ●●

●●

●●

●●

To what extent do regional and local governments differ with respect to revenue mobilization, public expenditures and the level and quality of public services? And to what extent is reducing such disparities an explicit aim of transfer programs? Are both the objectives of transfers and the basis on which they are allocated among subnational governments clearly specified? Is there a specific target in terms of reducing fiscal disparities, and, if so, is the target clearly defined? How publicly are such matters reported and discussed? To what extent do transfers succeed in achieving their stated objectives? How can they be altered to do better in this respect? Are changes needed in the process of design, such as more (or less) involvement of recipient as well as donor governments or a greater (or lesser) role for independent appraisal? Are there specific problems in implementation (e.g., lack of timely information flows across departments or between governments)? Is there a formal monitoring and evaluation program in place covering all these matters? Is it supported by an adequate data base on subnational government finance? How transparently is this process conducted?

Designing and implementing an intergovernmental transfer system is seldom a task assigned to any specific person or office. The transfers in place in any country at any time are usually the accumulated result of a wide range of past decisions, and many hands have stirred the pot in the process. The inertial weight of the past is so large that many programs live well past their times. To mention one small example, in Brazil state revenue sharing has been fixed for decades despite a constitutional requirement that it should be changed to reflect changes in the relative ability of different state governments to carry out their spending responsibilities. Similar concerns are often expressed in many other countries around the world. One way to try to keep a transfer system at least broadly in line with

M4439-BAHL_9781786435293_t.indd 326

01/03/2018 11:08



Intergovernmental transfers ­327

its supposed objectives may be to establish some sort of policy analysis unit, usually within the Ministry of Finance, to gather data and carry out a comprehensive analysis of proposed or needed changes in the intergovernmental fiscal system. An example is the Fiscal Support Unit of the Ministry of Finance in Colombia. Colombia is also a country that has long made excellent use of the resources of such international organizations as the World Bank, the IMF and the Inter-American Development Bank to provide information and analysis with respect to the impact of its fiscal programs – while of course listening at most with only one ear to the advice that outside experts tend to volunteer, whether asked or not. Colombia also is now sufficiently endowed with analytical capacity to have other government units – for example, the Central Bank and the National Planning Department – as well as think tanks (for example, Fedesarrollo) and a growing group of academic scholars who all contribute to the public debate. Some other middle-income countries have to some extent followed this path, though few poor countries are sufficiently endowed with the capacity (or data) to do so. Another (possibly complementary) approach to improving intergovernmental fiscal transfers that a few countries have followed is to empower some kind of at least semi-independent grants commission to make recommendations for adjustments in the system and (in one or two instances) to consider seriously, if not always to accept, these recommendations (see Box 7.5). Unfortunately, few countries at any income level currently have adequate systems in place to monitor and evaluate intergovernmental transfers. In some cases, initially establishing a transfer system was so difficult and complex politically that once it was done countries were inclined to move on to other issues, with little attention being subsequently paid to the system. There are many reasons why monitoring of the intergovernmental transfer system is essential. At the most basic level, for example, it is important to verify that each recipient government gets precisely the transfers to which it is entitled. In one Latin American country some years ago, for example, one of the most important single changes made in the transfer system was simply to ensure that every municipality had a bank account in which the central treasury unfailingly deposited its monthly grant allotment on the first banking day of every month. Until then, some areas had not received any funds on time for some years; after that, even when civil unrest or natural disaster intervened, the grants flowed as they were supposed to do, and a long-standing bone of contention between some regions and the center was no more. This change also made possible more sensible local budgeting and cash flow management. Even in this internet age not all countries have yet taken such simple steps to make transfers more reliable and effective.

M4439-BAHL_9781786435293_t.indd 327

01/03/2018 11:08

328

Fiscal decentralization and local finance in developing countries

BOX 7.5  FINANCE COMMISSIONS1 A few countries have established, sometimes constitutionally, special commissions responsible for important aspects of intergovernmental finance. Perhaps the best-known such body is the Commonwealth Grants Commission (CGC) in Australia, but somewhat similar commissions exist in several large and federal developing countries, notably India, Pakistan and South Africa. The basic idea is to have a body that is outside the normal structure of government but competently staffed and credible politically to be able to look at the changing context of intergovernmental fiscal relations with a fresh eye and to recommend changes that will be seriously debated and considered by lawmakers and by the public. Experience with this approach has been mixed. Australia The model for finance commissions, Australia’s CGC, was created in 1933 and is still important today. The CGC is charged with recommending how the revenues from the federal goods and services tax (GST) are distributed among Australia’s five states and two territories “such that, after allowing for material factors affecting revenues and expenditures, each would have the fiscal capacity to provide services and the associated infrastructure at the same standard, if each made the same effort to raise revenue from its own sources and operated at the same level of efficiency” (CGC, 2015, vol. 1, p. 2).2 The Commission is a permanent body, with at least three members appointed by the central government for a fixed (renewable) term. It has an annual budget and operates as part of the Treasury Department, but is not bound by government policies. Its annual reports to Parliament, which are usually comprehensive and supported by substantial technical material, are published and widely discussed. Although its recommendations are not binding, have not always been accepted, and have sometimes been very unpopular with certain states, the CGC continues to be an important and often definitive factor in shaping intergovernmental fiscal transfers in Australia. In addition, all Australian states have local government grants commissions. State governments are required to review the recommendations of these commissions as a condition for accessing federal grants for local authorities. Apart from special assignments related to local finance, the main task of the state commissions is to make recommendations for the distribution of federal grants to local authorities. Various criticisms have been made of these commissions. Some have noted that the equalization principles laid down by the CGC are sometimes interpreted differently by the states so that intergovernmental fiscal transfer regimes vary across the states. Others have said the annual increase in the size of the state grant pools does not match state variations in fiscal capacity (Lewis and Searle, 2011). Petchey and Levtchenkova (2007) stress that expenditure norms are endogenously set – based on actual expenditures – rather than exogenously set.3 Boadway and Shah (2009) warn that the system may be overly complicated and too academic in its approach.4

M4439-BAHL_9781786435293_t.indd 328

01/03/2018 11:08



Intergovernmental transfers ­329

India The Australian model was in some ways the model when India’s constitution established the Central Finance Commission (CFC) in 1951.5 The CFC’s mandate is to make recommendations on the vertical sharing arrangement for central government taxes, the allocation of shared taxes and certain grants among the states, and measures needed to supplement the budgets of state governments to provide resources to third-tier local governments – municipalities (urban) and panchayats (rural). Unlike Australia’s CGC, however, India’s CFC is appointed by the President every fifth year and dissolves after making its recommendations. The five members, usually all respected experts supported by professional staff, visit every state and issue a report that is generally comprehensive and supported by technical analysis. Although its recommendations are not binding, they have generally been accepted by Congress. Intergovernmental transfers in India have in effect been developed over the years following the periodic CFC reports, from the first report in 1952 to the fourteenth in 2015. However, not all transfers have come under the scrutiny of the CFC (Rao and Singh, 2006). Planning commission grants and the various line ministry grants to states were not encompassed in its mandate: Finance commission grants accounted for about 59 percent of the total amount of transfers (Rao, 2009). With the abolition of the Planning Commission in 2015, however, and the accompanying devolvement of many line ministry grants to the states, the relative importance of CFC grants has increased. In addition, for many years the basic CFC grant allocation basically ‘filled the gap’ between actual state expenditures and state revenues rather than attempting to assess the difference between capacity to raise revenues and the cost of providing a normal level of expenditures, creating a serious disincentive to good state fiscal management (Rao, 2009). An additional problem was that, as with many long-standing institutions, the dead hand of the past often weighed heavily on CFC decisions. Most changes recommended were only incremental, with the objectives of maintaining both stability in intergovernmental fiscal relations and political credibility. For example, the state share of the divisible pool of central taxes was set at 29.5 percent by the Eleventh Finance Commission, raised to 32 percent by the Twelfth and then nudged down to 30.5 percent share by the Thirteenth, with equally small marginal changes being made in the horizontal allocation. However, the most recent (Fourteenth) Commission broke the mold in 2015, recommending an increase in the vertical share to 42 percent by 2020, largely to compensate for the decentralization to the states of responsibility for many previously existing central programs (Reddy, 2015). India also has state finance commissions (SFCs) that are required by the constitution and independent of the CFC. The CFC neither receives SFC reports nor comments on their recommendations. Moreover, although the CFC may augment state resources for intergovernmental transfers and make recommendations for local finance reforms, these recommendations do not bind the SFCs. How local governments are financed in India (as in Canada) is decided entirely by the states. Many observers think that the SFCs have done little, hampered as they are by the absence of adequate professional staff, limited data on local government finances and the fact that their recommendations have generally been ignored by the states

M4439-BAHL_9781786435293_t.indd 329

01/03/2018 11:08

330

Fiscal decentralization and local finance in developing countries

(Alok, 2011).6 Even when a good report is prepared, it has often been submitted so late that it cannot be taken into account in preparing the state budget (Bahl et al., 2010a). Nonetheless, some SFCs have made useful recommendations on such matters as bottom-up planning, the abolition of discretionary grants by members of state legislatures, and the establishment of a more transparent (and higher) vertical share for third-tier local governments (Oommen, 2010). Pakistan Although Pakistan’s National Finance Commission (NFC) is similar in some ways to India’s CGC, it has a very different history and has in general been less successful. As in India, the NFC is supposed to decide every five years both the size of the divisible revenue sharing pool and the distribution of this amount among the four provinces. However, in recent years its decisions have effectively crippled the intergovernmental system. In 2009, the federal government agreed to the 7th NFC award, raising the share of subnational governments from 45 to 57.5 percent of all federally raised taxes and then passing a constitutional amendment in 2010 that significantly decentralized spending responsibilities. However, the provinces did not absorb most of the decentralized bureaucrats and did not accept all of the committed capital projects (Ghaus Pasha and Pasha, 2015), resulting in a worsened situation for the national budget – a situation that was not altered by the most recent (2016) NFC award. To make matters worse, the constitution now fixes the federal revenue share at a ceiling of 42.5 percent of federal tax revenues. This means the federal share can go down in future awards, but it cannot be increased. Since Pakistan’s governments in total get little revenue (about 10 percent of GDP), the federal government now must finance its expenditures (including some other transfers to the provinces) and meet its debt obligations with little more than 5 percent of GDP. Or, it could take on the long-overdue task of increasing revenue mobilization (though the higher subnational government revenue share complicates this situation). The sustainability, let alone efficiency or equity, of this situation is questionable. Much of the difficulty arises because the constitution mandates that the four provinces – despite their vast differences in wealth, needs and demographic ­conditions – must agree on the proposed formula. It is not surprising that consensus has been hard to reach, and in fact the last formal (binding) agreement was in 1996, which was later extended to 2006 and then replaced by a new arrangement decided by the President (Ahmad et al., 2007). In addition, the NFC is a much more explicitly political body than the CFC or CGC: five of its 10 members are ex officio and it has no staff of its own. Like their counterparts in India, Pakistan’s provincial finance commissions (PFCs) have generally not measured up to expectations. PFCs have no linkage with the NFC, take different approaches in each of the four provinces, and differ from each other and the NFC in structure (although all appear to be dominated by the provincial finance ministry), functions and powers (Cyan and Porter, 2007). Their principal role is to determine the distribution of grant revenues among (thirdtier) local governments in addition to bringing together the provincial and local governments to negotiate intergovernmental fiscal relations more generally. In

M4439-BAHL_9781786435293_t.indd 330

01/03/2018 11:08



Intergovernmental transfers ­331

practice, politics, the NFC’s problems at the national level and poor data have meant that PFCs have really done very little, although some of the existing ­variation across provinces in how local grants are distributed probably reflects an attempt by PFCs to respond to specific local situations and diversity (Bahl et al., 2008, 2014; Bahl and Cyan, 2009). South Africa The newest of the four grants commissions is the South African Fiscal and Finance Commission (FFC) established in the 1997 Constitution. The FFC, which is required to be impartial, is charged with making recommendations on the ‘equitable’ distribution of national resources. Its recommendations are presented to the Ministry of Finance and are not binding, although the minister must provide reasons to Parliament if he does not accept them. In practice, he often has not, and has instead largely supported competing recommendations made by the National Treasury. Although the FFC can and does comment on the actual allocations made by the government, it has been unsuccessful in getting much public support for its own views. Its detailed reports, however, have proved to be helpful in informing the political discourse, and hence making the competing claims more transparent. The FFC has a small professional staff and is a permanent body, like the Australian CGC. It has nine members with renewable terms: four appointed by the central government; three in consultation with the provincial government leadership; and two in consultation with local government organizations (since the FFC is also responsible for studying and making recommendations on local finance). As noted above, its main role at present appears to be as a (more or less) ‘inhouse’ research and advisory unit, although it is not obvious that its work has had much effect on the actions of the Treasury, which is now clearly in charge of intergovernmental fiscal relations in South Africa. Notes: 1. This draws on Bahl and Cyan (2009). 2. For an interesting (if not totally impartial) review of the first 50 years of the CGC, see Commonwealth Grants Commission (1983). See .cgc.gov.au for a fuller account of the Commission’s activities. 3. Some consider the same feature of the Canadian system – that the (implicit) expenditure norms of the equalization transfer are determined by what provinces do and not by central mandate – to be a strength, not a problem. 4. The same has often been said of Canada’s considerably simpler system: reportedly, a Minister of Finance once said that he had only one official who understood it – and he was about to retire! 5. For a discussion of the historical development of the recommendations of the Indian Finance Commissions, see Chelliah (2006). See also .fincomindia.nic.in. 6. As with most things in India, the quality and impact of SFC reports differ sharply from place to place and time to time. In Uttar Pradesh, for example, the 1996 SFC’s recommendations for state–local transfers were generally sensible and basically accepted by the state government.

M4439-BAHL_9781786435293_t.indd 331

01/03/2018 11:08

332

Fiscal decentralization and local finance in developing countries

The other side of a transfer also needs close attention: do subnational governments comply with the conditions attached to many transfers? Someone needs to collect and analyze the data to make sure that required matching has been done, that expenditures made from transfers are made for eligible categories, and that grant funds reached the intended recipients. At the same time, central governments generally impose some mandates on subnational governments – for example, to spend some minimum amount (or share) on education or not to exceed some maximum spending level (or ratio) on administration. Mandates are usually a bad idea; but if they must be part of the system, they should be tracked and their impacts evaluated. Analysis is also needed to evaluate the extent to which transfers tend to raise or lower local efforts to raise revenues, to determine their effect on equalization objectives and to evaluate how well ‘performance grants’ actually perform. Conditions in developing countries sometimes change quickly, and yesterday’s transfers may not fit tomorrow’s economy. The earmarked transfers so often favored by donor agencies seeking to ensure more funding for their favored activities, as well as by central governments wishing to control lower-level governments, can easily proliferate over time and become administratively burdensome.49 Conditionality may become outmoded and the costs of complying with ever-increasing reporting burdens overwhelming unless transfer systems are continually monitored and periodically modernized. New demands for this and that transfer and often for more conditionality to respond to this or that perceived ‘need’ continue to emerge. Both existing and potential new changes to transfer systems require close attention to improve outcomes and head off bad ideas. A good policy practice would be for all conditional grants to have a sunset at which time they would be reviewed and revised in structure, and continued (or not). Developing countries can ill afford mistakes in allocating their scarce fiscal resources. Of course, since human and political capacities are also always scarce in such countries it is all too easy to let things drift until some unavoidably  visible crisis comes along. Nonetheless there are some success stories: ●●

●●

In South Africa, for example, the Treasury Department has done a good job of monitoring and evaluating the intergovernmental transfer system, and issues a periodic report on the status of local governments. The Fiscal and Finance Commission also has a resident staff that produces regular analysis of intergovernmental fiscal issues. In India, every fifth year the Finance Commission produces a solid report based on a comprehensive data base on the performance

M4439-BAHL_9781786435293_t.indd 332

01/03/2018 11:08



Intergovernmental transfers ­333

●●

●●

of state governments, with recommendations for changes in the arrangements for vertical and horizontal sharing. The analysis is shared and enhanced by hearings in each of the states. Indonesia’s Directorate General of Fiscal Balance in the Ministry of Finance tracks and evaluates the performance of provincial and local governments, and in 2012 carried out a comprehensive and professional evaluation of the system (Directorate General of Fiscal Balance, 2012). Colombia’s Fiscal Support Division in the Ministry of Finance has assembled a good data base and presents periodic analyses of subnational performance.

As these examples suggest, developing countries can assemble a proper data base on subnational government finance and make good use of the information. While there is still much that can and should be done to improve matters in most countries, some progress has been made, and more can no doubt be expected in the future. International agencies and bilateral donors have long played a role in helping with the evaluation of intergovernmental fiscal systems. They have carried out or sponsored some of the most important work in this area, and have contributed to numerous improvements in the systems of some countries. Increasingly, however, countries like those mentioned in the preceding paragraph are moving to do such work themselves. This is as it should be. One must hope that the donor agencies will also move with the times and shift their resources more to ensuring better training and development of local experts, providing more support (including training and resources) for in-country think tanks (official and non-government), and helping to develop stronger data bases on intergovernmental finance. Of course, outside agencies may also provide assistance by responding to requests for technical assistance and bringing relevant international experience to the table.

CONCLUSION The best transfer system for any country is the one that best achieves the objectives set for it. What those objectives are, and how transfer systems are intended to achieve them, are matters determined more by political negotiation than by economic analysis. The outcome is almost always a hodge-podge of different transfers intended to appease and balance­ ­different interests. Nonetheless, although the details vary considerably from country to country and time to time, there is a critical set of ideas

M4439-BAHL_9781786435293_t.indd 333

01/03/2018 11:08

334

Fiscal decentralization and local finance in developing countries

that can be brought to bear in building any intergovernmental fiscal system. Transfer systems should be designed to achieve their intended objectives, and be monitored to evaluate this. Equalization grants should reduce unwanted fiscal disparities. Subnational governments should not be encouraged by the grant system to reduce their own tax efforts. Intergovernmental transfers intended to encourage this or that more specific objective – improved subnational administration, providing better services to poorer citizens – can and should be designed so that they are likely to achieve these objectives to at least some extent. Although it is seldom desirable to copy features from the grant system in other countries without fitting them carefully to the case at hand, there is much to be learned from experience elsewhere. No one gets it quite right at first, and most countries have trouble treading the fine line between revising transfers too often and letting the whole system fall out of touch with its underlying objectives. No set of intergovernmental transfers ever succeeds in doing everything that everyone wants it to do. The work of building a good transfer system, like the broader task of fiscal decentralization (or improving governance in general), is never really over. The best one can do is try to keep on top of what is going on and move the system as opportunity permits in what is thought be the right direction to the extent and when possible.

NOTES   1. To make the presentation simpler, we assume that all central government revenues are taxes. Many non-tax revenues (e.g., dedicated user charges) are seldom legal sources for financing transfers to local governments, but central revenues are usually fungible enough so that this assumption is not too far off the mark.   2. The countries are those for which the data were available in the source cited.  3. This issue has frequently come to the fore in Canada, for example, often (but not always) led by the linguistically distinct province of Quebec. For a brief account of some of the different possible interpretations of the contested concept of ‘fiscal balance’ in Canada, see Bird (2006), and for an attempt to estimate the very tenuous relation between the facts and some of the political posturing on this issue, see Vaillancourt and Bird (2007).   4. If ‘normal’ revenue exceeded minimum expenditures for a subnational government, it would not be eligible to participate in the grant system.   5. One needs to be careful with such jurisdictional comparisons, however, because such ratios are sensitive to redrawing boundaries or amalgamating or dividing jurisdictions. For example, in some countries in which (as discussed later) some transfers are distributed in equal shares (or a fixed amount) to each locality, the number of localities tends to increase over time – and so, unsurprisingly, does the disparity between the richest and the poorest units.   6. As we discuss below, one exception is when profitable natural resources are in otherwise poor regions and those regions receive some share of the profits.

M4439-BAHL_9781786435293_t.indd 334

01/03/2018 11:08



Intergovernmental transfers ­335

  7. See, for example, the analysis in Bird and Tarasov (2004) of the relative ineffectiveness of the equalization components of even large ‘vertical gap’ transfers like those in most developed federal countries, most of which have frequently changed their equalization systems.   8. Because an equalization grant in effect puts all localities in the same position at the margin when it comes to responding to the incentive of a conditional grant (i.e., the ‘tax price’ to them of increasing the favored expenditure is the same), separating the two types of grant arguably increases both the equity and the efficiency of each.   9. Canada had a somewhat similar system during the 1940s when the provinces gave up their constitutional power to tax income to the federal government in exchange for a ‘rental’ payment. After World War II, however, the provinces took their taxing power back, although over the next half century most (not all) agreed to let the federal government administer their principal taxes, though at rates set by the provinces. 10. For fuller discussions of vertical sharing, see Bahl and Wallace (2007) and Boadway and Shah (2009). 11. In some countries, the respective shares for provincial and third-tier local governments are earmarked, but in others, the regional governments have considerable freedom as to how they allocate this amount. 12. Shah (2013, p. 239) tells the story of an egregious experience with a capital transfer in Bangkok, “where central financing for a section of above ground metro was withdrawn, leaving poles that support no rails.” 13. Since such revenues include natural resource revenues, when oil prices rose sharply – although most revenues accrued to one province – the level of all provincial revenues was substantially increased, and so was the equalization payment pool. However, this pool was financed entirely from federal revenues which had increased considerably less (in part because some of the increased provincial revenues were deductible for federal tax purposes). The reaction of the federal government was to change the formula to reduce the impact of natural resource revenues but not to change the system, which continues to be driven by provincial rather than federal revenues. This may not seem particularly sensible but, as always, every country has its own history, its own style and its own peculiarities. 14. Until 2005, the sharing rate was 25 percent, but was increased to 50 percent to compensate for the revenue loss from the elimination of the local business tax. 15. For further discussion of various combinations of tax decentralization and administrative centralization in Canada, China, Germany and Spain, see Bird (2015). 16. The responsibility for tax administration was divided between a central and local tax bureau in each province in 1994 (Bahl, 1999). But with the dismantling of the business tax in 2015, the local tax bureau has much less responsibility for tax administration. 17. As McLure (1983) emphasized long ago, when tax revenues (or bases) are allocated by formula, it is the structure of the formula and not simply the structure of the tax that determines its economic incidence. For example, Canada allocates the corporate income tax base between provinces by a formula that assigns equal weight to the proportion of the corporate payroll in the province and the proportion of its sales; consequently, provincial corporate income taxes are really based on payroll and sales. On the other hand, unlike China, Canada allocates VAT collections to provinces based entirely on the share of taxable sales in the province, a procedure that both ensures that the provincial VAT remains a destination-based consumption tax, and does so without incurring the high compliance costs commonly incurred with respect to taxing interstate trade. 18. Well, nothing – even the arithmetic of grants – is completely transparent in the complex political world of intergovernmental transfers. For example, in this illustration per capita GDP may be measured as the ‘distance’ from average (mean) GDP or from median GDP, or as a ratio to the average; or some missing data might be interpolated or some subnational governments (metropolitan areas, natural resource rich areas) might be excluded altogether. When it comes to understanding transfers, the details are critical.

M4439-BAHL_9781786435293_t.indd 335

01/03/2018 11:08

336

Fiscal decentralization and local finance in developing countries

19. In South Africa, for example, the constitution provides that transfers will flow directly from the central government to municipalities. Since census data are not always available at the municipal level, the way in which funds are allocated is often driven by the answer to the question ‘What data are available?’ 20. A useful discussion of designing formula grants may be found in Martinez-Vazquez and Searle (2007); see also Kim and Lotz (2008) on the use of needs indicators in OECD countries. 21. The HDI is a composite statistic of life expectancy, education and per capita income indicators which is used to rank countries into four tiers of human development. 22. Consider, for example, the extensive reports of the Australian Commonwealth Grants Commission (2015), which in effect uses a variant of this approach in distributing GST revenues to the states (see also www.cgc.gov.au). 23. This was one reason this approach was not used in designing the first equalization transfers by the Fiscal and Financial Commission in South Africa, since Cape Town would have ended up with much more than the much poorer northern provinces. 24. See, for example, the cases in Martinez-Vazquez and Searle (2007) and Ahmad (1997). 25. As usual, matters are not quite this simple. An equalization formula like this will also tend to raise the total level of subnational taxes. Such formulas may also have differential effects on different regions when, as in most countries, some subnational governments receive grants and others do not, or when some jurisdictions are much larger than others and thus play a more important role in determining the average. Many other factors may also affect outcomes, including: the weight of factors in the formula other than the ratio between the average subnational tax rate and the rate in a particular jurisdiction; how tax bases are estimated; what tax bases (e.g. resource revenue) are included; and how the effective tax rate used in the formula is calculated. A useful recent review that touches on some of these questions is Boadway (2015). However, many aspects of equalization grants can only be fully understood by exploring them (e.g. through simulations) in the context of a specific country. For a broad recent review of Canada’s equalization system see, for example, Beland et al. (2017). Helpful additional discussion and country studies may be found in Martinez-Vazquez and Searle (2007). 26. In recent years, however, and notably since the major reform of state sales taxes with the introduction of the GST, Indian scholars have produced considerably more refined measures of the relationship between the composition of economic activity in states and indicators of state tax capacity and effort: see e.g. Garg et al. (2017) and Mukherjee (2017). 27. In one country, for example, simple simulations showed that simply plugging the (largely uncontested) measure of the ratio of rural to total population in a region into a formula produced almost the same distribution of grants as a much more refined set of measures (many rather controversial) that had been proposed by various interested parties primarily to direct more transfers to poorer regions. In this, as in other cases, our experience is that what the decision-makers really care about is essentially the result – who gets what – and not the ingredients of the recipe that produced the result. 28. In the past, distributing transfers could be a complex and slow process, sometimes subject to discretionary interventions by politicians. Now, all that is required to get the money to the intended recipient is to allocate the funds electronically to the appropriate regional and local accounts. Transparency has to some extent reduced political meddling. 29. As Fiszbein (1997) shows, such differences in local and central preferences are common. 30. For example, in Indonesia – which provides funding for subnational government employee salaries through its DAU general grant – the central government has some control over the number of employees in theory, but in practice most payroll is reimbursed. Shah (2012) argues that the result has been increased subnational government employment – and, perhaps, reduced local government efficiency. 31. For an interesting discussion of the relative merits of centralized vs. decentralized provision of disaster relief, see Goodspeed (2015).

M4439-BAHL_9781786435293_t.indd 336

01/03/2018 11:08



Intergovernmental transfers ­337

32. Defined as those with positive ‘Polity2’ scores (a scalar measure of the degree of democracy or autocracy employed by, e.g., Giavazzi and Tabellini, 2005). 33. For a more detailed discussion of this point, see McLure (1994) and Bahl and Tumennasan (2004). 34. The last two points mentioned have often been discussed in the literature on the ‘resource curse’ (e.g. Collier and Hoeffler, 2005) and on specific problems that have often arisen in sharing natural resource revenues (Natural Resource Governance Institute, 2016). This last publication, like many others, suggests many solutions to the problems to which such windfalls often lead. Perhaps the most popular is to establish a ‘heritage fund’ earmarked for long-lasting infrastructure investments intended to spread the wealth over both time and (usually) all parts of the country in one way or another. A model often cited is that of Norway: for a good example, see the suggestions for Uganda in Hartmark (2011). 35. For an example of how complex it can be to determine what is going on while the revenue sharing arrangements are being worked out, see the discussion in Bird (2012a) of a major change in Colombia’s natural resource revenue sharing regime. 36. For an interesting attempt to construct (and to some extent test) an explanation for the widely varying degree of reliance on such general interregional transfer regimes, see Beramendi (2012). Although this is not the major concern of his book, his argument is worth noting: essentially, that the widely different degrees of inequality and different mixtures of interpersonal and interregional transfers found in different countries may be considered as reflections of the varying extent to which different regions have different economic bases, the pattern of interregional mobility and the extent to which people see redistribution as an issue between regions or between individuals across regions. Moreover, the fiscal structures initially adopted to fit these circumstances in turn will themselves tend to alter the underlying conditions, and hence induce further changes over time. These are deep waters that need much further exploration. 37. In many low-income countries, international donors also often favor conditional grants related to those policy objectives in which they are interested. As Hirschman and Bird (1968) emphasized long ago, everything (good and bad) that can be said about conditionality with respect to transfers also applies to foreign aid. 38. For a discussion of how electoral and bureaucratic self-interest affects the design of grant systems, see Bahl and Linn (1992). 39. On the diverse roles that transfers may play with respect to nation-building, see Vaillancourt and Bird (2016) and Sorens (2015). 40. The problem may also reflect the efficiency with which subnational governments deliver services, an issue that was discussed in Chapters 3 and 4. 41. A recent review of the revenue substitution effect in Latin America is provided by the studies in Fretes Cibils and Ter-Minassian (2015). 42. See also the cases discussed in Martinez-Vazquez and Searle (2007) and Kim et al. (2010). 43. See, for example, Steffensen and Larsen (2006), Smoke and Lewis (2014) and Shah (2005, 2010). 44. In rich countries and poor countries alike, a common explanation why rich regions may support at least some equalization is that they consider it one way to slow down in-migration from poor regions (Beramendi, 2012). But no rich region is eager to dig too deeply into its own pockets to achieve this objective. 45. A task that is far from easy, even in data-rich developed countries, as Beland et al. (2017) show. 46. Occasionally, countries introduce ‘horizontal’ transfer regimes that not only allocate transfers to the ‘below-average’ regions but also finance those transfers in whole or part by taxing (imposing a negative grant) on ‘above average’ regions. Chile does this to some extent with its municipal community fund, and similar schemes are found in Germany and some ex-Soviet Union countries such as Russia, Kazakhstan, Lithuania, Ukraine and Belarus (Golovanova and Kurlyandskaya, 2011).

M4439-BAHL_9781786435293_t.indd 337

01/03/2018 11:08

338

Fiscal decentralization and local finance in developing countries

47. Bird (1986) suggests that even in countries such as Switzerland it appears that formulas have sometimes been altered primarily to ensure that outcomes remain unchanged: that is, what matters most in political terms is not the formula but the distribution it produces. 48. The thinking here is straightforward. If poorer places receive greater per capita grants than richer places, the grant system will narrow fiscal disparities. 49. In China, for example, the number of conditional grants was over 200 by 2010 (Bahl et al., 2014). There is of course little new under the sun: by the mid-nineteenth century, so many taxes had been earmarked (hypothecated) to specific expenditures (mostly debt service) in England in order to persuade skeptical voters that the state could be trusted to spend taxes on things that they accepted (and creditors that the interest on state loans would actually be paid) that sensible budgeting became impossible, until a series of major reforms eventually produced a consolidated and unified central budget (Daunton, 2001).

M4439-BAHL_9781786435293_t.indd 338

01/03/2018 11:08

PART IV

Summing Up

M4439-BAHL_9781786435293_t.indd 339

01/03/2018 11:08

M4439-BAHL_9781786435293_t.indd 340

01/03/2018 11:08

8.  Financing metropolitan areas* The most fundamental step in evaluating metropolitan fiscal performance is to diagnose in a broad-based and well-grounded way the match between the features of the national institutional and fiscal framework and a country’s objectives for metropolitan government and development. (Smoke, 2013, p. 80)

Few aspects of fiscal decentralization in low- and middle-income countries are more important than ensuring that their cities, especially the largest metropolitan cities, can adequately accommodate not only those already there but also the many millions who will soon join them.1 How well they do so will affect not only the quality of life for those living in these cities but also the level and pattern of national economic growth. As Glaeser (2011) and others have shown, there is ample evidence that productivity and incomes increase with city size and density, and that city growth and national economic growth are strongly linked. Bringing people together in the right way can create substantial agglomeration economies through what Duranton and Puga (2004) have called ‘matching’ (enabling labor and other markets to function more efficiently), ‘learning’ (moving up the learning curve is easier when teachers and examples are available) and ‘sharing’ (through better access to human and physical infrastructure). Eliminating barriers to capturing such agglomeration benefits is one path to increased productivity. Some cities have attempted to follow this path by making it easier for businesses to be established and expanded and by improving infrastructure. Some cities and countries have attempted to reduce the demand for local public services and infrastructure by slowing the movement of lowerincome people into urban areas and attempting to control and restrict the sorts of informal activities in which they often engage to survive. Such policies, sometimes intended to improve the security and the quality of life of the better-off, may have perverse effects – such as increasing inequality and social inclusiveness – while perhaps reducing security and sustainable growth.2 Few countries have created such explicit barriers to urban migration as China’s system of household registration (hukou), but many have restricted and penalized street traders and informal transport providers and harassed and evicted those living in informal slums. One important driver of such policies is the simple inability of most cities in developing 341

M4439-BAHL_9781786435293_t.indd 341

01/03/2018 11:08

342

Fiscal decentralization and local finance in developing countries

countries to finance an adequate level of public services. An important reason they cannot do so is because most urban local governments have, in effect, been thrown off the deep end by the failure of higher-level governments either to develop any clear urban strategy or to provide cities with fiscal instruments capable of doing that part of the job for which they should be responsible. The fiscal framework in most low- and middle-income countries features centralized tax systems, user charges that do not recover costs, intergovernmental transfer systems that sometimes discriminate against urban centers, and almost always severe limits on local budgetary autonomy. None of these features helps cities meet expenditure needs or capture the revenue opportunities that accompany urbanization. They are especially bad for the biggest cities, which generally have different fiscal needs and capacities than other localities (Bird and Slack, 2013). Some countries have taken steps towards filling the gap between what exists now and what is needed in such ways as: setting up new government structures; making changes in expenditure assignments; encouraging increased revenue mobilization by urban local governments; creating new targeted grant programs for large cities; and relaxing debt restrictions. Some cities have acted on their own, for example, resorting to land sales to support spending needs and trying to capture some of the land value increases that come with urbanization. But most such reforms, whether at the city or country level, have occurred in an ad hoc or piecemeal way rather than as part of any comprehensive metropolitan fiscal strategy. In this chapter, we ask and try to answer three questions that are central to the issue. Does urbanization call for a change in the fiscal balance between large cities and the rest of the country? We think it usually does. Are existing approaches to governance, financing and service delivery in metropolitan areas satisfactory? We think most of them are not. Are there new directions for fiscal and governance reforms in metropolitan areas that should be considered? We suggest some that seem promising.

METROPOLITAN AREAS IN THE NATIONAL SETTING People are drawn to cities largely because better economic opportunities are available. Two classes of migrant may be distinguished. Some are sufficiently educated or skilled to assimilate into the formal labor market. These people are looking for ways to advance their careers, to increase their human capital and to achieve a better quality of life. They want adequate housing and housing-related public services like electricity and water, good

M4439-BAHL_9781786435293_t.indd 342

01/03/2018 11:08



Financing metropolitan areas ­343

schools and health care, clean and environmentally friendly surroundings, a good public transportation system and, importantly, a satisfactory level of personal security. To attract and retain this component of the labor force, urban public services usually need not only to be expanded but also often upgraded and properly maintained. To do so requires a significant expansion of local government spending. The other group of migrants in developing countries generally does not come packaged with education, relevant work experience or financial resources.3 This group, sometimes numerically by far the largest, cannot be so easily absorbed. Often, city leaders and most of the existing population do not want such migrants and see them as a burden – a view that is widespread but not well supported by evidence (McGranahan, 2016). Treating such groups as people to be planned against rather than planned for usually leaves them to struggle for existence, often for generations, in inadequately served slums or informal settlements – an approach that does no one any good in the long run. Cities (and countries) that wish to develop without serious unrest and the associated problems need to meet the essential public service needs of all residents sufficiently for them – or at least their children – to have a reasonable prospect of accumulating sufficient human and financial capital to achieve a productive foothold in the metropolitan and national labor market. Existing urban residents often resent migrants because they feel crowded by them and see them as competing for housing and the use of roadways and other public amenities. Most of the labor force in large cities consists of existing residents who will demand upgraded public services as incomes increase.4 The growing private sector in metropolitan areas demands public expenditure on infrastructure to enhance their competitive position through: better information technology (IT) and transportation linkages; education to enhance human capital skills; public utilities that support the efficient production of goods and services; quality governance that leads to efficient delivery of services; and better fiscal planning to avert the need for counter-productive regulatory measures.5 The income, consumption and property tax bases of metropolitan areas should, like expenditure demands, increase with urbanization, as should the willingness to pay taxes. The ability of local administrations in cities to impose taxes should also improve with urbanization. Often, however, local governments are not allowed to access more productive tax bases. Even when they are permitted to do so, it is not easy to persuade local people and businesses that they should pay for more and better services even if they believe, as experience suggests they have little reason to do, that they will in fact see such benefits in the future. It is equally difficult to get other governments to agree to pay their fair share for services that

M4439-BAHL_9781786435293_t.indd 343

01/03/2018 11:08

344

Fiscal decentralization and local finance in developing countries

provide benefits even to non-residents. Demands for expenditure increases rise almost step by step with urbanization. But it usually takes some time for the revenue base to expand, thus exacerbating the fiscal gap. Getting agreement on new taxes (and transfer systems) and designing and especially implementing them can takes years, not months. Persuading people that they are getting services for their tax (and fee) dollars may take even longer. Rewriting the fiscal contract at the local level is no easier than at the national level.6 It takes time, often a long time, for people to become accustomed to and to accept new ways of financing local government. Urbanization and the Size of Cities Half the population of developing countries lives in urban areas. Much more urbanization lies ahead. While projections differ, most agree that the world population is likely to double in the next 30 years or so, with almost all the increase being absorbed by urban areas in developing countries (UN, 2008). Some developing countries, particularly in Latin America, have already gone through the big movement to cities and are well along on the learning curve for urban policy reform. But many others will see large flows of migrants to cities and major increases in urbanization in the next few decades. In India, for example, nearly 70 percent of the population still lives in rural areas, as does about two-thirds of the rapidly growing population of sub-Saharan Africa. The size of population growth will be most striking in the largest cities, and there will soon be many more such cities. By 2025 there will be perhaps 27 ‘mega-cities’ with populations greater than 10 million – 21 of them in less-developed countries – holding about 10 percent of the world’s population. In addition, there may be about 50 cities with populations in the 5–10 million range, most in developing countries, and many more cities with a million or more people (though the growth rate may be even higher in cities with less than a million). By 2030, for instance, half of Africa’s population is expected to be in urban areas. Urbanization is upon us, everywhere, and those who think about local government finance in developing countries – a subject that in the past was often focused largely on rural areas, villages and small towns – need to think bigger.7 Why cities matter to economic development is well known: agglomeration effects allow firms to capture economies of scale; more exchange of ideas increases labor productivity and innovation; access to a larger and more specialized labor market helps relax supply constraints and increases productivity; and a more advanced infrastructure and education system facilitates productivity increases. It is not uncommon for individual metropolitan areas to account for more than one-fourth of national gross

M4439-BAHL_9781786435293_t.indd 344

01/03/2018 11:08



Financing metropolitan areas ­345

domestic product (GDP) in OECD member countries (OECD, 2006a). The same economic dominance is true in developing countries, no matter how economic primacy is measured (Bahl, 2017). For example, 27 percent of Turkey’s GDP is generated in Istanbul (OECD, 2008) and over 50 percent of Argentina’s in Buenos Aires (Braun and Webb, 2012). In Ghana, 27 percent of GDP is produced in the most economically dense 5 percent of its land area (World Bank, 2009b). In Panama, about 80 percent of all consumption takes place in urban areas. One half of Brazil’s GDP is produced in 12 metropolitan areas where one-third of the population resides, with São Paulo alone being the location of 402 of 500 corporate headquarters (Arretche, 2013). The six metropolitan areas in South Africa account for 40 percent of the population but 60 percent of national GDP (Steytler, 2013). In almost all countries, developing as well as developed, the performance of the national economy is largely driven by urban areas. As many have argued, from Pirenne (1927) through Jacobs (1984) to Glaeser (2011), cities generally lead the way with respect to the development and restructuring of local economies, and indeed countries in general. As migration constraints are relaxed (for example, owing to reduced transportation costs) labor can move to more productive employment. Firms located in secondary cities increasingly supply raw material and intermediate goods to metropolitan industries. As diseconomies of scale increase production costs in larger urban centers, however, more production may begin to move to secondary centers. Firms may also be driven away from large cities by environmental degradation and congestion, poorly allocated infrastructure investments, a lack of public utilities and inappropriate land market regulations (Henderson, 2015). One way and another, the higher quality of life that first developed in large metropolitan areas should spread to secondary urban centers. More rural, agriculturalbased areas may also benefit not only from increased demand but also because the balance between land and labor is improved by out-migration. However, history suggests that such developments occur only slowly. Metropolitan cities may continue to be economically dominant for a long time. The rest of the country may thus often perceive more costs – e.g. through the brain drain as more talented workers move to the large cities to seek better opportunities – than benefits from big city growth for many years. The Political Economy of Metropolitan Growth This imbalance in perceived benefits makes it more difficult to achieve the necessary political support for this unbalanced spatial pattern, so it is not surprising that the treatment of metropolitan areas in the national fiscal

M4439-BAHL_9781786435293_t.indd 345

01/03/2018 11:08

346

Fiscal decentralization and local finance in developing countries

system often occasions national political debate. The better public services usually found in big cities may give rise to friction (Smoke, 2013; Sud and Yilmaz, 2013). Other parts of the country may feel unfairly treated and conclude that they are subsidizing those in the city. At the same time, those in the city may feel that they are being unfairly taxed to support areas that contribute less to the national well-being and have less economic potential. Such debates – between central cities and suburbs, between prosperous provinces and those less fortunate, and between big cities and smaller communities – are common.8 Even though much of what each side says in such debates often rests more on anecdote than hard evidence, with both sides appealing to ‘alternative facts’ as they perceive them, the outcome often influences political decisions. Here, we discuss briefly only three aspects of this issue: (a) whether metropolitan areas contribute enough to the support of non-metropolitan areas; (b) whether asymmetric treatment of metropolitan areas in the intergovernmental fiscal system is justified; and (c) whether strengthening subnational governments makes rational public policy less likely. Do metropolitan areas get too much? Large urban areas contain a disproportionately large share of the national tax base. For example, 80 percent of all Indian taxes are collected in urban areas (McKinsey Global Institute, 2010) and 75 percent of all Mexican taxes are collected in the Mexico City metropolitan area (OECD, 2004). At the same time, big cities usually provide their residents and businesses with the best public services in the country. Invariably, the question of whether the rest of the country receives its fair share of the revenues collected in the big cities is a contentious issue. Those responsible for metropolitan areas commonly argue that they are subsidizing the rest of the country while their own infrastructure and public servicing needs are going unmet – as indeed they often are when cities are denied access to broad own-source finance and must rely on national politicians to right the imbalance. However, the rest of the country – and most (usually territorially based) national and regional politicians – may see the transfer of resources away from urban areas as politically necessary, or perhaps even as an essential component of national development policy. In India, for example, where over 70 percent of the population resides in rural areas but less than 30 percent of expenditures are made by rural local governments (Bahl et al., 2010a; Oommen, 2010), many national (and state) politicians appear to believe that rich urban areas should provide even larger transfers to nonmetropolitan areas. The inter-regional flow of funds may sometimes play a role in arguments about potential secession (Vaillancourt and Bird, 2016). Richer regions

M4439-BAHL_9781786435293_t.indd 346

01/03/2018 11:08



Financing metropolitan areas ­347

that argue they contribute more through formal or informal equalization mechanisms to central government revenues than they should may be more willing to secede than poorer ones. But poorer regions that think they do not get enough may equally want out. In OECD countries, for example, Catalonia (in Spain) and Flanders (in Belgium) are rich regions that can argue they make a net contribution to the national treasury (which they do, because they contain relatively more high-income people subject to income taxes and also have higher levels of consumption subject to sales taxes than other regions). On the other hand, in the United Kingdom context, even though residents of Scotland clearly benefit substantially from generous national fiscal transfers, those supporting secession argue that until very recently this inflow was more than offset by the value of the oil and gas revenues received by the central government from deposits off the Scottish coast. Everyone involved in such disputes can usually find what they consider to be convincing arguments for their side. Many attempts have been made to estimate the flow of funds from urban tax bases to the rest of the country. The most straightforward approach is simply to calculate the difference between the amount of central (and regional) taxes collected in metropolitan areas and the sum of direct expenditures by higher-level governments in metropolitan areas and intergovernmental transfers to local governments in metropolitan areas. The difference between the outflows (taxes) and inflows (expenditures and transfers) is then usually interpreted as measuring the net contribution to the rest of the country. Such studies require making many questionable assumptions with respect to such issues as how to handle central government expenditures that cannot be allocated to specific local areas (e.g., defense), the appropriate treatment of deficit spending by the central or state governments, and even how best to measure incidence (e.g. through the conventional ‘differential’ approach, usually implicitly imposing a proportional income tax as the standard, or the ‘balanced-budget’ approach, which may make more sense on a spatial basis).9 Many such ‘flow of funds’ regional studies have been made in conflicted federal countries such as Canada;10 but most dodge many critical issues – for example, with respect to expenditure incidence and benefit incidence (De Wulf, 1975), the extent to which expenditure benefits (welfare gains) are equal to the amounts spent, the treatment of externalities, and the excess burdens of taxes – so that it is difficult to know what to make of them. Early studies of fiscal incidence attempting to some extent to take into account some of the factors just mentioned (e.g. the study of Korea in Bahl et al., 1986) are unsurprisingly inconclusive. More recent incidence studies like those summarized by Lustig (2016) draw on much richer data bases but inevitably still rest to some extent on sometimes tenuous and

M4439-BAHL_9781786435293_t.indd 347

01/03/2018 11:08

348

Fiscal decentralization and local finance in developing countries

contentious conceptual foundations. No solid empirical studies in developing countries (or in fact anywhere) seem to deal adequately with measuring and assessing the complex issue of net fiscal incidence either within metropolitan areas or between those areas and the rest of the country. It is much simpler to put forward and defend one’s own ‘facts’ about who gets what or to offer unsupported opinions about who should get what than it is to produce solid evidence about what is actually going on, let alone to secure agreement about what should be going on. Perhaps it is fortunate that inter-regional flow of funds analyses in the end seem to have had little effect on policy decisions. Still, from time to time this issue pops up in some country and stirs the debate again.11 Asymmetry in the intergovernmental fiscal system Metropolitan governments should receive an appropriate share of intergovernmental transfers to address spillovers and distributional concerns (see Chapter 7). Importantly, however, they should also have more leeway than other local governments in raising revenue as well as more autonomy in spending and arranging capital financing. Big urban local governments should be as fiscally self-supporting as possible. One way to achieve such a goal may be to establish a special fiscal regime like that in Bogotá, Colombia. However, few developing countries are willing to give up so much control over metropolitan finances. In Brazil, although the constitution entitles municipalities to the same rights and duties as states, state governments nonetheless have full responsibility for creating metropolitan structures (Wetzel, 2013; Arretche, 2013) – an arrangement that is unlikely to lead to the creation of special fiscal regimes for metropolitan governments. In India, where states are similarly responsible for controlling metropolitan area governments, many have used their power mainly to delay implementation of a 1974 constitutional amendment intended to provide a clear schedule of rights and duties of local governments (McKinsey, 2010; Mathur, 2013). In unitary countries, where local governments are governed by central law, their position is generally even less secure; however in a few such countries, for example, the Philippines, Indonesia and South Africa (Smoke, 2013), metropolitan local governments do have significant fiscal powers. One reason central and regional governments are generally reluctant to create a special regime for metropolitan local governments is because they want to ensure that all local governments march in step with central and/ or regional policies. Another reason is that they may fear that giving large local governments access to more broad-based taxes may crowd out some central government taxes.12 Higher-level governments have many ways to control metropolitan spending. They may impose conditional grants and

M4439-BAHL_9781786435293_t.indd 348

01/03/2018 11:08



Financing metropolitan areas ­349

mandates; they may establish rules and compensation with respect to local public employees; and they may override local governments completely and take over delivery of local services. These and other methods are often used. At some point, however, perhaps even the most controlling higher-level government may decide – or be forced to decide – that the sheer magnitude of metropolitan fiscal problems requires a special fiscal regime that will let them manage their own fiscal affairs and deal with their own problems. Fiscal power and political power Another major reason why countries have been reluctant to support the path of urban-led economic development is political: introducing a special regime for metropolitan areas may change the balance of political power. Big city mayors such as as Boris Yeltsin (Russia), Joko Widodo (Indonesia) and Virgilio Barco Vargas (Colombia) have sometimes been able to use their performance on the metropolitan stage to make their case for national office. Since media are largely based in big cities, prominent local politicians often have a highly visible and powerful platform from which to challenge the central (or state) government. Higher levels of government may react by cutting or holding back on transfers in addition to refusing to establish any special metropolitan fiscal regime or to devolve additional taxing power to metropolitan areas. Why make it easier for possible political opponents to do a better job? The lag with which electoral representation reflects real population shifts in many countries gives more political weight to non-metropolitan voters and may create yet another barrier to establishing stronger and more independent city governments.13 The reluctance of those who control the center to give up the keys to the treasury to potential rivals is likely one reason why many countries are often slow to focus on metropolitan fiscal problems as a nationally important concern, let alone to create more independent and responsible metropolitan governments.

GOVERNANCE AND FINANCE IN METROPOLITAN AREAS How metropolitan areas are financed is closely related to how they are governed. There is no single best way to structure governance within metropolitan areas. Some structures may be more successful in promoting economic development; some may lead to more equitable service levels within the metropolitan area; some may increase participation by residents in shaping the local services they receive; and some may more easily be controlled by higher-level governments. The right governance structure

M4439-BAHL_9781786435293_t.indd 349

01/03/2018 11:08

350

Fiscal decentralization and local finance in developing countries

for any metropolitan area depends in theory on the importance assigned to these and other objectives, with different levels of government likely attaching different weights to each. The decision made often depends largely on which level of government makes it. The economic literature on this question understandably focuses on economic efficiency. The core argument is the familiar decentralization theorem, which holds that each function should be assigned to the lowest level of government consistent with its efficient performance (Oates, 1972). This approach to assigning expenditure responsibilities within metropolitan areas provides a useful framework for answering some important questions about how best to govern big cities:14 ●● ●● ●● ●● ●● ●●

To what extent should smaller local governments in a fragmented metropolitan governance system drive expenditure decisions? Is a metropolitan area-wide government necessary for managing and financing at least some public services? Can public enterprises (or private enterprises subject to public regulation) do a better job of delivering certain services? What physical area should a metropolitan regional government encompass? What is the appropriate role for higher-level government transfer programs? Which level of government should be responsible (and accountable to residents) for financing local services?

Answering such questions is not so difficult in theory. The more emphasis placed on local decision-making power, the more that smaller local governments should be strengthened. On the other hand, economies of scale and externalities often call for more area-wide governance, and the more fragmented the governance of metropolitan areas, the less likely they are to be able to provide adequate financing for local services. The answer depends on how decision-makers balance these concerns. To illustrate, a major factor shaping expenditure assignments in metropolitan areas may be the desire to have local governments with sufficiently small populations so that residents feel that their vote and their participation matters in influencing fiscal decisions.15 People understandably want to have a say in what their government does. Within metropolitan areas  – perhaps especially those with multicultural residents who may speak different languages or follow different religions – some groups often favor more fragmented governance systems for similar reasons. Preferences may also vary with income levels: rich neighborhoods may want to exclude most informal sector activities (street traders, slums) and spend more on,

M4439-BAHL_9781786435293_t.indd 350

01/03/2018 11:08



Financing metropolitan areas ­351

say, parks and schools; poorer neighborhoods may prefer better drainage and sanitation to make their living conditions more palatable (and their working conditions more productive). If those with similar preferences are sufficiently organized on a territorial basis to be heard within a metropolitan area, then a jurisdictionally fragmented system or a two-tier metropolitan government structure with a strong bottom tier is more likely to satisfy their needs than either a unitary metropolitan government or the partial equivalent of a coordinated network of local public utility enterprises or service districts. An additional virtue of more local control from the perspective of economic efficiency is that such decentralized systems provide more leeway for jurisdictions within the larger metropolitan area to compete with one another for residents, businesses and tax base. However, those more concerned with equitable distribution may consider competition more a cost than a benefit because those who start with the most resources often come out on top. For reasons like these, even when most budgetary decisions are in the hands of larger area-wide governments, many countries have created one or more tiers of lower-level units to give a degree of fiscal autonomy to neighborhood units, or sometimes perhaps only to get their advice as an input into policy decisions.16 In the Philippines, for example, local governments include 1,496 municipalities, 138 cities and 42,025 barangays (neighborhood governments) (Lewis and Searle, 2011). However, as with similar ‘bottom-level’ governance institutions in some other low- and middle-income countries (for example, villages in China), barangays in the Philippines have neither significant taxing authority nor important discretionary expenditure powers (Devas, 2004). As we discussed in Chapters 3 and 4, some public services are characterized by economies of scale, especially such capital-intensive infrastructure as public utilities, solid waste disposal, sewerage and mass transit. Other local public services appear to less subject to scale economies, although the relation between unit cost and size is often obscured by differences in the quality of services. Interestingly, a study of Indian metropolitan areas (McKinsey Global Institute, 2010) found that the cost of delivering basic services was 30–50 percent cheaper in metropolitan areas than in sparsely populated areas. For example, the cost of delivering a liter of piped water was about 50 percent cheaper because cities can leverage common supply depots and cut distribution costs. Similarly, some types of infrastructure, such as airports, may only make sense once local populations reach a certain size: the study just mentioned estimated that given the relationship between flights to a city and its population, on average the capital cost per daily flight was about three times more for cities of less than a million than for cities with populations greater than 4 million. As a rule, however,

M4439-BAHL_9781786435293_t.indd 351

01/03/2018 11:08

352

Fiscal decentralization and local finance in developing countries

metropolitan areas are more than large enough to capture most economies of scale (Slack and Bird, 2013), although this cannot usually be said for all local governments in jurisdictionally fragmented metropolitan areas. Local decisions may affect others. Spillovers from local government expenditures are likely in a jurisdictionally fragmented metropolitan area with a number of local governments operating near one another. Central city governments might underspend on infrastructure to control pollution or solid waste disposal and adversely affect environmental conditions in other jurisdictions. Other local governments may underspend on, say, health facilities, leading their residents to utilize central city facilities. Even if every locality provides some transit facilities, they are unlikely to be as coordinated (or perhaps as cheap) as those that could be provided on a metro-wide basis, resulting in worse and more congested services. Spillover and coordination problems can often be resolved most simply by providing such services through a metropolitan government sufficiently large to internalize the externalities. An alternative (though perhaps less efficient) approach may be to establish formal or informal cooperative agreements between local governments to improve service delivery.17 It is seldom easy to determine the appropriate service area that needs to be included to internalize externalities. Even if a good guess is considered good enough (as it often must be in the hectic real world in which policies are usually decided), the political cost of designing and implementing such arrangements may be so high (and the resulting welfare gains so hard to estimate convincingly enough to be a selling point) that such plans tend to be more discussed than implemented. In Toronto, Canada, for example, although mass transit has long been the subject of intense political controversy and the province has established a metropolitan region transit authority, different local transit systems are still far from integrated. Although there have been various provincial initiatives to create metrowide governance institutions, none has yet been established. Economic efficiency and management considerations help drive the choice of government structures in metropolitan areas. But there are also important equity and social considerations. Families with similar income levels or ethnic backgrounds may cluster in the same area. Fragmented local government structures may result in significant fiscal disparities in spending levels and tax burdens, and these disparities might be amplified if local taxing autonomy is increased. A metropolitan area-wide government can erase formal tax and spending disparities between jurisdictions, though some neighborhood differences in service levels will usually persist, sometimes for a long period. Cape Town in South Africa moved from a jurisdictional fragmented model to a metropolitan model of governance largely to reduce such disparities.

M4439-BAHL_9781786435293_t.indd 352

01/03/2018 11:08



Financing metropolitan areas ­353

No country has found it easy to resolve major problems of metropolitan government and finance. There is no one best governance structure in either theory or practice that serves all circumstances. Every country has its own peculiar history, politics, culture and happenstance that shape how it now governs metropolitan areas and how big cities are financed. How governments are structured is always and everywhere a major determinant of the revenue and expenditure assignments to metropolitan local governments. In Chapter 3 we argued that finance should in principle follow function. Reality tells us that function is largely shaped by how governments at different levels and in different places are structured (Bahl, 2013). Alternative Governance Structures Governance structures may be broadly categorized in several ways. Bahl (2011, 2013, 2017), following Bahl and Linn (1992), outlines three basic approaches to metropolitan governance in terms of their apparent principal objective: jurisdictional fragmentation (local control), functional fragmentation (technical efficiency) and metropolitan government (coordination and internalizing externalities).18 In practice, balancing the advantages and disadvantages of variants of these three forms of metropolitan governance usually results in some compromise model. Once in place, even when the character of the metropolitan population changes dramatically over time, governance structures and jurisdiction boundaries tend to persist. However, the structure of metropolitan government is not written in stone. It can be changed, but such changes impose significant transition costs and can be achieved only with major political efforts. Jurisdictional fragmentation Jurisdictional fragmentation means that governance in the metropolitan area is primarily in the hands of many general-purpose local governments (municipalities) that operate with some degree of independence in choosing what they do and how they finance it. Although in some cases there may be an overlying metropolitan government, a region-wide special district or some cooperative agreements between municipalities, the main service providers are the lower-tier local governments. The advantage of this approach is obviously that it keeps government closer to the people. When local governments become submerged in a larger unit, local politicians and officials become less known and less accountable to those they are supposed to serve.19 Costs from failing to capture economies of scale and problems in dealing with externalities may offset any gains from increased accountability (or from the possibility of increased competition in taxation and service

M4439-BAHL_9781786435293_t.indd 353

01/03/2018 11:08

354

Fiscal decentralization and local finance in developing countries

provision). Since the tax bases of local governments are likely to vary widely, and since their expenditure needs and capacities to deliver services also usually differ, significant fiscal disparities among local governments often exist within metropolitan areas. In some cases, unit costs of service delivery may be higher in smaller jurisdictions owing to scale economies and perhaps also because of duplicate administrative services. In many lower- and middle-income countries, residents may not have any clear way (through elections or other means) to ensure local accountability. Accountability with respect to who is responsible for what and to whom becomes even weaker when people may live in one locality, work in another and shop in yet another. Despite such problems, jurisdictional fragmentation is a good description of how metropolitan areas in many industrial countries are governed (Bahl, 2011; Bird and Slack, 2013). In the United States, for example, most urban services are delivered by municipalities, counties and single-function special districts (Fox and Slack, 2011). Much the same is true in Europe. In Copenhagen, for example, the metropolitan population of 2.4 million is governed by 45 municipalities and an overlying capital region (Lotz, 2006). Local governance in the Paris urban agglomeration (population about 8 million) is by 80 municipalities, three departments and numerous companies that provide public services. The Stockholm metropolitan region includes 65 municipalities and five counties (OECD, 2006b). The Randstad (Netherlands) metropolitan region contains 50 municipalities (OECD, 2007a). Much the same is the case in some developing countries, as Box 8.1 illustrates. Functional fragmentation One way to coordinate the delivery of a single function (or a related group of functions) within a metropolitan setting is to assign it to a public company or special district government. Such functional fragmentation exists in almost all metropolitan areas, including those with only local governments as well as those with some form of area-wide metropolitan government. The arrangements vary widely, as does the extent to which this approach is employed. The main advantage of functional fragmentation is that serving a larger area and/or population may enable fuller realization of economies of scale and a higher level of technical efficiency and coordination in service delivery. It may also allow more independence from general government, and hence permit the attraction and retention of higher-quality management and staff. These advantages are multiplied if, as is often the case, the entity has access to a dedicated revenue stream such as an earmarked tax, a share of the budget of a higher-level government, a compulsory transfer from the city government, or user charges. A stronger

M4439-BAHL_9781786435293_t.indd 354

01/03/2018 11:08



Financing metropolitan areas ­355

BOX 8.1 JURISDICTIONAL FRAGMENTATION: EXAMPLES FROM DEVELOPING COUNTRIES There are many examples of jurisdictional fragmentation in the metropolitan areas of low-income countries: ●

Basic public services in the Manila metropolitan area are the responsibility of 11 cities and six municipalities, with wide disparities in their taxable capacities and expenditure levels (Manasan, 2009). These are overlaid by a Metropolitan Manila Development Authority (MMDA) that is responsible for planning, coordinating and delivering services with a regional impact. ● Services in the Mexico City metropolitan area are provided not only by the federal district (now a state) and its 16 municipal-like sub-units (boroughs), but also by the states of Mexico and Hidalgo (with their 59 municipalities) and the federal government. There has been little coordination of service delivery within the metropolitan area, and virtually no planning (Raich, 2013). ● The Kolkata metropolitan area is governed by three municipal corporations, 38 municipalities and 24 rural local governments. The Kolkata Metropolitan Development Authority (KMDA), an area-wide state agency with some elected local representatives on its board, is financed by grants from the federal and state governments, and is responsible for major infrastructure development in the metropolitan area (Sridhar and Bandyopadhyay, 2007). ● The São Paulo metropolitan region, with a population of about 18 million, includes 39 municipal governments with no overlapping metropolitan government (Wetzel, 2013). The failure to effectively coordinate the delivery of public services across the metropolitan area has long been a matter of concern (Rezende and Garson, 2006; World Bank, 2007a).

financial base often also provides greater access to capital markets than most general-purpose local governments have. The major drawback to the functional approach to providing urban services is that public companies and even special districts are less directly accountable to local voters than an elected municipal council. The extent to which this is true depends on how the board and the management of the autonomous agency function, with practice varying widely from local ownership and control to total higher-level government control (Berry, 2009). A functional authority finds it easier to deliver a specific service in a technically competent way if it has more control over its activities, staff and finances. It also finds it easier to pay less attention to the wishes and needs of those who are underserved or badly served. Public enterprises are thus often under less pressure to satisfy their customers (local residents) than are either private enterprises (provided there are potential competitors) or local governments (Bird, 1980). As Self (1972) puts it, the functional approach makes sense when the

M4439-BAHL_9781786435293_t.indd 355

01/03/2018 11:08

356

Fiscal decentralization and local finance in developing countries

goal is to achieve a well-specified task (goal effectiveness) but not when the need is either to decide what that task should be (goal articulation) or how to value its achievement relative to other possible uses of scarce public funds (goal coordination). The latter two are quintessentially political functions that require a more explicitly and openly political institutional home. Nonetheless, in some cases the functional approach to financing and delivering metropolitan services may be best. When the services delivered are amenable to pricing (e.g., public transportation, garbage collection), imposing user charges may provide not only a base level of revenues but also, if the charges imposed are well designed, a much-needed incentive to provide the right services (Bird, 2001) – that is, those for which people are willing to pay. When, as is often the case, local government proposals to impose even well-designed user charges are viewed by most as simply hidden taxes, creating a user-charge financed utility may be the best way to go from both an economic and a political perspective. In some cases, the sale of services may even be used to subsidize the city budget. In Stockholm, a holding company was organized to manage several city-owned companies that provided such services as public housing, real estate management, port operations and water utilities. When these public companies were in a surplus position, they paid dividends to the city budget. The same is true in the case of two energy companies in which the city of Oslo has held equity. The city of Lausanne in Switzerland fully incorporated the electricity company into its budget, and the company maintained a surplus position during the early 2000s (Bahl, 2011). Other examples of publicly owned businesses channeling funds to local governments may be found around the world both in developed countries like Japan and in lower-income countries like Colombia and Romania. Despite its evident popularity, however, basing general local government financing on profits from public businesses can also lead to problems. Sometimes the result may even be that the public sector ends up subsidizing business, and hence increasing rather than decreasing taxes on others (Bahl, 2011). In Eastern Europe, for example, Riga (the capital of Latvia) provided services through 42 companies in which it held ownership or had an equity stake. Most were self-supporting, but the transport enterprise at one point claimed about 10 percent of the city’s operating budget. In Zagreb, Croatia’s capital, most capital spending (and some current spending) has been the responsibility of a holding company that was created following the merger of 22 municipal companies. The holding company received a subsidy of over 15 percent of the city’s budget. As in Sofia and Budapest (Alm and Buckley, 1994), city budgets seem often to end up subsidizing the costs of public transit companies, which may or may not

M4439-BAHL_9781786435293_t.indd 356

01/03/2018 11:08



Financing metropolitan areas ­357

be economically sensible (Bird, 1980). In Paris, where the city government participates in (or is part owner of) enterprises that provide services ranging from transportation to social services, many of these companies – even when they impose user charges – also receive compulsory transfers from the city budget. Similarly, Madrid makes compulsory transfers to the two public companies that provide its transportation services; and in several large Italian cities transfers to companies providing transportation, waste collection and disposal, and water treatment services account for about 25 percent of expenditures. Most special district governments or public enterprises in metropolitan areas are single-purpose, e.g., transportation or water supply. However, some combine different functions: ●●

●●

●●

●●

In Canada, the Greater Vancouver Regional District consolidated all functions provided previously by special districts (most notably hospitals, water and sewer, capital expenditures, and solid waste management) and is financed primarily from user charges. Although the governing board of the regional district includes elected municipal government officials, municipalities can opt out of most district functions. In metropolitan Bogotá, where a district public utility company has long been responsible for providing several area-wide services, a separate public company was created to implement a comprehensive transportation plan that included the regulation of private providers of bus services. Although the operating costs of transit operations are fully financed from user charges and a surcharge on the gasoline tax, capital costs are financed by the central government. In metropolitan Mexico City and Rio de Janeiro, mass transit is the responsibility of many providers, and there is relatively little coordination on routes or fares. A joint venture company owned by the city of Buenos Aires and the (adjacent) province of Buenos Aires is responsible for the disposal of solid waste. India makes extensive use of parastatals, which are public companies operated by various departments of the state (or federal) government.20 Although an autonomous agency of the Mumbai municipal corporation is responsible for electricity and bus services, and appears to be well managed, there are 21 parastatals operating within Mumbai which account for a large share of total infrastructure planning and spending in the metropolitan area, and in some cases route their funds through various metropolitan agencies, thus improving coordination in service delivery (Pethe, 2013).

M4439-BAHL_9781786435293_t.indd 357

01/03/2018 11:08

358

Fiscal decentralization and local finance in developing countries

Metropolitan government Area-wide metropolitan governments may either provide all local government services themselves (the one-tier model) or share the task in different ways and to different degrees with lower-level governments (the two-tier model), or perhaps with public companies or special districts as discussed above. There are many forms and varieties of metropolitan government dominance. Some metropolitan governments may not encompass all the economically relevant metropolitan area as in the case of Colombia’s Special District (Bogotá). In other cases, very large cities such as Mumbai may not have any special (metropolitan) status but nonetheless are so dominant in their areas that they constitute a sort of de facto metropolitan government, and may sometimes be formally responsible for some planning and coordination responsibilities, and perhaps also the delivery of some regional services such as mass transportation and utilities (OECD, 2006a). The advantages of metropolitan government are much like those commonly associated with incorporating a firm: fuller internalization of spillover effects, simpler coordination in the delivery of functions, fuller advantage of economies of scale, and the possibility of capturing a broader customer (tax) base more efficiently. Resources can be allocated more efficiently when decision responsibility is centralized. They may also, if desired, be distributed more equally because the level of local service provision is no longer tied to the wealth of each local jurisdiction. In addition, bigger governments may be more able to secure subsidies and influence higher-level governments on fiscal issues. Some think these advantages are so great that in some countries every large urban area should have a unified large metropolitan government (McKinsey, 2010). Others are less keen to reduce the extent to which local voters can shape fiscal decisions. Problems may arise if local tensions are exacerbated by forcing existing local governments into a metropolitan structure or if tensions between metropolitan and higher-level governments increase when there are shared functions between the two, as in Mumbai and Manila, for example. It is also not always the case that service delivery on a metropolitan basis is cheaper. Eliminating duplication by going from, say, six fire departments to one may seem to be an obvious cost-saver. But costs may increase if, for example, it turns out to be impossible to terminate or demote anyone and the outcome is mainly to create a new layer of administration with wages leveled up to the previously highest level. Equally, the quality of service (e.g. response time to fires) may decrease in some areas with fewer, more centralized fire stations. Or perhaps the reduction of local fiscal competition in the metropolitan area may make it easier for budgets to expand and expenditures to rise. There is little hard evidence on

M4439-BAHL_9781786435293_t.indd 358

01/03/2018 11:08



Financing metropolitan areas ­359

such matters even in developed countries (Slack and Bird, 2013, Bahl and Campbell, 1976). Another problem may be that the boundaries of the metropolitan government are not set widely enough to cover the real metropolitan area, as in Bogotá, Mexico City and Toronto, to mention only a few examples. One way to address this problem is to adjust the boundaries through annexation or consolidation, as was done in South Africa (Ahmad, 1996; Cameron, 2005). However, few countries have the same opportunity to approach this issue from a fresh perspective as post-apartheid South Africa did. In Canada, although local governments have no formal constitutional status and some major reorganization has taken place, it has usually proved impossible to encompass entire metropolitan regions within an area-wide government structure.21 In principle a two-tier structure, with local governments providing specifically local services and perhaps some interjurisdictional equalization – as in Chile, where there has long been explicit redistribution of local taxes among municipal governments (World Bank, 2003)22 – may be the best approach conceptually in many cases (Bird and Slack, 2007). In practice, however, the significant political costs of amalgamation, annexation or mergers (OECD, 2006) and constitutional impediments often means that the best that can be done is to strive for voluntary intergovernmental agreements to facilitate coordination of services, though the effectiveness of such arrangements is at best likely to be variable. Box 8.2 illustrates further the wide variety of governance arrangements found in practice.

THE BUDGETARY PRESSURES FACING LARGE CITIES Providing adequate public services in big cities in poor countries poses unique challenges. Most have inadequate infrastructure and housing and considerable poverty, problems that are accentuated by continued in-migration. Even though the revenue potential in metropolitan areas is greater than elsewhere, the backlog in (and growing need for) public services means that marginal fixes are unlikely to resolve the basic public financing problem they face. And even marginal fixes are often difficult because of the prevailing fiscal arrangements and the political climate. More money is needed to do the job right, but money alone cannot solve the problems big cities face. Their governance structure must also be right. Ideally, it should facilitate determining the extent to which fiscal decentralization to and within the metropolitan area is needed; coordinate the work of many different government agencies at all levels in the area;

M4439-BAHL_9781786435293_t.indd 359

01/03/2018 11:08

360

Fiscal decentralization and local finance in developing countries

BOX 8.2 METROPOLITAN GOVERNANCE: EXAMPLES FROM DEVELOPING COUNTRIES In high-income countries (other than the United States), metropolitan governance on a region-wide basis is not uncommon. The models used vary, but common features are boundaries that to a greater or lesser extent include the labor market area and provision for region-wide delivery of public functions that require large service areas. Toronto, Tokyo and London are examples. Some metropolitan areas in low-income countries also feature region-wide governance. For example: ●

In 1996, the former 61 local government entities comprising Cape Town were consolidated into a two-tier system of six general-purpose governments and a metropolitan authority. In 2000, a single local authority, the ‘unicity’ of Cape Town, was created (OECD, 2008a). All local government functions were moved to the metropolitan level, although social services are a shared function with the province. ● The Istanbul metropolitan area is governed by both a provincial administration with an appointed leadership and a metropolitan municipality with an elected leadership. The metropolitan municipality performs most major urban functions, while the provincial administration performs some areawide functions and oversees coordination. Beneath the metropolitan municipality are 73 local-level municipalities that perform mostly neighborhood functions. The result in Istanbul is that most fiscal decisions are made at the metropolitan level by either the metropolitan municipality or the provincial administration level (OECD, 2008). ● Bangkok is a provincial-level city that governs the entire metropolitan area, overlaying 18 districts, each of which has an elected local council – although again most local government budget decisions and revenue raising powers reside at the metropolitan level.

upgrade the quality of the local government staff; and develop a viable plan for resource mobilization. None of these tasks is easy, but the stakes are high enough to make it worth trying to do them better. Failing to get metropolitan governance and finance right may severely damage the competitive position not only of the metropolitan area but also of the country. It may even threaten political stability if the right balance between addressing the needs of the rich, the growing middle class and the poor (both within the metropolitan region and between that region and the rest of the country) cannot be found. Cities may be and often are leaders in innovation, both economically and socially. But they can also be incubators of mass protest and social unrest. Countries that fail to respond properly to calls for societal change may reap further social discontent and economic damage (UN Habitat and ESCAP, 2015).

M4439-BAHL_9781786435293_t.indd 360

01/03/2018 11:08



Financing metropolitan areas ­361

In many countries, the window to capture the first-round agglomeration gains of urbanization may not be open for long. Those metropolitan areas that fail to accommodate social change and to provide adequate supporting public services for economic growth may lose out to those that do better.23 Partial, temporary quick fixes to urban problems may all too often become permanent features of the landscape. It is thus important for sustainable development that countries face up to and resolve as best as possible the problems facing their large urban areas. A bonus from attempting to do this is that the results will provide useful experience and knowledge suggesting how countries could develop and implement a more general system of intergovernmental fiscal relations that will be sustainable and flexible enough to accommodate future waves of population migration and private sector development. Public Servicing Needs Most metropolitan local governments spend more per person than other local governments (Bird and Slack, 2013). For example, the central cities of Buenos Aires and São Paulo spend twice as much per capita as other local governments in their metropolitan regions. Many factors may account for higher spending in big cities: ●●

●●

●●

●●

It costs more to provide some services, owing to higher factor costs (labor, land). Economies of scale may offset such cost ­differences to some extent, but there is little evidence to ­suggest that cost ­savings are ­significant beyond low levels of urban ­concentration (Slack and Bird, 2013).24 Large urban area governments incur expenditures that smaller communities do not face, such as larger and more complicated road networks, mass transit, more regulation and control of congestion, special firefighting equipment (e.g. for higher buildings) and the like (Bahl, 2017). Since those who live in big cities on average have higher incomes – after all, this is one of the main reasons many of them are there – they are as likely as their counterparts in OECD countries (Florida, 2002) to expect more and better local public services such as better recreation and park facilities and a cleaner and safer environment, as well as better health and education to the extent local governments are responsible for providing such services. Large cities in poor countries face the difficult and expensive task of providing public services (and to some extent perhaps also better housing) in large slum areas. As Freire (2013) discusses, even if only minimal urban services are provided, this can be costly.

M4439-BAHL_9781786435293_t.indd 361

01/03/2018 11:08

362 ●●

Fiscal decentralization and local finance in developing countries

Finally, the other side of the budgetary coin, which we discuss later in this chapter, is that metropolitan area local governments have more taxable capacity than others; and that metropolitan politicians, like others, are likely to spend more when they have more.

For these and other reasons, local governments in metropolitan areas usually have higher per capita expenditures than other local governments. But higher spending levels do not mean that local public services in most of the world’s largest cities are high quality. Often, they are well below any minimum standard anyone would likely set. As a recent survey in India noted, for example, most large cities fall “well short of not only the level of services to which international cities aspire but even below a ‘basic’ standard of living for their residents” (McKinsey, 2010, p. 54). India is not alone in this respect. Much the same can be said of big cities in other low- and middle-income countries. In China, for instance, the provision of social services to migrant workers in cities is viewed as “one of the most critical elements of the necessary reforms” (World Bank and Development Research Center of China, 2014, p. 186). In low-income countries, city governments often understandably give high priority to poverty reduction and direct resources towards this objective. The Problem of Public Management The poor record of service delivery by local governments in developing countries is a principal reason why many policy analysts have argued strongly that more centralized management of public expenditure is the best approach for developing countries (Bahl and Linn, 1992). Things have changed to some extent in recent years, however. Some local governments in presentday metropolitan areas can now manage effectively, at least when they are allowed to do so by higher-level governments (World Bank, 2009c; Van Ryneveld, 2006). But many are held back by weak staffing and management systems. India, for example, has been characterized as a country that has not faced up to the complexities of managing large cities, and continues to stay with weak administrative processes and weak leadership that leads to poor delivery choices (McKinsey, 2010). Nonetheless, cities such as Ahmedabad and Bengalaru (Bangalore) have managed to develop and implement some well-thought out policies to improve outcomes (Ahluwalia et al., 2014). In general, however (as Sud and Yilmaz, 2013 argue), in most low-income countries the institutional weaknesses of local governments – not only in terms of their generally low position in the intergovernmental system but also, more directly, their weak administrative capacity – often imposes a greater barrier to the provision of good services than does the shortage of resources.

M4439-BAHL_9781786435293_t.indd 362

01/03/2018 11:08



Financing metropolitan areas ­363

One important way to strengthen public management in metropolitan areas is to give more discretion to local governments to make their own decisions about service delivery and about managing their budgets. At present, higher-level governments in many countries have their hands all over local governments: they appoint chief local officers; control decisions about hiring, firing and promoting employees and how much they are paid; and keep close control over spending decisions, and especially the selection and design of capital projects. Letting at least the larger local governments make their own decisions on such matters, as well as giving them more revenue-raising powers, would launch a new era in which local governments would have much more power, and responsibility for what they do. Most countries seem reluctant to do anything like this because, as we discussed earlier, they have little trust in the ability of local governments and may perhaps fear strengthening potential political rivals. And, as almost any central (or state) government legislator will tell you, metropolitan areas are the golden revenue goose that needs to be kept inside the fence. Higher-level governments may also often be afraid that subnational governments will mess up and they will then be called upon to clean up the mess. A key aspect of improving urban management is to ensure increased accountability of service providers to their ultimate clients – local residents (including businesses). How exactly such accountability is established will vary with the political and administrative system and culture. Doing so may involve any or all of the following: increased political oversight by elected officials and local councils; community and business advisory councils; citizen report cards (Paul, 2014); more transparent budgeting; requirements for published (audited) financial reports covering both expenditures financed by transfers and those financed from local revenues; and explicit and transparent intergovernmental contractual agreements (Bird, 2000). Without improved accountability, neither public nor private providers have strong incentives to improve the management and delivery of metropolitan services. Of course, since even the best-run, best-staffed and best-intentioned local governments can run into trouble, an explicit procedure for dealing with fiscal problems at the local level should also be in place, perhaps along the lines discussed in Chapter 3. More autonomy in making service delivery decisions is usually a key part of the answer to how local governments can be made more efficient at delivering local public services. Although it is not the whole answer, developing a more autonomous governance structure for metropolitan areas so that they have both the resources and the incentive to tackle the task of how to finance and deliver area-wide services more efficiently, effectively and equitably is perhaps the best way for countries to find that answer.

M4439-BAHL_9781786435293_t.indd 363

01/03/2018 11:08

364

Fiscal decentralization and local finance in developing countries

Infrastructure As we discussed in Chapter 4, keeping up with capital facility needs is perhaps the major financial challenge facing growing metropolitan cities in developing countries. This underlines the need to pay special attention to the capital side of the budget. Ingram et al. (2013) estimate that lowincome countries may need to spend about 2.8 percent of GDP for new infrastructure investment in urban areas in addition to another 2 percent of GDP for maintenance.25 This level of expenditure is well beyond the fiscal capacity of developing countries, given that on average they raise only about 17 percent of GDP in revenues (Bahl, 2014). Individual country studies suggest a similarly bleak outlook. A recent estimate for India, for example, is that capital investment in new public infrastructure (mainly transport) is planned to double to a sustained level of over 1.1 percent of GDP for the next few decades (Ahluwalia et al., 2014). This may not sound like much, but the total revenues (including transfers) of all urban local governments in India would have to more than double to finance this level of investment (Mohanty et al., 2007). Slums Nearly 1 billion people, or about one-third of the urban population in developing countries, live in slums. Freire (2013) estimates that the number of slum dwellers may double over the next two decades as rapid urbanization takes place in sub-Saharan and South Asian countries and formal markets continue to fail to serve the market for low-income shelter. In Africa, for instance, the population living in urban slums more than doubled from 1990 to 2014. Collier and Venables (2015, p. 415) characterize the typical housing as a low-cost shack, probably self-built and self-financed, and located in a shantytown. Although slums in developing countries are dysfunctional in many ways, one of their most common features is that the public sector usually provides minimal services to slum dwellers. Nonetheless, not all who live there are at the bottom of the income pyramid: one estimate is that at least a quarter of them have sufficient income to afford more decent (low-cost) housing (Baker and McClain, 2009). There is more to the problem of urban slums than simple poverty, as Freire (2013) discusses. The cost of providing shelter to slum dwellers varies greatly from place to place and is not easily estimated. The UN Millennium Project estimated that, from 2005 to 2020, upgrading 100 million slum dwellers would cost $67 billion over the 15 years. Expanding this to encompass the current 1 billion slum dwellers would require about $60 billion a year, or six times the current amount being spent on slum upgrading. At present, for

M4439-BAHL_9781786435293_t.indd 364

01/03/2018 11:08



Financing metropolitan areas ­365

example, local governments in Nairobi (Kenya) spend only about US$7 per capita in total, and in Lagos (Nigeria) the equivalent figure is about US$2, compared to Habitat’s estimated unit cost of US$500 for slum upgrading alone (Freire, 2013). Slum improvement focuses on three activities: investment in infrastructure and public service amenities, improvement of shelter, and security. Although the two latter activities must in many ways be led by higher-level governments, the security (of persons and to some extent property rights) that is essential to improvement requires very local and ‘hands-on’ action not just at the metropolitan but also at the neighborhood level. Similarly, the improvement of local public services in slum areas inevitably requires close involvement of both metropolitan and local governments even though the cost will almost always be well beyond the limits of their present access to finance. In India, for example, over half of the population of Mumbai lives in slums, with little access to water or sanitation let alone health and education. In urban India, just 78 percent of slum dwellers use tap water; only 37 percent use communal toilet facilities; 24 percent walked 0.2–0.5 km to latrine facilities, and there is little by way of solid waste disposal; and only 84 percent of slums had approach roads that would service motor vehicles (Bandyopadhyay and Rao, 2009). In most slums, residents have very limited property rights (de Soto, 1990, 2000) and usually receive little by way of services from governments at any level. There is much to be done and little fiscal capacity to do it. As always, when it comes to making the local public sector work better in developing countries, money – and the lack of it – matters. Even if one assumes that those in charge of the largest and richest cities in most low- and middleincome countries want to focus on improving the lives of their poorer residents, the major financial problems such cities face simply to keep functioning at a minimal level make it difficult for them to do much.

FINANCING METROPOLITAN GOVERNMENTS In some ways, the poverty of metropolitan governments is a bit strange. As economic activity migrates to urban areas and infrastructure expands, so does the revenue capacity of metropolitan areas. More money should thus be flowing into local budgets to meet the growing need for local public expenditure. Much more. However, in most low- and middle-income countries local governments cannot access such productive tax bases as consumption and income. Moreover, few have done as much as they could and should to exploit the revenue sources available to them, such as property taxes and user charges.

M4439-BAHL_9781786435293_t.indd 365

01/03/2018 11:08

366

Fiscal decentralization and local finance in developing countries

One reason for both problems may again lie in the governance structure. If the metropolitan area is fragmented and served by many local jurisdictions, broad-based sales or income taxes may not work well because shoppers and workers are mobile; and even when they are not, businesses are, so that exporting tax burdens (and thus breaking the accountability connection between those who pay and those who benefit) is all too simple. Similar problems may occur with business licenses and property taxes to some extent. Property taxes, licenses, user charges and intergovernmental transfers tend to be the main revenues of local governments, even in some of the very biggest metropolitan areas. When the principal local government covers all or most of the metropolitan area, however, broad-based taxes on production, consumption or payrolls are more technically feasible and less economically distorting – provided that higher-level governments permit their imposition. When they do have access to such broader tax bases, big cities can raise substantial revenue. Indeed, as we discuss below, they may well do so even while they fail to exploit more economically efficient (but politically often more costly) revenue sources such as the property tax. But even when cities have potentially productive taxes they are unlikely to exploit them fully if they are able instead to squeeze more transfers out of the central government. It is always easier for local officials to please their constituents by securing increased transfers from above than by subjecting them to increased local taxes or user charges. Consumption and Production Taxes Most consumption and production taxes imposed by big cities in developing countries are gross receipts (turnover) taxes, levied at the point of sale (unlike the taxes on specific activities noted in Box 8.3). Economists are not keen on gross receipts taxes because they distort private consumption decisions by increasing taxes more on goods with longer supply chains (often referred to as ‘cascading’); private investment decisions by favoring vertically integrated firms; the accountability of local governments by permitting tax burdens to be exported; and budgetary planning owing to such administrative (and political) problems as allocating revenues when a headquarters (HQ) located in one city pays taxes for branches operating in other cities. Local politicians may, of course, see most of these defects as virtues: tax cascading increases the tax base; tax exporting reduces the burden on local taxpayers; and nobody apart from the occasional economist seems to worry much about such matters as the distorting effects of increasing vertical integration and the messy HQ problem.

M4439-BAHL_9781786435293_t.indd 366

01/03/2018 11:08



Financing metropolitan areas ­367

BOX 8.3  SELECTIVE LOCAL TAXES Many countries impose selective sales taxes on certain services and license fees on certain types of businesses. Many (e.g. billboards, minor license fees for barbers, etc.) seldom produce much revenue and serve little regulatory purpose either, so we discuss only a few of the more important such levies here.* Electricity and Telecommunications Metropolitan local governments in Istanbul, Delhi and Jakarta tax electricity bills. Others have at times imposed similar taxes on telecommunications services (such as telephone bills, TV antennas, etc.). Such taxes are usually simple to administer if added to an existing utility bill and probably progressive in their incidence, especially if applied only to residential consumers. However, to the extent they apply to business use also, as they often do, such discriminatory service taxes may not only discourage investment in general but also the adoption of new technology in ­particular – something which Roller and Waverman (2001) demonstrated was often a key factor in the development of low- and middle-income countries. Electricity and phone services are often good proxies for the use of local public services, have significant revenue potential and are characterized by low price elasticity of demand (Martinez-Vazquez, 2013); but whether they should be singled out for special taxation is an issue requiring close examination. Entertainment, Hotels, Restaurants Most countries have local taxes on these activities both because they are considered ‘luxury’ (inessential, perhaps even ‘bad’) expenditures and because they are often seen mainly as taxes on visitors or non-residents rather than on local people. The incidence of such taxes is likely progressive, but their administration is often costly. A general sales tax covering services, like a value-added tax (VAT), should of course apply to all such services; but it is a quite different question whether the revenues received from selective taxes are worth the administrative and compliance costs involved. A striking feature about such levies is that no one seems to have looked closely at their net fiscal or economic impact. Often, what seems to have happened is that at some point someone in authority thought it was a good idea to tax this or that activity, so a tax was imposed. And there such taxes tend to stay, whether or not they still make sense (or ever did). Betting and Gambling Jakarta once had a tax on gambling that produced so much local revenue that the central government noticed and took it over as its own. Not all such local taxes (and often licenses) on betting and gambling activities have been as successful in producing revenue; but many – like the numerous regional lotteries in Colombia  – have been around for decades and continue to pull in customers and produce some revenue for subnational activities. Gambling is addictive for some and is not

M4439-BAHL_9781786435293_t.indd 367

01/03/2018 11:08

368

Fiscal decentralization and local finance in developing countries

confined to the wealthy, so such taxation may sometimes be surprisingly regressive in its impact. Moreover, in part because much of the proceeds must be repaid to keep the customers coming back, gambling taxes and levies can be costly to administer. Still, since the considerable amount of money often goes on gambling and such taxes tend to be accepted by most, local governments are likely to continue to tap into this revenue source when they can. Note:  * For further discussion, see e.g. Bird (1991) as well as Corthay (2009).

What matters most to local decision-makers is that the gross receipts tax not only produces a lot of revenue at a low rate but is relatively cheap to administer on a metropolitan basis, which avoids the shifting of tax burdens across jurisdiction lines and the increases in disparities that would arise with purely municipal government administration. A gross receipts tax is not only demonstrably more acceptable to voters than a property tax; it is also inherently much more income-elastic, yielding more revenue as the level of economic activity (and/or prices) rises. Whether the efficiency gains from the additional local public services that can be financed with such a technically bad tax, perhaps combined with some gain in accountability from the greater fiscal autonomy resulting from having access to a more elastic tax base, can offset the welfare costs imposed by a distortive tax is an interesting question to which no one as yet has an answer. One way of avoiding the efficiency costs of gross receipts taxes is to fold such a local gross receipts tax into the (considerably less distorting) national value added tax (VAT), preferably while leaving some control over the local tax rate to the local government (Bird and Slack, 2013; Fretes Cibils and Ter-Minassian, 2015). China recently did this with its local business tax (though not with local rate autonomy), and India has discussed doing much the same with certain state and local sales taxes. Such reforms are never easy, given the obvious transition costs and the fact that some localities (or provinces) would lose revenue with a straight swap at a national average rate and will likely require some subsidy to do so. But they are always worth considering carefully in countries like those mentioned in Box 8.4.26 Income and Payroll Taxes Although income and payroll taxes are widely used by subnational governments in OECD countries, they are rarely used by local governments in developing countries. One reason may simply be the limited reach of even national income taxes in many developing countries (Bird and Zolt, 2005),

M4439-BAHL_9781786435293_t.indd 368

01/03/2018 11:08



Financing metropolitan areas ­369

BOX 8.4 METROPOLITAN LOCAL GOVERNMENT SALES TAXES Argentina Argentina’s provinces, including the metropolitan city of Buenos Aires (which has provincial status), imposes a turnover tax on total sales revenues and has discretion to change the rate or base of the tax. Buenos Aires city derives about 70 percent of its tax revenues from the turnover tax; but with the dominance of the metropolitan area in the country and about two-thirds of the metropolitan population living in the (separate) province of Buenos Aires, much of the tax burden is no doubt shifted to non-city residents – although the lower rate applied to manufacturing reduces the importance of this factor. As Artana et al. (2015) discuss, replacing this tax with a local tax (with locally established rates) piggybacked on the national VAT would make good economic sense, certainly in the metropolitan area. The importance of the financial sector in the city’s tax base, and the revenue uncertainty inherent in any transition, has so far deterred any reform. Brazil Brazil’s local service activity tax (ISS) is imposed on all services except communications and intercity transportation at rates that vary between 2 and 5 percent of total revenues, most of which is collected by the largest cities (Rezende and Garson, 2006). This tax raises about twice as much revenue as the local property tax. Colombia The municipal industry and commerce tax accounts for about 40 percent of local tax revenues in Colombia, with most being collected by the larger cities. Within certain limits, local governments can determine both the tax rate and the tax base (which in practice is usually some estimate of annual gross receipts). Rates vary widely among different industries and activities. Bogotá, with a population of over 10 million, in general levies higher rates – perhaps exploiting ‘agglomeration rents’ to some extent – and reaps much more in revenue from this source than anywhere else in Colombia (Fretes Ciblis and Ter Minassian, 2015). The tax is popular with both people and politicians, but acts as a barrier discouraging business from transitioning to the formal sector of the economy (Bird, 2012a). India The Mumbai Municipal Corporation until recently raised about half of its ownsource revenues from the octroi, a tax on goods entering the city which is collected at special entry stations. This tax, which has a complicated rate schedule and is imposed in a cumbersome way, has long been criticized for imposing heavy compliance and administrative costs, distorting the allocation of resources and opening the door to significant corruption. Although such taxes were abolished in Pakistan

M4439-BAHL_9781786435293_t.indd 369

01/03/2018 11:08

370

Fiscal decentralization and local finance in developing countries

over a decade ago – and also in India except in Maharashtra state, where Mumbai is located – the major barrier to reform in Mumbai was reportedly that the size of the grant required to replace its revenues “would be of unimaginable magnitude” (Pethe, 2013, p. 253). The national and state goods and services taxes, or GSTs (VATs) currently being put in place in India will supplant the octroi. Philippines The 17 cities and municipalities in metropolitan Manila impose a business tax on total sales at the point of sale, which puts them in competition with one another for tax base, distorts business decisions and may result in significant inequities across local governments. In 2008, the variation in per capita business tax revenue across local governments in the metropolitan region was from US$5 to US$169 (Nasehi and Rangwala, 2011), and business tax revenues on average finance nearly 40 percent of total local government expenditures. South Africa South Africa had a local tax imposed partly on sales and partly on payrolls – the regional service council levy (RSC) – that raised a significant share of local revenue. Like many local business taxes, the RSC was seriously flawed in terms of its administration (Bahl and Solomon, 2003). It was abolished in 2006. Although it was replaced by a compensating grant, revenues from the grant have not proved to be as income-elastic as the former RSC revenues, and local fiscal autonomy has been substantially reduced. China In China, as in some transitional countries in Eastern Europe, a share of the national VAT was assigned to regional governments, with the shares being calculated (inevitably somewhat arbitrarily) on a derivation basis. Because subnational governments have no authority to change the rate or base of the tax, this revenue sharing is more like a transfer than a local tax. Such arrangements may perhaps sometimes be considered a sort of step toward subnational government taxation because the basic infrastructure of the administration (tax roll assessment, collection) is in place and because people are used to paying such taxes. In addition to a share of the national VAT, local governments in China were permitted to impose a local business tax on a wide range of service activities. Although both the rate and base of this tax were set centrally, it produced about 30 percent of subnational government tax revenues. In 2014, the central government began to fold this local tax into the central VAT, compensating the revenue loss by raising the provincial share of national VAT collections to 50 percent (Bahl et al., 2014). Some local leaders, however, think the previous earmarked business tax was a more protected transfer than the new shared VAT, which will presumably be affected by future central government tax structure reforms. Good tax policy and good intergovernmental fiscal policy do not always march hand in hand.

M4439-BAHL_9781786435293_t.indd 370

01/03/2018 11:08



Financing metropolitan areas ­371

which means there is not much base to reach. Paradoxically, the longerterm revenue potential of income taxes may be another reason central governments normally want to keep their hands on this tax base. Similarly, central governments often want to reserve the payroll tax base, or part of it, to finance social insurance programs (Bird and Smart, 2014). The Federal District of Mexico and the two other states that overlap metropolitan Mexico City all levy a tax on payrolls.27 They are free to choose the tax rate and base and to administer the tax. Although the tax is not without flaws – it can discourage formal employment (Bird, 1982) and is flawed because it is based on place of work, and hence not clearly related to financing of public services where workers live – it is relatively simple to administer even at the local level. However, its revenue potential is not realized to any great extent in Mexico, where it yields the equivalent of only about 0.3 percent of GDP, although it proved a relatively stable revenue source even in the recession of 2009 (Pineda et al., 2015). Another interesting case is the piggyback income tax imposed to cover municipal costs in some large cities in Eastern Europe.28 Such ‘local’ income taxes, which in practice are essentially payroll taxes, are usually imposed as central taxes with centrally determined rates but are often shared, as in Bucharest (Romania), on a derivation basis (the local share of tax revenue stays where it is collected). In other cases, like Zagreb (Croatia), the city government is empowered to levy a surtax on the central income tax. Unlike the Mexican case mentioned above, however, local revenues from such sources fell sharply in many Eastern European cities during the 2009 recession. Where the metropolitan government area covers most of the relevant labor market, a payroll tax may provide an economically and politically acceptable source of local revenue, especially if imposed as a surcharge on a central income or payroll tax. Such taxes have been successful, for example, in Nordic countries where the tax base is broad and the tax administration is good (Lotz, 2006). In most developing countries, however, the income tax has a narrow base and is not well administered. Moreover, the existence of a substantial informal workforce that is outside the scope of the tax causes problems. Taxes on consumption or production – the general sales (and gross receipts) taxes discussed above and the selective sales taxes discussed below – seem more likely to be both more feasible and productive sources of metropolitan revenue in low- and middle-income countries and less likely to be economically distorting than either income or payroll taxes. Property and Land Taxes Rapidly rising land values generally accompany rapid urbanization, and form an especially attractive potential revenue source for metropolitan

M4439-BAHL_9781786435293_t.indd 371

01/03/2018 11:08

372

Fiscal decentralization and local finance in developing countries

governments in many developing countries. The earlier discussion in Box  6.6 outlines a number of ways in which countries have attempted to secure at least some of the increased values, especially to help finance needed investment in new infrastructure. We shall not repeat this discussion here, but it is worth underlining the basic argument of Chapter 6 that, from many perspectives, the property tax is an ideal choice for financing local government services, including those in metropolitan areas. The tax has (or at least should have) a broad base and can generate significant revenue at low nominal rates. Particularly in metropolitan areas, the capacity of local governments has improved significantly in many countries, making the implementation of a modern property tax both feasible and a high priority. It is not too difficult to design and implement this tax in a way that reduces the extent to which it burdens the poor; and, given the impact of improved local services on property values, property owners should be increasingly willing to pay as their property values are lifted by the combination of the rising urban tide and sensible public investment in infrastructure. In addition to producing revenue for local government, a good property tax is not only economically efficient but also can and should promote better use of land in metropolitan areas.29 In practice, however, the property tax is all too often a relatively minor source of revenue even in the largest and fastest-growing cities of most developing countries. A survey of 30 large metropolitan areas carried out by McCluskey and Franzsen (2013) tells us two things. First, as one might expect, most property taxes are collected in metropolitan areas. For example, metropolitan Manila accounts for 20 percent of the Philippine population but for nearly half of all property tax collections. Second, the revenue performance of the property tax varies widely from city to city, with some showing growth and others experiencing real per capita decline. As usual, when it comes to comparing local fiscal issues across countries there is no single or simple story to be told. Investment in new infrastructure in urban areas – in transport, drainage, water supply and the like – certainly stimulates property value increases and may reduce resistance to paying increased taxes. As noted earlier, close to 5 percent of GDP apparently needs to be devoted to urban infrastructure spending in low- and middle-income countries (Ingram et al., 2013). If this investment were made, and only 20 percent of it captured in increased property taxes, revenue from this source would be more than doubled. The mixed performance of property taxes in recent years is thus disappointing, although not surprising given the apparently inherent unpopularity of the tax. An additional reason for the relative lack of emphasis on property and land taxes may be that some large cities – like Buenos Aires, São Paulo and Bogotá in Latin America – have access to broad-based sales taxes that are

M4439-BAHL_9781786435293_t.indd 372

01/03/2018 11:08



Financing metropolitan areas ­373

more productive, more elastic, more easily administered and less politically difficult than the property tax. In other cases, as in Mexico City, metropolitan governments have been able to avoid facing the political struggle of collecting more in property tax, largely because they are financed by grants from higher-level governments – the same governments that make it difficult for local governments to impose productive property taxes by constraining the extent to which they control the base, rates and sometimes even the administration of the tax. Whether hampered from above or not, weak administration is another factor that limits property tax revenue in many developing countries. Often, cities fail to tax many properties at all. In Delhi (India) for example, only 38 percent of eligible properties are on the tax roll (Mathur et al., 2009).30 Even when properties are assessed for tax purposes, values are often grossly underassessed and seldom adjusted even when real estate values are increasing rapidly. Even when they are assessed, too often the taxes are not fully collected. All four of the legs of the property tax discussed in Chapter 6 – identifying the base, valuing it, assessing the tax and collecting it – are thus frequently too weak to produce much revenue even in the largest metropolitan areas of many developing countries. Nonetheless, despite its many problems sometimes the property tax does yield significant revenues, as in large South African cities. For other metropolitan areas to follow the same path – as economic arguments, revenue needs and the better functioning land markets developing in many countries suggest they should – countries might consider the following four-point reform strategy: 1. Metropolitan local governments should be able to set property tax rates and administer the tax. As experience in countries as diverse as Indonesia and Colombia suggests, greater local fiscal autonomy improves the prospects for a more productive property tax. 2. If governance in the metropolitan area is fragmented, to be successful the property tax should have a uniform base throughout the region, and probably be administered at the metropolitan level to capture economies of scale in property tax administration and adopt the new technologies that are now available. Often, as discussed in Chapter 6, property taxes are now administered by local governments in jurisdictionally fragmented metropolitan areas. But local governments could still set their own tax rates under a metropolitan tax administration, which could even be owned jointly by the local governments and financed out of the revenues collected.31 3. Ideally, property transfer taxes should be merged with the property tax and administered at the metropolitan level (or possibly at the state

M4439-BAHL_9781786435293_t.indd 373

01/03/2018 11:08

374

Fiscal decentralization and local finance in developing countries

level, depending on how subnational governments are structured). Doing so would provide both better scrutiny of property transfer declarations and evidence that could be used for better valuations. Other taxes related to land and property, such as the land value capture approaches, should also be the responsibility of the unified administration. 4. Finally, as we set out in Chapters 5 and 7, a hard budget constraint should be imposed on metropolitan local governments, with intergovernmental transfers being limited to those needed to address national priorities, and perhaps also to provide the desired degree of fiscal equalization between local governments in the metropolitan areas as well as between the metropolitan areas and the rest of the country. Taxing Motor Vehicles There is a strong case for subnational taxation of motor vehicles (Bahl and Linn, 1992; Bird, 2010). This case is especially strong in metropolitan areas. Roads are expensive to build and maintain. In most developing countries, the number of motor vehicles has been growing faster than the population and the roadway infrastructure. Motor vehicle ownership is a sign of prosperity in developing countries, and vehicles and their use are easily taxed. Driving generates negative externalities and costs – traffic management, parking, accidents, pollution, congestion. Such costs differ from place to place but are usually highest in metropolitan areas: in part reflecting choices people make about where they live and work and how they get around; in part choices businesses make about where they will locate; and in part choices governments at different levels make about the kind of public transport network they are willing to provide, and how they deal with housing and land markets. Almost nowhere, however, are residents or businesses asked to pay for the social costs they generate when they use motor vehicles. They should pay these costs. Developing and implementing a family of taxes on motor vehicle ownership and use that will improve resource allocation, raise significant revenue and be administratively feasible has not been taken seriously enough by most metropolitan areas, even though they are hampered in doing so by the actions of higher-level governments (Bahl and Linn, 1992; Bird 2005; Martinez-Vazquez, 2013). Vehicles and their use can be taxed in many ways. Automobile ownership can be taxed through registration fees, personal property taxes and sales taxes; automobile usage can be taxed by taxing fuel consumption and license fees; road usage can be affected by tolls, parking taxes and restricted use licenses. In most developing countries, however, where fuel taxes and sales taxes are

M4439-BAHL_9781786435293_t.indd 374

01/03/2018 11:08



Financing metropolitan areas ­375

generally imposed by higher-level governments, the main subnational tax (license fee) on motor vehicles is usually that on vehicle registration. One reason most fuel taxes are imposed by central governments is that they have considerable revenue potential. Revenues from this source are usually income elastic, increasing with prices (if levied on an ad valorem basis) and with the growth of the number of vehicles. Fuel use is obviously related both to road use and to such external effects of vehicle use as accidents, pollution and congestion. One survey estimated the external cost of congestion to be 2–3 percent of GDP and the external costs of vehicle accidents to be 1–2 percent of GDP (Ley and Boccardo, 2010). But these relationships are usually too complex to capture in any precise way with a simple tax on fuel (Newbery and Santos, 1999). However, ‘precise’ is a word that can rarely be attached to subnational taxation in low- and middle-income countries, and these external costs are too large to be ignored. What can be done relatively effectively and efficiently to reduce such costs should be done. Since these costs are likely to be highest in metropolitan areas, which also incur substantial costs for road investment and maintenance, there is a good case for a locally imposed tax on motor fuels in addition to any national tax. If a metropolitan government (or province or special district) encompasses all or most of the metropolitan region, it can probably successfully administer a locally determined surcharge on a central fuel tax at the distribution level (the pump) without too much leakage. Alternatively, since central fuel taxes are generally imposed at the refinery or wholesale level (and then paid by the collecting agent to the province to which the shipment is destined), it should also be possible to administer such a differentiated tax at that level, with the refiner or wholesaler acting as a collection agent for the states/provinces and remitting taxes in accordance with the destination of fuel shipments.32 One reason such ideas have been little discussed is perhaps because higher-level governments do not want local governments to encroach on this important tax base. Another is that fuel prices are a sensitive political issue in many countries, central governments already have enough headaches from fluctuations in world market prices. In some cases, countries have substantial fuel subsidies, which are not only usually inadvisable (McLure, 2014) but complicate the taxation of motor fuels. Fear of political resistance is no doubt an important reason few metropolitan governments seem to have seriously pursued the prospect of locally differentiated fuel taxes. Nonetheless, some examples of subnational fuel taxes may be found. For example, in Brazil, states impose a differentially higher VAT rate on motor fuels, as do Colombian cities more indirectly by applying a higher gross receipts tax rate on motor fuels.

M4439-BAHL_9781786435293_t.indd 375

01/03/2018 11:08

376

Fiscal decentralization and local finance in developing countries

In contrast to the few cities in the business of taxing fuel, many more impose charges for motor vehicle registration and licensing, although few seem to get anything near the potential revenue yield of such charges (Martinez-Vazquez, 2011a). There are two general approaches to levying registration taxes (or charges). One is by imposing an annual tax (sometimes called a ‘personal property tax’) based on the depreciated value of the motor vehicle. Such taxes often, as in Argentine provinces, impose higher rates on higher-valued vehicles, usually for distributive reasons. Of course, the value of an automobile is not usually correlated to its fuel consumption or its undesirable carbon emissions. The lower taxes imposed on commercial vehicles are even more difficult to justify on any grounds (Artana et al., 2015). The other approach to taxing registration is to vary the tax rate with such features as the age and engine size of a vehicle (older and larger cars generally contribute more to pollution), the registered location of the vehicle (cars in cities add more to pollution and congestion) and axle weight (heavier vehicles do exponentially more damage to roads and require roads that are costlier to build) (Bird and Slack, 2013). The economic effects of even the best-designed annual license fee are limited. Annual charges may perhaps have some effect on the decision of whether to own a motor vehicle or on the type of vehicle to buy, but at the levels common in most big cities in developing countries, such effects seem unlikely to be significant. More importantly, once a vehicle is owned, special taxes on its purchase and annual license fees are fixed costs and unlikely to affect road usage much, if at all. Finally, it is not always easy to administer even something as simple as an annual vehicle license fee. Politicians in developing countries often tread lightly when dealing with those prosperous enough to have vehicles – a group likely to include their families and friends. A potentially promising aspect of motor vehicle registration in countries with better administration is the possible use of a system to ration road use in congested urban areas. Singapore’s pioneering program with a restricted license based on congestion levels and peak hour commuting patterns is well known, though no other developing country (and few developed ones) has since emulated it.33 A less targeted but simpler approach may be to use the license system to limit the number of motor vehicles on the road. In China, for example, some metropolitan municipalities have set a cap on vehicle registrations and established a quota for newly registered license plates. Some countries in Latin America allow only cars with certain licenses (usually ending in an odd or even number) on the road on certain days. Although such crude measures are not targeted to particularly congested areas or to rush hours in particular – and the rich can get around them by buying another car – this approach not only appears

M4439-BAHL_9781786435293_t.indd 376

01/03/2018 11:08



Financing metropolitan areas ­377

to be ­politically acceptable but also appears to have had a small effect on congestion (Rivasplata, 2012). A more effective (if perhaps counter-intuitive) way to deal with congestion may sometimes be to charge more for parking. Studies in some cities show that 20 percent or more of traffic in congested urban areas arises from people looking for free parking spaces. As every driver knows, one person double parking for a few minutes (or trying to squeeze into a parking space) can create a major traffic snarl in a congested area. Charging properly for parking involves not merely enforcing parking meters and parking rules but also designing and implementing an appropriate system of parking regulations in the first place, including the amount and type of parking spaces provided by businesses and residents (Barter, 2010). Enforcing such rules is not costless and certainly not popular, perhaps especially with small businesses (such as restaurants) that depend to a large extent on informal parking ‘directors’ who work the street in front for gratuities and make the most of every inch of available space. In Indian and Pakistani cities, for example, such people can and do park cars even more tightly than in the stacked automated parking racks found in some European city centers. However, with the right set of rules governing private parking (including taxing it) and good rules and acceptable levels of enforcement on the street, parking could become a source not only of at least some net public revenue but also a partial solution to rather than a contributor to urban congestion (Bird and Slack, 2013). At present, however, most developing metropolitan cities are as far away from this ideal as they are from congestion or road pricing.34 The best way to price roads, both to pay for them and to reduce congestion, is of course to do so directly. Electronically monitored road access to congested areas or road tolls (for example, to access airports, or tunnels and bridges) are technically feasible, used in some developed countries, and can both improve travel times and provide funds to cover the cost of borrowing to build such capital facilities in the first place. While such systems are expensive to install and maintain, and perhaps a step too far into the future for many low-income countries at present, even in such countries large cities should begin to consider this option. Constant gridlock is one way to slow down the growth of big cities. But this result is tied to a price that is then paid by reducing economic growth, so a better solution is needed. One approach worth exploring may be to skip the century or so of painful learning during which developed countries have for the most part ignored this problem and begin exploring the introduction of a real road-pricing system along the lines first set out in theory by Vickrey (1963) but now technologically practical.35 Another approach, less desirable but perhaps more likely to be adopted, is to attempt to deal with the problem

M4439-BAHL_9781786435293_t.indd 377

01/03/2018 11:08

378

Fiscal decentralization and local finance in developing countries

through increasingly rigid (and almost certainly largely ineffective) restrictive regulations.36 User Charges Economists like user charges – provided of course that they are well designed and implemented. As with prices in general, user charges should be set close to full cost recovery levels, thus improving the efficiency of service delivery (because people will only pay for what they want and because by doing so they will inform public suppliers about what and how much they want), and putting less strain on local taxes (for example, by helping finance infrastructure investment, and hence reducing substantially the claim of cities on intergovernmental transfers). However, because most people seem to like user charges about as little as they like taxes, political leaders who are usually reluctant to face their constituents with new taxes are equally dubious about the idea of charging them for what they get. Even residents in metropolitan areas who are accustomed to paying not-very-transparent indirect taxes for not-very-good services may balk at receiving an increased direct charge for something they likely think they either deserve to get for nothing or do not think is worth what it costs. Despite the widespread reluctance to bite the user charge bullet, as we noted in Chapter 5 many services typically provided by metropolitan area governments are amenable to pricing, e.g., water and sewerage, electricity, mass transit, road use and much more. Many cities do impose charges on some of these activities either as direct user fees (water, electricity) or an earmarked ‘benefit’ tax for services like garbage collection and solid waste disposal, or special assessments to cover the cost of new public investments. In developed countries like the US and Switzerland, user charge revenues constitute a significant share of local revenues (Bird and Slack, 2017). But even in these countries, charges are seldom imposed either at levels sufficient to cover full costs or in the form of the ‘marginal cost prices’ required to achieve the most efficient use of scarce public funds.37 Many reasons are offered for the failure of local public governments around most of the world to make good use of good charges even though they are often the best potential way to finance many local public services in terms of both economic efficiency and political accountability. Sometimes, users cannot be identified. Sometimes, the cost of imposing user charges is too high. But the most common reason offered for not charging properly is because full-cost pricing (like market-based pricing in general) is inequitably regressive – a result most think is particularly undesirable when it comes to public services. The fact that the public sector considers something sufficiently important to take charge of its provision is even

M4439-BAHL_9781786435293_t.indd 378

01/03/2018 11:08



Financing metropolitan areas ­379

assumed by some to imply that in principle everyone should have access to such services on an equal (which almost invariably means subsidized) basis. Although better ways are usually available to protect poor families than by distorting public prices, the view that everyone is entitled to equal access to public services seems difficult to overcome. The fact that someone must pay is simply ignored by many who assert that even congestible government services provided to identifiable beneficiaries are an entitlement that should not be paid for with cost-recovery prices. These arguments ignore considerable evidence that imposing the costs of such services on those who do not benefit and subsidizing those who do is a proven way to ensure that there is less for all to share. For example, since rich families consume much more water than poor families, subsidizing water usage is usually regressive (e.g. Katzman, 1978, on Malaysia). The poor can only benefit from subsidies when they have access to the subsidized service, and some studies have found that subsidizing usage sometimes make utilities more dependent on ‘connection fees’ for revenues, thus encouraging them to expand services to those who can pay such fees (e.g. those buying new houses in urban sprawl areas) rather than to the poor who are those supposed to benefit from underpricing water.38 As many developing countries have learned, failing to charge users the right prices all too often means that services are simply not extended to large segments of the population, with the poor (as always) the major losers. Poor pricing also distorts urban development patterns (Slack, 2002) and wastes resources. Intergovernmental Transfers Metropolitan governments are usually the richest and largest local governments in any country. An appropriate financial target for them is to ensure that, if properly managed, the resources over which they have control are sufficient to pay for the public services they choose to provide to residents. Higher-level governments should be responsible for financing services that they mandate local governments to provide. Non-residents should pay for benefits they may receive from local services. There are many ways they can do so: through some broad-based taxes on, for example, payrolls or sales (catching commuters and shoppers); intergovernmental contractual arrangements (for instance when suburban residents attend central city schools or use central health facilities) or appropriate user charges (perhaps for transit services or utilities provided by the metropolitan authority; and perhaps even access to and use of congested sectors of the road system). Establishing this kind of self-sufficiency target would be difficult; but thinking it through at the least provides a useful guide to how

M4439-BAHL_9781786435293_t.indd 379

01/03/2018 11:08

380

Fiscal decentralization and local finance in developing countries

to shape urban public sector finance in area-wide metropolitan government structures as exist for example in Bangkok or Shanghai. However, such a standard also could be applied to such fragmented metropolitan local government structures as Manila or Kolkata only if key aspects were dealt with at the regional rather than local level (e.g. through a two-tier structure). If this were not done, requiring such self-sufficiency at the local level would likely exacerbate fiscal disparities within the metropolitan area. Though little hard evidence is available, local governments in metropolitan areas do generally fund more of their budgets from own sources than other local governments. Shah (2013) found that on average in a sample of 17 metropolitan areas that intergovernmental transfers accounted for less than half (42 percent) of their total revenues. This average does not mean much because, in addition to the usual problems in comparing figures across countries, the range within the sample was great, from less than 10 percent in Addis Ababa and Pretoria to over 75 percent in Istanbul. Even if the numbers were solid, however, grant dependence is not necessarily a good way to measure fiscal self-sufficiency. One city might rely little on grants but not be responsible for much in the way of services, while another might rely more heavily on grants but also be responsible for roads, primary education and local health services. Since power does not automatically come accompanied by money, the second city mentioned may in fact have so little control over how the grant money is spent that the funds nominally flowing into its coffers from transfers are in fact central spending in disguise.39 When it comes to interpreting local fiscal figures, everything is in the details. A few examples illustrate the range of variation. São Paulo provides a broad range of functions to 20 million people and finances 44 percent of this amount from own-source revenues (Wetzel, 2013). In Delhi, transfers account for about 36 per cent of total revenues (Bandyyopadhyay and Rao, 2009). In Johannesburg (where the main functions are water, sanitation, waste removal, solid waste and electricity), intergovernmental transfers account for only 24 percent of revenues as compared to 48 percent from user charges, with the balance from local taxes (Steytler, 2013). On the other hand, in Cairo and Mexico City almost all revenues come from intergovernmental transfers. Most developing countries do not have a special regime for large urban areas. They treat metropolitan local governments the same way as they treat any local government (Shah, 2013). There are some exceptions. Some countries provide additional resources to accommodate special needs in metropolitan areas – for example, national capital districts (Slack and Chattopadhyay, 2009) – or give them special ‘regional’ status as in Mexico City and Bogotá.40 Others exclude the largest cities from certain

M4439-BAHL_9781786435293_t.indd 380

01/03/2018 11:08



Financing metropolitan areas ­381

grant programs. In Indonesia, for example, the Jakarta metropolitan area is excluded from both general revenue sharing and conditional grant programs on the grounds that it already has a fiscal surplus. At the same time, however, Jakarta receives a special share of national personal income tax revenues. Countries that share central government revenues on a derivation basis (i.e., return shared tax revenues to where they are collected) favor wealthier metropolitan local governments. In China, for instance, the four largest metropolitan governments (which also have provincial status) receive significantly larger transfers per capita (Bahl et al., 2014). Metropolitan Bangkok similarly receives significant revenues from centrally-determined surcharges on VAT and excises that are shared on an origin basis (Shah, 2012). As discussed in Chapter 7, India has used a separate transfer program for urban local governments – the Jawaharlal Nehru National Urban Renewal Mission (JNNURM) launched in 2005 (and ended in 2014). The primary objective of this transfer was to finance public infrastructure on a sustainable basis, so grants were earmarked for infrastructure and further conditioned by requiring certain reforms to improve urban governance.41 As Ahluwalia et al. (2014) note, the program was hampered by slow release of funds, cost over-runs, inadequate capacity to absorb grants at the local level, problems in monitoring the progress with urban management reforms and enforcing the conditionality, and the inability of state and local governments to back the JNNURM with their own financial resources. Nonetheless, not only have there been some infrastructure improvements, chiefly around water supply and drainage, but at least in some cases the program also appears to have made some Indian cities raise their sights in terms of what is possible. When grants are allocated by formula, how much big cities receive depends on the formula as well as the size of the program. When factors like urbanization rate and population size are weighted heavily in the formula, the outcome often favors metropolitan areas. They also tend to benefit when transfers are intended to provide an incentive for some local government action because they are more likely to have the resources to buy into the program. On the other hand, countries like Japan and Korea that distribute transfers with an equalization formula that finances all or some part of the gap between estimated expenditure needs and estimated revenue capacity (as discussed in Chapter 7) usually tend to distribute relatively more to poorer non-metropolitan governments. Significant interjurisdictional disparities within metropolitan areas are common. In metropolitan Manila, for instance, per capita local government expenditures range from US$292 per capita in rich Makati to US$40

M4439-BAHL_9781786435293_t.indd 381

01/03/2018 11:08

382

Fiscal decentralization and local finance in developing countries

per capita in Pateros city. However, since the variation in per capita revenue sharing is much less, the transfer system does little to correct the situation (Nasehi and Rangwala, 2011). In Istanbul, lower-level governments (regardless of their income level) are required to transfer 35 percent of their tax sharing revenues to the metropolitan municipality to compensate for services provided. In São Paulo, where the central city accounts for 56 percent of the population and 66 percent of all local government expenditure, it receives a larger share of its revenues from intergovernmental transfers than suburban municipalities (Wetzel, 2013). In contrast, the situation in the Santiago metropolitan area in Chile is different because of the redistributive impact of an inter-municipal transfer program (the Fondo Comun). As these varied examples suggest, countries can create any sort of regime they want with respect to intergovernmental transfers in general, within the metropolitan area, and between it and the rest of the country. At present, most metropolitan areas in developing countries are heavily dependent on intergovernmental transfers to finance local public services. This is seldom a healthy or sustainable situation. Metropolitan areas should finance most local public services with revenues that are raised from beneficiaries (i.e., from local taxes and user charges), with intergovernmental transfers being limited to those compensating for benefit spillovers and intended to alleviate such distributional concerns as upgrading slums. Moving to a more self-sufficient, locally financed system will take time. For local governments to make effective use of more fiscal autonomy, central governments must support and assist them in building up the skills and capacity needed to utilize their new fiscal powers sensibly (see Chapter 3). They also often need to restructure the transfer system to provide them the incentives to do so (see Chapter 7). Shifting toward explicit local taxation and charges and away from grants received from above (which are perceived by many as roughly equivalent to manna from heaven) is painful for local citizens. It is also hard for local politicians and officials, who must reorient their focus from squeezing money out of higher-level governments to persuading their fellow citizens to hand it over to the local government – a task that usually requires persuading people that the funds are well spent. Such drastic changes in institutions and attitudes do not take place quickly. Developing the needed legal and administrative structure (and culture) that may make them possible also takes time. It is difficult to imagine how such changes can come about without strong leaders who can sell the vision of a better future sufficiently well to build the supportive and sustainable coalition needed to see the process through. Such things are much easier for academics to say than for anyone to do. But the reality is that the present metropolitan finance structures in many countries do not provide sustainable support for continued development, and will in any

M4439-BAHL_9781786435293_t.indd 382

01/03/2018 11:08



Financing metropolitan areas ­383

case soon be swamped by continuing growth. At least some countries may soon choose (or be forced) to move in this direction. Borrowing As metropolitan areas gain more responsibility for providing infrastructure they must be empowered to borrow to finance capital improvements (Chapter 4). This is in principle not a problem, particularly if metropolitan governments are made more responsible for their own fiscal future as we have suggested they should be. However, problems arise if higher-level governments retain final responsibility for repaying local government loans. The terms under which a metropolitan government can access capital markets should largely be governed by its own characteristics: the strength of the local economic base; its rate of growth; the extent of local control over potential revenues sufficient to generate a safe margin of coverage over assigned expenditure responsibilities and other fixed commitments; sufficient fiscal discretion to be able to raise tax revenues and user charges without seeking permission from a higher-level government; and well-managed debt with no large amounts outstanding and no recent history of continuing fiscal deficits.42 Whether there is a general regime for local borrowing or a specific one for metropolitan governments only, the rules should be such that those governments that are not creditworthy along these lines cannot borrow. Again, it may well be the case that it would be best in many metropolitan areas, even those in which smaller local governments continue to be the main service providers, if most borrowing (particularly for infrastructure projects) was done at the metropolitan or regional level.43

HOW TO REFORM METROPOLITAN FINANCE The public finances of metropolitan areas are ripe for reform almost everywhere. The contribution of successful big cities to national economic growth has been widely celebrated. But the huge and growing gap in the level and quality of local governance in the big cities of the developing world endangers this success. One path to sustainable national economic development is for a nation’s cities, like its businesses, to gain a competitive position in regional and world and local markets. A key element to doing so is often to provide better public services at affordable tax rates. Developing a good metropolitan fiscal strategy should be a priority concern of governments wishing to improve the lives of the people for whom and to whom they are responsible.

M4439-BAHL_9781786435293_t.indd 383

01/03/2018 11:08

384

Fiscal decentralization and local finance in developing countries

Such sentiments are not new: World Bank (2009b), Glaeser and JoshiGhani (2015) and many others have stressed that not only should migration to cities not be discouraged but also that other barriers standing in the way of capturing agglomeration effects, such as regulatory costs ­hampering interregional and international trade and inadequate transportation networks, should also be eliminated. However, surprisingly little attention has been paid to how to make the increasing number of large cities in developing countries not only more manageable but also more effective economic engines for growth. Most analysts (if not yet all governing elites) now accept that the common twentieth-century perception that the way to deal with rapid urban growth is to discourage it was misconceived. A common policy was to provide fiscal and other incentives for firms that located outside metropolitan areas, often in regions selected for their political importance or on the basis of criteria that were more appealing to planners than to investors. Such policies were almost always dictated from above and almost never involved any degree of fiscal decentralization, let alone local participation. Other policies were to level slums and let those living there fend for themselves wherever they could, or even to move the national capital to a ‘green field’ site (Brasilia, Astana). None of these policies proved to be the answer to accommodating public service needs in large urban areas. Both the best and the most feasible ways forward – which are seldom the same – may differ considerably in different countries at different times. There is no single or simple answer to how best to mesh intergovernmental fiscal relations with urban or other development strategies. But there is much to be said for beginning reforms in this area by focusing on the big cities.44 Not only are these cities big and growing, they are also critical to the development of the whole country. Getting metropolitan government and finance right matters not only for those who live (or will live) in metropolitan areas but also to everyone else in the country. Another reason for a ‘metro first’ policy in local government reform is that, although metropolitan reforms are often complex to design and politically difficult to implement, they are also more likely to yield perceptible results for a given (technical and political) effort than a general attempt to reform all local governments. More opportunities for immediate and visible improvement often exist in big cities, and there of course many more local resources in both public and private sectors with which to work. A third reason for focusing on metropolitan reform is that the greater visibility of what goes on in the big city makes it in many ways the ideal forum in which to pilot new ideas and see if they can be made to work and actually produce beneficial changes. If they do, they can then be generalized. If they do not, something else can be tried that may be better. Other

M4439-BAHL_9781786435293_t.indd 384

01/03/2018 11:08



Financing metropolitan areas ­385

countries can learn much from China’s extensive experience in piloting institutional changes in one or a few subnational areas to make sure it works before rolling it out to the whole country. Of course, not only is every country different, but so is every city even within the same country. The first rule of metropolitan reform is thus that no one rule fits all cities. Success inevitably turns on choices made by decision-makers not only at the center but also in each city; and since no one city is likely to be exactly like any other in this respect, reforms need to be carefully tailored to local conditions.45 However, a second general rule can and should be applied to all cities: the rule of responsibility. If local governments in metropolitan areas (or elsewhere) are expected to play a positive role in development, they must as far as possible be made fully responsible for their own success, or failure. This means first that the hard budget constraint (discussed in Chapter 5) should definitely apply. Cities should be expected to finance local expenditures with locally raised revenues, in effect charging a tax price that covers the marginal cost of providing local benefit services (Bird and Slack, 2014; Bahl, 2013). To achieve this goal in most developing countries, most metropolitan governments require additional revenue-raising powers, even if one result may be to make central government efforts to increase its own revenue mobilization a bit more difficult. Similarly, central governments should not impose unfunded mandates or other controls on local government spending. If they use local governments as their agents in delivering such services as education and health, they must finance fully their share of the costs. In short, if central governments want their big cities to succeed – as they should in the interests of national development – they must often recognize that they will need perhaps to ease up a bit when it comes to achieving some of their own immediate ambitions, something that no one ever likes to do. Metropolitan fiscal reform must usually be led by reform at the central government level (or at the regional government level in some federations). Politicians in higher-level governments like to be free to make whatever changes they think are needed in intergovernmental fiscal relations, for example, by slashing transfers to deal with central deficits. However, for sustainable success in reforming metropolitan finance the central government may need to institutionalize the national debate on metropolitan finance and governance. To begin with, for example, a country might consider establishing a commission to study the feasibility of a special regime for metropolitan area finances, with the scope of the inquiry to include metropolitan governance, the assignment of expenditure responsibilities and revenue raising powers (taxation, user charges, borrowing), and of course the implications for the transfer system (Bahl, 2013). In addition, the central government should not only establish clear and detailed

M4439-BAHL_9781786435293_t.indd 385

01/03/2018 11:08

386

Fiscal decentralization and local finance in developing countries

­ rovisions for improved information, transparency and accountability at p all levels of government; it should also ensure that the necessary resources (including training) and institutional structures are set up to make the system work. Such structures – intergovernmental committees, reporting and approval systems and so on – need to incorporate key players at all levels of government sufficiently to ensure their understanding and acceptance of the future changes that will almost inevitably be required in a changing world.46 Most metropolitan areas now have some formal urban planning system, although it is almost never integrated with a fiscal plan or strategy, in part because few developing countries have a metropolitan fiscal strategy. But good plans do not guarantee reform, and we do not underestimate the difficulty of changing the intergovernmental fiscal system. In Brazil, for example, the possibility of implementing a workable metropolitan strategy is remote given the existing constitutional and legislative system. India did provide for at least the possibility of developing such a strategy in a major constitutional amendment in 1992, but little has been done owing to the huge inertia of the traditional approach to local governance and finance. Arguably, the post-apartheid reforms in South Africa are the best recent example of implementing a reform that recognized the special place of metropolitan areas in the fiscal system. But even here the system has not gone far enough, especially on the local revenue side. One reason for this general state of affairs is because few countries think there is a metropolitan finance problem they should worry about. Not only is the quality of local public services already much better in metropolitan areas than elsewhere, but also most big city governments do in one way or another manage to finance more of their spending from their own resources than do other local governments. Moreover, since the electorate in the city is usually more educated and active than in other areas, arguably even the accountability process may work reasonably well – or at least better than in the rest of the country. It is seldom clear to most central governments why they should spend scarce political capital on changing a system that seems to be working. Another reason the case for reforming metropolitan governance and finance is neglected is because higher-level governments are understandably reluctant to tamper with what is often their main cash cow. Metropolitan areas generate a significant share of the revenues flowing to the central (and state) governments, and the latter have little interest in diverting any of this revenue to be spent by a metropolitan area government that may become not only a revenue competitor but also a political competitor. National and regional politicians, like local politicians in fragmented metropolitan governance structures, are unlikely to be keen to reduce their

M4439-BAHL_9781786435293_t.indd 386

01/03/2018 11:08



Financing metropolitan areas ­387

own influence and to build up the power and status of potential rivals in a strong metropolitan government. Metropolitan fiscal reform (like fiscal decentralization in general) thus usually has many actual and potential enemies and few champions, as we discuss further in Chapter 9. The continued growth of urban populations and urban economies and the challenges of global competition may however weaken the constraints to reform in at least some low- and middle-income countries, leading them to consider more seriously the need for a more coherent metropolitan fiscal strategy. Some countries may become willing to consider establishing a distinct regime for spending, taxing and borrowing in large metropolitan areas.47 Financing and providing adequate public service levels to foster growth and satisfy the population is a task that often cannot be successfully accomplished within the existing jurisdictional boundaries and fiscal powers of many central cities. A new approach to deal with the governance and finance of the big cities is needed, one that we have argued should include more taxation and service delivery at the regional level, coupled with more focus on local accountability and revenue self-sufficiency. Metropolitan area governments need to be responsible for more than planning and land use regulation. Cities need broader responsibility and autonomy with respect to budgeting and service delivery, more taxing and borrowing powers, and a firmer political basis (e.g. an elected government). On the other hand, they also need less higher-level intervention in their operations through budgetary controls, mandates and conditionality. The other side of this coin is that they should be far less dependent on transfers (or subsidized borrowing) for support than other local governments because they have broader tax bases and more ability not only to impose taxes to finance local public services but also to raise and service debt finance of new capital infrastructure. In short, for big cities to ‘grow up’ and do what they can and should, countries need to begin to treat them more like adults, who can make their own decisions within their assigned sphere of activity and deal with their own problems when they make mistakes, as they will. One approach to reforming metropolitan finance may be to create a strong unified (one-tier) metropolitan government as in Cape Town, or perhaps a ‘city-province’ or ‘provincial city’ such as Beijing or Bangkok. This may perhaps be done by relaxing annexation laws and making it easier to expand municipal boundaries. Alternatively, an overarching metropolitan regional government may be combined in a two-tier structure, with local governments charged with providing more specifically local services financed by local user charges and taxes (sometimes piggybacked on a regional tax or even administered regionally or nationally, though

M4439-BAHL_9781786435293_t.indd 387

01/03/2018 11:08

388

Fiscal decentralization and local finance in developing countries

preferably with local determination of rates) in order to strengthen local accountability and allow for differences in local preferences. Yet another approach might be to convert central or state agencies with metropolitan regional responsibilities (e.g. for transportation or water and drainage) into more autonomous (and preferably elected or at least represented) local bodies. Whatever approach is adopted, one key feature of any substantial metropolitan reform must be to raise tax prices in metropolitan areas to a level commensurate with the cost of providing services. Big cities are costly. Those who want to live there should pay the price for doing so. Many roads may be followed to the goal of metropolitan fiscal self-sufficiency, but all require that tolls be paid. In some countries, an important part of any new metropolitan government financing regime may be the establishment, or recognition, of a metropolitan area government (preferably an elected one) as the senior level of local government. Responsibility for regional expenditure functions with significant autonomy should be assigned to this level, together with some broad-based taxes and perhaps regional property tax administration (see Chapter 6) as well as borrowing powers for capital projects (Chapter 4). A lower-tier municipal government structure could be created (or continued) to provide local services, with funding from local property taxes (or surtaxes), user charges and licenses. Whether or not there is an adequate national equalization system, which, as Chapter 7 argues, is generally needed to maintain minimum standards throughout the country, the regional (metropolitan) government may also perhaps be responsible for designing, financing and implementing an intra-metropolitan transfer system to address unwanted fiscal disparities within the metropolitan region. Metropolitan local governments should also have substantial autonomy in hiring, firing and compensating employees, as well as in making cooperative arrangements and contracting with other governments at all levels, or with private enterprises, as they see fit (within, of course, national procurement rules). They should be held to a hard budget constraint and be subject to clear national rules with respect to financial reporting, auditing and transparency. Within a two-tier structure, although local governments would operate under the same rules, clear arrangements should be made to ensure that each local government as well as the regional government is treated in the same way.48 Most importantly, all governments within the metropolitan area should have taxing powers commensurate with the expenditure responsibility assigned to them. They should, for example, be permitted to impose higher property and land rates, as well as higher rates of tax on motor vehicle

M4439-BAHL_9781786435293_t.indd 388

01/03/2018 11:08



Financing metropolitan areas ­389

ownership and use, if they choose to do so. If they are responsible for expensive public services like education or health to any significant extent, they should also be able to impose at least one broad-based tax for general purposes on sales, payrolls or perhaps some form of local business tax, in some cases perhaps as a surcharge on a national tax at a rate of their choosing. User charges should of course be employed whenever possible to cover costs for services provided in the metropolitan area, including public utilities, transit and at least some roads and bridges. To encourage good fiscal behavior in all these respects, metropolitan areas, although they would remain eligible for certain conditional transfers, should no longer receive general intergovernmental transfers None of the reforms mentioned will be easy to achieve in any developing country. But each can be done to some extent in at least some countries. And all deserve careful consideration in any country whose leaders want better lives for their people: they should be considering carefully how best to finance and govern the big cities in which more and more of the people are going to be working and living so that they function as well as possible and support rather than drag down national development.

NOTES  *  This chapter draws on and expands our previous published work on this subject: see especially Bahl and Linn (1992), Bahl (2011, 2011a, 2013), and Bird and Slack (2008a, 2013). We also draw heavily on Bahl (2017), but note that the views expressed in the present book are those of the authors and should not necessarily be considered as reflecting or carrying the endorsement of the UN.   1. Some analysts have long been concerned with how best to finance large cities in developing countries: for two useful early reviews, see Hicks (1974) and Smith (1974).   2. For a useful survey of the extensive literature on all these aspects of urbanization, see McGranahan (2016).   3. In some countries, many migrants in this second group arrive not by choice but by circumstance, driven out of more rural areas or even out of their native countries by ­conflict and violence. When migrant populations are ethnically, linguistically or religiously different from the existing population, additional problems may arise ­ (although we do not discuss this question further in this chapter).   4. A global sample of urban areas in developing countries reports that 44 percent of the population increase was due to natural growth, 25 percent to migration and 30 percent to reclassification of the urban boundaries (UN Habitat and ESCAP, 2015).  5. See, for example, the extensive discussions in such studies as Yusuf (2013), Glaeser and Gottlieb (2009), Venables (2005), Ahluwalia et al. (2014), World Bank (2009b), Henderson (2010), and UN Habitat and ESCAP (2015).  6. On fiscal contracting at the national level, see Timmons (2005) and Bird and Zolt (2015a).   7. Many years ago, one of us wrote a book on taxing agriculture, which was then the dominant activity in most developing countries (Bird, 1974). That book began by arguing that the (then common) idea that the path to sound development finance was to extract the ‘surplus’ from agriculture through taxation was not only far too simple but often

M4439-BAHL_9781786435293_t.indd 389

01/03/2018 11:08

390

  8.  9. 10. 11. 12. 13. 14. 15. 16.

17. 18.

19.

20. 21.

Fiscal decentralization and local finance in developing countries simply wrong, and that a much more careful examination of local circumstances and possibilities and of the interdependence between economic sectors was necessary before determining who can and should pay how much and for what. Much the same caveats apply to the current discussion – for example, Krugman (1991), Henderson (2015), and Glaeser and Joshi-Ghani (2015) among others – about large cities being the engines of growth and the source of potentially taxable surplus: every case is different, and each needs separate study. Such issues are often especially contentious in federal countries, as discussed in Bird and Vaillancourt (2006), Bahl et al. (2013) and Mohanty (2014), but they also arise in such unitary but decentralized countries as Colombia (Bird, 1984). On the policy relevance of different approaches to incidence analysis, see especially Break (1974). For a relatively sophisticated (though still crude) attempt along these lines for Canada, see Vaillancourt and Bird (2007). See e.g. the studies in Bosch et al. (2010); see also Bird (2006). Such fears are not always well founded, as discussed in Chapter 2. For an interesting review of the effects of legislative malapportionment on tax policy in general, see Ardanaz and Scartascini (2011). Interestingly, the original empirical work testing the Tiebout model was on the effects of fiscal factors on location choices within US metropolitan areas (Oates, 1969; Hoyt, 2006; Fischel, 2006). In the early decentralization discussions in post-Soviet Eastern and Central Europe, for example, many favored very small local governments for such reasons, perhaps partly in reaction to the previous highly centralized governance system. For example, the 21 districts (with councils including both elected and appointed members) within the city of Madrid have been delegated administrative functions in such areas as urban parks, health and licensing. In 2007, these districts managed about 12 percent of the city budget. Similarly, the Netherlands has elected district councils that operate at a level below the elected municipal councils (OECD, 2007, 2007a). Some (e.g. Desgagné, 2013) have suggested that the voluntary approach may even be more successful. However, the evidence on the relative merits of the various approaches is sparse, as Spicer (2015) notes in a recent review. Bird and Slack (2008a) focus on much the same issues under somewhat different labels, discussing both the jurisdictional and metropolitan approaches as ‘one-tier’ systems, but paying more attention to more explicitly ‘federal’ (two-tier) metropolitan structures as well as to the pseudo-federal arrangements (under the label ‘voluntary cooperation’) sometimes made by local governments to achieve more efficient and coordinated provision of urban public services. This study also deals briefly with the ‘functional fragmentation’ approach under the label of ‘special purpose districts.’ Still other, less simplified but useful classifications may be found in such sources as OECD (2006), Klink (2008), Lefevre (2008) and Shah (2013). When what is now the City of Toronto was created, for example, the number of people represented by each councilor in one suburban municipality changed from 7,300 to 54,214 compared to the much smaller change from 41,850 to 54,214 in the former city of Toronto (Slack, 2000). Other countries also have parastatals: for a general discussion, as well as a somewhat out-of-date country example, see Bird (1984). A particularly striking case is Montreal, where the metropolitan region was first (partly) amalgamated when 27 small local governments on the island of Montreal were unified in 2001, and then de-amalgamated in 2006 (after a change of government at the provincial level) and reorganized with other municipalities on the north and south shores of the St. Lawrence River into a new, looser regional ‘community’ consisting (now) of 82 local governments, counting the 27 boroughs (mainly following previous city boundaries) on the island of Montreal that manage most local services but have little independent revenue power. The structure is more complex than can be quickly

M4439-BAHL_9781786435293_t.indd 390

01/03/2018 11:08



22. 23. 24.

25. 26.

27. 28. 29. 30.

31. 32.

33. 34.

Financing metropolitan areas ­391 summarized here, but it is perhaps worth quoting the conclusion of a recent study attempting to evaluate the effects of these drastic shifts in organizational form on the efficiency and equity of local expenditure: “Our results reinforce the conclusion of previous studies concerning institutional reforms at the local level. These reforms have little impact in terms of efficiency and equity for metropolitan areas. Future initiatives would do better if they focus on policies based on voluntary cooperation instead of compulsory amalgamation” (Desgagné, 2013, p. 25). For a more recent discussion and analysis, see Henriquez Diaz et al. (2011). For an interesting discussion of cities that have taken greatest advantage of the urbanization era, and those that have lagged, see Yusuf (2013). When services are supplied by several local governments in a metropolitan area, costs may be higher because of, for instance, administrative duplication or less buying power with respect to inputs; or they may be lower because of interlocal competition. Again, there is little evidence either way about such matters. These estimates are derived from an income-driven model based on a sample of 30 low-income countries. See also the earlier discussion, in Box 5.1, of an alternative and economically preferable form of local business tax. Bird (2015a) argues that although VAT rate autonomy may be feasible at the regional level, as in Canada (where most provinces received central subsidies to induce them to move to VAT), at the local level – except perhaps in large metropolitan city-provinces – such a system is impracticable. He suggests that it would usually be best to take advantage of the information base that most countries have already developed to support their national VAT by imposing the more rational business tax in the form of an origin-based production VAT suggested in Box 5.1. Such a VAT, in varying forms, exists as a local tax in Japan, Italy and France, and has at times also been used in Germany and some US states, and has been considered (though not accepted) in Canada and South Africa, as discussed in Bird (2014). Although the Federal District of Mexico City has now formally become Mexico’s 32nd state, its new constitution had not been completed at the time of writing, and many important fiscal aspects remained unclear (Rios, 2016). For an early discussion of local income taxes in Hungary, see Bird and Wallich (1992). A recent discussion of the relative merits of piggyback income taxes and payroll taxes is Martinez-Vazquez (2013). For an early study along these lines, see Oldman et al. (1967). Many others have of course since trodden this path, not least the numerous studies carried out under the auspices of the Lincoln Institute of Land Policy (www. lincolninst.edu). Similar problems are of course found in cities of all sizes. To mention only two other cases personally encountered by the authors, some years ago the medium-sized city of Santa Fé in Argentina failed to tax about half its potential property tax base, and in the extreme case of post-conflict Liberia only a handful of business properties in the capital of Monrovia were taxed at all. This is, for example, how the Municipal Property Assessment Corporation (MPAC), a provincial agency which carries out valuations and prepares the tax roll, operates in Ontario, Canada. Such a proposal was put forward in Bahl and Linn (1992) and has recently been discussed, but not acted on, in Bolivia (Brosio, 2012). For a good early study of how to design and implement this and other taxes on motor vehicle taxes in a developing country (Jamaica), see Smith (1991), and for a useful overview see Smith (2006). On the Singapore experience, see Chin (2010). One way to deal with congestion might perhaps be to impose an additional tax on businesses that create undue amounts of congestion (say, bars and banks?), perhaps earmarking the proceeds to improve public parking facilities and enforcement. Since the two activities mentioned also often impose additional policing costs (to control fighting and theft, respectively) one might perhaps think also about factoring such ­considerations into any annual business license fees they pay.

M4439-BAHL_9781786435293_t.indd 391

01/03/2018 11:08

392

Fiscal decentralization and local finance in developing countries

35. India’s recent decision to introduce a biometric personal identification system is a bold example of taking the high-technology road in a country where many are still illiterate. 36. This paragraph in part repeats arguments made some years ago with respect to urban development in China (Bird, 2005a). Of course, as anyone would likely have predicted, the rush to the automobile, a symbol and signal of prosperity, has continued in China, as in most countries. 37. There are some thorny issues to be sorted out regarding what is meant by ‘full cost recovery.’ For example, should users of a modern mass transit system in one metropolitan area be required to pay for full cost recovery when there are vertical externalities in play, such as the increased national growth rate that might occur? 38. For other older (but unfortunately not out-of-date) examples, see e.g. Bird and Miller (1989); a more recent examination of problems in water pricing may be found in OECD (2009). The question of how to overcome the reluctance to price public services properly cannot be discussed in detail here: for some preliminary thoughts, though not ­specifically in the developing country context, see Bird (2017). 39. For example, substantial central transfers for education are received by Colombian departments (regional governments) that have little control over how the funds are spent, although the large urbanized departments in practice exercise more control than others (Acosta and Bird, 2005). 40. See also the case studies in Slack and Chattopadhyay (2013). 41. The mandatory reforms included adoption of double-entry accounting; e-governance based on GIS; adoption of management information systems; property tax reform using GIS; and recovery of operations and maintenance expenditures with user charges. 42. As a final condition, though one not usually within the control of the local government itself, countries should also provide a financial framework for local government that includes well thought out and comprehensive rules about how and when local governments can become ‘bankrupt’ and an acceptable institutional framework setting out how outstanding debts will be repaid in these circumstances – preferably, of course, from local sources. 43. For instance, in Canada, where every province has its own rules governing local borrowing, some provinces pool all local borrowing at the provincial level and some do so at the ‘regional’ level both when there is a formal ‘two-tier’ governance structure (as in Ontario’s regions) and when there is not (as in British Columbia). Much could still be done to improve municipal borrowing in Canada, however, as Hanniman (2015) notes. 44. Bahl and Linn (1992, p. 478) ended their book with a strong pitch for a metropolitan strategy. Since then, one of us has, perhaps unsurprisingly, not changed his mind on this point; the other has gradually, following decades of experience, come to agree – though still favoring more the two-tier than the one-tier version of metropolitan government. 45. An excellent example of the importance of this ‘local effect’ may be seen at a much smaller scale in the interesting comparative study by Jibao and Prichard (2015) of the very different outcomes of similar policy initiatives in three cities in Sierra Leone. 46. Most of these elements are discussed further in earlier chapters, especially Chapters 3 and 5. For a brief earlier discussion of how to institutionalize the process of adjusting intergovernmental fiscal arrangements over time to accommodate changes, whether motivated by internal or external factors – that is, how to combine the stability needed for sound intergovernmental finance with the flexibility needed to accommodate change – see Bird (2001a). 47. As mentioned earlier, voluntary cooperation among local governments in the metropolitan area may to some extent be a substitute. Such structures have in some instances worked in developed countries such as Italy and Canada (Bird and Slack, 2008a) with well-established, democratic and generally well-run local governments and stable intergovernmental regimes. This alternative seems less likely to perform well in the conditions of most low- and middle-income countries, although, as Klink (2008) shows in his discussion of how large cities in Latin America are governed, much more research is needed on this question.

M4439-BAHL_9781786435293_t.indd 392

01/03/2018 11:08



Financing metropolitan areas ­393

48.  For example, while each local government should finance its own operating costs, the metropolitan region may impose a sales tax at a rate that covers the ‘average’ employee costs, and transfer to each local government the same amount per employee as well as allowing them to impose an additional tax surcharge (if they wish). If the metropolitan area chooses to introduce more horizontal sharing than provided by the national equalization system, it may of course adjust its transfers to local governments accordingly.

M4439-BAHL_9781786435293_t.indd 393

01/03/2018 11:08

9.  Giving decentralization a chance A large and growing number of countries are reexamining the roles of various levels of government and their partnership with the private sector and civil society to create governments that work and serve their people. The overall thrust of these changes manifests a trend to either devolution (empowering people) or localization (decentralization). (Boadway and Shah, 2009, p. 545)

This book is about fiscal decentralization, the empowerment of local populations through the empowerment of their local governments. We have argued that fiscal decentralization is in principle a good way to ensure that the public services provided are those that people want. Successful decentralization may be accompanied by faster and more acceptable urbanization; more innovation in governance; and even by a higher rate of revenue mobilization if, when people get what they want, they are more willing to pay for it. In most developing countries, however, fiscal decentralization has had little chance to prove itself for a variety of reasons: some understandable, though not necessarily well founded, such as fears about its macroeconomic consequences and the weaknesses of local governments; and some less commendable, such as paternalism, the desire of central officials to hold on to power and, of course, the political maneuvering by which ‘big men’ try to stay big. This chapter is about how decentralization may have a better chance to succeed if some of the constraints that have held it back in many countries can be overcome. Although we think many countries could better utilize scarce public resources by making better use of local governments, as we emphasized throughout this book every case is different, and every country must determine for itself how and to what extent it should decentralize. But some key features are common. The average level of taxation in developing countries sits at much the same low level as it did 30 years ago – about 15 percent of GDP (Bahl, 2014).1 Local public services are seldom very good and almost always unequally delivered, and the infrastructure gap remains large. Subnational governments spend on average only about 5 percent of GDP, with elected local governments usually controlling at most a fraction of this amount. Central governments seldom consider subnational government fiscal reform a priority issue. Nonetheless, as we argue in Chapter 8, rapid urbanization may soon induce (or force) some countries to consider 394

M4439-BAHL_9781786435293_t.indd 394

01/03/2018 11:08



Giving decentralization a chance ­395

anew many of the issues discussed in this book. A few are already some distance down the decentralization path, with some local governments (especially in metropolitan areas) being relatively well managed. Others may follow as local populations become more involved with their local governments, especially where accountability is stronger.

ARE THERE STILL DANGERS OF DECENTRALIZATION? Many hesitate to decentralize for reasons that were articulated some time ago when Prud’homme (1995) and Tanzi (1996) warned about the dangers of decentralization, such as out-of-control borrowing by subnational governments compromising central government finances, opening new (local) doors to corruption and wasteful local spending. They were not the only respected experts to doubt the efficacy of fiscal decentralization as a development strategy. For example, the International Monetary Fund (IMF) and the World Bank were sufficiently concerned that public services might collapse in the aftermath of Indonesia’s ‘big bang’ decentralization that the establishment of a contingency fund was seriously considered. Many countries that did decentralize to a limited extent did so cautiously, often introducing such restraints on local spending as expenditure mandates and conditional grants to reduce the likelihood that the result would be ill-advised and even runaway levels of spending. These concerns were in part motivated by the debt crisis associated with extensive subnational government borrowing in several large Latin American countries in the late 1980s. Defaults by regional governments in Brazil and Argentina resulted in national economic crises that forced central government bailouts. As we discussed in Chapter 2, however, the problem in these countries was due less to fiscal decentralization as such, and more to the absence of control over local debt issuance and the failure of higher-level governments to enforce a hard budget constraint (HBC) at the subnational level. One result of this experience was that many countries, including the biggest offenders, have subsequently addressed the debt problem, with the result that many middle- and low-income countries now have sustainable decentralized borrowing regimes.2 Concerns about corruption were also warranted to some extent. Corruption remains a serious problem in many countries, and no doubt new modes of stealing and other wrongdoing did emerge as subnational governments became more autonomous. However, as with other ‘hidden’ activities, corruption remains an issue about which we still know far too little; and what research we have on this question remains inconclusive

M4439-BAHL_9781786435293_t.indd 395

01/03/2018 11:08

396

Fiscal decentralization and local finance in developing countries

about whether corruption is favored more by decentralization or centralization (see Chapter 2 and Martinez-Vazquez et al., 2007). We now know more about how well local governments do in delivering public services. Although the evidence is again mixed, the worst outcomes envisaged by some have not materialized and some good things seem to have occurred. Indonesia’s big bang transition to a more decentralized expenditure regime seems to have worked at least moderately well, as has Colombia’s more gradual approach. In neither country is there now much support for the view that ‘the center did (or can do) it better.’ Of course, strengthening local government finances has not always been a success story. A few countries, notably Russia, have reversed direction and gone back to their centralized ways. Some African countries (for example, Uganda) seem to have moved faster toward decentralization than their local governments could manage, and have slowed and pulled back decentralization to some extent. And of course, decentralized or not, the level of local public services in most low-income countries remains well below any normatively acceptable level. A striking feature even in the most decentralized countries is that the share of local governments in national expenditures did not rise significantly. Moreover, few countries decentralized revenues to the same extent as expenditures, and almost none focused on extending local autonomy and increasing local accountability to local people. Even in those countries that did increase subnational government revenues and expenditures to a significant extent – such as Argentina, Brazil, Colombia and South Africa – there are, as many have noted, flaws in the design and implementation of the decentralization regime. Only a few countries have pushed decentralization very far, and almost none has made fiscal decentralization an integral part of a sustainable development strategy. Even the few that have decentralized to some extent still have much that needs to be improved before decentralization can really be considered a successful component of a coherent approach to national development. In this final chapter, we suggest a few guidelines that may help countries get the best results possible from decentralization.

FISCAL DECENTRALIZATION: THE NEXT ROUND We argue in Chapter 1 that good fiscal decentralization can make a significant contribution to economic and social development in most developing countries. Some may not accept this argument, for various reasons. They may argue that giving too much fiscal power to subnational governments will create macroeconomic problems or that subnational governments are

M4439-BAHL_9781786435293_t.indd 396

01/03/2018 11:08



Giving decentralization a chance ­397

simply not capable of taxing and spending sensibly. They may think that only the central government can run things the way they should be run – or at least the way they think they should be run. They may simply be unwilling to give up any power to regional or local governments. Those who dislike decentralization for whatever reason, even if they cannot block it, may manage to insert so many constraints into the system that they ensure it cannot fully succeed. Sensible decentralization requires close attention to how to deal with the inevitable problems as poorly managed subnational governments fall prey to excessive corruption, have poor fiscal discipline and fail to deliver basic public services properly. But dealing with such real (or potential) problems does not require such commonly found features as bestowing inadequate taxing power on subnational governments, placing substantial restrictions on local budget autonomy and accountability, and failing to give voice to local populations. Even countries like China, where there is no hint that there is going to be any formal move away from the existing highly centralized system of control over subnational government finance, could do more to realize more benefits from decentralization, as is discussed in Box 9.1. In the balance of this chapter, we draw on the experience in the last three decades to lay out some guidelines that may help countries do better in designing, improving and implementing fiscal decentralization. Of course, as we have already said, perhaps too often, there is no one best way to do fiscal decentralization. There is no silver bullet to solve all problems and no experience in any country that can simply be copied in another country. What works well in Denmark might be rotten in Nepal. Local context is a big part of the story. Nonetheless, we think that there are several principles that, if followed, are likely to result in more successful fiscal decentralization in any country.3 Guideline 1: Fiscal Decentralization Is a System Fiscal decentralization is a system, not a collection of unrelated bits and pieces. To get it right, decentralization must be thought of in systemic terms, not as a series of one-off reforms. Ideally, all the pieces in this system need to be on the table and then fitted together to ensure that they reinforce each other rather than have (unintentional) offsetting impacts. A major impediment to successful fiscal decentralization has often been a missing piece. If the local revenue piece is left out, full accountability is not realized; if subnational governments do not have autonomy to make budget decisions, local preferences will not be adequately addressed; if intergovernmental transfers are used to fill budget gaps, hard budget constraints will not be obeyed, and so on.

M4439-BAHL_9781786435293_t.indd 397

01/03/2018 11:08

398

Fiscal decentralization and local finance in developing countries

BOX 9.1 DECENTRALIZATION IS NOT FOR EVERYONE: THE CASE OF CHINA China is committed to centralized governance and finance. If the main goal of the government has been to improve the quality of life with economic development, one cannot argue with the results. The growth rate in GDP since the mid-1980s has been staggering; the economy has modernized; 500 million people have been lifted out of poverty; the quality of public services has been ratcheted up; and revenue mobilization has reached the level of other middle-income countries. China is in fact substantially fiscally decentralized in terms of expenditure. But the key feature of decentralization in China is that the direct accountability of appointed local officials is upward to higher-level government: what China’s provincial, county, township and other subnational governments do is essentially shaped by central priorities and directives. Given the country’s size and diversity, and the fact that over 80 percent of total government expenditures are made by provincial and local governments, it is not surprising that there is nonetheless some degree of discretion at the local level. However, all taxing power resides at the central level; subnational governments are financed mainly through transfers from the center in the form of shared taxes, unconditional grants and conditional grants; and until very recently subnational governments had no direct borrowing powers. In recent years, China has made some significant changes in its intergovernmental fiscal system. These changes do not amount to fiscal decentralization as discussed in most of this book because there is no direct accountability to local voters; but they have had major implications for the regional distribution of revenues, the allocation of public resources and the delivery of services. Viewed from the perspective of 2017, it is not clear what will come next with respect to financing of provincial and local governments in China. A dose of real decentralization could perhaps prove useful in dealing with some of the following questions: ● ● ● ● ● ●

How to find the right balance between shared taxes and grants in financing subnational governments? How to move toward more cost recovery and better user charge financing of public services? How to reassign expenditure responsibility for social insurance programs and public enterprises more appropriately? How to use tax policy in addressing environmental and congestion problems? How to develop a metropolitan area strategy for public finance? How to identify and respond to the demands of local constituencies for public services?

To some extent of course, it may prove possible to resolve the first four of these issues to a considerable degree under the present centralized fiscal regime, which is most likely the path that will be chosen. The fifth problem (metropolitan finance) would appear to require that some local governments have more taxing powers than others, but presumably that too could be achieved by the present centralized regime. Although real decentralization would seem to be required to respond to the last issue listed, much else would of course have to change in China before this issue becomes a priority.

M4439-BAHL_9781786435293_t.indd 398

01/03/2018 11:08



Giving decentralization a chance ­399

Another source of failure arising from ignoring the systemic nature of decentralization is that the policy views held by different actors in the process are not coordinated. If revenue raising powers are assigned independently of the assignment of expenditure responsibilities, the result is often significant inefficiency. If the intergovernmental transfer system consists largely of ad hoc distribution arrangements, the result is almost certainly a soft budget constraint. A piecemeal reform of any one component of the system (e.g. a local property tax reform) may have some good effects but is unlikely to contribute much to local self-empowerment. One way to avoid such problems is to design the desired system as a whole, and then to lay out a plan for how and when each element of the system should be changed to get to the desired result. Fiscal decentralization involves a lot more than fiscal matters, and resolving complex political and administration matters such as the electoral system and civil service arrangements is often as or more important than getting taxes, spending, transfers and borrowing right. Moving gradually through a series of different phases may often be the right (or only feasible) way to go, but everyone should know where they are going. To do so requires an underlying comprehensive, overall plan, including provisions to deal with the inevitable transition problems during phase-in. While life is uncertain and no battle plan ever survives the first encounter with the enemy – or an unexpected flash flood in the middle of the battlefield – a good leader (or policy analyst), like a good general, needs to have thought through in detail both the steps needed to achieve the main policy objective and how to react to various possible problems before they arise. The key elements/components of a system of fiscal decentralization are described in the first column of Table 9.1. The remaining three columns of the table indicate some of the possible compromises as one moves from a desirable to a second-best and then to a less-desirable system. The key point is that any comprehensive approach to fiscal decentralization needs to ensure these key elements fit together. The most crucial element of a decentralized system is all too often the most neglected one. Ideally, local governments should be locally elected, preferably by popular vote of the local population. If the local leadership is appointed by higher levels of government, their accountability will be upwards and not down to the local population. The efficiency gains that are at the heart of fiscal decentralization strategies will not be captured if local governments cannot make budget choices on behalf of their constituents.4 It is almost as important that the local council appoint the local chief officers (treasurer, chief education officer and so on). Otherwise, implementation will not be locally directed, and accountability for the quality of services delivered will be upward to the central or state ­government. Other

M4439-BAHL_9781786435293_t.indd 399

01/03/2018 11:08

400

Fiscal decentralization and local finance in developing countries

Table 9.1  Components of a successful system of fiscal decentralization Component

Desirable

Second best

Least desirable

Local  accountability

Popular election

Indirect election

Chief officers

Locally appointed

Budget

Local approval; hard constraint Significant local control over how money is spent

Central secondment Local approval; soft constraint Local autonomy with significant limits

Appointment by higher-level government Direct central provision Central approval; soft constraint Local government is a spending agent of the higher-level government Little or no revenue raising power Mostly conditional grants

Expenditure  discretion Own revenue

Significant local powers

Some local power

Inter­  governmental transfers

Mostly generalpurpose grants

Borrowing  powers

Broad powers and hard budget constraint

Civil service

Locals governments hire fire and determine compensation Clear central rules; adequate information system; public reporting; institutional framework for intergovernmental forum

Balanced combination of conditional and unconditional grants Restricted borrowing powers and hard budget constraint Local government employment fully financed by higherlevel government Central rules and monitoring of use of central funds; informal intergovernmental meetings

Monitoring   and review

No borrowing powers Local governments have no power to hire, fire or compensate Center approves everything that matters

necessary conditions for fiscal decentralization are significant expenditure responsibilities, a significant amount of taxing power, budget making autonomy, transparency and a hard budget constraint. As we argue in Chapter 5, the latter should force local governments to live within their

M4439-BAHL_9781786435293_t.indd 400

01/03/2018 11:08



Giving decentralization a chance ­401

means and local officials to be accountable for the hard choices they must make. Not everyone believes that decentralization needs to be viewed comprehensively. Judging by the evidence, most countries (and donor agencies) appear to think that fiscal decentralization can be accomplished by a one-off revision of the transfer system, an upgrading of the property tax administration or a change in the local election system. Government leaders seeking to build a supporting political coalition for decentralization may find this ‘one dimension’ approach easier to deal with. Foreign experts in donor agencies also often seem to find it easier to just get on with pushing this or that piecemeal reform. The gradual adoption of a series of related policies may indeed be more feasible and produce some visible (and politically saleable) results. But pushing a piecemeal reform program without having a clear and comprehensive objective in mind is all too much like taking a trip without a road map (or a GPS): who knows where you might end up? The road to successful decentralization may take a thousand steps, but to get there one must first know where one is going. Guideline 2: Accountability Requires Significant Local Fiscal Autonomy The principal objective of sound and sustainable fiscal decentralization is to ensure that elected (or even appointed) local officials are accountable to their local constituents, that is, governors and provincial legislatures to those who live in the province; mayors and councils to those who live in the city; special district managers and board members to those who live in the service area and so on. Where this accountability is weakened, even if justified by some externality or equity or management concern, fiscal decentralization is weakened. On the expenditure side, the key to achieving this objective is for central governments to limit the autonomy of local governments only where significant efficiency or equity concerns dictate they should. Those who make the rules need to make rules for themselves if they want others to follow the rules. Specifically: ●●

●● ●●

Unfunded expenditure mandates should be eliminated. If a mandate is necessary, then it should be funded (e.g. with a conditional grant), or the function should be assumed by the higher-level government. The use of conditional grants should be limited to supporting expenditures where externality concerns are demonstrably important. Subnational government budgets should be approved by the relevant subnational councils, and should be comprehensive.

M4439-BAHL_9781786435293_t.indd 401

01/03/2018 11:08

402 ●●

●●

Fiscal decentralization and local finance in developing countries

Local budgets should be balanced according to a clear legal framework, with appropriate penalties for violations. Year-end final accounts should also be balanced (except for short-term cash flow problems) according to clear definitions, with appropriate penalties. All budgets and financial reports should be readily available public documents.

The revenue side of the budget is equally important for accountability, although it is all too often seriously weak at the subnational level in many countries. Voters will hold elected officials more accountable if local public services are financed to a significant extent from locally imposed taxes and charges rather than by central government transfers (often seen as ‘other people’s money’). Local taxes should be visible to local voters, and large enough to impose a noticeable burden that cannot be easily exported to residents outside the jurisdiction who are not the beneficiaries of local services. The mishmash of minor taxes and nuisance levies allotted to the local level in many developing countries will not do the trick. It is often alleged that there are no good revenue-productive local government taxes available. As we discuss in Chapters 5, 6 and 8, this is not true. Individual income or payroll taxes, property taxes, user charges, business licenses, taxes on the use of motor vehicles and an array of subnational ‘piggybacks’ on central taxes are all viable options in many circumstances. The failure of higher-level governments to permit, encourage (through the design of transfers) and even force (through strict hard budget constraints) subnational governments to make good use of such revenue sources is an important reason fiscal decentralization in many countries has not been successful. Accountability is central to good fiscal decentralization. Often, however, matters are not cut and dried, and trade-offs must be made. For example, as we discuss in Chapter 8, moving to a metropolitan area-wide government may move decisions further from local communities, but may also make it possible to impose more productive taxes and to take externalities more into account when making expenditure decisions. The net effects on welfare are far from clear, but countries might reasonably encourage moves in this direction. As another example, using special districts to deliver services may improve efficiency but again reduce the influence of local citizens on spending decisions. Again, choices must be made. Similarly, as we argue in Chapter 5, user charge financing may not only increase efficiency but also accountability; however, in some cases social policy concerns may reasonably lead communities to opt for general tax financing. Alternatively, a softer approach may be used, for instance: by complementing metropolitan government with a bottom tier of ­governance that

M4439-BAHL_9781786435293_t.indd 402

01/03/2018 11:08



Giving decentralization a chance ­403

maintains some degree of local rule; by requiring special district boards to include elected local officials; and by designing user charges to insulate users at risk from adverse effects. Guideline 3: Impose a Hard Budget Constraint Subnational governments should be required to balance their annual budgets ex ante and ex post. Local governments must have good reason to believe that they are on their own. The result will be more prudent expenditure decisions, more accountability to voters, more local revenue mobilization (if this route is open to subnational governments, as it should be) and more involved (and possibly more informed) local voters. However, imposing a hard budget constraint on subnational governments is easier said than done. Many issues need to be resolved. For example: ●●

●●

●●

●●

●●

As described in Chapter 5, the HBC requires two budgets to be balanced: the recurrent expenditure budget and the capital budget. This raises questions about which expenditures qualify for inclusion in the capital budget, and which sources of revenue qualify as recurrent.5 Clear rules are needed on subnational financial accounting. Higher-level governments must in effect certify that ‘balance’ is achieved, i.e., the legal requirements are met. But they should not have any further say on the content of the budget. Formal budget approval should rest with the subnational government council. Higher-level governments should discontinue practices that soften the budget constraint, such as deficit grants (i.e., year-end grants to cover revenue shortfalls), bailouts on delinquent debt and direct coverage of year-end shortfalls on certain items of expenditure. Intergovernmental transfers should be distributed on an objective and transparent basis, though certain capital project grants and those to meet disaster relief and other extraordinary situations might continue to be ad hoc. Specific routes should be specified with respect to what happens if a subnational government overspends. Although much depends on the situation, reactions may include: the imposition of a higher property tax rate or utility surcharge by the higher-level government; withholding a share of intergovernmental transfers; placing legal liability for deficits on the elected council; suspending certain expenditures and imposing a moratorium on others; imposing receivership for continued violations and so on. The package of sanctions needs to be spelled out in detail.

M4439-BAHL_9781786435293_t.indd 403

01/03/2018 11:08

404

Fiscal decentralization and local finance in developing countries

Sometimes local governments may get into trouble, but the root cause might be the higher level of government itself. If the central government abruptly discontinues a subnational government tax, fails to deliver an expected (predetermined) transfer, imposes an unfunded mandate, suspends a capital grant program while a project is underway or imposes wage increases on local governments – all actions we have observed in several countries – local deficits may result. Imposing penalties on the local governments for breaching the HBC is unwarranted in all these cases. Central governments make the rules, and they need to abide by them. When they do not, they should not punish the victims. Guideline 4: Big Bang or Gradualism? Either Can Work When decentralization became a major issue in Eastern and Central Europe in the early 1990s, there was much discussion about which was the best approach: a ‘big bang’ (do it all in one fell swoop) or gradualism (do one thing at a time). Both in the transitional countries and in the developing world more generally, some countries have followed one path and some the other. The big bang approach may have several advantages – for example, taking advantage of a particularly propitious moment to launch decentralization and getting implementation underway before the opposition is organized enough to block things. But it also has disadvantages: rushing to action may not lead to doing the right things; some important issues may get lost in the hurried shuffle before implementation; and many aspects of good implementation (drawing up regulations, strengthening administrative capacity, developing an information and monitoring structure) may fall well behind in a rapid rollout, leaving a mess that has to be cleared up after the fact. Indonesia is probably the most celebrated example of the big bang approach. In 2001, it implemented a decentralization program that introduced sweeping changes in expenditure assignment and a new revenue sharing program. However, it left the question of local government revenue assignment untouched. China also introduced a sweeping change in its subnational government financing system in 1994, but it left the question of expenditure responsibility untouched and stayed with its commitment to centralized fiscal decision-making (Table 9.2). In both cases, the reforms were broadly successful in achieving their intended primary objectives: Indonesia in transferring responsibility for nearly 30 percent of all government expenditures to local government; and China in restoring balance in the revenue allocation between central and local governments. But neither was complete, or completely successful. As Table 9.2 also outlines, other countries have decentralized more

M4439-BAHL_9781786435293_t.indd 404

01/03/2018 11:08

405

M4439-BAHL_9781786435293_t.indd 405

01/03/2018 11:08

N N Y

N Y Y N N Y Y

N, Local revenue autonomy reduced Y Y

Yb N Y

Y

N

Nc N N

N

Y

Y N Y

Y Y

Y (Some)

N

Y Revenue; transfers; accountability

Colombia (1980–2016)

Yd

Y N, but later  reform restructured transfers N N N

Y, but not implemented N

Ya Local government strengthening

India (1992)

Notes: a Only covered third-tier local governments under a constitutional amendment. b Local governments were divided into metropolitan, municipal and rural councils with different taxing powers for each. c Four large cities have provincial status, but there is no differentiation in revenue or expenditure powers. d Local and State Councils elected, Chief Administrative Officer appointed. In contrast to most countries, Colombia’s fiscal decentralization has taken place gradually, with relevant changes in local revenues, local governance, central transfers and the constitution taking place from time to time over the last few decades. For India, 1992 was the year of constitutional amendment.

Asymmetric structure Comprehensive reform Comprehensive monitoring  system Local elections

More expenditure  autonomy Local revenue  autonomy Accountability down Transfer system reform

Y Revenue reform

Y Expenditure assignment; transfers Y

Y Expenditure assignment; accountability Y

China (1994)

Policy plan Emphasis

Indonesia (2001)

South Africa (1996)

What was included?

Table 9.2  Contrasting approaches to intergovernmental fiscal reform

406

Fiscal decentralization and local finance in developing countries

gradually. Colombia began decentralizing in the 1980s by strengthening local revenues and implementing new intergovernmental transfer systems. It then established an expert committee to develop a more complete program for fiscal decentralization and, after substantial public and political discussion, proceeded to introduce major constitutional revisions in the early 1990s to improve the system, including both the introduction of an improved local electoral system and the establishment of much stronger monitoring capacity at the central level. Subsequently, following several more expert reviews and further modifications, yet another constitutional change was made a few years ago to reform resource revenue sharing. An interesting feature of Colombia’s decentralization is that it has always to some extent been applied asymmetrically in different regions and localities, depending in part on their revenue base and administrative capacity. South Africa in a sense combined both the big bang and gradual approaches. Following the end of apartheid, it immediately established a new system of spheres (vs. tiers) of government, and differentiated the fiscal roles of various types of local government. Subsequently, however, the central government redrew boundaries for metropolitan local governments and municipalities, reduced the number of local governments and reformed the local finance system. In the end, South Africa backtracked on its initial commitment to increase local revenue autonomy. Colombia too continues to impose a variety of direct controls on local expenditures. As these examples suggest, the real question is not whether the ‘right’ approach to decentralization is via the big bang or gradual routes, but rather whether the central government seriously wants full fiscal decentralization, has laid out all the elements needed to achieve it and can implement the full package successfully. By this (strong) standard, no one has yet done the job in full, so it is not surprising that the ‘promised land’ some advocates of decentralization have at times envisaged has not yet been attained anywhere in the developing world. No central government has yet gone all the way; nor does any seem likely to do so soon. Nonetheless, many countries can gain from moving some distance in this direction, and some may have to do so for domestic political reasons whether their leaders like it or not. It does not matter whether they do so in a leap or by small steps, provided they do it right – i.e. viewing the process as a whole and focusing on the key issues set out above. Guideline 5: An Asymmetric Approach Is Usually Advisable Most developing countries are characterized by broad differences in the fiscal capacity, fiscal needs and fiscal potential of their local governments. India is a good example. Local governments range from huge sprawling

M4439-BAHL_9781786435293_t.indd 406

01/03/2018 11:08



Giving decentralization a chance ­407

metropolitan areas to small remote rural villages, and everything in between. It is simply not possible that any uniform subnational government fiscal structure will fit them all equally well. Different local and regional governments have very different capabilities to deliver and finance services as well as different capabilities to borrow. The better part of wisdom is often to recognize reality, and to structure a system in which different subnational governments are classified and treated differently in terms of their fiscal powers and responsibilities. India does this to some extent at the regional (state) level by treating ‘special category states’ differently; it also has different structures for urban and rural local governments. Nonetheless, as in most countries, much more could be done (especially at the local government level) to differentiate localities in terms of capacity and needs, to define the different categories and to specify how an area might ‘graduate’ from one class to the next. In addition, most countries lack an appropriate equalization grant system (or an alternative, such as a national support program) to ensure that at least basic local public service levels are provided in even the weakest local governments (see Chapter 7). At the other end of the spectrum, as we argued in Chapter 8, there is also a strong case for metropolitan local governments to be treated differently. Almost by definition, developing countries are interested in economic growth, and today’s main engines of growth are generally found in the large urban areas. The discussion in Chapter 8 suggests that in most cases the best approach is to give more expenditure autonomy and taxing power to large urban governments while reducing their dependence on national transfers – to let them float on their own bottoms. In contrast, smaller and poorer rural local governments are generally going to remain heavily dependent on intergovernmental transfers and to have less spending and tax autonomy. Between these two extremes, there is usually a range of small towns and cities that may over time ‘graduate’ through an intergovernmental system that systematically weans them off transfer dependence and gradually increases their fiscal independence. Countries from Bangladesh to Brazil have designated metropolitan or other ‘special’ cities (e.g., the national capital) or urban vs. rural governments. However, few have established clear rules setting out the conditions under which localities move from one category to the other, providing incentives for doing so, gone very far in decentralizing fiscal powers to even the strongest and best-run subnational governments, or established an adequate equalization system. The South African approach is an interesting model (Bahl and Smoke, 2003a). Subnational governments are categorized into three classes. First, the six metropolitan area governments have more extensive fiscal powers and raise a significant portion of their revenues from local sources. Second,

M4439-BAHL_9781786435293_t.indd 407

01/03/2018 11:08

408

Fiscal decentralization and local finance in developing countries

over 200 municipalities are classified based on their fiscal capacity and expenditure delivery capacity. These local governments have some taxing powers but are more dependent on intergovernmental transfers to finance their budgets. The third class, consisting of about 50 rural local governments with relatively less capacity, is almost fully dependent on transfers. The system (like all real-world systems) is far from ideal in some respects, but it provides a solid framework on which to build towards a better future. Guideline 6: Monitoring Matters Successful fiscal decentralization requires countries to track the fiscal performance of their local governments on a continuous basis. Over time, disparities among regions within a country change; the quality of the basic infrastructure changes; priority areas for investment change; the technical capacities of local governments change; natural resource discoveries alter the fiscal landscape and so on. Central and local governments must be able to adjust to such changes. A key concern in designing and implementing fiscal decentralization is how to strike the right balance between the stability essential to permit governments at all levels to make good fiscal decisions and the flexibility needed when circumstances change – or when decisions turn out to be bad. The only solution is to keep a close eye on what is going on, to have an early warning system to signal when things seem to be going wrong, and to have clear rules about what should be done in different circumstances. To do these things, countries need both information and an institutional framework that receives, analyzes and acts on this information. Less-developed countries and countries in transition are for the most part characterized by centralized systems of government that control, on average, about 80 percent of direct expenditures and an even greater share of revenues. These dominant central governments are almost always in full control of subnational governments. This situation is unlikely to change quickly, so fiscal decentralization in most countries is almost invariably largely controlled and regulated from the center. To get decentralization right, the central government clearly must know what is going on down there: its ability to monitor local fiscal affairs is crucial. While every country is different, examples of what is needed include: ●● ●●

A uniform structure of subnational government accounts that are regularly and properly audited. As subnational governments begin to move toward debt financing of capital improvements, there must be clear disclosure requirements and (usually) borrowing limits.

M4439-BAHL_9781786435293_t.indd 408

01/03/2018 11:08



Giving decentralization a chance ­409 ●●

●● ●● ●●

Central governments must be able to monitor the fiscal performance of local governments, identify those in financial difficulties (and be able to determine the underlying causes), and enforce the rules for a hard budget constraint. Similar monitoring and enforcement of the terms of conditional grants, expenditure mandates and taxing limits is needed. Regular reviews of central transfers and other policies affecting subnational finance, with adjustments when needed.6 Provision of training and support, especially to smaller local governments, in areas such as accounting, treasury, tax administration, data processing and project evaluation.

The information that the central government requires to do such things is also needed by regional and local governments so they can track and understand their own fiscal performance. They need reliable information for effective budget preparation and execution – for example, forecasting revenues and expenditures, appraising possible tax reliefs or new programs, seeking new central support for certain activities or comparing their performance with that of neighboring governments. Two ingredients are essential to carry out monitoring. The most basic is a data system that provides the necessary information in a timely and transparent fashion to all interested parties. To establish such a system, as well as to make use of its results, the central government needs a fiscal analysis unit with adequate trained staff to do the job. Similarly, the local governments that are the principal source of data – data that they need to do their own job properly – also generally need support in doing so. Often (as in South Africa, Indonesia and Colombia), the central unit is placed in the Ministry of Finance, where it can more easily coordinate with those responsible for other aspects of fiscal policy such as taxation and borrowing. Another option is to create an independent unit whose primary duty is policy research and advisory, like South Africa’s Fiscal and Finance Commission and Uganda’s Local Government Finance Commission. Other countries, like India and Pakistan, have periodic finance commissions (see Chapter 6). Still others, like Colombia, appoint special commissions from time to time to examine intergovernmental financial matters. In some countries such as Brazil and Argentina ‘think tanks’ have developed outside government that can give independent views of the performance of the intergovernmental finance system. Although such outside views are seldom welcomed by those in power, they can play a very useful role in informing and shaping opinion and perhaps in some cases policy – in effect, replacing to a considerable extent a role long played in many developing countries by international agencies and foreign advisors.

M4439-BAHL_9781786435293_t.indd 409

01/03/2018 11:08

410

Fiscal decentralization and local finance in developing countries

The hardest task is usually to set up an adequately comprehensive and reliable data system. Ideally, a comprehensive census of government finances, reporting the actual financial outcomes for individual subnational governments, is essential information if the performance of the intergovernmental system is to be monitored. Few developing countries (and not all developed countries) have such an up-to-date information system that describes the finances of subnational governments in detail. It is not surprising that none yet seems to have a solid fiscal analysis model that is used to track the performance of local government finances.7 Although even the smallest local government (provided it has some degree of autonomy) needs at least the capability to monitor and track expenditure outcomes, it is perhaps only the larger (regional and metropolitan) subnational governments that can play on their own in this league. Guideline 7: Real Fiscal Decentralization Requires a Champion In a fascinating interview, a former President of Bolivia discussed why, when he achieved power, he immediately introduced a ‘law of popular participation’ that in effect transferred most of the responsibility for health, education and many other services to local (municipal) governments. He did so in part for political reasons. By decentralizing to the municipal level, he bypassed and defused the strong political pressures to ‘regionalize’ – that is, support the existing regional political elites which opposed him. More interestingly, however, he went on to say: I realized that strong democracies are decentralized because by decentralizing, you push the problems that are really important to people to their level, where they can do something about them, if you’re willing to give them the resources. If you keep power centralized, you suffer the risk that the discontents and frustrations of their daily lives put the stability of the state in jeopardy. (Sánchez de Lozada and Faguet, 2015, pp. 43–4)

Not only is this an excellent summary of the basic argument for decentralization, it is also perhaps the best example in recent years of a real ‘champion’ of decentralization – one who both knew what he was doing and why he was doing it, as well as how to get it done. Fiscal decentralization seldom has such an enthusiastic and powerful champion. The preferences of political leaders and bureaucrats may be very different from those of most people; and when it comes to such key decisions as the degree to which local governments should have taxing power, politicians at all levels seem less than enthusiastic. If decentralization as defined in this book is to succeed, it needs a strong internal champion who understands the costs and benefits of establishing such a

M4439-BAHL_9781786435293_t.indd 410

01/03/2018 11:08



Giving decentralization a chance ­411

program, and continues to push for its improvement. In most countries, key leaders are often understandably ambivalent in their support for stronger local government. And even when they go all out, as Sánchez de Lozada did in Bolivia, those who come after may do their best to weaken and alter the decentralized system.8 Table 9.3 depicts who one might expect in principle to be most likely to support decentralization (Bahl, 2002). To the extent decentralization is a grass-roots, ‘bottom-up’ movement, one might expect both most citizens and most elected politicians, who depend on those citizens for votes, to be natural supporters. However, to the extent fiscal decentralization is thought to conflict with macroeconomic stabilization policy, support from the top – the President – is likely to be less firm because hyperinflation or recession is much more of a threat to re-election than is the absence of a more decentralized system of governance and finance. Since national legislators like to be elected, if voters embrace decentralization their representatives may also support it. However, most represent a specific constituency and are most interested in how decentralization may benefit their own constituency, and in the extent to which they will be able to claim credit for such benefits. They are therefore likely to be considerably less keen than policy analysts about the need for transparency. They are also most unlikely to be enthusiastic about increased taxation. Presumably, most local governments will favor decentralization; but the rich and poor (or urban and rural) have very different views about which decentralization is best. The more well-off local areas may favor increased fiscal discretion and a laissez-faire approach to fiscal decentralization; the poorer ones are more likely to prefer a redistributive system based on a guaranteed revenue flow. The central ministry in charge of local government would also like a guaranteed revenue flow, but of course one that would strengthen rather than weaken its control over local government affairs. Outside the domestic political system, some external donors and international agencies may be supporters. For example, in many countries the World Bank and the regional development banks have often seen decentralization as part of a development strategy that will lead to a more satisfactory and balanced growth. At times, the United States Agency for International Development (USAID) and some other bilateral aid agencies have also been strong supporters of decentralization, often focusing on the democracy or redistributive aspects. Given its focus on avoiding instability, the IMF has usually tended to take a more cautious and qualified view. Sometimes, external input, especially when accompanied by potential funding, has caught the attention of the central government and led it to look more carefully at decentralization. However, such looks are unlikely to lead to meaningful reform unless there is strong domestic

M4439-BAHL_9781786435293_t.indd 411

01/03/2018 11:08

412

Fiscal decentralization and local finance in developing countries

Table 9.3  Champions of fiscal decentralization Comments Potentially strong supporters The people and  their elected representatives President

Parliament or  Congress Urban local  governments External donors

Demand for more control over local services and for participation in governance at the local level. Decentralization may be popular with the electorate but the President/Head of State must worry about the impact of decentralization on stabilization, since inflation and unemployment may threaten his/her position. Decentralization in general may be popular, but politicians like be able to claim credit for specific local projects, and hence are likely to favor less transparent and less structured decentralization. Bigger cities are most likely to be concerned with achieving more spending autonomy and more access to the tax base. Donors may be cheerleaders and can provide technical assistance to get the process underway, but they should never be a substitute for an in-country champion.

Less keen Ministry of Finance Ministry of Economy Line ministries

In part for stabilization purposes, likely to support strict limits to decentralization, particularly on the taxation and borrowing side. Concerned about controlling public investment (both what and where) and about focusing on programs with benefits that are wider than the local area. Want to retain control over the standards of public service delivery, and usually also to continue to have approval control over local spending decisions.

Ambivalent Ministry of Local  Government Weaker local  governments

M4439-BAHL_9781786435293_t.indd 412

Will support a bigger (and preferably guaranteed) share for local governments, but would also like a major voice in controlling the distribution of those resources. Would like a guaranteed transfer of resources from the urban and wealthier local governments to the rest. More interested in a transfer system than a local taxing system.

01/03/2018 11:08



Giving decentralization a chance ­413

support, and outside comments are usually quickly forgotten when the money is gone. Invariably, decentralization proposals also have enemies. One is usually the Ministry of Finance, which is invariably reluctant to lose control over subnational government fiscal behavior. When it does go along, it usually favors an approach featuring limited freedom for local governments to set tax rates for any major taxes, strictly controlled borrowing powers, and enforcement of a hard budget constraint. Experience suggests that it is often right to do so; and, as we have argued, such features are not necessarily inconsistent (and may sometimes even be necessary) with attaining an efficient, equitable and sustainably decentralized regime. Another common enemy is the Ministry of Economy (or Planning, as they used to be called). As a rule, a key function of this ministry in many developing countries is to control public investment decisions, and it is unlikely to welcome losing any of this control. Local direction of investment may be thought to compromise national planning with respect to how capital expenditures should be distributed by function and by location. The latter is often politically important because bestowing projects on favored localities has long been a means of generating support for the central regime. Most spending (line) ministries are likely to oppose decentralization on grounds that are more paternalistic. Their view is often that local governments do not have the technical capacity to deliver services or to plan resource allocation, and need strong central direction. Of course, line ministry officials also like their jobs and do not want to lose them to subnational governments. If there must be decentralization, they tend to be most comfortable with conditional grants and mandated expenditure requirements that let them stay in their offices with their hands on the controls. What all this comes down to is that a one-off major decentralization reform is unlikely without strong and continuing support from someone at the top who can build a supportive political coalition and sustain it through an inevitably lengthy transitional process during which the legislative framework is put into place and the many institutional changes required accomplished – someone like Sánchez de Lozada in Bolivia, for example. In the absence of such a champion, those who think decentralization is good for their country must try to launch a national discussion on what can and should be done, and then build on this to develop and promote a decentralization program which will then often need to be standing by for some time before it may begin to be implemented in pieces over time as and when the political and economic stars are in line.

M4439-BAHL_9781786435293_t.indd 413

01/03/2018 11:08

414

Fiscal decentralization and local finance in developing countries

Guideline 8: Donors Should Rethink their Role One last guideline is not for countries that may contemplate proceeding down the path of fiscal decentralization, but rather for the external organizations that have played an important and mostly positive role in keeping fiscal decentralization on the front burner of government policy in many less-developed countries. Indeed, were it not for frequent prodding, and considerable funding, from international agencies, central government politicians and officials in some countries might long ago have buried the whole idea. External aid agencies have contributed in the form of technical assistance, funding development budgets in poor countries and continually insisting – sometimes to the irritation of central governments – on emphasizing such concerns as poverty alleviation, anti-corruption programs, training a new cadre of public administrators to lead fiscal decentralization, and the budgetary and financial accounting reforms needed to improve subnational governments. In some cases, external agencies have also served as a useful stalking horse to try out a new idea on the population and the power structure. Not every idea pushed from abroad has been good or has had good effects. In the urban areas, for example, Kharas and Linn (2013) note that although donors seem usually to focus on the right issues, there is a glaring disconnect between what is proposed and what is implemented. Others have suggested that sometimes outsiders have used the leverage from large loans to meddle in national affairs, that the wrong capital projects might sometimes have been supported, that some ‘experts’ may have been unqualified and so on. Sometimes these criticisms have some merit, sometimes they may just reflect different views about what should be done, and sometimes they may be politically motivated. While this is not the place to go into such matters, donors could usefully do some new thinking about fiscal decentralization. Specifically, experience and observation over several decades of work in this field in every region of the world suggests at least four areas where new approaches seem needed. First, and most importantly, outside agencies cannot and should not lead the way in this highly political area. They may sometimes usefully provide technical assistance in jumpstarting new budgeting or tax administration practices; but in doing so much more attention needs to be paid to the local environment and less weight put on often not very relevant foreign experience. A key argument for decentralization is to permit and encourage local governments to take charge of their own affairs in a competent and accountable way. There is no room for paternalistic approaches in this most political of areas. At most, the role of foreigners may be to provide technical advice when asked and, perhaps (if requested), some

M4439-BAHL_9781786435293_t.indd 414

01/03/2018 11:08



Giving decentralization a chance ­415

financial support in certain areas when it fits with their own objectives and is provided and accounted for in a transparent way. Sustainable decentralization depends on both central and local governments being motivated to succeed and being helped, on request, to develop the needed capacity to plan and manage for success. It cannot and should not be pushed on unwilling clients. Second, everyone involved usually needs more ‘training’ to do the job right – ranging from formal educational courses to ‘hands-on’ on-the-job experience. If those involved domestically recognize this need, and foreign agencies want to help, there are many ways to do so. They might help improve the quality of local training institutions and provide more relevant training opportunities both in general and for such specialized functions as valuation. Funds allocated to foreign or domestic training of government officials at all levels are far more likely to affect outcomes than financing expensive resident technical assistance missions or strings of ‘fly-in, flyout’ experts on this or that – and we speak as frequent ‘fliers.’9 Third, it may be time to think about some new modes of assistance, or at least new areas of assistance. To mention only a few that we have discussed earlier, there might, for example, be more focus on the growing – and, we argue, critical – problem of improving metropolitan area governance and finance; on helping countries develop subnational data bases and effective monitoring activities; and on encouraging the development of independent think tanks that can address local finance issues and perhaps also be focal points for some of the training mentioned in the preceding paragraph. Finally, it is long past time for some international agency (or agencies) to take on the difficult but essential task of designing and implementing a fiscal data base for subnational government finances in developing countries. Often, to do so may require providing substantial help to some countries to improve their own data collection. Better data are not the answer to all the problems of subnational finance. But without better data, countries – let alone the international community – will not be able to understand the nature or size of the task or to assess the extent to which it is being dealt with.

IMPLEMENTING FISCAL DECENTRALIZATION10 To the extent countries become fiscally decentralized, the move is likely to be phased in over a period of years (as we suggested above). Even if a country has a clear idea of where it wishes to go, the order in which the pieces of the system are introduced is important. In practice, politics and administrative constraints usually rule the process. However, here we

M4439-BAHL_9781786435293_t.indd 415

01/03/2018 11:08

416

Fiscal decentralization and local finance in developing countries

Table 9.4  Sequencing fiscal decentralization: a normative approach Sequence

Activity

Step 1

Carry out a national debate on the issues related to decentralization policy Do the policy design and develop the White Paper Pass the decentralization law Develop the implementation regulations Implement the decentralization program Monitor, evaluate and retrofit

Step 2 Step 3 Step 4 Step 5 Step 6

Source:  Adapted from Bahl and Martinez-Vazquez (2006a).

outline a normative approach to sequencing decentralization that might serve as a useful baseline against which to compare real-world practice. A textbook approach to sequencing fiscal decentralization may be structured in six steps, as shown in Table 9.4. As we discussed earlier, ideally the process might begin with a national debate involving the key stakeholders, for example, in the context of a national election campaign where different parties put forward different approaches, or perhaps initiated by a national commission. Without such a national debate, it is unlikely that any serious decentralization proposal will be able to assemble sufficient ‘buy-in’ to stay on track through what is likely to be a process requiring at least several years of development, implementation and adjustment. The second step is to set out the design of the proposed reform in an official policy paper. Such a paper may, as noted above, be the starting point of the discussion; but what is needed in this second critical phase is a formal government proposal, which may even be a new draft law, showing exactly how the government proposes to accomplish what it considers to be the ‘wishes of the people’ with respect to decentralization. Such a proposal would, for example, outline the main components of the proposed reform, and set out a timetable for implementation. The key question is of course, what comes first? As we have stressed earlier, it is critical that ‘finance follows function’; that is, that first the role of each level of government in each expenditure function is clearly set out, and then revenue-raising powers and the intergovernmental transfer regime are worked out to make that allocation work as well as possible.11 Step 3 is then to draft and pass the decentralization law(s). These laws may stand alone or may require some prior constitutional revision (as in, for example, India and Colombia). Achieving the right balance of stability and flexibility is of course never easy but, as many countries have learned, it is not wise to put detailed fiscal decentralization provisions in

M4439-BAHL_9781786435293_t.indd 416

01/03/2018 11:08



Giving decentralization a chance ­417

the constitution. Some federal constitutions enshrine fiscal rules that may have seemed wise or necessary when the constitution was framed but that have subsequently turned out to be considerably less sensible in the face of changing circumstances and technological change. To mention only one example, the provision in the South African Constitution that prohibits local governments from levying any sales or income taxes makes it difficult to devise a sensible fiscal structure for metropolitan governments. Once the law is in place, there is still much work to be done to develop and adopt the regulations that lay out in detail what is to be done (Step 4). For example, the law may require the transfer of some civil servants from central to subnational governments. Regulations are required to spell out in detail how this is to be done – to describe how pensions will be handled, to specify how seniority issues are to be dealt with, to provide clarity on exceptions and so on. Step 5 is implementation. Once a (hopefully clear) set of rules and regulations is in place, the new system must be put in place – with central and subnational governments operating with their new tasks, new structures, new institutions, and often new personnel and responsibilities. In a specific fiscal year, subnational governments must begin delivering the set of functions they have been assigned and covering the cost with the new package of resources they have been given. The central government simultaneously has to transfer (usually increased) revenues to subnational governments, often using a new formula, while devolving designated expenditure management and delivery responsibilities. At both levels, new people – sometimes transferred from the other level – must play new roles. None of this is simple, and many different problems may arise in the transition process. The final stage in the sequence (Step 6) is therefore to ensure that, from the beginning, there is in place a well-designed and operational system of monitoring and evaluation by the central government and a clear and uniform accountability system at the subnational level. The higher-level government must carefully track the progress of the fiscal decentralization, measure this against the goals of the decentralization, and have a process in place for making necessary policy and administrative adjustments. Preferably, those adjustments must also be first discussed and largely agreed with the subnational governments themselves, for example, in a formal intergovernmental forum that meets periodically to review progress and discuss what adjustments may be needed. Meanwhile, the necessary institutions need to be in place for the decentralized units of government to become accountable to their constituencies as well as for the inevitably complex and often very specific intergovernmental negotiations that will be required to make this complex new machinery function properly.

M4439-BAHL_9781786435293_t.indd 417

01/03/2018 11:08

418

Fiscal decentralization and local finance in developing countries

These six steps may seem formidable, but all are necessary to introducing and implementing successfully sustainable fiscal decentralization strategy. The sequence suggested is logical and allows each step to build on the necessary prerequisites, thus minimizing the chances that the whole process will come to a crashing halt. In the real world, of course, no country is likely to be able to follow the path laid out in Table 9.4. Sometimes, departures from this path may be necessary. They need not block progress unduly – for example, one can pass a law and get started before having all the details of implementation set out. Sometimes, however, getting policies out of sequence may be almost guaranteed to lead to failure, as in the many countries that have approached fiscal decentralization by tinkering with revenues and transfers rather than getting the expenditure side clear in the first place.

SUSTAINING FISCAL DECENTRALIZATION Designing and implementing a decentralized fiscal system is a complex task, but so is almost any aspect of public policy in a complex society. Formulating feasible and acceptable policies is difficult. Conveying to those who must implement the policies what they are to do and ensuring that they do what they are supposed to do are equally difficult. Getting feedback from consumers (the people) and suppliers (officials) on what is going on and what should be done to change outcomes and processes to accommodate changing conditions, possibilities and needs is a neverending and difficult task. The system needs to be corrected when it fails, strengthened when unanticipated problems show themselves, changed to accommodate the changing economy and upgraded when required. Matters are especially complex when it comes to sustaining decentralization reforms in developing countries that are highly centralized because most higher-level officials are at best weak supporters of fiscal decentralization. In such circumstances, with many eager to find and publicize faults, sustaining decentralization is often as difficult as implementing it in the first place. There is no simple way to deal with any of these matters. Central to sustaining any decentralization reform is a certain degree of commitment at the central government level. In addition to obeying the rules itself, the center must be capable of and prepared to monitor outcomes at the subnational level, and to discipline those who do not comply. It is not enough to have the right legislation in place; it must also be enforced. As discussed in Chapter 5, the most important rule that local governments must obey is to balance the budget according to the rules that should be clearly set out in legislation. If a local government does not comply, the law might call for a

M4439-BAHL_9781786435293_t.indd 418

01/03/2018 11:08



Giving decentralization a chance ­419

variety (or sequence) of reactions such as intercepts of intergovernmental transfers, imposition of a special tax, curtailment of expenditures or, in the extreme, the appointment of a special receivership. Other important rules also need to be monitored and enforced at both central and subnational levels, including: ●●

●●

●●

●●

Certain fiscal measures such as preferential tax treatments and even conditional grants, whether imposed at the local level or above, should have a sunset provision and must be evaluated before they continue for another period. The central government should not be allowed to fail to pay grant entitlements in a timely fashion or to impose unfunded mandates. Higher-level governments make the rules and should be required to obey the rules they make. Subnational governments should be penalized if they fail to comply with the requirements of conditional grants or uniform accounting and reporting rules. But higher-level governments should be prepared to provide training and support to help poorer localities comply, and may consider perhaps asymmetrically devolving expenditures to match the capacity of localities to comply with the rules. Corruption by politicians and officials at all levels should be dealt with by investigative units (at higher levels when appropriate) as well as by the courts.

As was discussed above, a monitoring unit for subnational government finances should be established in the higher-level government and should be empowered with a comprehensive data base on local government finances. It should report annually on the current state of subnational government fiscal affairs in the country. This report should evaluate the revenue and expenditure performance of subnational governments, identify those that are stressed and focus on what might be done to reduce unwanted disparities. The report should be widely disseminated. This is a crucial dimension of the maintenance of a decentralized system, and in effect provides an early warning system for problems. Subnational government administrative capacity in some countries, and especially in larger urban areas, has improved markedly since the mid-1990s, as has their ability to absorb new technologies. In most cases, however, the ability of local governments to effectively manage and plan is still weak. Making decentralization sustainable requires focusing on strengthening this capacity with continuous training to upgrade skills. An institutional framework for the training of local government officers is invariably a key component of any sustainable decentralization program.

M4439-BAHL_9781786435293_t.indd 419

01/03/2018 11:08

420

Fiscal decentralization and local finance in developing countries

SOME FINAL THOUGHTS In most developing countries, fiscal decentralization has not amounted to much. The advantages of centralization and the political power of those controlling the central government have been too strong. But the world has changed, and the case for decentralization is arguably becoming more difficult to ignore. Macroeconomic instability may slow matters down, but decentralization may perhaps be about to have its day. Governments in many developing countries are now elected, often on platforms that at least talk about increased citizen participation in governance; economic growth may to some extent have eroded some of the traditional arguments in favor of fiscal centralization; and the service delivery capabilities of at least the larger urban local governments have improved dramatically in many countries. But those who think increasing local autonomy is often one good way to go still face substantial obstacles. A major roadblock in many countries is that some past moves towards decentralization have been poorly conceived and implemented. Design must match objectives, and implementation must face up to the many dimensions of decentralization. Each country must face its own problems, largely on its own. However, as we have argued in this book, the problems encountered around the world share many common elements; and the solutions to them, though inevitably and correctly shaped and attuned to local circumstances, also have common elements. Everyone’s problem is different. But everyone faces similar problems in some respects, and the solutions to those problems in many cases to some extent lie along the path we have laid out in this book. We do not have ‘the answer.’ No one does. But we hope that those who are looking for answers to their problems in this area will find some useful guidance in this book.

NOTES   1. IMF (2011) shows a small increase in the tax-to-GDP ratio for resource-rich countries in the Middle East and North Africa over the 1980–2008 period, but little movement in the rest of the developing world.   2. For a good discussion of the issues, see Canuto and Liu (2013).   3. One of us once wrote a paper on the rules for successful decentralization that received a fair amount of attention (Bahl, 2002). While he was in Indonesia on other matters, the mission economist from the World Bank asked that he present to an expert group of Indonesian officials and experts. The consensus of the group was that many of these rules were not followed in the Indonesia reform but that the decentralization had nonetheless been successful. There are several possible takeaways from this experience: country experts are better equipped than foreign advisors at making the rules; the same

M4439-BAHL_9781786435293_t.indd 420

01/03/2018 11:08



  4.

  5.

  6.   7.

  8.

  9.

10. 11.

Giving decentralization a chance ­421 general rules for decentralization do not fit all countries equally well; politics trumps economics; or Indonesia may not yet be finished with its decentralization and the rules may yet be on the mark. The other author – not present at either the creation or this meeting – thinks all these observations are valid to some extent: the rules do not quite fit any country; locals know more about local circumstances; politics always dominates; and Indonesia still has far to go. Nonetheless, guidelines like those proposed here may provide a useful template for all engaged in or thinking about fiscal decentralization in developing countries. Of course, as many have noted, locally elected governments often do not go along with the wishes of their constituents (Weingast, 2009; Lockwood, 2006; Hettich and Winer, 1999). However, imperfect as the electoral system undoubtedly is, on average it seems plausibly more likely to reflect the wishes of local constituents than any other system that comes to mind. If a more comprehensive, unified budget approach (perhaps even one on an accrual basis) is chosen, as some have urged, the central government still needs to set out and enforce clear rules on subnational government budgeting and accounting. As noted in Chapter 3, however, we think this approach is at present seldom advisable at the local level in developing countries. Ideally, such reviews should not just be at the central government level, but also by intergovernmental as well as independent agencies. Matters are often a bit different in federal countries where detailed information is often available on federal-regional and on regional finances, in part because it is required constitutionally and is a key ingredient in the ongoing negotiations that usually characterize federations. However, in most federations, since local finance is left almost entirely to regional governments, it can be surprisingly difficult to obtain complete, up-to-date or even uniform local financial information for the whole country. As Faguet (2014) and Sánchez de Lozada and Faguet (2015) stress, the most critical aspects of decentralization are those related to governance, and the basic empowerment of the local government level in Bolivia has turned out to be the most lasting and resilient element of the 1994 decentralization reform. To a lesser extent the same is true in Colombia, where the most critical and lasting reform was arguably the popular election of local government leaders. Interestingly, this issue had been the only major reform on which the expert committee, whose report (DNP, 1981) was arguably the impetus for the major decentralization reform a decade later, could not reach agreement because of the  strong dissent from some conservative members. A few years later, however, when  the political party favoring such elections took office, it pushed this measure through; within a few years local popular government became an established part of the political landscape. Some international agencies might also perhaps do well to reconsider their own staff development policies, which have all too often seemed to shift staff from areas which they were beginning to understand and where they could work effectively to areas where they knew as little as anyone else. Agencies seem to have a deep fear of staff ‘going native,’ which they seem to interpret as agreeing too much with locals and not being fully on board in pushing whatever the latest top-down change in agency policy happens to be. They would often do better to listen to what the better of these at least partly acclimatized ‘natives’ have to tell them about the real problems and conditions in the countries they are supposedly concerned with helping. This section draws heavily on Bahl and Martinez-Vazquez (2006a). As is discussed in Chapter 8, in the case of metropolitan areas – where various approaches to structuring governance in the metropolitan area are possible – this rule may be stated more accurately as ‘finance follows function follows government structure.’

M4439-BAHL_9781786435293_t.indd 421

01/03/2018 11:08

References Acosta, Olga Lucia and Richard Bird (2005) “The Dilemma of Decentralization in Colombia,” in Richard Bird, James Poterba, and Joel Slemrod, eds, Fiscal Reform in Colombia: Problems and Prospects (Cambridge MA: MIT Press), pp. 47–86. Adam, Antonis, Manthos Delis and Pantelis Kammas (2008) “Fiscal Decentralization and Public Sector Efficiency: Evidence from OECD Countries,” CESifo Working Paper No. 2364, August. ACIR (1994) Changing Public Attitudes on Government and Taxes (Washington DC: Advisory Commission on Intergovernmental Relations). Afonso, José Roberto and Rafael Barroso (2007) “Brazilian Tax Affairs,” Berkeley Program in Law and Economics, Paper 043007, University of California. Afonso, José Roberto Rodrigues and Erika Amorim Araújo (2006) “Local Government Organization and Finance: Brazil,” in Anwar Shah, ed., Local Governance and Finance in Developing Countries (Washington DC: World Bank), pp. 381–447. Afonso, José Roberto Rodrigues, Erika Amorim Araújo and Geraldo Biasoto Júnior (2005) “Fiscal Space and Public Sector Investments in Infrastructure: A Brazilian Case-Study,” Texto para discussão no. 1141, Prepared for World Bank, Brasilia, December. Ahluwalia, Isher Judge, Ravi Kanbur and P.K. Mohanty, eds (2014) Urbanisation in India: Challenges, Opportunities and the Way Forward (New Delhi: Sage). Ahmad, Ehtisham, ed. (1997) Financing Decentralized Expenditures: An International Comparison of Grants (Cheltenham, UK and Lyme NH, USA: Edward Elgar Publishing). Ahmad, Ehtisham, Giorgio Brosio and Vito Tanzi (2008) “Local Service Provision in OECD Countries: Do Decentralized Operations Work Better?,” Working Paper 08/67, International Monetary Fund, Washington DC. Ahmad, Ehtisham, Jose Gonzalez Anaya, Giorgio Brosio, Mercedes Garcia-Escribano, Ben Lockwood and Ernesto Revilla (2007) “Why Focus on Spending Needs Factors: The Political Economy of Fiscal 422

M4439-BAHL_9781786435293_t.indd 422

01/03/2018 11:08



References ­423

Transfer Reforms in Mexico,” Working Paper, WP/07/252, International Monetary Fund, Washington DC. Ahmad, Ehtisham, Ben Lockwood and R. Singh (2004) “Taxation Reforms and Changes in Revenue Assignments in China,” IMF Working Paper WP/04/125, International Monetary Fund, Washington DC. Ahmad, Ehtisham and Bob Searle (2006) “On the Implementation of Transfers to Subnational Governments,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Fiscal Federalism (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 381–404. Ahmad, Junaid (1996) “The Structure of Urban Governance in South African Cities,” International Tax and Public Finance, 3: 193–213. Akerlof, George and Robert Shiller (2015) Phishing for Phools: The Economics of Manipulation and Deception (Princeton NJ: Princeton University Press). Albalate, Daniel, Germà Bel and Xavier Fageda (2012) “Beyond the Efficiency-Equity Dilemma: Centralization as a Determinant of Government Investment in Infrastructure,” Papers in Regional Science, 91 (3): 599–615. Alegre, Juan Gonzalez, Andreas Kappeler, Atanas Kolev and Timo Välilä (2008) “Composition of Government Investment in Europe: Some Forensic Evidence,” EIB Papers, 13 (1): 22–55. Aleksiene, Albino and Arvydas Bagdonavicius (2008) “Lessons from the Introduction of Value-Based Property Taxes in Lithuania,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Making the Property Tax Work: Experiences in Developing and Transitional Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 411–35. Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat and Romain Wacziarg (2003) “Fractionalization,” Journal of Economic Growth, 8: 155–94. Alexeev, Michael and Galina Kurlyandskaya (2003) “Fiscal Federalism and Incentives in a Russian Region,” Journal of Comparative Economics, 31: 20–33. Ali, Merima, Odd-Helge Fjeldstad, Boqian Jiang and Abdulaziz Shifa (2015) “Colonial Legacy, State-Building and the Salience of Ethnicity in Sub-Saharan Africa,” CMI Working Paper WP 2015:16, Bergen. Alm, James (2010) “Municipal Financing of Urban Infrastructure: Knowns and Unknowns,” Working Paper 19, Wolfensohn Center for Development, Brookings Institution, Washington DC. Alm, James, Patricia Annez and Arbind Modi (2004) “Stamp Duties in Indian States: A Case for Reform,” Policy Research Working Paper 3413, World Bank, Washington DC. Alm, James and Jameson Boex (2008) “The Role of Economic versus

M4439-BAHL_9781786435293_t.indd 423

01/03/2018 11:09

424

Fiscal decentralization and local finance in developing countries

Political Factors in the Incidence of Intergovernmental Transfers in Nigeria,” Journal of Developing Areas, 42 (1): 1–19. Alm, James and Robert Buckley (1994) “Decentralization, Privatization, and the Solvency of Local Governments in Reforming Economies: The Case of Budapest,” Environment and Planning C: Government and Policy 12 (3): 333–46. Alm, James and Jorge Martinez-Vazquez (2009) “South Africa’s Provincial Equitable Share: An Assessment of Issues and Proposals for Reform,” Working Paper 09-04, Andrew Young School of Policy Studies, Georgia State University. Alm, James, Jorge Martinez-Vazquez and Sally Wallace, eds (2004) Taxing the Hard to Tax: Lessons from Theory and Practice (Amsterdam: Elsevier). Alm, James and Sally Wallace (2004) “Payroll Taxes and Contributions in Jamaica, Jamaica Tax Reform Project,” Working Paper 6, Andrew Young School of Policy Studies, Georgia State University. Almy, Rich (2004) “IAAO Course 402: Property Tax Policy,” Section 6: The Property Tax. Alok, V.N. (2011) “Role of Panchayat Bodies in Rural Development since 1959,” Theme Paper for the 55th Members’ Annual Conference, Indian Institute of Public Administration, New Delhi. Alterman, Rachelle (2007) “More than Land Assembly: Land Readjustment for the Supply of Urban Services,” in Yu-Hung Hong and Barrie Needham, eds, Analyzing Land Readjustment: Economics, Law and Collective Action (Cambridge MA: Lincoln Institute of Land Policy), pp. 57–88. Amatong, Juanita (1968) “Taxation of Capital Gains in Developing Countries,” IMF Staff Papers, 5 (2): 344–86. Anderson, John (2009) “A Review of the Evidence,” in Richard Dye and Richard England, eds, Land Value Taxation: Theory, Evidence and Practice (Cambridge MA: Lincoln Institute of Land Policy), pp. 99–128. Andres, Luis Alberto, Dan Biller and Jordan Schwartz (2016) “The Infrastructure Gap and Decentralization,” in Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy (London: Routledge), pp. 33–53. Andrew Young School of Policy Studies (2005) “State Fiscal Management Reform Assessment: Final Report for USAID/India,” Andrew Young School of Policy Studies, Georgia State University. Annett, Anthony (2000) “Social Fractionalization, Political Instability, and the Size of Government,” Working Paper WP/00/82, International Monetary Fund, Washington DC. Annez, Patricia Clarke (2007) “Urban Infrastructure Finance from Private

M4439-BAHL_9781786435293_t.indd 424

01/03/2018 11:09



References ­425

Operators: What Have We Learned from Recent Experience?,” in George Peterson and Patricia Clarke Annez, eds, Financing Cities: Fiscal Responsibility and Urban Infrastructure in Brazil, China India, Poland and South Africa (Washington DC: World Bank), pp. 307–38. Araujo, M. Caridad, Francisco H.G. Ferreira, Peter Lanjouw and Ozler Berk (2006) “Local Inequality and Project Choice: Theory and Evidence from Ecuador,” World Bank Policy Research Working Paper No. 3997. Ardanaz, Martin and Carlos Scartascini (2011) “Why Don’t We Tax the Rich? Inequality, Legislative Malapportionment, and Personal Income Taxation around the World,” IDB Working Paper Series No. IDB-WP282, Inter-American Development Bank, November. Arretche, Marta (2013) “Governance and Finance in Two Brazilian Metropolitan Areas,” in Enid Slack and Rupak Chattopadhyay, eds, Finance and Governance of Metropolitan Areas in Federal Systems (Toronto: Oxford University Press), pp. 49–76. Artana, Daniel, Sebastian Auguste, Marcela Cristini Cynthia Moskovits and Ivana Templado (2015) “Two to Tango: Argentina’s Vertical Imbalance and Revenue Mobilization,” in Vicente Fretes Ciblis and Teresa Ter-Minassian, eds, Decentralizing Revenue in Latin America: Why and How (Washington DC: Inter-American Development Bank), pp. 41–78. Arzaghi, Mohammad and Vernon Henderson (2005) “Why Countries are Fiscally Decentralizing,” Journal of Public Economics, 89: 1157–89. Arze del Granado, Javier, Jorge Martinez-Vazquez and Renata Simatupang (2008) “Local Fiscal Competition in Developing Countries: The Case of Indonesia,” Urban Public Economics Review, 8: 13–45. Asthana, Anand (2003) “Decentralisation and Supply Efficiency: The Case of Rural Water Supply in Central India,” Journal of Development Studies, 39 (4): 148–59. Azfar, Omar and J. Livingston (2002) “Federalist Disciplines or Local Capture? An Empirical Analysis of Decentralization in Uganda,” IRIS Working Paper 256, University of Maryland. Azfar, Omar, Satu Kahkonen and Patrick Meagher (2001) “Conditions for Effective Decentralized Governance: A Synthesis of Research Findings,” IRIS Center, University of Maryland, March. Bach, S., H. Blöchliger and D. Wallau (2009) “The Spending Power of SubCentral Governments: A Pilot Study,” OECD Economics Department Working Paper No. 705. Backhaus, Jurgen (2000) “Germany,” in Robert Andelson, ed., LandValue Taxation around the World, 3rd edn (Malden MA: Blackwell), pp. 221–38.

M4439-BAHL_9781786435293_t.indd 425

01/03/2018 11:09

426

Fiscal decentralization and local finance in developing countries

Bahl, Roy (1972) “A Representative Tax System Approach to Measuring Tax Effort in Developing Countries,” IMF Staff Papers, 19 (1): 87–124. Bahl, Roy, ed. (1979) The Taxation of Urban Property in Less Developed Countries (Madison: University of Wisconsin Press). Bahl, Roy, ed. (1991) The Jamaican Tax Reform (Cambridge MA: Lincoln Institute of Land Policy). Bahl, Roy (1994) “Revenues and Revenue Assignment: Intergovernmental Fiscal Relations in the Russian Federation,” in Christine I. Wallich, ed., Russia and the Challenge of Fiscal Federalism (Washington DC: World Bank). Bahl, Roy (1999) Fiscal Policy in China: Taxation and Intergovernmental Fiscal Relations (South San Francisco: 1990 Institute). Bahl, Roy (2002) “Implementation Rules for Fiscal Decentralization,” in Govinda Rao, ed., Development, Poverty and Fiscal Policy: Decentralization of Institutions (New Delhi: Delhi University Press), pp. 253–77. Bahl, Roy (2004) “Property Transfer Tax and Stamp Duty,” Jamaica Comprehensive Tax Study, Paper No. 2, Andrew Young School of Policy Studies, Georgia State University. Bahl, Roy (2008) “Opportunities and Risks of Fiscal Decentralization: A Developing Country Perspective,” in Gregory Ingram, ed., Achieving Decentralization Objectives (Cambridge MA: Lincoln Institute of Land Policy), pp. 19–37. Bahl, Roy (2009) Property Tax Reform in Developing and Transition Countries (Washington DC: USAID). Bahl, Roy (2010) “Jordan: The Taxation of Real Property,” Report submitted to USAID/Jordan Fiscal Reform Project II, September. Bahl, Roy (2010a) “Conditional vs Unconditional Grants in Developing Countries,” in Junghun Kim, Jørgen Lotz and Niels Jørgen Mau, eds, General Grants versus Earmarked Grants Theory and Practice: The Copenhagen Workshop 2009 (Albertslund: Korea Institute of Public Finance and Danish Ministry of Interior and Health), pp. 126–48. Bahl, Roy (2011) “Financing Metropolitan Areas,” in Jorge MartinezVazquez and Paul Smoke, eds, Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 285–308. Bahl, Roy (2011a) “Financing Subnational Governments with Decentralized Taxes,” in Emilio Albi and Jorge Martinez-Vazquez, eds, The Elgar Guide to Tax Systems (Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing), pp. 328–59. Bahl, Roy (2013) “The Decentralization of Governance in Metropolitan Areas,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing

M4439-BAHL_9781786435293_t.indd 426

01/03/2018 11:09



References ­427

Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 85–106. Bahl, Roy (2014) “A Retrospective on Taxation in Developing Countries: Will the Weakest Link be Strengthened?,” in Richard Bird and Jorge Martinez-Vazquez, eds, Taxation and Economic Development: The Weakest Link (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 405–33. Bahl, Roy (2017) “Metropolitan City Finances in the Asia and Pacific Region, Background Document for 4th High Level Dialogue on Financing for Development in Asia and the Pacific,” United Nations and ESCAP, Bangkok. Bahl, Roy and Richard Bird (2008) “Subnational Taxes in Developing Countries: The Way Forward,” Public Budgeting and Finance, 28 (4): 1–25. Bahl, Roy and Richard Bird (2016) “Principles and Practice,” in Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 101–45. Bahl, Roy and Alan K. Campbell (1976) “The Implications of Local Government Reform: Efficiency, Equity, and Administrative Dimensions,” in Roy Bahl and Alan K. Campbell, eds, State and Local Government: The Political Economy of Reform (New York: Free Press), pp. 185–203. Bahl, Roy and Musharraf Cyan (2009) “Egypt: Proposal for a Grants Commission at the Central and Governorate Levels,” USAID and Andrew Young School of Policy Studies, International Studies Program, Georgia State University, Working Paper, April 21. Bahl, Roy and Musharraf Cyan (2011) “Tax Assignment: Does the Practice Match the Theory” Environment and Planning C: Government and Policy, 29 (2): 264–80. Bahl, Roy, Musharraf Cyan and Sally Wallace (2011) “Challenges to Intergovernmental Fiscal Relations in Pakistan: The Revenue Assignment Dimension,” in Jorge Martinez-Vazquez and François Vaillancourt, eds, Decentralization in Developing Countries (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing), pp. 131–54. Bahl, Roy, Musharraf Cyan and Sally Wallace (2015) “The Potential of Provincial Taxation,” in Jorge Martinez-Vazquez and Musharraf Cyan, eds, The Role of Taxation in Pakistan’s Economic Revival (Karachi: Oxford University Press), pp. 488–548. Bahl, Roy, Musharraf Cyan and Sally Wallace (2015a) “Underutilized Land and Property Taxation,” in Jorge Martinez-Vazquez and Musharraf Cyan, eds, The Role of Taxation in Pakistan’s Economic Revival (Karachi: Oxford University Press), pp. 387–487. Bahl, Roy and William Duncombe (1991) Economic Growth and Fiscal Planning: New York in the 1990s (New Brunswick NJ: Rutgers).

M4439-BAHL_9781786435293_t.indd 427

01/03/2018 11:09

428

Fiscal decentralization and local finance in developing countries

Bahl, Roy and Hernando Garzon (2010) “Panama Property Tax Report,” Report prepared for the World Bank, Latin America Department, June. Bahl, Roy, Chor-Ching Goh and Baoyun Qiao (2014) Reforming the Public Finance System to Fit a More Urbanized China (Beijing: China Financial and Economic Publishing House). Bahl, Roy, Chuk Kyo Kim and Chong Kee Park (1986) Public Finances During the Korean Modernization Process (Cambridge MA: Harvard University Press). Bahl, Roy and Johannes Linn (1983) “The Assignment of Local Government Revenues in Developing Countries,” in Charles McLure, ed., Tax Assignment in Federal Countries (Canberra: Australian National University Press), pp. 177–99. Bahl, Roy and Johannes Linn (1992) Urban Public Finance in Developing Countries (New York: Oxford University Press). Bahl, Roy and Johannes Linn (2015) “Delivering and Financing Public Services in Metropolitan Areas,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilevel Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 334–57. Bahl, Roy, Johannes Linn and Deborah Wetzel, eds (2013) Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy). Bahl, Roy and Jorge Martinez-Vazquez (2006) “Fiscal Federalism and Economic Reform in China,” in Jessica Wallack and T.N. Srinivasan, eds, Federalism and Economic Reform: International Perspectives (Cambridge: Cambridge University Press), pp. 249–300. Bahl, Roy and Jorge Martinez-Vazquez (2006a) “Sequencing Fiscal Decentralization,” Policy Research Paper 3914, World Bank, Washington DC. Bahl, Roy and Jorge Martinez-Vazquez (2008) “The Property Tax in Developing Countries: Current Practice and Prospects,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Making the Property Tax Work: Experiences in Developing and Transitional Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 35–57. Bahl, Roy, Jorge Martinez-Vazquez and Sally Wallace (2002) “State and Local Government Choices in Fiscal Redistribution,” National Tax Journal, 55 (4): 723–42. Bahl, Roy, Jorge Martinez-Vazquez and Joan Youngman, eds (2008) Making the Property Tax Work: Experiences in Developing and Transitional Countries (Cambridge MA: Lincoln Institute of Land Policy). Bahl, Roy, Jorge Martinez-Vazquez and Joan Youngman, eds (2010) Challenging the Conventional Wisdom about the Property Tax (Cambridge MA: Lincoln Institute of Land Policy).

M4439-BAHL_9781786435293_t.indd 428

01/03/2018 11:09



References ­429

Bahl, Roy and Shyam Nath (1986) “Public Expenditure Decentralization in Developing Countries,” Environment and Planning C: Government and Policy, 4: 405–18. Bahl, Roy and Geeta Sethi, (2012) “Fiscal Decentralization in Latin America: An Overview,” Public Sector and Governance Unit (LCSPS), Poverty Reduction and Economic Management Unit, Latin America and the Caribbean, World Bank, Washington DC (draft, June). Bahl, Roy, Geeta Sethi and Sally Wallace (2010a) “Fiscal Decentralization to Rural Local Governments in India: A Case Study of West Bengal,” Publius: The Journal of Federalism, 40 (2): 312–37. Bahl, Roy and Paul Smoke, eds (2003) Restructuring Local Government Finance in Developing Countries: Lessons from South Africa (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing). Bahl, Roy and Paul Smoke (2003a) “Overview of Fiscal Decentralization in South Africa,” in Roy Bahl and Paul Smoke, eds, Restructuring Local Government Finance in Developing Countries: Lessons from South Africa (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 1–22. Bahl, Roy and David Solomon (2003) “The Regional Service Council Levy,” in Roy Bahl and Paul Smoke, eds, Restructuring Local Government Finance in Developing Countries: Lessons from South Africa (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 127–72. Bahl, Roy and Bayar Tummenasan (2004) “How Should Revenues from Natural Resources Be Shared in Indonesia?,” Oil, Gas and Energy Law (OGEL), 3, www.ogel.org/article.asp?key51290. Bahl, Roy and Sally Wallace (2005) “Public Financing in Developing and Transition Countries,” Public Budgeting and Finance, 25 (4): 83–98. Bahl, Roy and Sally Wallace (2007) “Intergovernmental Transfers: The Vertical Sharing Dimension,” in Jorge Martinez-Vazquez and Bob Searle, eds, Fiscal Equalization: Challenges in the Design of Intergovernmental Transfers (New York: Springer), pp. 205–50. Bahl, Roy, Sally Wallace and Musharraf Cyan (2008) “Provincial Taxation in Pakistan,” International Studies Program, Georgia State University. Bahl, Roy and Christine Wallich (1992) “Intergovernmental Fiscal Relations in China,” Policy Research Working Paper No. 863, World Bank, Washington DC. Baker, Judy L. and Kim McClain (2009) “Private Sector Initiatives in Slum Upgrading,” Urban Papers UP-8, World Bank, May. Bandyopadhyay Simanti and Govinda Rao (2009) “Fiscal Health of Selected Indian Cities,” Policy Research Working Paper No. 4863, World Bank, Washington DC.

M4439-BAHL_9781786435293_t.indd 429

01/03/2018 11:09

430

Fiscal decentralization and local finance in developing countries

Banfield, Edward and James Wilson (1963) City Politics (New York: Vintage). Barankay, Iwan and Ben Lockwood (2006) “Decentralization and the Productive Efficiency of Government: Evidence from Swiss Cantons,” IZA DP No. 2477. Bardhan, Pranab and Dilip Mookherjee (2000) “Capture and Governance at Local and National Levels,” American Economic Review, 90 (2): 135–9. Bardhan, Pranab and Dilip Mookherjee, eds (2006) Decentralization and Local Governments in Developing Countries: A Comparative Perspective (Cambridge MA: MIT Press). Baretti, Christian, Bernd Huber and Karl Lichtblau (2002) “A Tax on Tax Revenue: The Incentive Effects of Equalizing Transfers: Evidence from Germany,” International Tax and Public Finance, 9 (6): 631–49. Barlow, John and Manfred Rober (1996) “Steering not Rowing: Co-ordination and Control in the Management of Public Services in Britain and Germany,” International Journal of Public Sector Management, 9 (5/6): 73–89. Barter, Paul (2010) Parking Policies in Asian Cities (Manila: Asian Development Bank). Bartle, John, R., W. Bartley Hildreth and Justin Marlowe, eds (2012) Management Policies in Local Government Finance, 6th edn (Washington DC: ICMA Press). Bartolini, David, Sibylle Stossberg and Hansjörg Blöchliger (2016) “Fiscal Decentralisation and Regional Disparities,” OECD Economics Department Working Paper No. 1330. Baskaran, Tushyanthan, Lars Feld and Jan Schnellenbach (2016) “Fiscal Federalism, Decentralization and Economic Growth: A Meta-Analysis,” Freiburger Diskussionspapiere zur Ordnunsokonomik No. 16/02, available at http://citeseerx.ist.psu.edu/viewdoc/download?doi510.1.1.399.1 915&rep5rep1&type5pdf. Basurto, Pia, Pascaline Dupas and Jonathan Robinson (2017) “Decentralization and Efficiency of Subsidy Targeting: Evidence from Chiefs in Rural Malawi,” NBER Working Paper No. 23383. Bednar, Jenna (2009) The Robust Federation: Principles of Design (Cambridge: Cambridge University Press). Begović, Boris, ed. (2004), “Property Taxation in Serbia: Ownership Rights and Interests in Real Property Subject to Taxation, Fiscal Decentralization Initiative for Central and Eastern Europe,” Open Society Institute, Budapest. Bel, Germa (2013) “Local Government Size and Efficiency in CapitalIntensive Services: What Evidence Is There of Economies of Scale, Density and Scope?,” in Santiago Lago-Peñas and Jorge Martinez-Vazquez,

M4439-BAHL_9781786435293_t.indd 430

01/03/2018 11:09



References ­431

eds, The Challenge of Local Government Size (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 148–70. Bel, Germà and Xavier Fageda (2009) “Preventing Competition Because of Solidarity: Rhetoric and Reality of Airport Investments in Spain,” Applied Economics, 41 (22): 2853–65. Beland, Daniel, André Lecours, Gregory P. Marchildon, Haizhen Mou and M. Rose Olfer (2017) Fiscal Federalism and Equalization Policy in Canada: Political and Economic Dimensions (Toronto: University of Toronto Press). Bell, Michael, John Bowman and Jerome German (2009) “The Assessment Requirement for a Separate Tax on Land,” in Richard Dye and Richard England, eds, Land Value Taxation: Theory, Evidence and Practice (Cambridge MA: Lincoln Institute of Land Policy). Beramendi, Paolo (2012) The Political Geography of Inequality: Regions and Redistribution (Cambridge: Cambridge University Press). Bernstein, Thomas and Xioabo Lu (2003) Taxation Without Representation in Contemporary Rural China (Cambridge: Cambridge University Press). Berry, Christopher R. (2009) Imperfect Union: Representation and Taxation in Multilevel Governments (New York: Cambridge University Press). Besley, Timothy and Anne Case (1995) “Incumbent Behavior: VoteSeeking, Tax-Setting, and Yardstick Competition,” American Economic Review, 85: 25–45. Bhattacharyya, Sambit, Louis Conradie and Rabah Arezki (2016) “Resource Discovery and the Politics of Fiscal Decentralization,” CSAE Working Paper WPS/2016-05, Centre for the Study of African Economies, University of Oxford, March. Bird, Richard (1967) “Stamp Tax Reform in Colombia,” Bulletin for International Fiscal Documentation, 21: 247–55. Bird, Richard (1970) Taxation and Development: Lessons from Colombian Experience (Cambridge MA: Harvard University Press). Bird, Richard (1970a) The Growth of Government Spending in Canada (Toronto: Canadian Tax Foundation). Bird, Richard (1974) Taxing Agricultural Land in Developing Countries (Cambridge MA: Harvard University Press). Bird, Richard (1976) Charging for Public Services: A New Look at an Old Idea (Toronto: Canadian Tax Foundation). Bird, Richard (1976a) “Assessing Tax Performance in Developing Countries: A Critical Review of the Literature,” Finanzarchiv, 34 (2): 244–65. Bird, Richard (1980) Central-Local Fiscal Relations and the Provision of Urban Public Services (Canberra: Centre for Research on Federal Financial Relations, ANU).

M4439-BAHL_9781786435293_t.indd 431

01/03/2018 11:09

432

Fiscal decentralization and local finance in developing countries

Bird, Richard (1982) “Taxation and Employment in Developing Countries,” Finanzarchiv, 40 (2, 1982): 211–39. Bird, Richard (1983) “The Allocation of Taxing Powers in Papua New Guinea,” Institute of National Affairs Discussion Paper No. 15, Port Moresby, PNG. Bird, Richard (1984) Intergovernmental Finance in Colombia (Cambridge MA: Harvard Law School International Tax Program). Bird, Richard (1986) Federal Finance in Comparative Perspective (Toronto: Canadian Tax Foundation). Bird, Richard (1991) “The Taxation of Services,” in Roy Bahl, ed., The Jamaican Tax Reform (Cambridge MA: Lincoln Institute of Land Policy), pp. 587–600. Bird, Richard (1993) “Threading the Fiscal Labyrinth: Some Issues in Fiscal Decentralization,” National Tax Journal, 46 (2): 207–27. Bird, Richard (1994) “A Comparative Perspective on Federal Finance,” in Keith Banting, Douglas Brown and Thomas Courchene, eds, The Future of Fiscal Federalism (Kingston, Ont: School for Policy Studies, Queen’s University), pp. 293–322. Bird, Richard (1995) “Decentralizing Infrastructure: For Good or for Ill?,” in Antonio Estache, ed., Decentralizing Infrastructure: Advantages and Limitations (Washington DC: World Bank), pp. 22–51. Bird, Richard (1995a) “Financing Local Services: Patterns, Problems, and Possibilities,” Major Report 31, Centre for Urban and Community Studies, University of Toronto, February. Bird, Richard (2000) “Rethinking Subnational Taxes: A New Look at Tax Assignment,” Tax Notes International, 20 (May 8): 2069–96. Bird, Richard (2000a) “Fiscal Decentralization and Competitive Governments,” in G. Galeotti, P. Salmon and R. Wintrobe, eds, Competition and Structure: The Political Economy of Collective Decisions. Essays in Honor of Albert Breton (Cambridge: Cambridge University Press), pp. 129–49. Bird, Richard (2001) “User Charges in Local Government Finance,” in Richard Stren and Maria Emilia Freire, eds, The Challenge of Urban Government (Washington DC: World Bank Institute, 2001), pp. 171–82. Bird, Richard (2001a) Intergovernmental Fiscal Relations in Latin America: Policy Designs and Policy Outcomes (Washington DC: Inter-American Development Bank). Bird, Richard (2005) “Evaluating Public Expenditures: Does It Matter How They Are Financed?,” in Anwar Shah, ed., Fiscal Management (Washington DC: World Bank, 2005), pp. 83–108. Bird, Richard (2005a) “Getting It Right: Financing Urban Development in China,” Asia-Pacific Tax Bulletin, 11 (2): 107–17.

M4439-BAHL_9781786435293_t.indd 432

01/03/2018 11:09



References ­433

Bird, Richard (2006) “Fiscal Flows, Fiscal Balance, and Fiscal Sustainability,” in Richard Bird and François Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington DC: World Bank), pp. 81–97. Bird, Richard (2006a) “Local Business Taxes,” in Richard Bird and François  Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington, DC: World Bank), pp. 225–46. Bird, Richard (2009) “Tax Assignment Revisited,” in John Head and Richard Krever, eds, Tax Reform in the 21st Century: A Volume in Memory of Richard Musgrave (Austin TX: Wolters Kluwer), pp. 441–70. Bird, Richard (2010) “Subnational Taxation in Developing Countries: A Review of the Literature,” Policy Research Working Paper 5450, World Bank, Washington DC. Bird, Richard (2011) “Subnational Taxation in Developing Countries: A Review of the Literature,” Journal of International Commerce, Economics and Policy, 2 (1), 139–61. Bird, Richard (2011a) “Are There Trends in Local Finance? A Cautionary Note on Comparative Studies and Normative Models of Local Government Finance,” IMFG Papers on Municipal Finance and Governance No. 1, Toronto. Bird, Richard (2012) “Subnational Taxation in Large Emerging Countries: BRIC Plus One,” IMFG Papers on Municipal Finance and Governance No. 6, Toronto. Bird, Richard (2012a) “Fiscal Decentralization in Colombia: A Work (Still) in Progress,” Working Paper 1223, International Studies Program, Andrew Young School of Public Policy, Georgia State University. Bird, Richard (2014) “A Better Local Business Tax: The BVT,” IMFG Papers on Municipal Finance and Governance No. 18, Toronto. Bird, Richard (2015) “Tax Decentralization and Decentralizing Tax Administration: Different Questions, Different Answers,” Working Paper 1509, International Center for Public Policy, Andrew Young School of Public Policy, Georgia State University. Bird, Richard (2015a) “Below the Salt: Decentralizing Value-Added Taxes,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing), pp. 291–333. Bird, Richard (2017) “Why We Should but Don’t Pay the Right Prices for Urban Infrastructure,” in Richard Bird and Enid Slack, eds, Financing Infrastructure: Who Should Pay? (Montreal: McGill-Queen’s University Press), pp. 238–71. Bird, Richard, Loren Brandt, Scott Rozelle and Linxiu Zhang (2011) “Fiscal Reform and Rural Public Finance in China,” in Joyce Yangyn

M4439-BAHL_9781786435293_t.indd 433

01/03/2018 11:09

434

Fiscal decentralization and local finance in developing countries

Man and Yu-Hung Hong, eds, China’s Local Public Finance in Transition (Cambridge, MA: Lincoln Institute for Land Policy), pp. 227–43. Bird, Richard and Duanjie Chen (1998) “Federal Finance and Fiscal Federalism: The Two Worlds of Canadian Public Finance,” Canadian Public Administration, 43 (1), 51–74. Bird, Richard, Bernard Dafflon, Claude Jeanrenaud and Gebhard Kirchgassner (2003) “Assignment of Responsibilities and Fiscal Federalism,” in Raoul Blindenbacher and Arnold Koller, eds, Federalism in a Changing World: Learning from Each Other (Montreal: McGillQueen’s University Press), pp. 351–72. Bird, Richard, Robert Ebel and Christine Wallich, eds (1995) Decentralization of the Socialist State: Intergovernmental Finance in Transition Economies (Washington DC: World Bank). Bird, Richard and Robert Ebel, eds (2007) Fiscal Fragmentation in Decentralized Countries (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing). Bird, Richard and Ariel Fiszbein (1998) “Fiscal Decentralization in Colombia: The Central Role of the Central Government,” in Richard Bird and François Vaillancourt, eds, Fiscal Federalism in Developing Countries (Cambridge: Cambridge University Press), pp. 172–205. Bird, Richard and Pierre-Pascal Gendron (2007) The VAT in Developing and Transitional Countries (Cambridge: Cambridge University Press). Bird, Richard and Pierre-Pascal Gendron (2010) “Sales Taxes in Canada: The GST-HST-QST-RST ‘System,’” Tax Law Review, 63 (3): 517–82. Bird, Richard and Joosung Jun (2007) “Earmarking in Theory and Korean Practice,” in Stephen L.H. Phua, ed., Excise Taxation in Asia (Singapore: National University of Singapore), pp. 49–86. Bird, Richard and Kenneth McKenzie (2001) “Taxing Business: A Provincial Affair?,” C.D. Howe Institute, Toronto, Commentary No. 154, November. Bird, Richard and Barbara Miller (1989) “Taxation, Pricing and the Poor,” in Richard Bird and Susan Horton, eds, Government Policy and the Poor in Developing Countries (Toronto: University of Toronto Press), pp. 49–80. Bird, Richard and Barbara Miller (1989a) “The Incidence of Indirect Taxation on Low-Income Households in Jamaica,” Economic Development and Cultural Change, 37 (2): 393–409. Bird, Richard and Edgard R. Rodriguez (1999) “Decentralization and Poverty Alleviation: International Experience with reference to the Philippines,” Public Administration and Development, 19 (3): 199–219. Bird, Richard and Enid Slack (2004) International Handbook of Land and Property Taxation (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing).

M4439-BAHL_9781786435293_t.indd 434

01/03/2018 11:09



References ­435

Bird, Richard and Enid Slack (2007) “An Approach to Metropolitan Governance and Finance,” Environment and Policy C: Government and Policy, 25 (5): 729–55. Bird, Richard and Enid Slack (2008) “Property Tax and Rural Local Finance,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Making the Property Tax Work (Cambridge, MA: Lincoln Institute of Land Policy), pp. 103–26 Bird, Richard and Enid Slack (2008a) “Fiscal Aspects of Metropolitan Governance,” in Eduardo Rojas, Juan R. Cuadrado-Roura and José Miguel Fernández Güell, eds, Governing the Metropolis: Principles and Cases (Washington DC: Inter-American Development Bank), pp. 193–259. Bird, Richard and Enid Slack (2013) “Metropolitan Public Finance: An Overview,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Metropolitan Government Finance in Developing Countries (Cambridge, MA: Lincoln Institute of Land Policy), pp. 135–58. Bird, Richard and Enid Slack (2014) “Local Taxes and Local Expenditures in Developing Countries: Strengthening the Wicksellian Connection,” Public Administration and Development, 34 (4): 359–69. Bird, Richard and Enid Slack (2015) Is Your City Healthy? Measuring Urban Fiscal Health. (Toronto: Institute on Municipal Finance and Governance). Bird, Richard and Enid Slack, eds (2017) Financing Infrastructure: Who Should Pay? (Montreal: McGill-Queen’s University Press). Bird, Richard, Enid Slack and Almos Tassonyi (2012) A Tale of Two Taxes: Reforming the Property Tax in Ontario (Cambridge MA: Lincoln Institute of Land Policy). Bird, Richard and Michael Smart (2002) “Intergovernmental Fiscal Transfers: Lessons from International Experience,” World Development, 30 (6): 899–912. Bird, Richard and Michael Smart (2014) “Financing Social Expenditures in Developing Countries: Payroll or Value-Added Taxes?,” in Markus Frölich, David Kaplan, Carmen Pagés, Jamele Rigolini and David Robalini, eds, Social Security, Informality and Labor Markets: How to Protect Workers While Creating Good Jobs (Oxford University Press, 2014), pp. 411–41. Bird, Richard and Andrey Tarasov (2004) “Closing the Gap: Fiscal Imbalances and Intergovernmental Transfers in Developed Federations,” Environment and Policy C: Government and Policy, 22: 77–102. Bird, Richard and Almos Tassonyi (2001) “Constraints on Provincial and Municipal Borrowing in Canada: Markets, Rules, and Norms,” Canadian Public Administration, 44 (1): 84–109.

M4439-BAHL_9781786435293_t.indd 435

01/03/2018 11:09

436

Fiscal decentralization and local finance in developing countries

Bird, Richard and Almos Tassonyi (2003) “Constraining Subnational Fiscal Behavior in Canada: Different Approaches, Similar Results?,” in Jonathan A. Rodden, Gunnar S. Eskeland and Jennie Litvack, eds, Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA: MIT Press), pp. 85–132. Bird, Richard and Thomas Tsiopoulos (1997) “User Charges for Public Services: Potentials and Problems,” Canadian Tax Journal, 45: 25–86. Bird, Richard and François Vaillancourt, eds (1998) Fiscal Decentralization in Developing Countries (Cambridge, UK: Cambridge University Press). Bird, Richard and François Vaillancourt (2006) “Changing with the Times: Success, Failure and Inertia in Canadian Federal Arrangements, 1945–2002,” in Jessica S. Wallack and T.N. Srinivasan, eds, Federalism and Economic Reform: International Perspectives (Cambridge, UK: Cambridge University Press), pp. 189–248. Bird, Richard and Sally Wallace (2004) “Is it Really So Hard to Tax the Hard to Tax? The Context and Role of Presumptive Taxes,” in James Alm, Jorge Martinez-Vazquez and Sally Wallace, eds, Taxing the Hard to Tax: Lessons from Theory and Practice (Amsterdam: Elsevier), pp. 121–58. Bird, Richard and Christine Wallich (1992) “Financing Local Government in Hungary,” Policy Research Working Paper WPS 869, World Bank, Washington DC. Bird, Richard and Thomas Wilson (2016) “The Corporate Tax in Canada: Does Its Past Foretell Its Future?,” School of Public Policy University of Calgary SPP Research Papers, vol. 9, issue 38, December. Bird, Richard and Eric Zolt (2005) “Redistribution via Taxation: The Limited Role of the Personal Income Tax in Developing Countries,” UCLA Law Review, 52 (6): 1627–95. Bird, Richard and Eric Zolt (2008) “Technology and Taxation in Developing Countries: From Hand to Mouse,” National Tax Journal, 61 (4): 791–821. Bird, Richard and Eric Zolt (2015) “Taxes, Spending and Inequality in Canada and the United States: Two Stories or One?,” Osgoode Hall Law Journal, 52 (2): 401–27. Bird, Richard and Eric Zolt (2015a) “Fiscal Contracting in Latin America,” World Development, 67: 323–35. Bizioli, Gianluigi and Claudio Sacchetto, eds (2011) Tax Aspects of Fiscal Federalism: A Comparative Analysis (Amsterdam: IBFD). Blöchliger, Hansjörg (2015) “The Challenge of Measuring Fiscal Decentralization,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilevel Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 617–32.

M4439-BAHL_9781786435293_t.indd 436

01/03/2018 11:09



References ­437

Blöchliger, Hansjörg and David King (2006) “Less Than You Thought: The Fiscal Autonomy of Sub-Central Governments,” OECD Economic Studies No. 43, Paris. Blöchliger, Hansjörg and Camila Vammalle (2010) “Intergovernmental Grants in OECD Countries: Trends and Some Policy Issues,” in Junghun Kim, Jørgen Lotz and Niels Jørgen Mau, eds, General Grants versus Earmarked Grants Theory and Practice: The Copenhagen Workshop 2009 (Albertslund: Korea Institute of Public Finance and Danish Ministry of Interior and Health), pp. 167–90. Boadway, Robin (2015) “Intergovernmental Transfers: Rationale and Policy,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 410–36. Boadway, Robin and Anwar Shah (2009) Fiscal Federalism: Principles and Practices of Multiorder Governance (Cambridge: Cambridge University Press). Boex, Jameson and Jorge Martinez-Vazquez (2006) Local Government Finance Reform in Developing Countries: The Case of Tanzania (New York: Palgrave Macmillan). Boex, Jameson and Renata R. Simatupang (2008) “Fiscal Decentralisation and Empowerment: Evolving Concepts and Alternative Measures,” Fiscal Studies, 29 (4): 435–65. Borge, Lars-Erik, Jan K. Brueckner and Jorn Rattsø (2012) “Partial Fiscal Decentralization and Public-Sector Heterogeneity: Theory and Evidence from Norway,” CESifo Working Paper No. 3954, August. Borrero Ochoa, Oscar, Esperanza Durán, Jorge Hernández and Magda Montaña (2011) “Evaluating the Practice of Betterment Levies in Colombia: The Experience of Bogota and Manizales,” Working Paper, Lincoln Institute of Land Policy, Cambridge MA. Bosch, Nuria, Marta Espasa and Albert Sole Olle (2010) The Political Economy of Inter-Regional Fiscal Flows (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing). Bossert, Thomas J. (2015) “Empirical Studies of an Approach to Decentralization: ‘Decision Space’ in Decentralized Health Systems,” in Jean-Paul Faguet and Caroline Pöschl, eds, Is Decentralization Good for Development: Perspectives from Academics and Policy Makers (Oxford: Oxford University Press), pp. 277–98. Bowles, Samuel and Herbert Gintis (2002) “Social Capital and Community Governance,” Economic Journal, 112: 419–36. Bowman, John (2002) “Current Property Tax Laws,” in Michael Bell and John H. Bowman, eds, Property Taxes in South Africa: Challenges in the

M4439-BAHL_9781786435293_t.indd 437

01/03/2018 11:09

438

Fiscal decentralization and local finance in developing countries

Post-Apartheid Era (Cambridge MA: Lincoln Institute of Land Policy), pp. 45–58. Brandt, Loren and Thomas Rawski, eds (2008) China’s Great Economic Transformation (Cambridge, UK: Cambridge University Press). Braun, Miguel and Steven Webb (2012) “Fiscal Federalism in Argentina: Recent Trends and Future Challenges,” Public Sector and Governance Unit (LCSPS), Poverty Reduction and Economic Management Unit, Latin America and the Caribbean (World Bank: Washington DC) (draft version, June). Brautigam, Deborah, Odd-Helge Fjeldstad and Mick Moore (2008) Taxation and State-Building in Developing Countries: Capacity and Consent (Cambridge, UK: Cambridge University Press). Break, George (1974) “The Incidence and Economic Effects of Taxation,” in Alan Blinder et al., The Economics of Public Finance (Washington DC: Brookings Institution), pp. 119–237. Brennan, G. and J.M. Buchanan (1980) The Power to Tax: Analytical Foundations of a Fiscal Constitution (Cambridge, UK: Cambridge University Press). Breton, Albert (1996) Competitive Governments (Cambridge, UK: Cambridge University Press). Breton, Albert (2004) “Federalization (and Not Decentralization) as an Empowerment Mechanism,” in Richard Bird, ed., Who Decides? Government in the New Millennium (Toronto: C.D. Howe Institute), pp. 107–25. Breton, Albert, Alberto Cassone and Angela Fraschini (1998) “Decentralization and Subsidiarity: Toward a Theoretical Reconciliation,” University of Pennsylvania Journal of Economic Law, 19 (1): 21–51. Breton, Albert and Anthony Scott (1978) The Economic Constitution of Federal States (Toronto: University of Toronto Press). Breuss, Fritz and Markus Eller (2004) “Fiscal Decentralisation and Economic Growth: Is There Really a Link?,” CESifo DICE Report, 2 (1): 3–9. Briggs, Asa (1996) Victorian Cities (London: Folio Society). Brodjonegoro, Bambang and Jorge Martinez-Vazquez (2005) “An Analysis of Indonesia’s Transfer System: Recent Performance and Future Prospects,” in Jorge Martinez-Vazquez, Sri Mulyani Indrawati and James Alm, eds, Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 159–98. Brosio, Giorgio (2012) “Reducing Reliance on Natural Resource Revenues and Increasing Subnational Tax Autonomy in Bolivia,” IDB Working Paper No. 198, Washington DC.

M4439-BAHL_9781786435293_t.indd 438

01/03/2018 11:09



References ­439

Brosio, Giorgio (2015) “The Sharing of Natural Resource Revenues between Levels of Government,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 506–45. Brosio, Giorgio and Juan Pablo Jimenez (2012) “The Intergovernmental Allocation of Resources from Natural Resources: Finding a Balance between Centripetal and Centrifugal Pressures,” in Giorgio Brosio and Juan Pablo Jimenez, eds, Decentralization and Reform in Latin America: Improving Intergovernmental Relations (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 290–320. Brown, Mimi (2013) Property Rates in Hong Kong: Assessment, Collection and Administration, 2nd edn (Hong Kong: Rating and Valuation Department). Brueckner, Jan (2003) “Strategic Interaction among Governments: An Overview of Empirical Studies,” International Regional Science Review, 26 (2): 175–88. Brueckner, Jan K. (2006) “Fiscal Federalism and Economic Growth,” Journal of Public Economics, 90 (10–11): 2107–20. Buehn, Andreas, Christian Lessmann and Gunther Markwardt (2013) “Decentralization and the Shadow Economy: Oates Meets AllinghamSandmo,” Applied Economics, 45 (18): 2567–78. Burkhead, Jesse and Jerry Miner (1971) Public Expenditure (London: Macmillan). Burki, Shahid Javed and Guillermo Perry, eds (1999) Decentralization and Accountability of the Public Sector: Annual World Bank Conference on Development in Latin America and the Caribbean 1999 (Washington DC: World Bank). Byrnes, Joel and Brian Dollery (2002) “Do Economies of Scale Exist in Australian Local Government? A Review of the Research Evidence,” Urban Policy and Research, 20 (4), 391–414. Cameron, Robert (2005) “Metropolitan Restructuring (and More Restructuring) in South Africa,” Public Administration and Development, 25: 329–39. Campos, Jose Edgardo and Joel S. Hellman (2005) “Governance Gone Local: Does Decentralization Improve Accountability?,” in Roland White and Paul Smoke, eds, East Asia Decentralizes: Making Local Government Work (Washington DC: World Bank), pp. 237–52. Campos, Ed and Sanjay Pradhan (1996) “Budgetary Institutions and Expenditure Outcomes,” World Bank Policy Research Working Paper No. 1646. Canavire-Bacarreza, Gustavo Javier, Cristian Sepulveda and Jorge Martinez-Vazquez (2015) “Fiscal Surf and Turf: Peru’s Efficiency and

M4439-BAHL_9781786435293_t.indd 439

01/03/2018 11:09

440

Fiscal decentralization and local finance in developing countries

Equity in Revenue Mobilization,” in Vicente Fretes Ciblis and Teresa Ter-Minassian, eds (2015) Decentralizing Revenue in Latin America: Why and How (Washington DC: Inter-American Development Bank), pp. 199–228. Canuto, Otaviano and Lili Liu, eds (2013) Until Debt Do Us Part: Subnational Debt, Insolvency, and Markets (Washington DC: World Bank). Capuno, J.J., T.C. Manuel and M.B Salvador (2001) “Estimating the IRA, Centrally Provided Local Public Goods and Services, and Other Central Fiscal Transfers to Local Governments,” Report Prepared for National Economic and Development Authority with support from the Australian Government through the Philippines-Australian Governance Facility. Capuno, Joseph J., Stella A. Quimbo, Aleli D. Kraft, Carlos Antonio R. Tan, Jr. and Vigile Marie B. Fabella (2015) “Does Yardstick Competition Influence Local Government Fiscal Behavior in the Philippines?,” in Jean-Paul Faguet and Caroline Pöschl, eds, Is Decentralization Good for Development: Perspectives from Academics and Policy Makers (Oxford: Oxford University Press), pp. 227–52. Cavanagh, Joe, Suzanne Flynn and Delphine Moretti (2016) Implementing Accrual Accounting in the Public Sector (Washington DC: International Monetary Fund). Centeno, Miguel Angel (2002) Blood and Debt: War and the Nation-State in Latin America (University Park: Pennsylvania State University Press). Chaparro, Juan Camilo, Michael Smart and Juan Gonzalo Zapata (2005) “Intergovernmental Transfers and Municipal Finance in Colombia,” in Richard Bird, James Porterba and Joel Slemrod, eds, Fiscal Reform in Colombia: Problems and Prospects (Cambridge MA: MIT Press), pp. 287–318. Charbit, Claire and Catherine Gamper (2016) “Coordination of Infrastructure Investment across Levels of Government,” in Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 352–78. Chelliah, Raja (2006) “A Tract on Reform of Federal Fiscal Relations in India,” Monograph 1/2006 Madras School of Economics, Chennai, India. Cheng, Bejamin S. (1999) “Causality between Taxes and Expenditures: Evidence from Latin American Countries,” Journal of Economics and Finance, 23 (2): 184–92. Chin, Kian-Keong (2010) “The Singapore Experience: The Evolution of Technologies, Costs and Benefits, and Lessons Learnt,” Discussion Papers Joint Transport Research Centre Round Table, February 4–5, Paris. Chowdhury, S., F. Yamauchi and R. Dewina (2009) “Governance,

M4439-BAHL_9781786435293_t.indd 440

01/03/2018 11:09



References ­441

Decentralization and Local Infrastructure Provision in Indonesia,” IFPRI Discussion Paper 00902, International Food Policy Research Institute, Washington DC. Cline, Howard (1962) Mexico: Revolution to Evolution 1940–1960 (London: Oxford University Press). Cnossen, Sijbren (1996) “VAT Treatment of Immovable Property,” in Victor Thuronyi, ed., Tax Law Design and Drafting, vol. 1 (Washington DC: International Monetary Fund), pp. 231–45. Collier, Paul and Anke Hoeffler (2005) “Resource Rents, Governance, and Conflict,” Journal of Conflict Resolution, 49 (4): 625–33. Collier, Paul and Anthony Venables (2015) “Housing Markets and Urbanization in Africa: Unleashing a Formal Market Process,” in Edward Glaeser and Abha Joshi-Ghani, eds, The Urban Imperative: Towards Competitive Cities (New Delhi: Oxford University Press), pp. 413–36. Comisión de Expertos para la Equidad y la Competitividad Tributaria (2015) “Informe Final Presentado al Ministro de Hacienda y Crédito Público,” Bogotá, December. Commonwealth Grants Commission (CGC) (1983) Equality in Diversity: Fifty Years of the Commonwealth Grants Commission (Canberra: Australian Government Publishing Service). Commonwealth Grants Commission (2015) Report on GST Revenue Sharing Relativities, 2015 Review (Canberra: CGC). Congleton, Roger (2015) “Asymmetric Federalism: Constitutional and Fiscal Exchange and the Political Economy of Decentralization,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 124–43. Corthay, Laurent (2009) “Local Taxes, Regulations, and the Business Environment,” Investment Climate in Practice No. 5, FIAS, World Bank-International Finance Corporation, April. Costa-Font, Joan and Gilberto Turati (2016) “Regional Health Care Decentralization in Unitary States: Equal Spending, Equal Satisfaction?,” CESifo Working Paper No. 5973, June. Cremer, Jacques, Antonio Estache and Paul Seabright (1995) “The Decentralization of Public Services: Lessons from the Theory of the Firm,” in Antonio Estache, ed., Decentralizing Infrastructure: Advantages and Limitations (Washington DC: World Bank). Cropper, Maureen and Soma Bhattacharya (2012) “Public Transport Subsidies and Affordability in Mumbai, India,” Urban Studies Research, 2012: Article ID 865972. Crook, Richard and Alan Sverrisson (2001) “Decentralisation and Poverty-Alleviation in Developing Countries: A Comparative Analysis

M4439-BAHL_9781786435293_t.indd 441

01/03/2018 11:09

442

Fiscal decentralization and local finance in developing countries

or, is West Bengal Unique?,” IDS Working Paper 130, Institute of Development Studies, February. Cui, Wei (2011) “Fiscal Federalism in Chinese Taxation,” World Tax Journal, 3: 455–80. Curry, Bill (2016) “Agency Warns of Bid-Rigging as Ottawa Set to Spend on Infrastructure, The Globe and Mail (Toronto), May 30. Cyan, Musharraf and Doug Porter (2007) “From Central to Local: Promoting Local Accountability and Service Delivery through Grant Systems,” Pakistan Resident Mission, Asian Development Bank, Islamabad. Cyan, Musharraf, Canisius Karuranga and François Vaillancourt (2013) “Local Government Revenue Potential in Rwanda,” International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University, August. Dabla-Norris, Era (2005) “Issues in Intergovernmental Fiscal Relations in China,” IMF Working Paper WP/05/30, International Monetary Fund, Washington DC. Dabla-Norris, Era, Florian Misch, Duncan Clearly and Munawer Khwaja (2017) “Tax Administration and Firm Performance: New Data and Evidence for Emerging Market and Developing Economies,” IMF Working Paper WP/17/95, International Monetary Fund, Washington DC. Dafflon, Bernard, ed. (2002) Local Public Finance in Europe (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Dafflon, Bernard (2011) “Conceptual Problems in the Assignment of Functions in SEE: The Case of Albania,” in Jorge Martinez-Vazquez and François Vaillancourt, eds, Decentralization in Developing Countries: Global Perspectives on the Obstacles to Fiscal Devolution (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 23–82. Dafflon, Bernard and S. Daguet (2012) “Local Environmental User Charges in Switzerland: Implementation and Performance,” EuroEconomica, 5 (31): 75–87. Dahlby, Bev (1996) “Fiscal Externalities and the Design of Intergovernmental Grants,” International Tax and Public Finance, 3 (3): 397–412. Dahlby, Bev (2008) The Marginal Cost of Public Funds (Cambridge MA: MIT Press). Das, Jishnu et al. (2013) “School Inputs, Household Substitution, and Test Scores,” American Economic Journal: Applied Economics, 5 (2): 29–57. Daunton, Martin (2001) Trusting Leviathan: The Politics of Taxation in Britain, 1799–1914 (Cambridge, UK: Cambridge University Press).

M4439-BAHL_9781786435293_t.indd 442

01/03/2018 11:09



References ­443

Davey, Kenneth (1988) “Municipal Development Funds and Intermediaries,” Working Paper WPS 32, World Bank, Washington DC, July. Davoodi, Hamid and Heng-fu Zou (1998) “Fiscal Decentralization and Economic Growth: A Cross-Country Study,” Journal of Urban Economics, 43 (2): 244–57. De, Indrandil and Tirthankar Nag (2016) “Dangers of Decentralisation in Urban Slums: A Comparative Study of Water Supply and Drainage Service Delivery in Kolkata, India,” Development Policy Review, 34 (2): 253–76. De Cesare, Claudia (2004) “General Characteristics of Property Tax Systems in Latin America,” paper presented at 7th International Conference on Optimizing Property Tax Systems in Latin America, Guadalajara, Jalisco, Mexico. De Cesare, Claudia (2012) “Overview of Property Tax in Latin America,” Working Paper WP10CD1, Lincoln Institute of Land Policy, Cambridge MA. De Cesare, Claudia and Les Ruddock (1999) “The Property Tax System in Brazil,” in William McCluskey, ed., Property Tax: An International Comparative Review (Aldershot: Ashgate), pp. 266–82. De Mello, Luiz (2007) “Fiscal Responsibility and Fiscal Adjustment: The Case of Brazilian Local Governments,” in George Peterson and Patricia Clarke Annez, eds, Financing Cities (Washington DC: World Bank), pp. 40–73. De Mello, Luiz (2011) “Does Fiscal Decentralization Strengthen Social Capital? Cross-Country Evidence and the Experiences of Brazil and Indonesia,” Environment and Planning C: Government and Policy, 29 (2): 281–96. De Mello, Luiz (2012) “Fiscal Decentralisation and Public Investment: The Experience of Latin America,” in Giorgio Brosio and Juan Pablo Jimenez, eds, Decentralization and Reform in Latin America: Improving Intergovernmental Relations (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 112–39. De Mello, Luiz and Matias Barenstein (2001) “Fiscal Decentralization and Governance: A Cross-Country Analysis,” IMF Working Paper WP/01/71, May. De Ree, Joppe, Karthik Muralidharan, Menno Pradhan and Halsey Rogers (2015) “Double for Nothing? Experimental Evidence on the Impact of an Unconditional Teacher Salary Increase on Student Performance in Indonesia,” NBER Working Paper 21806. De Silva, Migara, Galina Kurlyandskaya, Elena Andreeva and Natalia Golovanova (2009) Intergovernmental Reforms in the Russian Federation: One Step Forward, Two Steps Back? (Washington DC: World Bank).

M4439-BAHL_9781786435293_t.indd 443

01/03/2018 11:09

444

Fiscal decentralization and local finance in developing countries

De Soto, Hernando (1990) The Other Path (New York: Harper & Row). De Soto, Hernando (2000) The Mystery of Capital (New York: Basic Books). De Wulf, Luc (1975) “Fiscal Incidence Studies in Developing Countries: Study and Critique,” IMF Staff Papers, 22 (1): 61–131. Deaton, Angus and Nancy Cartwright (2016) “Understanding and Misunderstanding Randomized Control Trials,” NBER Working Paper 22595, September. Deininger, Klaus and Paul Mpuga (2004) “Does Greater Accountability Improve the Quality of Delivery of Public Services?,” World Bank Policy Research Working Paper 3277. Dell’Anno, Roberto and Désirée Teobaldelli (2015) “Keeping Both Corruption and the Shadow Economy in Check: The Role of Decentralization,” International Tax and Public Finance, 22 (1): 1–40. DeLog and LPSI Secretariat (2015) Localizing Public Services and Development: The Local Public Sector’s Role in Achieving Development Goals in Health and Education (Washington DC: Urban Institute). Desgagné, Maxime (2013) “Fiscal Convergence and Municipal Reforms in the Montreal Metropolitan Region between 1996 and 2011,” Research Report, Université de Montréal. Devas, Nick (2004) Urban Governance, Voice and Poverty in the Developing World (London: Earthscan). Devas, Nick and Roy Kelly (2001) “Regulation or Revenues? An Analysis of Local Business Licences, with a Case Study of the Single Business Permit Reform in Kenya,” Public Administration and Development, 21 (5): 381–91. Devkota, K.L. (2014) “Impact of Fiscal Decentralization on Economic Growth in the Districts of Nepal,” Working Paper 14-20, International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University. Dillinger, William (1988) “Urban Property Tax Reform: The Case of the Philippines Real Property Tax Administration Project,” Report INU-16, Infrastructure and Urban Development Department, World Bank, Washington DC, May. Dillinger, William (1991) Urban Property Tax Reform (Washington DC: World Bank). Dillinger, William (1994) “Decentralization and its Implications for Urban Service Delivery,” Urban Management Program Discussion Paper No. 16, World Bank, Washington DC. Dillinger, William, Guillermo Perry and Steven Webb (2003) “Is Fiscal Stability Compatible with Decentralization? The Case of Latin America, in Jorge Martinez-Vazquez and James Alm, eds, Public

M4439-BAHL_9781786435293_t.indd 444

01/03/2018 11:09



References ­445

Finance in Developing and Transitional Countries (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing), pp. 232–60. Diokno, Benjamin (2009) “Decentralization in the Philippines after Ten Years: What Have We Learned?,” in Shinichi Ichimura and Roy Bahl, eds, Decentralization Policies in Asian Development (Singapore: World Economic Press), pp. 161–92. Directorate General of Fiscal Balance, Ministry of Finance, Republic of Indonesia (2012) Fiscal Decentralization in Indonesia: A Decade after Big Bang (Jakarta: University of Indonesia Press). DNP (Departamento Nacional de Planeación) (1981) Las Finanzas intergubernamentales en Colombia: Informe Final de la Misión (Bogotá: Departamento Nacional de Planeación). Doebele, William (1979) “Land Readjustment as an Alternative to Taxation for the Recovery of Betterment: The Case of South Korea,” in Roy Bahl, ed., The Taxation of Urban Property in Less Developed Countries (Madison: University of Wisconsin Press), pp. 163–90. Doebele, William A. (1998) “The Recovery of ‘Socially Created’ Land Values in Colombia,” Land Lines, 10 (4): 5–7. Dollar, David and Bert Hofman (2008) “Intergovernmental Fiscal Reforms toward Realizing a Harmonious Society,” in Lou Jiwei and Shuilin Wang, eds, Public Finance in China: Reform and Growth for a Harmonious Society (Washington DC: World Bank), pp. 39–52. Domingos, Omar Pinto (2011) “Implementing Local Property Tax Reform in Brazil,” Land Lines, 23 (1): 2–8. Dornfest, Alan (2010) “In Search of Optimal Revaluation Policy: Benefits and Pitfalls,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Challenging the Conventional Wisdom about the Property Tax (Cambridge MA: Lincoln Institute of Land Policy), pp. 75–107. Downing, Paul (1973) “User Charges and the Development of Urban Land,” National Tax Journal, 26 (4): 631–37. Dreher, Axel and Merle Kreibaum (2015) “Weapons of Choice,” CESifo Working Paper No. 5419, June. Dunn, Jonathan and Deborah Wetzel (2000) “Fiscal Decentralization in Former Socialist Economies: Progress and Prospects,” in Proceedings of the Ninety-Second Annual Conference on Taxation (Washington DC: National Tax Association). Duranton, Gilles and Diego Puga (2004) “Micro-Foundations of Urban Agglomeration Economies,” in J.V. Henderson and J.F. Thisse, eds, Handbook of Regional and Urban Economics (Amsterdam: Elsevier), vol. 4, pp. 2063–117. Dye, Richard and Richard England, eds (2009), Land Value Taxation:

M4439-BAHL_9781786435293_t.indd 445

01/03/2018 11:09

446

Fiscal decentralization and local finance in developing countries

Theory, Evidence and Practice (Cambridge MA: Lincoln Institute of Land Policy). Eaton, Kent, Kai Kaiser and Paul Smoke (2010) The Political Economy of Decentralization Reforms: Implications for Aid Effectiveness (Washington DC: World Bank). Ebel, Robert and Yameng Wang (2017) “The Role of User Charges in Financing the Flow of U.S. Infrastructure Services,” in Richard Bird and Enid Slack, eds, Financing Infrastructure: Who Should Pay? (Montreal: McGill-Queen’s University Press), pp. 136–77. Ebel, Robert and Serdar Yilmaz (2003) “On the Measurement and Impact of Fiscal Decentralization,” in Jorge Martinez-Vazquez and James Alm, eds, Public Finance in Developing and Transitional Countries (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 101–20. Eccleston, Richard and Richard Krever (2017) The Future of Federalism: Intergovernmental Financial Relations in an Age of Austerity (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Eckert, Joseph (2008) “Computer-Assisted Mass Appraisal Options for Developing Countries,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Making the Property Tax Work: Experiences in Developing and Transitional Countries (Cambridge, MA: Lincoln Institute of Land Policy), pp. 207–28. Edwards, M. and Y. Xiao (2009) “Annexation, Local Government Spending, and the Complicating Role of Density,” Urban Affairs Review, 45 (2): 147–65. Ekpo, Akpan (2007) “Federal Republic of Nigeria,” in Anwar Shah, ed., The Practice of Fiscal Federalism: Comparative Perspectives (Montreal: McGill-Queen’s University Press), pp. 204–34. Ellis, Simon and Aurelio Menendez (2016) “Rural Roads: The Challenge of Decentralized Implementation,” in Jonas Frank and Jorge MartinezVazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 279–301. Enahoro, John and Jayeola Olabisi (2012) “Tax Administration and Revenue Generation of Lagos State Government, Nigeria,” Research Journal of Finance and Accounting, 3 (5): 133–9. Engel, E., R. Fischer and A. Galetovic (2010) “The Economics of Infrastructure Finance: Public Private Partnerships versus Public Provision,” Documento de Trabajo Serie Economía No. 276, Centro de Economía Aplicada, Ingeniería Industrial, Universidad de Chile. Engelschalk, Michael (2004) “Creating a Favorable Tax Environment for Small Business,” in James Alm and Sally Wallace, eds, Taxing the Hardto-Tax (Amsterdam: Elsevier), pp. 275–311.

M4439-BAHL_9781786435293_t.indd 446

01/03/2018 11:09



References ­447

Enikolopov, R. and E. Zhuravskaya (2003) “Decentralization and Political Institutions,” Journal of Public Economics, 83 (3): 325–46. Estache, Antonio (2006) “Infrastructure: A Survey of Recent and Upcoming Issues,” World Bank, Infrastructure Vice-Presidency, and Poverty Reduction and Economic Management Vice-Presidency Paper, April. Estache, Antonio (2010) “Infrastructure Finance in Developing Countries: An Overview,” EIB Papers, 15 (2): 61–88. Estache, Antonio and S. Sinha (1995) “Does Decentralization Increase Spending on Infrastructure?,” World Bank Policy Research Working Paper 1457. Estrin, Saul and Adeline Pelletier (2015) “Privatisation in Developing Countries: What Are the Lessons of Experience,” EPSPEAKS Topic Guide, London. Faguet, Jean-Paul (2004) “Does Decentralization Increase Government Responsiveness to Local Needs? Evidence from Bolivia,” Journal of Public Economics, 88 (3–4): 867–93. Faguet, Jean-Paul (2005) “The Effects of Decentralisation on Public Investment: Evidence and Four Lessons from Bolivia and Colombia,” Crisis States Research Centre Working Papers Series 1, 62. Faguet, Jean-Paul (2011) “Decentralization and Governance,” Economic Organisation and Public Policy Discussion Paper Series No. 27, Suntory Centre, London School of Economics, June. Faguet, Jean-Paul (2014) “Decentralization and Governance,” World Development, 53 (C): 2–13. Faguet, Jean-Paul and Caroline Pöschl, eds (2015) Is Decentralization Good for Development? Perspectives from Academics and Policy Makers (Oxford: Oxford University Press). Fan, C.S., C. Lin and D. Treisman (2009) “Political Decentralization and Corruption: Evidence from around the World,” Journal of Public Economics, 93 (1–2): 14–34. Farvacque-Vitkovic, Catherine and Mihaly Kopanyi, eds (2014) Municipal Finances: A Handbook for Local Governments (Washington DC: World Bank). Feld, Lars, Gebhard Kirchgässner and Christoph Schaltegger (2003) “Decentralized Taxation and the Size of Government: Evidence from Swiss State and Local Governments,” CESifo Working Paper Series 1087. Fengler, Wolfgang and Bert Hofman (2009) “Managing Indonesia’s Rapid Decentralization: Achievements and Challenges,” in Shinichi Ichimura and Roy Bahl, eds, Decentralization Policies in Asian Development (Singapore: World Scientific Press), pp. 245–62.

M4439-BAHL_9781786435293_t.indd 447

01/03/2018 11:09

448

Fiscal decentralization and local finance in developing countries

Fernández Llera, Roberto (2012) “Descentralización, inversión pública y consolidación fiscal: hacia una nueva geometría del triángulo,” paper prepared for Seminario Tributación y crecimiento con equidad, CEPAL, Santiago, August. Fischel, William (2006) “Footloose at Fifty: An Introduction to the Tiebout Anniversary Essays,” in William Fischel, ed., The Tiebout Model at Fifty: Essays in Honor of Wallace Oates (Cambridge MA: Lincoln Institute of Land Policy), pp. 1–20. Fischer, Stanley and William Easterly (1990) “The Economics of the Government Budget Constraint,” World Bank Research Observer, 5 (2): 127–42. Fisman, Raymond and Roberta Gatti (2002) “Decentralization and Corruption: Evidence across Countries,” Journal of Public Economics, 83: 325–45. Fiszbein, Ariel (1997) “The Emergence of Local Capacity: Lessons from Colombia,” World Development, 25 (7): 1029–43. Fjeldstad, Odd-Helge, Gérard Chambas and Jean-François Brun (2014) “Local Government Taxation in Sub-Saharan Africa: Review and an Agenda for Research,” CMI Working Paper WP 2014:2, Christian Michelsen Institute, Bergen, March. Fletcher, Jason and Matthew Murray (2008) “What Factors Influence the Structure of the State Income Tax?,” Public Finance Review, 36 (4):475–96. Florida, Richard (2002) The Rise of the Creative Class (New York: Basic Books). Found, Adam (2012) “Economies of Scale in Fire and Police Services,” IMFG Papers on Municipal Finance and Governance, No. 12, Toronto. Fox, William and Tami Gurley (2006) “Will Consolidation Improve SubNational Governments?,” World Bank Policy Research Working Paper 3913. Fox, William and Matthew Murray (2016) “The Challenge of Operating and Maintaining Infrastructure,” in Jonas Frank and Jorge MartinezVazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 305–23. Fox, William and Enid Slack (2011) “North America,” in Jorge MartinezVazquez and Paul Smoke, eds, Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 257–82. Fox, William and Christine Wallich (2007) “Fiscal Federalism in BosniaHerzegovina: Subsidiarity and Solidarity in a Three-Nation State,” in Richard Bird and Robert Ebel, eds, Fiscal Fragmentation in Decentralized Countries: Subsidiarity, Solidarity and Asymmetry (Cheltenham,

M4439-BAHL_9781786435293_t.indd 448

01/03/2018 11:09



References ­449

UK and Northampton MA, USA: Edward Elgar Publishing), pp. 399–434. Frank, Jonas and Jorge Martinez-Vazquez, eds (2016) Decentralization and Infrastructure in the Global Economy: From Gaps to Solutions (New York: Routledge). Frank, Jonas and Jorge Martinez-Vazquez, (2016a) “Decentralization and Infrastructure: From Gaps to Solutions,” in Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy: From Gaps to Solutions (New York: Routledge), pp. 1–30. Franzsen, Riël (1998) “Commentary on Land Taxes and Property Taxes in Developing Countries by Roy Bahl,” in Dick Netzer, ed., Land Value Taxation: Can It and Will It Work Today? (Cambridge MA: Lincoln Institute of Land Policy), pp. 172–82. Franzsen, Riël (2009) “International Experience,” in Richard Dye and Richard England, eds, Land Value Taxation: Theory, Evidence and Practice (Cambridge MA: Lincoln Institute of Land Policy), pp. 27–48. Franzsen, Riël and William J. McCluskey (2008) “The Feasibility of Site Value Taxation,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Making the Property Tax Work: Experiences in Developing and Transitional Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 268–306. Franzsen, Riël and William J. McCluskey (2017) Property Tax in Africa: Status, Challenges, and Prospects (Cambridge MA: Lincoln Institute of Land Policy). Freire, Maria E. (2013) “Slum Upgrading,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 367–92. Fretes Ciblis, Vicente and Teresa Ter-Minassian, eds (2015) Decentralizing Revenue in Latin America: Why and How (Washington DC: InterAmerican Development Bank). Frey, Bruno and Rainer Eichenberger (2004) The New Democratic Federalism for Europe (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Gainer, Maya (2017) “From the Ground Up: Developing Jamaica’s National Land Agency, 2000–2016,” Innovations for Successful Societies, Princeton University, http://successfulsocieties.princeton.edu. Galasso, Emanuela and Martin Ravallion (2005) “Decentralized Targeting of an Anti-Poverty Program,” Journal of Public Economics, 89 (4): 705–27. Garcia, Manuel Jaen (2012) “The Revenues-Expenditures Nexus: A Panel

M4439-BAHL_9781786435293_t.indd 449

01/03/2018 11:09

450

Fiscal decentralization and local finance in developing countries

Data Analysis of Spain’s Regions,” International Journal of Academic Research in Economics and Management Sciences, 1 (1): 24–38. Garcia-Kilroy, Catiana and Heinz Rudolph (2017) Private Financing of Public Infrastructure through PPPs in Latin America and the Caribbean (Washington DC: World Bank). Garg, Sandhya, Ashima Goyal and Rupayan Pal (2017) “Why Tax Effort Falls Short of Tax Capacity in Indian States,” Public Finance Review, 45 (2): 232–59. Garzon, Hernando and Jaime Vazquez-Caro (2004) “Colombia: La Perspectiva del Impuesto Predial en Areas Rurales,” Chemonics International, Washington DC. Gasmi, F., A. Maingard, P. Norumba and C. Virto (2011) “Empirical Evidence on the Impact of Privatization of Fixed-Line Operators on Telecommunications Performance: Comparing OECD, Latin American, and African Companies,” TSE Working Paper, No. 11-241, Toulouse School of Economics, February. George, Henry ([1879] 1958) Progress and Poverty (New York: Robert Schalkenbach Foundation). Ghaus-Pasha, Aisha and Hafiz Pasha (2015) Fiscal Implications of the 18th Amendment and the 7th NFC Award: The Outlook for Provincial Finances (Islamabad: World Bank). Giavezzi, F. and G. Tabellini (2005) “Economic and Political Liberalizations,” Journal of Monetary Economics, 52: 1297–330. Glaeser, Edward (2011) Triumph of the City (New York: Penguin). Glaeser, Edward and Abha Joshi-Ghani (2014) “The Urban Imperative: Toward Shared Prosperity,” World Bank Policy Research Working Paper 6875. Glaeser, Edward and Abha Joshi Ghani, eds (2015) The Urban Imperative: Towards Competitive Cities (New York: Oxford University Press). Glaeser, Edward and Joshua Gottlieb (2009) “The Wealth of Cities: Agglomeration Economies and Spatial Equilibrium in the United States,” Journal of Economic Literature, 47 (4): 983–1028. Gloudemans, R.J. (2000) “Implementing a Land Value Tax in Urban Residential Communities,” Working Paper WP00RG1, Lincoln Institute of Land Policy, Cambridge MA. Goel, Rajeev K. and James W. Saunoris (2016) “Government Decentralization and Prevalence of the Shadow Economy,” Public Finance Review, 44 (2): 263–88. Goel, Rajeev, Ummad Mazhar, Michael Nelson and Rati Ram (2016) “Government Decentralization and International Government Performance,” CESifo Working Paper No. 6013, July. Golovanova, Natalia and Galina Kurlyandskaya (2011) “Eurasia,” in Jorge

M4439-BAHL_9781786435293_t.indd 450

01/03/2018 11:09



References ­451

Martinez-Vazquez and Paul Smoke, eds, Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 113–48. Gomez-Reino, Juan Luiz and Jorge Martinez-Vazquez (2013) “An International Perspective on the Determinants of Local Government Fragmentation,” International Center for Public Policy Working Paper 68, Andrew Young School of Public Policy, Georgia State University. Gomez Sabaini, Juan and Juan Pablo Jimenez (2012) “The Financing of Subnational Governments in Latin America,” in Giorgio Brosio and Juan Pablo Jimenez, eds, Decentralization and Reform in Latin America: Improving Intergovernmental Relations (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Goodspeed, Timothy J. (2015) “Decentralization and Natural Disasters,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilevel Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 567–84. Gramlich, Edward (1976) “The New York City Fiscal Crisis: What Happened and What Is to Be Done?,” American Economic Review, 66 (2): 415–29. Granados Vergara, Claudia Marcela and Fabio Jose Sánchez Torres (2008) “Do the Reforms Contribute to Social Progress? The Case of Water Supply and Sewerage Systems and Child Mortality in Colombia, 1990–2004,” Universidad de Los Andes, School of Economics, January. Graser, Dina and Pamela Robinson (2016) “A Recipe for Fiscal Trust,” IMFG Perspectives No. 13, Institute on Municipal Finance and Government, Munk School of Global Affairs, University of Toronto. Gravelle, Jennifer and Sally Wallace (2009) “Overview of the Trends in Property Tax Base Erosion,” in Nancy Augustine, Michael Bell, David Brunoi and Joan Youngman, eds, Erosion of the Property Tax Base: Trends, Causes and Consequences (Cambridge MA: Lincoln Institute of Land Policy), pp. 17–46. Grindle, Merrilee and Pilar Domingo, eds (2003) Proclaiming Revolution: Bolivia in Comparative Perspective (Cambridge MA: Harvard University Press). Guevara, Milweda (2004) “Real Property Taxation in the Philippines,” in Richard Bird and Enid Slack, eds, International Handbook of Land and Property Taxation (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 152–8. Habitat III Policy Paper Framework (2015) “Municipal Finance and Local Fiscal Systems,” paper for UN Conference on Housing and Sustainable Urban Development, Quito, December, available at http://habitat3.org/

M4439-BAHL_9781786435293_t.indd 451

01/03/2018 11:09

452

Fiscal decentralization and local finance in developing countries

wp-content/uploads/PU5-HABITAT-III-POLICY-PAPER-FRAME WORK.pdf. Haldenwang, Christian von, Alice Elfert, Tobias Engelmann, Samuel Germain, Gregor Sahler and Amelie Stanzel Ferreira (2015) The Devolution of the Land and Building Tax in Indonesia (Bonn: German Development Institute). Hankla, Charles and William Downs (2010) “Decentralisation, Governance and the Structure of Local Political Institutions: Lessons for Reform?,” Local Government Studies, 36 (6): 759–83. Hanniman, Kyle (2015) “Can Credit Markets Promote Municipal Fiscal Health?,” in Richard Bird and Enid Slack, eds, Is Your City Healthy? Measuring Urban Fiscal Health (Toronto: Institute on Municipal Finance and Governance), pp. 43–60. Harberger, Arnold C. (1989) “Lessons of Tax Reform from the Experiences of Uruguay, Indonesia, and Chile,” in Malcolm Gillis, ed., Tax Reform in Developing Countries (Durham NC: Duke University Press), pp. 27–43. Harjowiryono, M. (2011) “Development of Indonesia’s Intergovernmental Financing System,” Fiscal Decentralization in Indonesia a Decade after the Big Bang, International Conference on Fiscal Decentralization in Indonesia, Ministry of Finance, Jakarta, September. Hartmark AS (2011) “Petroleum Revenue Management – Uganda: Capacity Needs Assessment Report,” November. Hausmann, Ricardo (1998) “Fiscal Institutions for Decentralising Democracies: Which Way to Go?,” in Kuchiro Fukasaku and Ricardo Hausmann, eds, Democracy, Decentralisation and Deficits in Latin America (Washington DC: Inter-American Development Bank and OECD), pp. 13–32. Henderson, Vernon (2010) “Cities and Development,” Journal of Regional Science, 50 (1): 515–40. Henderson, Vernon (2015) “Urbanization and the Geography of Development,” in Edward Glaeser and Abha Joshi Ghani, eds, The Urban Imperative: Towards Competitive Cities (Oxford: Oxford University Press), pp. 51–88. Henriquez Diaz, Jose Marcelo, Javier Fuenzalida Aguirre and Francisca Del Fierro Torres (2011) “Compensado la desigualdad de ingresos locales: El Fond Común Municipal (FCM) en Chile,” Centro de Sistemas Públicos, Universidad de Chile. Herrero-Alcalde, Ana, Jorge Martinez-Vazquez and Encarnacion MurilloGarcia (2011) “The Challenge of Designing Capital Equalization Transfers: An Application to Spanish Regions,” Publius: The Journal of Federalism, 42 (2): 260–88.

M4439-BAHL_9781786435293_t.indd 452

01/03/2018 11:09



References ­453

Herstein, Lesley (2012) “Adding Value: Recognizing the Link between Engineers and Municipal Finance and Governance,” Presentation at the Institute of Municipal Finance and Governance, Munk School of Global Affairs, University of Toronto, May. Hettich, Walter and Stanley Winer (1999) Democratic Choice and Taxation: A Theoretical and Empirical Analysis (New York: Cambridge University Press). Hicks, Ursula (1961) Development from Below (Oxford: Clarendon). Hicks, Ursula (1974) The Large City: A World Problem (New York: Wiley). High Powered Expert Committee for Estimating the Investment Requirements for Urban Infrastructure Services (2011) Report on Indian Urban Infrastructure and Services (Delhi: Ministry of Urban Development, Government of India). Hirschman, Albert and Richard Bird (1968) Foreign Aid: A Critique and a Proposal (Princeton: Princeton University Press). Hirschman, Albert and Charles Lindblom (1962) “Economic Development, Research and Development, Policy Making: Some Converging Views,” Behavioral Science, 7 (2): 211–22. Hofman, Bert and Susana Cordeiro Guerra (2007) “Ensuring InterRegional Equity and Poverty Reduction,” in Jorge Martinez-Vazquez and Bob Searle, eds, Fiscal Equalization: Challenges in the Design of Intergovernmental Transfers (New York: Springer), pp. 31–60. Hofman, Bert and Kai Kaiser (2004) “The Making of the ‘Big Bang’ and Its Aftermath: A Political Economy Perspective,” in James Alm, Jorge Martinez-Vazquez and Sri Mulyani Indrawati, eds, Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 15–46. Hofman, Bert, Kadjatmiko, Kai Kaiser and Bambang Suharnoko Sjahrir (2006) “Evaluating Fiscal Equalization in Indonesia,” Policy Research Working Paper 3911, World Bank, Washington DC. Holland, Daniel (1979) “Adjusting the Property Tax for Growth, Equity, and Administrative Simplicity: A Proposal for La Paz, Bolivia,” in Roy Bahl, ed., The Taxation of Urban Property in Less Developed Countries (Madison: University of Wisconsin Press), pp. 119–34. Hong, Yu Hung (2007) “Law, Reciprocity and Economic Incentives,” in Yu Hung Hong and Barrie Needham, eds, Analyzing and Readjustment: Economics, Law, and Collective Action (Cambridge MA: Lincoln Institute of Land Policy), pp. 183–94. Hong, Yu Hung and Barrie Needham, eds (2007) Analyzing and Readjustment: Economics, Law, and Collective Action (Cambridge MA: Lincoln Institute of Land Policy).

M4439-BAHL_9781786435293_t.indd 453

01/03/2018 11:09

454

Fiscal decentralization and local finance in developing countries

Hoyt, William (2006) “Imperfect Competition Between Communities, Politics, and Capitalization,” in William Fischel, ed., The Tiebout Model at Fifty: Essays in Honor of Wallace Oates (Cambridge MA: Lincoln Institute of Land Policy), pp. 127–57. Hu, Ha and Robert Ebel (2016) “Multi-Tier Monitoring of Infrastructure: Top Down and Bottom Up,” in Jonas Frank and Jorge MartinezVazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 324–51. Hull, Catherine and Bob Searle (2007) “The Impact of Equalization on Service Delivery,” in Jorge Martinez-Vazquez and Bob Searle, eds, Fiscal Equalization: Challenges in the Design of Intergovernmental Transfers (New York: Springer), pp. 61–94. Hunter van Ryneveld (2008) “Potential Additional Sources of Revenue,” Report to the City of Johannesburg, June. Huther, Jeff and Anwar Shah (1998) “Applying a Simple Measure of Good Governance to the Debate on Fiscal Decentralization,” Policy Research Working Paper 1894, World Bank, Washington DC. Ikawa, Hiroshi (2008) “Trinity Reform of Local Fiscal System in Japan,” in Shinichi Ichimura and Roy Bahl, eds, Decentralization Policies in Asian Development (Singapore: World Scientific Press). Ingram, Gregory, Zhi Liu and Karin Brandt (2013) “Metropolitan Infrastructure and Capital Finance,” in Roy Bahl, Johannes Linn and Deborah Wetzel, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 339–66. Inman, Robert (1996) “How to Have a Fiscal Crisis: Lessons from Philadelphia,” American Economic Review, 85 (2): 378–83. Inman, Robert (2003) “Transfers and Bailouts: Enforcing Local Fiscal Discipline with Lessons from U.S. Federalism,” in Jonathan Rodden, Gunnar Eskeland and Jennie Litvack, eds, Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA: MIT Press), pp. 35–83. Inman, Robert (2009) “Local Deficits in Times of Economic Crisis: Any Role?,” World Report on Fiscal Federalism’09 (Barcelona: Institut d’Economia de Barcelona), pp. 28–37. Inman, Robert and Daniel Rubinfeld (1997) “Rethinking Federalism,” Journal of Economic Perspectives, 11 (4): 43–64. Institute on Municipal Finance and Governance (2012) “Moving Our Region: Funding Transportation for our Future,” Forum, No. 1 (Toronto). Institute on Municipal Finance and Governance (2015) “Requiem for an Institution: The End of the Indian Planning Commission,” IMFG Forum (Toronto), No. 5/2015.

M4439-BAHL_9781786435293_t.indd 454

01/03/2018 11:09



References ­455

International Finance Corporation (2011) Avoiding the Fiscal Pitfalls of Subnational Regulation (Washington DC: World Bank Group). IMF (annual) Government Finance Statistics (Washington DC: International Monetary Fund). IMF (2011) “Revenue Mobilization in Developing Countries,” Fiscal Affairs Department, March. IMF (2014) “Making Public–Private Partnerships More Effective” (video), October 8, available at www.worldbank.org/en/events/2014/10/06/makingpublic-private-partnerships-more-effe​ctive. IMF (2014a) Government Finance Statistics Manual 2014 (Washington DC: International Monetary Fund). IMF-World Bank Group (2016) “Public–Private Partnerships Fiscal Risk Assessment Model User Guide,” April. Ioannidis, John, T.D. Stanley and Hristos Doucouliagos (2016) “The Power of Bias in Economics Research,” Deakin University, Faculty of Business and Law, Economics Series SWP 2016/1. Irarrazaval, Ignacio (2004) “Property Tax in Chile,” in Richard Bird and Enid Slack, eds, International Handbook on Land and Property Taxation (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing), pp. 159–64. Iregui, Ana Maria (2005) “Decentralised Provision of Quasi-Private Goods: The Case of Colombia,” Economic Modelling, 22: 683–706. Ivanya, Maksym and Anwar Shah (2011) “Decentralization and Corruption: New Cross-Country Evidence,” Environment and Planning C: Government and Policy, 29 (2): 344–62. Jacobs, Divina (2009) “Capital Expenditures and the Budget,” published on IMF blog http://blog-pfm.imf.org. Jacobs, Jane (1984) Cities and the Wealth of Nations (New York: Random House). Jalan, Joytsna and Martin Ravallion (2002) “Geographic Poverty Traps? A Micro Model of Consumption Growth in China,” Journal of Applied Econometrics, 17 (4): 329–46. James, Kathryn (2015) The Rise of the Value-Added Tax (Cambridge, UK: Cambridge University Press). Jibao, Samuel S. and Wilson Prichard (2015) “The Political Economy of Property Tax in Africa: Explaining Reform Outcomes in Sierra Leone,” African Affairs, 114 (456): 1–28. Jin, Jing and Heng-Fu Zou (2003) “Soft Budget Constraints and Local Government in China,” in Jonathan Rodden, Gunnar Eskeland and Jennie Litvack, eds, Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA: MIT Press), pp. 289–324. Johansson, A., C. Heady, A. Arnold, B. Brys and L. Vartia (2008) “Tax

M4439-BAHL_9781786435293_t.indd 455

01/03/2018 11:09

456

Fiscal decentralization and local finance in developing countries

and Economic Growth,” Economics Department Working Paper No. 620, OECD, Paris. Josie, Jay, Garry MacDonald and Jeffrey Petchey (2008) “A Dynamic Equalization Model for Economic and Social Capital Grants with a South African Example,” Journal of Development Studies, 44 (8): 1169–89. Juul, K. (2006) “Decentralization, Local Taxation and Citizenship in Senegal,” Development and Change, 37 (4): 821–46. Kaiser, Kai, Jean Mabi Mulumba and Tony Verheijen (2011) “The Political Economy of Decentralization in the Democratic Republic of Congo (DRC),” in Jorge Martinez-Vazquez and François Vaillancourt, eds, Decentralization in Developing Countries: Global Perspectives on the Obstacles to Fiscal Devolution (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 251–302. Kappeler, Andreas, Albert Solé-Ollé, Andreas Stephan and Timo Välilä (2012) “Does Fiscal Decentralization Foster Regional Investment in Productive Infrastructure?,” KTH/CESIS Working Paper Series in Economics and Institutions of Innovation No. 273, April. Katzman, M.T. (1978) “Progressive Public Utility Rates as an Income Redistribution Device in Developing Countries: The Case of Municipal Water,” in John Toye, ed., Taxation and Economic Development (London: Frank Cass), pp. 174–92. Keen, Michael (1998) “Vertical Tax Externalities in the Theory of Fiscal Federalism,” IMF Staff Papers, 45 (3): 454–85. Keen, Michael and Christos Kotsogiannis (2004) “Federal Tax Competition and the Welfare Consequences of Decentralization,” Journal of Urban Economics, 56 (3): 397–407. Keen, Michael and Jack Mintz (2004) “The Optimal Threshold for a Value-Added Tax,” Journal of Public Economics, 88: 559–76. Kelly, Roy (1995) “Property Tax Reform in Indonesia: Applying a Collection Led Strategy” Bulletin for Indonesian Economic Studies, 29: 85–104. Kelly, Roy (2000) “Designing a Property Tax Strategy for Sub-Saharan Africa: An Analytical Framework Applied to Kenya,” Public Budgeting and Finance, 20 (4): 36–51. Kelly, Roy (2014) “Implementing Sustainable Property Tax Reform in Developing Countries,” in Richard Bird and Jorge Martinez-Vazquez, eds, Taxation and Development: The Weakest Link? (Cheltenham UK and Northampton MA, USA: Edward Elgar Publishing), pp. 326–63. Kelly, Roy (2015) “Strengthening the Revenue Side,” in Directorate General of Fiscal Balance, Fiscal Decentralization in Indonesia a Decade After the Big Bang (Jakarta: University of Indonesia Press), pp. 168–200.

M4439-BAHL_9781786435293_t.indd 456

01/03/2018 11:09



References ­457

Kelly, Roy, Johan Bastin and Karsono Surywibowo (1993) “Regional Government Finance in Indonesia: An Examination of Ongoing Reforms,” Harvard International Tax Program Development Discussion Paper No. 461, Tax Research Series No. 16, July. Kenyon, Daphne and Adam Langley (2016) “Non-Profit PILOTS (Payments in Lieu of Taxes),” Policy Brief, Lincoln Institute of Land Policy, Cambridge MA, November. Khagram, S., P. de Renzio and A. Fung, eds (2013) Open Budgets (Washington DC: Brookings Institution). Kharas, Homi and Johannes Linn (2013) “External Assistance for Urban Finance Development Needs, Strategies and Implementation,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 393–419. Khemani, Stuti (2015) “Political Capture of Decentralization: Vote Buying through Grants to Local Governments,” in Jean-Paul Faguet and Caroline Pöschl, eds, Is Decentralization Good for Development? Perspectives from Academics and Policy Makers (Oxford: Oxford University Press), pp. 196–223. Khumalo, Bongani and Renosi Mokate (2007) “Republic of South Africa,” in Anwar Shah, ed., The Practice of Fiscal Federalism: Comparative Perspectives (Montreal: McGill-Queen’s University Press), pp. 262–86. Kim, Junghun (2008) “The Use of Expenditure Needs: Equalization or Regional Policies,” in Junghun Kim and Jørgen Lotz, eds, Measuring Local Government Expenditure Needs, The Copenhagen Workshop 2007 (Seoul and Copenhagen: Korea Institute of Public Finance and Danish Ministry of Social Welfare), pp. 87–105. Kim, Junghun and Jørgen Lotz, eds (2008) Measuring Local Government Expenditure Needs, The Copenhagen Workshop 2007 (Seoul and Copenhagen: Korea Institute of Public Finance and Danish Ministry of Social Welfare). Kim, Junghun, Jørgen Lotz and Niels Jorgen Mau, eds (2010) General Grants versus Earmarked Grants Theory and Practice: The Copenhagen Workshop 2009 (Albertslund: Korea Institute of Public Finance and Danish Ministry of Interior and Health). King, Elisabeth (2013) “A Critical Review of Community-Driven Development Programmes in King Conflict-Affected Contexts,” International Rescue Committee. Kitchen, Harry and Douglas Auld (1995) Financing Education and Training in Canada (Toronto: Canadian Tax Foundation). Kitchen, Harry and Enid Slack (2006) “Providing Public Services in Remote

M4439-BAHL_9781786435293_t.indd 457

01/03/2018 11:09

458

Fiscal decentralization and local finance in developing countries

Areas,” in Richard Bird and François Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington DC: World Bank), pp. 123–39. Klein, Michael (2012) Infrastructure Policy: Basic Design Options (Washington DC: World Bank). Klink, Jeroen (2008) “Recent Perspectives on Metropolitan Organization, Functions, and Governance,” in Eduardo Rojas, Juan R. CuadradoRoura and José Miguel Fernández Güell, eds, Governing the Metropolis: Principles and Cases (Washington DC: Inter-American Development Bank), pp. 77–134. Kornai, Janos (1986) “The Soft Budget Constraint,” Kyklos, 39 (1): 3–30. Kravchuk, Robert (2008) “Political Decentralization in Transition Countries: Eastern Europe, the Former Soviet Union, and the New States of the Former Yugoslavia,” International Studies Program, Georgia State University and PADCO, April. Krugman, Paul (1991) “Increasing Returns and Economic Geography,” Journal of Political Economy, 99 (3): 483–99. Kurachi, Shintaro (2015) “Local Income Taxation and Intergovernmental Relationships in Denmark,” Keio-IES Discussion Paper No. 2015-004, Keio University Institute for Economics Studies, available at http:// econpapers.repec.org/paper/keodpaper/2015-004.htm. Kurlyandskaya, Galina (2005) “Equalization Transfers to Sub-National Governments in the Russian Federation,” Presentation to the 3rd International Conference on Federalism, Brussels, March 5. Kyriacou, Andreas and Oriol Roca-Sagalés (2011) “Fiscal and Political Decentralization and Government Quality,” Environment and Planning C: Government and Policy, 29 (2): 204–23. Labonne, Julien and Rob Chase (2007) “Who’s at the Wheel When Communities Drive Development?,” Social Development Papers No. 107, World Bank, Washington DC. Lago-Peñas, Santiago and Jorge Martinez-Vazquez, eds (2013) The Challenge of Local Government Size (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Lago-Peñas, Santiago, Jorge Martinez-Vazquez and Agnese Sacchi (2016) “The Impact of Fiscal Decentralization: A Survey,” Journal of Economic Surveys, DOI 10.1111/joes.12182. Lamont, James, Muhammad Handry Imansyah and Erny Marniasth (2012) “Overview,” in Directorate of Fiscal Balance, Fiscal Decentralization in Indonesia: A Decade after the Big Bang (Jakarta: Republic of Indonesia), pp. 1–34. Landau, Martin (1969) “Redundancy, Rationality, and the Problem of Duplication and Overlap,” Public Administration Review, 29 (4): 346–58.

M4439-BAHL_9781786435293_t.indd 458

01/03/2018 11:09



References ­459

Le Blanc, David (2007) “Providing Water to the Urban Poor in Developing Countries: The Role of Tariffs and Subsidies,” Sustainable Development Innovation Briefs, United Nations Department of Economic and Social Affairs, Issue 4, October. Lefevre, Christian (2008) “Democratic Governability of Metropolitan Areas: International Experience and Lessons for Latin American Cities,” in Eduardo Rojas, Juan R. Cuadrado-Roura and José Miguel Fernández Güell, eds, Governing the Metropolis: Principles and Cases (Washington DC: Inter-American Development Bank), pp. 137–92. Lennon, Sasha and William O’Neil (2003) “Facilitating Regional Economic Development: Local Government Perceptions and What It Can Do,” paper presented at the ANZRSAI 27th Annual Conference, Regional Development: Who Owns it, Fremantle, Western Australia, September 28–October 3. Letelier, Leonardo (2005) “Explaining Fiscal Decentralization,” Public Finance Review, 33 (2): 155–83. Levin, Jonathan (1991) “Measuring the Role of Subnational Governments,” International Monetary Fund Working Paper WP/91/8, Washington DC. Lewis, Arthur (1967) “Planning Public Expenditure,” in Max Millikan, ed., National Economic Planning (New York: Columbia University Press), pp. 210–27. Lewis, Blane (2006) “Local Government Taxation: An Analysis of Administrative Cost Inefficiency,” Bulletin of Indonesian Economic Studies, 42 (2): 213–33. Lewis, Blane and Bob Searle (2011) “Asia-Pacific,” in Jorge MartinezVazquez and Paul Smoke, eds, Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 61–112. Lewis, Blane and Paul Smoke (2012) “Incentives for Better Local Service Delivery,” in Directorate General of Fiscal Balance, Fiscal Decentralization in Indonesia a Decade after the Big Bang (Jakarta: University of Indonesia Press), pp. 250–84. Lewis-Faupel, Sean, Yusuf Neggers, Benjamin A. Olken and Rohini Pnade (2016) “Can Electronic Procurement Improve Infrastructure Provision? Evidence from Public Works in India and Indonesia,” American Economic Journal: Economic Policy, 8 (3): 258–83. Ley, Eduardo and Jessica Boccardo (2010) “The Taxation of Motor Fuel: International Comparison,” World Bank Policy Research Working Paper No. 5212, Washington DC. Li, Bingqin and Yongmei Zhang (2015) “Area-Based Competitions and Awards as a Motivational Tool for Public Service Provision: The Experience

M4439-BAHL_9781786435293_t.indd 459

01/03/2018 11:09

460

Fiscal decentralization and local finance in developing countries

of Xining, China,” in Jean-Paul Faguet and Caroline Pöschl, eds, Is Decentralization Good for Development? Perspectives from Academics and Policy Makers (Oxford: Oxford University Press), pp. 253–76. Lin, Justin Yifu and Zhiqiang Liu (2000) “Fiscal Decentralization and Economic Growth in China,” Economic Development and Cultural Change, 49 (1): 1–22. Lindaman, Kara and Kurt Thurmaier (2002) “Beyond Efficiency and Economy: An Examination of Basic Needs and Fiscal Decentralization,” Economic Development and Cultural Change, 50 (4): 915–34. Lindert, Peter (2004) Growing Public: Social Spending and Economic Growth since the Eighteenth Century, 2 vols. (Cambridge, UK: Cambridge University Press). Litvack, Jennie, Junaid Ahmad and Richard Bird (1998) Rethinking Decentralization (Washington DC: World Bank). Liu, Lili and Michael Waibel (2010) “Managing Subnational Credit and Default Risk,” Policy Research Working Paper 5362, World Bank, Washington DC. Liu, Lili and Steven Webb (2011) “Laws for Fiscal Responsibility for Subnational Discipline: International Experience,” Policy Research Working Paper 5587, World Bank, Washington DC. Llach, Lucas (2011) “Financing Public Services in the Buenos Aires Metropolitan Area,” paper prepared for the World Bank, unpublished, April. Lockwood, Ben (2006) “The Political Economy of Decentralization,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Fiscal Federalism (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 33–60. López Laborda, Julio, Jorge Martinez-Vazquez and Carlos Monasterio Escudero (2006) “The Practice of Fiscal Federalism in Spain,” Working Paper 0623, International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University. Lotz, Jørgen (2006) “Local Government Organization and Finance: Nordic Countries,” in Anwar Shah, ed., Local Governance in Industrial Countries (Washington DC: World Bank), pp. 223–64. Lotz, Jørgen (2012) “You Get What You Pay For: How Nordic Cities are Financed,” IMFG Paper on Municipal Finance and Governance No. 7, Toronto. Loyalka, Prashant, Sean Sylvia, Chengfang Liu, James Chu and Yaojiany Shi (2016) “Paper by Design: Teacher Performance Pay Design and the Distribution of Student Achievement,” available at http://ss.com/ abstract52775461.

M4439-BAHL_9781786435293_t.indd 460

01/03/2018 11:09



References ­461

Lustig, Nora (2016) “Fiscal Policy, Inequality and the Poor in the Developing World,” CEQ Working Paper No. 23, Tulane University, New Orleans, October. Ma, Xiang (2016) “Economics of Scale and Heterogeneity in the Provision of Public Goods: Evidence from School Consolidation in China,” available at file:///C:/Users/R/Downloads/department_of_economics_-_ economies_of_scale_and_heterogeneity_in_provision_of_public_goods_ evidence_from_school_consolidation_in_china_-_201.pdf. Majumder, Rajarshi (2012) “Removing Poverty and Inequality in India: The Role of Infrastructure,” MRPA Paper No. 40941, Munich Personal RePEc Archive, http://mpra.ub.uni-muenchen.de/40941/. Management Systems International (2008) Democracy and Governance Assessment: USAID/Southern Africa (Washington DC: Coffee International Development). Manasan, Rosario (2009) “Local Public Finance in the Philippines: Balancing Autonomy and Accountability,” in Shinichi Ichimura and Roy Bahl, eds, Decentralization Policies in Asian Development (Singapore: World Scientific Press), pp. 333–88. Manor, James (1999) The Political Economy of Democratic Decentralization (Washington DC: World Bank). Manor, James (2013) “When Local Government Strikes It Rich,” Research Report No. 1, Swedish International Centre for Local Democracy. Mansuri, Ghazala and Vijayendra Rao (2004) “Community-Based and -Driven Development: A Critical Review,” World Bank Research Observer, 19 (1): 1–39. Mansuri, Ghazala and Vijayendra Rao (2012) Localizing Development: Does Participation Really Work? (Washington DC: World Bank). Marcelo, Darwin, Cledan Mandri-Perrott, Schuyler House and Jordan Schwartz (2016) “Prioritizing Infrastructure Investment: A Framework for Government Decision Making,” Policy Research Working Paper No. 7674, World Bank, Washington DC. Marcesse, Thibaud (2016) “Public Policy and Informal Institutions: A New Moral Economy of Corruption in Rural India,” available at http:// papers.ssrn.com/sol3/papers.cfm?abstract_id52802655. Martinez-Vazquez, Jorge (1994) “Expenditures and Expenditure Assignment,” in Christine Wallich, ed., Russia and the Challenge of Fiscal Federalism (Washington DC: World Bank), pp. 96–128. Martinez-Vazquez, Jorge (2011) The Impact of Fiscal Decentralization: Issues in Theory and Challenges in Practices (Manila: Asian Development Bank). Martinez-Vazquez, Jorge (2011a) “Latin America,” in Jorge MartinezVazquez and Paul Smoke, eds, Local Government Finance: The Challenges

M4439-BAHL_9781786435293_t.indd 461

01/03/2018 11:09

462

Fiscal decentralization and local finance in developing countries

of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 189–230. Martinez-Vazquez, Jorge (2013) “Fiscal Decentralization in Peru: A Perspective on Recent Developments and Future Challenges,” Working Paper 13-24, International Center for Public Policy, Andrew  Young  School of Policy Studies, Georgia State University, November. Martinez-Vazquez, Jorge (2013a) “Local Nonproperty Revenues,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 183–211. Martinez-Vazquez, Jorge (2015) “Tax Assignments at the Regional and Local Level,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 358–88. Martinez-Vazquez, Jorge, Javier Arze del Granado and Jameson Boex (2007) Fighting Corruption in the Public Sector (Amsterdam: Elsevier). Martinez-Vazquez, Jorge, Juan Luis Gomez and Guevara Yao (2006a) “Fiscal Decentralization in the People’s Democratic Republic of Lao,” Working Paper 06-22, International Studies Program, Andrew Young School of Policy Studies, Georgia State University, September. Martinez Vazquez, Jorge, Charles McLure and François Vaillancourt (2006b) “Revenues and Expenditures in an Intergovernmental Framework,” in Richard Bird and François Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington DC: World Bank), pp. 15–34. Martinez-Vazquez, Jorge and Robert McNab (2001) “Fiscal Decentralization and Economic Growth,” International Studies Program Working Paper 01-1, Andrew Young School of Policy Studies, Georgia State University. Martinez-Vazquez, Jorge, Mark Rider and Sally Wallace (2008) Tax Reform in Russia (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Martinez-Vazquez, Jorge and Bob Searle, eds (2007) Fiscal Equalization: Challenges in the Design of Intergovernmental Transfers (New York: Springer). Martinez-Vazquez, Jorge and Paul Smoke, eds (2011) Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Martinez-Vazquez, Jorge and Wayne Thirsk (2011) Fiscal Decentralization in Ukraine: Accomplishments and Challenges in the Transition (New York: Nova Science). Martinez-Vazquez, Jorge and Andrey Timofeev (2009) “Decentralization

M4439-BAHL_9781786435293_t.indd 462

01/03/2018 11:09



References ­463

Measures Revisited,” Working Paper 09-13, International Studies Program, Andrew Young School of Policy Studies, Georgia State University, April. Martinez-Vazquez, Jorge and Andrey Timofeev (2010) “Choosing between Centralized and Decentralized Models of Tax Administration,” International Journal of Public Administration, 33 (12–13): 601–19. Martinez-Vazquez, Jorge and Andrey Timofeev (2011) “Decentralizing Egypt: Not Just Another Economic Reform,” in Jorge MartinezVazquez and François Vaillancourt, eds, Decentralization in Developing Countries: Global Perspectives on the Obstacles to Fiscal Devolution (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 389–431. Martinez-Vazquez, Jorge and Andrey Timofeev (2016) “Capital Infrastructure  and Equity Objectives in Decentralized Systems,” in  Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy: From Gaps to Solutions (New York: Routledge), pp. 171–216. Martinez-Vazquez, Jorge, Andrey Timofeev and Jameson Boex (2006c) Reforming Regional–Local Finance in Russia (Washington DC: World Bank). Martinez-Vazquez, Jorge and Francois Vaillancourt, eds (2011) Decentralization in Developing Countries: Global Perspectives on the Obstacles to Fiscal Devolution (Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing). Martinez-Vazquez, Jorge and Violeta Vulovic (2016) “How Well Do Subnational Borrowing Regulations Work?,” Working Paper 16-01, International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University. Martinez-Vazquez, Jorge and Ming-Hung Yao (2009) “Fiscal Decentralization and Public Sector Employment: A Cross-Country Analysis,” Working Paper 0903, International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University. Masagca, Jimmy, Manrico Masagca and M. Chunxiang (2009) “Enhancing Governance of the Barangay: Reflections of Academics,” Journal of Administration and Governance (JOAAG), 4 (2): 13–27. Mascagni, Giulia (2016) “Is Fiscal Decentralisation Delivering on its Promises? A Review of the Theory and Practice in Developing Countries,” IDS Working Paper 266, Brighton, UK. Mathur, Om Prakash (2012) “Intergovernmental Transfers in Local Government Finance: Decentralization and Strengthening of Local Authorities,” Report submitted to UN-Habitat, Nairobi. Mathur, Om Prakash (2013) “Metropolitan Governance and Finance in

M4439-BAHL_9781786435293_t.indd 463

01/03/2018 11:09

464

Fiscal decentralization and local finance in developing countries

India’s Federal Structure,” in Enid Slack and Rupak Chattopadhyay, eds, Governance and Finance of Metropolitan Areas in Federal Systems (Toronto: Oxford University Press), pp. 49–76. Mathur, Om Prakash, Debdulal Thakur and Nilesh Rajadhyasksha (2009) “Urban Property Tax Potential in India,” National Institute of Public Finance and Policy, New Delhi. Mauro, P. (1995) “Corruption and Growth,” Quarterly Journal of Economics, 110 (3): 681–712. May, R.J. (1969) Federalism and Fiscal Adjustment (Oxford: Clarendon). Mbiti, Isaac M. (2016) “The Need for Accountability in Education in Developing Countries,” Journal of Economic Perspectives, 30 (3): 109–32. McCarten, William and Vinod Vyasulu (2004) “Democratic Decentralization and Poverty Reduction in Madhya Pradesh: Searching for an Institutional Equilibrium,” Development in Practice, 14 (6): 733–40. McCluskey, William, Michael Bell and Lay Cheng Lim (2010) “Rental Value and Capital Value: Alternative Bases for the Property Tax,” in Roy Bahl, Jorge Martinez-Vazquez and Joan Youngman, eds, Challenging the Conventional Wisdom about the Property Tax (Cambridge MA: Lincoln Institute of Land Policy). McCluskey, William and Riel Franzen (2013) “Property Taxes in Metropolitan Cities,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 159–82. McCluskey, William and Riel Franzsen (2017) “Tanzania,” in Riel Franzsen and William McCluskey, eds, Property Tax in Africa: Status, Challenges, and Prospects (Cambridge MA: Lincoln Institute of Land Policy), pp. 408–28. McCluskey, William, Riel Franzsen and Roy Bahl (2017) “Policy and Practice,” in Riel Franzsen and William McCluskey, eds, Property Tax in Africa: Status, Challenges, and Prospects (Cambridge MA: Lincoln Institute of Land Policy), pp. 29–104. McCluskey, William, Riel Franzsen and Roy Bahl (2017a) “Challenges, Prospects and Recommendations,” in Riel Franzsen and William McCluskey, eds, Property Tax in Africa: Status, Challenges, and Prospects (Cambridge MA: Lincoln Institute of Land Policy), pp. 551–92. McCord, Anna (2012) “The Politics of Social Protection: Why are Public Works Programmes so Popular with Governments and Donors?,” Overseas Development Institute (ODI) Background Note, September. McGranahan, Gordon (2016) “The Role of Cities and Urbanisation in Achieving Development Goals,” Evidence Report No. 192, Institute of Development Studies, Brighton, UK, May.

M4439-BAHL_9781786435293_t.indd 464

01/03/2018 11:09



References ­465

McKinsey Global Institute (2010) India’s Urban Awakening: Building Inclusive Cities, Sustaining Economic Growth (Delhi and Mumbai: McKinsey & Company). McLure, Charles (1967) “The Interstate Exporting of State and Local Taxes: Estimates for 1962,” National Tax Journal, 20: 49–77. McLure, Charles (1983) “Assignment of Corporate Income Taxes in a Federal System,” in Charles McLure, ed., Tax Assignment in Federal Countries (Canberra: Centre on Federal Financial Relations, Australian National University), pp. 101–24. McLure, Charles (1994) “The Sharing of Taxes on Natural Resources and the Future of the Russian Federation,” in Christine Wallich, ed., Russia and the Challenge of Fiscal Federalism (Washington DC: World Bank), pp. 181–217. McLure, Charles (1995) “The Dangers of Decentralization: Comment,” World Bank Research Observer, 10 (2): 221–6. McLure, Charles (1997) “Topics in the Theory of Revenue Assignment: Gaps, Traps, and Nuances,” in Mario Blejer and Teresa Ter-Minassian, eds, Macroeconomic Dimensions of Public Finance (London: Routledge), pp. 94–109. McLure, Charles (1998) “The Revenue Assignment Problem: Ends, Means and Constraints,” Journal of Public Budgeting, Accounting and Financial Management, 9 (4): 652–83. McLure, Charles (2000) “Tax Assignment and Subnational Fiscal Autonomy,” Bulletin for International Fiscal Documentation, 54 (12): 626–35. McLure, Charles (2014) “Reforming Subsidies for Fossil Fuel Consumption: Killing Several Birds with One Stone,” in Richard Bird and Jorge Martinez-Vazquez, eds, Taxation and Development: The Weakest Link? (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 238–83. McLure, Charles and Wayne Thirsk (1978) “The Inequity of Taxing Iniquity: A Plea for Reduced Sumptuary Taxation in Developing Countries,” Economic Development and Cultural Change, 26 (3): 497–503. Medeiros, Marcelo, Tatiana Britto and Fabio Veras Soares (2006) “Targeted Cash Programmes in Brazil: BPC and the Bolsa Familia,” Working Paper 26, International Poverty Centre, June. Ménard, Claude (2013) “Is Public–Private Partnership Obsolete? Assessing the Obstacles and Shortcomings of PPP,” in Piet de Vries and Etienne B. Yehoue, eds, The Routledge Companion to Public–Private Partnerships (Abingdon, UK: Routledge), pp. 149–74. Merk, Olaf et al. (2012) “Financing Green Urban Infrastructure,” OECD Regional Development Working Papers 2012/10, Paris. Mikesell, John (2007) “Developing Options for the Administration of

M4439-BAHL_9781786435293_t.indd 465

01/03/2018 11:09

466

Fiscal decentralization and local finance in developing countries

Local Taxes: An International Review,” Public Budgeting and Finance, 27 (1): 41–68. Ministry of Urban Development (2011) “Report on Indian Urban Infrastructure and Services,” Report of the High-Powered Expert Committee for Estimating the Investment Requirements for Urban Infrastructure Services, New Delhi, March. Mintz, Jack and Michael Smart (2006) “Incentives for Public Investment under Fiscal Rules,” Policy Research Working Paper 3860, World Bank, Washington DC. Mogues, Tewodaj and Alvina Erman (2016) “Institutional Arrangements to Make Public Spending Responsible to the Poor: (Where) Have They Worked?,” International Food Policy Research Institute Discussion Paper 01519, Washington DC, April. Mohanty, P.K., B.M. Misra, Rajan Goyal and P.D. Jeromi (2007) “Municipal Finance in India: An Assessment,” Development Research Group Study No. 28, Department of Economic Analysis and Policy, Reserve Bank of India, Mumbai. Mohanty, Prasanna K. (2014) Cities and Public Policy: An Urban Agenda for India (New Delhi: Sage). Mohmand, Shandana Khan (2016) “Informal Local Governance Institutions: What They Do and Why They Matter,” IDS Working Paper 2016, No. 468, Brighton, UK, May. Moisio, Antti, Heikki Loikkanen and Lasse Oulasvirta (2010) “Public Services at the Local Level: The Finnish Way,” Government Institute for Economic Research, Policy Reports 2, June. Monari, Lucio (2002) “A Reality Check on Subsidizing Power for Irrigation in India,” World Bank Group Private Sector and Infrastructure Network, Note No. 244, Washington DC, April. Moore, Mick (2017) “The Political Economy of Long-Term Revenue Decline in Sri Lanka,” Working Paper 65, International Centre for Taxation and Development, Brighton, UK. Morgan, Steven, Jamie Daw and Michael Law (2013) “Rethinking Pharmacare in Canada,” C.D. Howe Institute Commentary No. 384, Toronto. Mountfield, Edward and Christine P.W. Wong (2005) “Public Expenditure on the Frontline: Toward Effective Management by Subnational Governments,” in Roland White and Paul Smoke, eds, East Asia Decentralizes: Making Local Government Work (Washington DC: World Bank), pp. 85–106. Muellbauer, John (2005) “Property Taxation and the Economy after the Barker Review,” Economic Journal, 115: C99–C117. Mukherjee, Sacchidananda (2017) “Changing Tax Capacity and Tax Effort of Indian States in the Era of High Economic Growth 2001–2014,”

M4439-BAHL_9781786435293_t.indd 466

01/03/2018 11:09



References ­467

Working Paper, National Institute of Public Finance and Policy, New Delhi. Musgrave, Richard (1959) The Theory of Public Finance (New York: McGraw-Hill). Musgrave, Richard (1969) Fiscal Systems (New Haven CT: Yale University Press). Musgrave, Richard (1981) Fiscal Reform in Bolivia (Cambridge MA: Law School of Harvard University). Musgrave, Richard A. (1983) “Who Should Tax, Where and What?,” in Charles McLure, ed., Tax Assignment in Federal Countries (Canberra: Centre for Research on Federal Financial Relations, Australian National University), pp. 2–19. Musgrave, Richard and Peggy Musgrave (1976) Public Finance in Theory and Practice, 3rd edn (New York: McGraw-Hill). Nasehi, Carly and Rahil Rangwala (2011) “Structures of Metropolitan Governance and Finance: A Case Study on Manila, Philippines,” Lincoln Institute of Land Policy Working Paper, Cambridge MA. Nath, Shyam (1987) “Is Residential Property Tax Equitable: A Case Study of Calcutta,” Economic and Political Weekly, 22 (29): 1190–96. National Democratic Institute for International Affairs (2004) Decentralization and Local Government in the Caribbean (Guyana: National Democratic Institute for International Affairs). National Institute of Urban Affairs (2011) “Ahmedabad Bus Rapid Transit System (Delhi),” available at www.niua.org/projects/tpt/AHM​ EDABAD%20BRTS.pdf. Natural Resource Governance Institute (2016) “Natural Resource Revenue Sharing,” available at https://resourcegovernance.org/analysis-tools/publi​ cations/natural-resource-revenue-sharing. Nauges, Céline and Caroline van den Berg (2008) “Spatial Heterogeneity in the Cost Structure of Water and Sanitation Services: A CrossCountry Comparison of Conditions for Scale Economies,” available at www.researchgate.net/publication/227622138. Nelson, M. (2013) “Corruption and the Size of Local Governments: Are They Related?,” in Santiago Lago-Peñas and Jorge Martinez-Vazquez, eds, The Challenge of Local Government Size (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 87–124. Netzer, Dick (1966) Economics of Property Tax (Washington DC: Brookings Institution). Netzer, Dick, ed. (1998) Land Value Taxation: Can It and Will It Work Today? (Cambridge MA: Lincoln Institute of Land Policy). Newbery, David and Georgina Santos (1999) “Road Taxes, Road User Charges and Earmarking,” Fiscal Studies, 20 (2): 103–32.

M4439-BAHL_9781786435293_t.indd 467

01/03/2018 11:09

468

Fiscal decentralization and local finance in developing countries

Neyapti, Bilin (2006) “Revenue Decentralization and Income Distribution,” Economics Letters, 92 (2006): 409–16. Nilekani, Nandan (2008) Imagining India: Ideas for the New Century (New Delhi: Penguin and Allen Lane). Norregaard, John (2013) “Taxing Immovable Property: Revenue Potential and Implementation Challenges,” IMF Working Paper 13/129, International Monetary Fund, Washington DC. Oates, Wallace (1969) “The Effects of Property Taxes and Local Public Spending on Property Values: An Empirical Study of Tax Capitalization and the Tiebout Hypothesis,” Journal of Political Economy, 77 (6): 957–71. Oates, Wallace (1972) Fiscal Federalism (New York: Harcourt Brace Jovanovich). Oates, Wallace (1981) “On Local Finance and the Tiebout Model,” American Economic Review, 71: 93–8. Oates, Wallace (1985) “Searching for Leviathan: An Empirical Study,” American Economic Review, 75: 748–57. Oates, Wallace (1988) “On the Evolution of Fiscal Federalism: Theory and Institutions” National Tax Journal, 61 (2): 313–34. Oates, Wallace (1996) ‘Taxation in a Federal System: The Tax-Assignment Problem,” Public Economics Review, 1: 35–60. Oates, Wallace (1999) “An Essay on Fiscal Federalism,” Journal of Economic Literature, 37 (3): 1120–49. Oates, Wallace (2008) “On the Evolution of Fiscal Federalism: Theory and Institutions,” National Tax Journal, 61 (2): 313–34. Oates, Wallace and Robert Schwab (1997) “The Impact of Urban Land Taxation: The Pittsburgh Experience,” National Tax Journal, 50 (1): 1–21. O’Brien, Kevin and Lianjiang Li (2006) Rightful Resistance in Rural China (Cambridge, UK: Cambridge University Press). OECD (1999) Taxing Powers of State and Local Government (Paris: OECD). OECD (2004) OECD Territorial Reviews: Mexico City (Paris: OECD). OECD (2006) “Intergovernmental Transfers and Decentralized Public Spending,” OECD Journal of Budgeting, 5 (4): 23–43. OECD (2006a) Competitive Cities in the Global Economy (Paris: OECD). OECD (2006b) OECD Territorial Reviews: Sweden (Paris: OECD). OECD (2007) OECD Territorial Reviews: Madrid, Spain (Paris: OECD). OECD (2007a) OECD Territorial Reviews: Randstadt Holland, Netherlands (Paris: OECD). OECD (2008) OECD Territorial Reviews: Istanbul, Turkey (Paris: OECD).

M4439-BAHL_9781786435293_t.indd 468

01/03/2018 11:09



References ­469

OECD (2008a) OECD Territorial Reviews: Cape Town, South Africa (Paris: OECD). OECD (2009) “Alternative Ways of Providing Water: Emerging Options and Their Policy Implications,” Advance copy for 5th World Water Forum, available at http://www.oecd.org/env/resources/42349741.pdf. OECD (2015) Tax Administration 2015 (Paris: OECD). OECD (2015a) Revenue Statistics 1965–2014 (Paris: OECD). OECD (2016) Fiscal Federalism 2016: Making Decentralisation Work (Paris: OECD). OECD (2017) Revenue Statistics in Latin America and the Caribbean (Paris: OECD). Oldman, Oliver, Henry Aaron, Richard Bird and Stephen Kass (1967) Financing Urban Development in Mexico City (Cambridge MA: Harvard University Press). Oldman, Oliver and Mary Teachout (1979) “Some Administrative Aspects of Site Value Taxation: Defining ‘Land’ and ‘Value’: Designing a Review Process,” in Roy Bahl, ed., The Taxation of Urban Property in Less Developed Countries (Madison: University of Wisconsin Press), pp. 207–37. Olken, Benjamin (2008) “Direct Democracy and Local Public Goods: Evidence from a Field Experiment in Indonesia,” NBER Working Paper 14123, Cambridge MA. Olson, Mancur (1969) “The Principle of ‘Fiscal Equivalence’: The Division of Responsibilities among Different Levels of Government,” American Economic Review, 59 (2): 479–87. Oommen, M.A. (2010) “Have State Finance Commissions Fulfilled Their Constitutional Mandates?,” Economic and Political Weekly, 65 (30): 39–44. Osborne, David and Ted Gaebler (1992) Reinventing Government (Reading MA: Addison-Wesley). Owoye, Oluwole (1995) “The Causal Relationship between Taxes and Expenditures in the G7 Countries: Cointegration and Error-Correction Models,” Applied Economics Letters, 2 (1): 19–22. Palanisami, K., Kadiri Mohan, Mark Giordano and Chris Charles (2011) “Measuring Irrigation Subsidies in Andhra Pradesh and Southern India: An Application of the GSI Method for Quantifying Subsidies,” Global Subsidies Initiative of the International Institute for Sustainable Development, Geneva. Pang, Kenneth (2006) Property Rates in Hong Kong: Assessment, Collection and Administration, 2nd edn (Hong Kong: Hong Kong SAR Government). Panizza, Ugo (1999) “On the Determinants of Fiscal Centralization: Theory and Evidence,” Journal of Public Economics, 74: 97–139.

M4439-BAHL_9781786435293_t.indd 469

01/03/2018 11:09

470

Fiscal decentralization and local finance in developing countries

Pattanayak, Sailendra (2016) “Expenditure Control: Key Features, Stages, and Actors, Technical Notes and Manuals,” Fiscal Affairs Department, International Monetary Fund, March. Paul, Samuel (2014) “Generating Demand for Good Governance,” in Isher Judge Ahluwalia, Ravi Kanbur and P.K. Mohanty, eds, Urbanisation in India: Challenges, Opportunities and the Way Forward (New Delhi: Sage), pp. 233–59. Peacock, Alan and Jack Wiseman (1967) The Growth of Public Expenditure in the United Kingdom (London: Allen & Unwin). Petchey, Jeffrey and Sophia Levtechenkova (2007) “Fiscal Capacity Equalization and Economic Efficiency: The Case of Australia,” in Jorge Martinez-Vazquez and Bob Searle, eds, Fiscal Equalization: Challenges in the Design of Intergovernmental Transfers (New York: Springer), pp. 13–30. Peterson, George E. and Elisa Muzzini (2005) “Decentralizing Basic Infrastructure Services,” in Roland White and Paul Smoke, eds, East Asia Decentralizes: Making Local Government Work (Washington DC: World Bank), pp. 209–36. Pethe, Abhay (2013) “Metropolitan Public Finances: The Case of Mumbai,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 243–72. Phares, Donald (2005) “Tax Exportation,” in Joseph Cordes, Robert Ebel and Jane Gravelle, eds, The Encyclopedia of Taxation and Tax Policy (Washington DC: Urban Institute Press), pp. 381–3. Piffano, Horacio (2005) “Fiscal Decentralization, Tax Competition, and Federal Tax Administration: A Note from the Argentine Experience,” Revista de Economía y Estadstica (Córdoba), 43: 7–28. Pineda, Emilio, Arturo Ramirez Verdugo and Alejandro Rasteletti (2015) “A Mariachi Medley: Mexico’s Long Road to Federal Fiscalism Reform,” in Vicente Fretes Ciblis and Teresa Ter-Minassian, eds, Decentralizing Revenue in Latin America: Why and How (Washington: Inter-American Development Bank), pp. 239–77. Pinhanez, Monica (2008) “Reinventing Vat Collection: Industry Vertical Assessment, Revenue Increase, and Public Sector Reliability,” PhD dissertation, MIT. Pirenne, Henri (1927) Medieval Cities: Their Origins and the Revival of Trade (Princeton NJ: Princeton University Press). Plamondon, R. and D. Zussman (1998) “The Compliance Costs of Canada’s Major Tax Systems and the Impact of Single Administration,” Canadian Tax Journal, 46 (4): 761–85. Pommerehne, Werner W. (1977) “Quantitative Aspects of Federalism:

M4439-BAHL_9781786435293_t.indd 470

01/03/2018 11:09



References ­471

A  Study of Six Countries,” in Wallace Oates, ed., The Political Economy of Fiscal Federalism (Lexington MA: Heath-Lexington), pp. 275–346. Pommerhene, Werner and Friedrich Schneider (1983) “Does Government in a Representative Democracy Follow a Majority of Voter’s Preferences? An Empirical Examination,” in Horst Hanusch, ed., Anatomy of Government Deficiencies (Berlin: Springer), pp. 61–88. Poterba, James (1996) “Budget Institutions and Fiscal Policy in the U.S. States,” American Economic Review, 86 (2): 395–400. Pradhan, Menno, et al. (2014) “Improving Educational Quality through Enhancing Community Participation: Results from a Randomized Field Experiment in Indonesia,” American Economic Journal: Applied Economics, 6 (2): 105–26. Pradhan, Sanjay (1996) Evaluating Public Spending: A Framework for Public Expenditure Reviews (Washington DC: World Bank). Prichard, Wilson (2015) Taxation, Responsiveness and Accountability in Sub-Saharan Africa (Cambridge, UK: Cambridge University Press). Prichard, Wilson, Alex Cobham and Andrew Goodall (2014) “The ICTD Government Revenue Dataset,” IDS Working Paper 19, Institute of Development Studies, Brighton, UK. Prohl, Silke and Friedrich Schneider (2009) “Does Decentralization Reduce Government Size? A Quantitative Study of the Decentralization Hypothesis,” Public Finance Review, 37 (6): 639–64. Prud’homme, Remy (1995) “On the Dangers of Decentralization,” World Bank Research Observer, 10 (2): 201–26. Prud’homme, Remy (1996) “Comments on ‘Conflicts and Dilemmas of Decentralization’ by Rudolf Hommes,” in M. Bruno and B. Pleskovic, eds, Annual World Bank Conference on Development Economics 1995 (Washington DC: World Bank), pp. 354–9. Prud’homme, Remy (2006) “Local Government Organization and Finance: France,” in Anwar Shah, ed., Local Governance in Industrial Countries (Washington DC: World Bank), pp. 83–116. Qian, Yinyi and Barry Weingast (1997) “Federalism as a Commitment to Preserving Market Incentives,” Journal of Economic Perspectives, 11 (4): 83–92. Qiao, Baoyun and Lezheng Liu (2013) Intergovernmental Fiscal Relationship and Risk Sharing Function of Public Finance (Beijing: China Financial and Economic Publishing House) [in Chinese]. Qiao, Y., J. Martinez-Vazquez and J. Xu (2008) “The Trade-off between Growth and Equity in Decentralization Policy: China’s Experience,” Journal of Development Economics, 86 (1): 112–28.

M4439-BAHL_9781786435293_t.indd 471

01/03/2018 11:09

472

Fiscal decentralization and local finance in developing countries

Raich, Uri (2013) Unequal Development: Decentralization and Metropolitan Finance in Mexico City (Saarbrucken: VDM Verlag Dr. Muller). Rajaram, Anand, Tuan Minh Le, Nataliya Biletska and Jim Brumby (2010) “A Diagnostic Framework for Assessing Public Investment Management,” Policy Research Working Paper WPS 5397, World Bank, Washington DC. Rajaraman, Indira (2004) “Taxing Agriculture in a Developing Country: A Possible Approach,” in James Alm, Jorge Martinez-Vazquez and Sally Wallace, eds, Taxing the Hard to Tax (Amsterdam: Elsevier), pp. 245–72. Rao, M. Govinda (2003) “Challenges of Fiscal Decentralization in Developing and Transitional Economies,” in Jorge Martinez-Vazquez and James Alm, eds, Public Finance in Developing and Transitional Countries (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 35–62. Rao, M. Govinda (2009) “Fiscal Federalism in India: Trends and Reforms,” in Shinichi Ichimura and Roy Bahl, eds, Decentralization Policies in Asian Development (Singapore: World Scientific Press), pp. 107–40. Rao, M. Govinda and Richard Bird (2011) “Coping with Change: The Need to Restructure Urban Governance and Finance in India,” IMFG Papers on Municipal Finance and Governance No. 4, Toronto. Rao, M. Govinda and Richard Bird (2014) “Governance and Fiscal Federalism,” in Isher Judge Ahluwalia, Ravi Kanbur and P.K. Mohanty, eds, Urbanisation in India: Challenges, Opportunities and the Way Forward (New Delhi: Sage), pp. 203–30. Rao, M. Govinda and Raja Chelliah (1991) Survey of Research on Fiscal Federalism in India (New Delhi: National Institute of Public Finance and Policy). Rao, Govinda, Amar Nath and B.K. Vani (2004) “Fiscal Decentralization in Karnataka,” in Geeta Sethi, ed., Fiscal Decentralization to Rural Governments in India (New Delhi: Oxford University Press), pp. 43–106. Rao, Govinda and Nirvikar Singh (2006) Political Economy of Federalism in India (Oxford: Oxford University Press). Rao, Vasanth (2008) “Is Area-Based Assessment an Alternative, an Intermediate Step, or an Impediment to Value-Based Taxation,” in Roy Bahl, Jorge Martinez-Vasquez and Joan Youngman, eds, Making the Property Tax Work (Cambridge MA: Lincoln Institute of Land Policy), pp. 241–306. Rattsø, Jørn (2002) “Fiscal Controls in Europe: A Summary,” in Bernard Dafflon, ed., Local Public Finance in Europe (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 277–90.

M4439-BAHL_9781786435293_t.indd 472

01/03/2018 11:09



References ­473

Reddy, G.B. (2015) “Finance Commission Proposes, the Union Disposes,” Economic and Political Weekly, 50 (26–27): 27–30. Reinikka, Ritva and Jakob Svensson (2004) “Local Capture: Evidence from a Central Government Transfer Program in Uganda,” Quarterly Journal of Economics, 11 (2): 679–705. Revilla, Ernesto (2012) “Mexico’s Fiscal Federalism,” Public Sector and Governance Unit, Poverty Reduction and Economic Management Unit, Latin America and the Caribbean, World Bank, Washington DC, draft, June. Rezende, Fernando (2007) “Federal Republic of Brazil,” in Anwar Shah, ed., The Practice of Fiscal Federalism: Comparative Perspectives (Montreal: McGill-Queen’s University Press), pp. 73–97. Rezende, Fernando. and S. Garson (2006) “Financing Metropolitan Areas in Brazil,” Revista de Economia Contemporânea, 10 (1): 5–34. Rhoads, William and Richard Bird (1967), “Financing Urbanization in Developing Countries by Benefit Taxation: Case Study of Colombia,” Land Economics, 42 (4): 403–12. Rincke, Johannes (2008) “Yardstick Competition and Public Sector Innovation,” International Tax and Public Finance, 16 (3): 337–61. Rios, Viridiana (2016) “Mexico City Will Become a State,” Mexico Institute, Woodrow Wilson Center, Washington DC, June. Rivasplata, Charles (2012) “Urban Vehicle Restriction and the Prospects for Congestion Pricing in Latin America,” Conference CODATU XV, Addis Ababa, October. Rodden, Jonathan (2003) “Federalism and Bailouts in Brazil,” in Jonathan Rodden, Gunnar Eskeland and Jennie Litvack, eds Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA: MIT Press), pp. 213–48. Rodden, Jonathan (2003a) “Reviving Leviathan: Fiscal Federalism and the Growth of Government,” International Organization, 57: 695–729. Rodden, Jonathan (2003b) “Soft Budget Constraints and German Federalism,” in Jonathan Rodden, Gunnar Eskeland and Jennie Litvack, eds, Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA MIT Press), pp. 161–86. Rodden, Jonathan, Gunnar Eskeland and Jennie Litvack, eds (2003) Fiscal Decentralization and the Challenge of Hard Budget Constraints (Cambridge MA: MIT Press). Rodríguez-Pose, A. and A. Bwire (2003) “The Economic (In)Efficiency of Devolution,” Environment and Planning A, 36 (11): 1907–28. Rodríguez-Pose, A. and R. Ezcurra (2011) “Is Fiscal Decentralization Harmful for Economic Growth?,” Journal of Economic Geography, 11 (4): 619–43.

M4439-BAHL_9781786435293_t.indd 473

01/03/2018 11:09

474

Fiscal decentralization and local finance in developing countries

Rodrik, Dani (2015) Economics Rules (New York: Norton). Roller, Lars-Hendrik and Leonard Waverman (2001) “Telecommunications Infrastructure and Economic Development: A Simultaneous Approach,” American Economic Review, 91 (4): 909–23. Romeo, Leonardo and Paul Smoke (2016) “The Political Economy of Local Infrastructure Planning,” in Jonas Frank and Jorge MartinezVazquez, eds, Decentralization and Infrastructure in the Global Economy: From Gaps to Solutions (New York: Routledge), pp. 381–417. Romualdez, Edwardo, Angel Yoingco and Antonio Casem (1973) Philippine Public Finance (Manila: GIC Enterprises). Saavedra, Pablo A. (2010) “A Study of the Impact of Decentralization on Access to Service Delivery,” PhD dissertation, Georgia State University, available at http://scholarworks.gsu.edu/pmap_diss/40. Salmon, Pierre (2012) “Decentralization and Growth: Has the Cross-Country Approach Met a Dead End?,” Cahir du FARGO No. 1120604, June. Sánchez de Lozada, Gonzalo and Jean-Paul Faguet (2015) “Why I Decentralized Bolivia,” in Jean-Paul Faguet and Caroline Pöschl, eds, Is Decentralization Good for Development? (Oxford: Oxford University Press), pp. 31–67. Sánchez Torres, Fabio, Irina España Eljaiek and Jannet Zenteno Gonzalez (2015) “Fiscal Cumbia: Colombia’s Revenue Dynamism,” in Vicente Fretes Ciblis and Teresa Ter-Minassian, eds, Decentralizing Revenue in Latin America: Why and How (Washington DC: Inter-American Development Bank), pp. 155–98. Sandroni, Paulo (2010) “A New Financial Instrument of Value Capture in São Paulo: Certificates of Additional Construction Potential,” in Gregory Ingram and Yu-Hung Hong, eds, Municipal Revenues and Land Policies (Cambridge MA: Lincoln Institute of Land Policy), pp. 218–36. Sansom, Graham (2009) “Canberra, Australia,” in Enid Slack and Rupak Chattopadhyay, eds, Finance and Governance of Capital Cities in Federal Systems (Montreal: McGill-Queen’s University Press), pp. 11–32. Schenk, Alan, Victor Thuronyi and Wei Cui (2015) Value Added Tax: A Comparative Approach, 2nd edn (Cambridge, UK: Cambridge University Press). Schick, Allen (1998) “Why Most Developing Countries Should Not Try New Zealand Reforms,” World Bank Research Observer, 12: 123–31. Schick, Allen (2012) “Basics First is Best Practice!,” posted July 27, available at blog-pfm.imf.org/pfmblog/2012/07/basics-first-is-best-practice.html. Schroeder, Larry (2003) “Municipal Powers and Functions: The Assignment Question,” in Roy Bahl and Paul Smoke, eds, Restructuring Local Government Finance in Developing Countries (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 23–70.

M4439-BAHL_9781786435293_t.indd 474

01/03/2018 11:09



References ­475

Scicluna, Edward, David Foot and Richard Bird (1982) “Productivity Measurement in the Public Sector: The Case of Police Services in Ontario,” in Robert Haveman, ed., Public Finance and Public Employment (Detroit MI: Wayne State University Press), pp. 257–74. Self, Peter (1972) Administrative Theories and Policies (London: Allen & Unwin). Sepulveda, Cristian and Jorge Martinez-Vazquez (2011) “The Consequences of Fiscal Decentralization on Poverty and Income Inequality,” Environment and Planning C: Government and Policy, 29 (2): 321–43. Serebrisky, Tomas, Andres Gomez-Lobo, Nicolas Estupinan and Ramon Munoz-Raskin (2009) “Affordability and Subsidies in Public Urban Transport: What Do We Mean, What Can Be Done?,” Transport Reviews, 29 (6): 715–39. Sethi, Geeta, ed. (2005) Fiscal Decentralization to Rural Local Governments in India (Washington DC: World Bank). Sewell, David (1996) “The Dangers of Decentralization According to Prud’homme: Some Further Aspects,” World Bank Research Observer, 11 (1): 143–50. Shah, Anwar (2005) “Fiscal Decentralization and Fiscal Performance,” Policy Research Working Paper No. 3786, World Bank, Washington DC. Shah, Anwar (2007) “Introduction: Principles of Fiscal Federalism,” in Anwar Shah, ed., The Practice of Fiscal Federalism: Comparative Perspectives (Montreal: McGill-Queen’s University Press), pp. 3–42. Shah, Anwar (2010) “Autonomy with Accountability: The Case for Performance Oriented grants,” in Junghun Kim, Jørgen Lotz and Niels Jørgen Mau, eds, General Grants versus Earmarked Grants Theory and Practice: The Copenhagen Workshop 2009 (Albertslund: Korea Institute of Public Finance and Danish Ministry of Interior and Health), pp. 74–106. Shah, Anwar (2012) “Options for Financing SNGs,” in Directorate General of Fiscal Balance, Fiscal Decentralization in Indonesia a Decade after the Big Bang (Jakarta: University of Indonesia Press), pp. 217–49. Shah, Anwar (2013) “Grant Financing of Metropolitan Areas: A Review of Principles and Worldwide Practices,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 213–42. Shah, Anwar (2016) “Decentralized Provision of Public Infrastructure and Corruption,” in Jonas Frank and Jorge Martinez-Vazquez, eds,

M4439-BAHL_9781786435293_t.indd 475

01/03/2018 11:09

476

Fiscal decentralization and local finance in developing countries

Decentralization and Infrastructure in the Global Economy: From Gaps to Solutions (New York: Routledge), pp. 418–54. Shoup, Carl S. (1969) Public Finance (Chicago IL: Aldine Publishing). Sidik, Machfud and Kadjatmiko (2004) “Combining Expenditure Assignment, Revenue Assignment and Grant Design in Indonesia’s Fiscal Decentralization,” in James Alm, Jorge Martinez-Vazquez and Sri Mulyani Indrawati, eds, Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 134–58. Siemiatycki, Matti and Jonathan Friedman (2012) “The Trade-Offs of Transferring Demand Risk on Urban Transit Public–Private Partnerships,” Public Works Management and Policy, 17: 283–302. Sjoquist, David (2005) “The Land Value Tax in Jamaica: An Analysis and Options for Reform,” Bulletin for International Fiscal Documentation, 59 (11): 489–97. Slack, Enid (2000) “A Preliminary Assessment of the New City of Toronto,” Canadian Journal of Regional Science, 23 (1): 13–29. Slack, Enid (2002) “Municipal Finance and the Pattern of Urban Growth,” C.D. Howe Institute Commentary No. 160, Toronto, February. Slack, Enid (2004) “Property Taxation in the United Kingdom,” in Richard Bird and Enid Slack, eds, International Handbook on Land and Property Taxation (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 81–90. Slack, Enid (2006) “Alternative Approaches to Taxing Land and Property,” in Richard Bird and François Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington, DC: World Bank), pp. 197–223. Slack, Enid (2007) “Managing the Coordination of Service Delivery in Metropolitan Areas: The Role of Metropolitan Governance,” World Bank Policy Research Working Paper WPS 4317, August. Slack, Enid and Richard Bird (2012) “Merging Municipalities: Is Bigger Better?,” in Antti Moisio, ed., Rethinking Local Government: Essays on Municipal Reform (Helsinki: Government Institute for Economic Research), pp. 83–130. Slack, Enid and Richard Bird (2013) “Does Municipal Amalgamation Strengthen the Financial Viability of Local Government? A Canadian Example,” Public Finance and Management, 13 (2): 99–123. Slack, Enid and Richard Bird (2014) “The Political Economy of Property Tax Reform,” OECD Working Papers on Fiscal Federalism No. 18, Paris, April. Slack, Enid and Richard Bird (2015) “How to Reform the Property Tax: Lessons from Around the World,” IMFG Papers on Municipal Finance and Governance, No. 21, Toronto.

M4439-BAHL_9781786435293_t.indd 476

01/03/2018 11:09



References ­477

Slack, Enid and Rupak Chattopadhyay (2009) Finance and Governance of Capital Cities in Federal Countries (Montreal: McGill-Queen’s University Press). Slack, Enid and Rupak Chattopadhyay (2013) Governance and Finance of Metropolitan Cities in Federal Countries (Toronto: Oxford University Press). Slack, Enid, John LaFevre and Ihor Shpak (1998) “Property Tax in Ukraine: Third Attempt,” Budget and Fiscal Review, Second Quarter 1998, Fiscal Analysis Office, Verkhovna Rada, Kyiv. Slack, Enid and Almos Tassonyi (2017) “Financing Urban Infrastructure in Canada: Overview, Trends, and Issues,” in Richard Bird and Enid Slack, eds, Financing Infrastructure: Who Should Pay? (Montreal: McGill-Queen’s University Press), pp. 21–53. Smart, Michael and Richard Bird (2010) “Earmarked Grants and Accountability in Government,” in Junghun Kim, Jørgen Lotz and Niels Jørgen Mau, eds, General Grants versus Earmarked Grants Theory and Practice: The Copenhagen Workshop 2009 (Albertslundk: Korea Institute of Public Finance and Danish Ministry of Interior and Health), pp. 40–73. Smith, Ernest (1998) Federal-Provincial Tax Sharing and Centralized Tax Collection in Canada (Toronto: Canadian Tax Foundation). Smith, Roger (1974) “Financing Cities in Developing Countries,” IMF Staff Papers, 21 (2): 329–88. Smith, Roger (1991) “Motor Vehicle Taxation,” in Roy Bahl, ed., The Jamaican Tax Reform (Cambridge MA: Lincoln Institute of Land Policy), pp. 557–86. Smith, Stephen (2006) “Taxes on Road Transport,” in Sijbren Cnossen, ed., Excise Tax Policy and Administration in Southern African Countries (Pretoria: University of South Africa), pp. 117–50. Smoke, Paul (1994) Local Government Finance in Developing Countries: The Case of Kenya (Nairobi: Oxford University Press). Smoke, Paul (2013) “Metropolitan Cities in the National Fiscal and Institutional Structure,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 57–84. Smoke, Paul (2014) “Why Theory and Practice Are Different: The Gap between Principles and Reality in Subnational Revenue Systems,” in Richard Bird and Jorge Martinez-Vazquez, eds, Taxation and Development: The Weakest Link? (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 287–325. Smoke, Paul, Eduardo J. Gomez and George E. Peterson, eds (2006) Decentralization in Asia and Latin America: Towards a Comparative

M4439-BAHL_9781786435293_t.indd 477

01/03/2018 11:09

478

Fiscal decentralization and local finance in developing countries

Interdisciplinary Perspective (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing). Smoke, Paul and Blaine Lewis (2014) “Incorporating Subnational Performance Incentives in the Indonesian Intergovernmental Framework,” Proceedings: Annual Conference on Taxation and Minutes of the Annual Meeting of the National Tax Association, 101: 418–26. Smoke, Paul and Joanne Morrison (2011) “Decentralization in Cambodia: Consolidating Central Power or Building Accountability from Below?,” in Jorge Martinez-Vazquez and François Vaillancourt, eds, Decentralization in Developing Countries: Global Perspectives on the Obstacles to Fiscal Devolution (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 327–58. Smoke, Paul, William Muhumuza and Emmanuel Ssewankambo (2010) Comparative Assessment of Decentralization in Africa: Uganda Desk Study (Washington: USAID). Smolka, Martim (2013) Implementing Value Capture in Latin America: Policies and Tools for Urban Development (Cambridge MA: Lincoln Institute of Land Policy). Smolka, Martim and Claudia M. De Cesare (2006) “Property Taxation and Informality: Challenges for Latin America,” Land Lines, 18 (3): 14–19. Soifer, Hillel David (2015) State Building in Latin America (Cambridge, UK: Cambridge University Press). Sokoloff, Kenneth L. and Eric M. Zolt (2006) “Inequality and Taxation: Evidence from the Americas on How Inequality May Influence Tax Institutions,” Tax Law Review, 59 (2): 201–76. Solomon, David with Ndumiso Gola, Mbulelo Ngxesha and Wycliffe Ndlela (2002) “Tribal Land and the Property Tax,” in Michael E. Bell and John H. Bowman, eds, Property Taxes in South Africa: Challenges in the Post-Apartheid Period (Cambridge MA: Lincoln Institute of Land Policy), pp. 195–214. Sorens, Jason P. (2015) “Secession Risk and Fiscal Federalism,” Publius: The Journal of Federalism, 46 (1): 25–50. Sow, Mousse and Ivohasina F. Razafimahefa (2015) “Fiscal Decentraliza­ tion and the Efficiency of Public Service Delivery,” IMF Working Paper WP/15/59, Washington DC. Spahn, Paul Bernd (1997) “Decentralized Government and Macroeconomic Control,” Infrastructure Notes FM-12, World Bank, Washington DC. Spahn, Paul Bernd (2015) “Contract Federalism,” in Ehtisham Ahmad and Giorgio Brosio, eds, Handbook of Multilateral Finance (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 144–60.

M4439-BAHL_9781786435293_t.indd 478

01/03/2018 11:09



References ­479

Spahn, Paul Bernd and Wolfgang Föttinger (1997) “Germany,” in Teresa Ter-Minassian, ed., Fiscal Federalism in Theory and Practice (Washington DC: International Monetary Fund), pp. 226–48. Spicer, Zachary (2015) “Cooperation and Capacity: Inter-Municipal Agreements in Canada,” IMFG Papers on Municipal Finance and Governance No. 19, Toronto. Spratt, Stephen (2012) “Environmental Taxation and Development: A Scoping Study,” ICTD Working Paper 2, International Centre for Taxation and Development, Brighton, UK, June. Sridhar, Kala and Simanti Bandyopadhyay (2007) “Improving Fiscal Health of Indian Cities: A Pilot Study of Kolkata,” National Institute of Public Finance and Policy, New Delhi. Steffensen, Jesper (2006) “Local Government Organization and Finance: Uganda,” in Anwar Shah, ed., Local Governance in Developing Countries (Washington DC: World Bank), pp. 93–141. Steffensen, Jesper and Henrick Larsen (2005) “Conceptual Basis for Performance Based Grant Systems and Selected International Experience,” United Nations Development Programme, New York. Stegarescu, Dan (2005) “Public Sector Decentralization: Measurement, Concepts and Recent International Trends,” Fiscal Studies, 26 (3): 301–33. Stegarescu, Dan (2009) “The Effects of Economic and Political Integration on Fiscal Decentralization: Evidence from OECD Countries,” Canadian Journal of Economics, 42 (2): 694–718. Stein, Ernesto (1998) “Fiscal Decentralisation and Government Size in Latin America,” in Kuchiro Fukasaku and Ricardo Hausmann, eds, Democracy, Decentralisation and Deficits in Latin America (Washington DC: Inter-American Development Bank and OECD), pp. 95–120. Steytler, Nico (2013) “Governance and Finance of Two South Africa Metropolitan Areas,” in Enid Slack and Rupak Chattopadhyay, eds, Governance and Finance of Metropolitan Areas in Federal Systems (Toronto: Oxford University Press), pp. 189–222. Stossberg, Sibylle, David Bartolini and Hansjörg Blöchliger (2016) “Fiscal Decentralisation and Income Inequality: Empirical Evidence from OECD Countries,” Economics Department Working Paper No. 1331, OECD, Paris. Stren, Richard (1992) “Large Cities in the Third World,” in Metropolitan Planning and Management in the Developing World: Abidjan and Quito (Nairobi: United Nations Centre for Human Settlements). Su, Ming and Quanhou Zhao (2006) “The Fiscal Framework and Urban Infrastructure Finance in China,” Policy Research Paper 4051, World Bank, Washington DC, November.

M4439-BAHL_9781786435293_t.indd 479

01/03/2018 11:09

480

Fiscal decentralization and local finance in developing countries

Sud, Inder and Serdar Yilmaz (2013) “Institutions and Politics of Metropolitan Management,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 107–34. Sutherland, Douglas, Robert Price and Isabelle Joumard (2005) “Fiscal Rules for Sub-Central Governments: Design and Impact,” Economics Department Working Paper No. 465, OECD, Paris. Tanzi, Vito (1996) “Fiscal Federalism and Decentralization: A Review of Some Efficiency and Macroeconomic Aspects,” in M. Bruno and B. Pleskovic, eds, Annual World Bank Conference on Development Economics (Washington DC: World Bank), pp. 295–316. Tanzi, Vito and Hamid Davoodi (1997) “Corruption, Public Investment, and Growth,” IMF Working Paper No. 97/139, Washington DC. Tassonyi, Almos, Richard Bird and Enid Slack (2015) “Can GTA Municipalities Raise Property Taxes? An Analysis of Tax Competition and Revenue Hills,” IMFG Papers on Municipal Finance and Governance, No. 20, Toronto. Tata Institute of Social Sciences and University of Mumbai (2001) Rationalisation of Property Tax in Brihanmumbai: A Proposed Capital Value-Based Model (Mumbai: Brihanmumbai Municipal Corporation). Taylor, Zack (2016) “Good Governance at the Local Level: Meaning and Measurement,” IMFG Papers on Municipal Finance and Governance, No. 26, Toronto. Ter-Minassian, Teresa, ed. (1997) Fiscal Federalism in Theory and Practice (Washington DC: International Monetary Fund). Ter-Minassian, Teresa (2015) “Elusive Goal: Brazil’s Subnational Revenue System Reform,” in Vicente Fretes Ciblis and Teresa Ter-Minassian, eds, Decentralizing Revenue in Latin America: Why and How (Washington DC: Inter-American Development Bank), pp. 139–80. Ter-Minassian, Teresa and Jon Craig (1997) “Control of Subnational Government Borrowing,” in Teresa Ter-Minassian, ed., Fiscal Federalism in Theory and Practice (Washington DC: International Monetary Fund), pp. 156–72. Terra, Rafael and Enlinson Mattos (2015) “Accountability and Yardstick Competition in the Public Provision of Education,” Working Paper 387, FGV São Paulo School of Economics, April. Thirsk, Wayne (1997) Tax Reform in Developing Countries (Washington DC: World Bank). Tiebout, Charles (1956) “A Pure Theory of Local Expenditures,” Journal of Political Economy, 64 (5): 416–24. Timmons, Jeffrey (2005) “The Fiscal Contract: States, Taxes, and Public Services,” World Politics, 57: 530–67.

M4439-BAHL_9781786435293_t.indd 480

01/03/2018 11:09



References ­481

Togo, Hisatake (1995) “The Metropolitan Strategies of Tokyo: Toward the Restoration of Balanced Growth,” in L. Sharpe, ed., The Government of World Cities: The Future of the Metro Model (Chichester, UK: Wiley), pp. 177–202. Tokyo Metropolitan Government (2010), available at www.metro.tokyo.jp/ ENGLISH/PROFILE/index.htm. Torgler, Benno, Friedrich Schneider and Christoph Schaltegger (2010) “Local Autonomy, Tax Morale and the Shadow Economy,” Public Choice, 144: 293–321. Tosun, Mehmet Serkan and Serdar Yilmaz (2008) “Centralization, Decentralization, and Conflict in the Middle East and North Africa,” World Bank Policy Research Working Paper; No. 4774. Treisman, Daniel (1999) After the Deluge: Regional Crises and Political Consolidation in Russia (Ann Arbor: University of Michigan Press). Treisman, Daniel (2000) Decentralization and the Quality of Government (Washington DC: International Monetary Fund). Treisman, Daniel (2007) The Architecture of Government: Rethinking Political Decentralization (Cambridge, UK: Cambridge University Press). Ulbricht, Alexander (2008) “The Decentralization of Tax Administration in Germany: Consequences,” in Núria Bosch and José M. Durán, eds, Fiscal Federalism and Political Decentralization (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 193–208. United Nations (2008) World Urbanization Prospects: The 2007 Revision, Economic and Social Affairs (New York). UNCDF (2013) Performance-Based Grant Systems: Concept and International Experience (New York: United Nations Capital Development Fund). UNDP (2006) Human Development Report 2006: Beyond Scarcity: Power, Poverty and the Global Water Crisis (New York: United Nations Development Programme). UN Habitat (2011) Land and Property Tax Policy Guide (Nairobi: United Nations). UN Habitat and ESCAP (2015) The State of Asian and Pacific Cities 2015: Urban Transformations, Shifting form Quantity to Quality (New York: United Nations). USAID (2006) “Reforming Property Taxation in Southeast Europe: A Comparative Review,” Fiscal Decentralization Initiative for Central and Eastern Europe, Open Society Institute, Budapest. Ushie, Vanessa (2012) “Political Decentralization and Natural Resource Governance in Nigeria,” Research Report, North-South Institute, Ottawa, December.

M4439-BAHL_9781786435293_t.indd 481

01/03/2018 11:09

482

Fiscal decentralization and local finance in developing countries

Vaillancourt, François (2006) “Budgeting, Financial Management, and Financial Markets in an Intergovernmental Context,” in Richard M. Bird and François Vaillancourt, eds, Perspectives on Fiscal Federalism (Washington DC: World Bank), pp. 35–56. Vaillancourt, François and Richard Bird (2007) “The Interregional Incidence of Central Budgets in Federations: Some Evidence from Canada,” Public Budgeting and Finance, 27 (1): 1–19. Vaillancourt, François and Richard Bird (2016) “Decentralization in European and MENA Countries: Glue or Solvent?,” in M. Mustafa Erdoğlu and Bryan Christiansen, eds, Comparative Political and Economic Perspectives on the MENA Region (Hershey PA: Information Science Reference/IGI Global), pp. 1–27. Van Dijk, Meine Pieter (2012) “Sanitation in Developing Countries: Innovative Solutions in a Value Chain Framework,” available at www. researchgate.net/publication/221929703. Van Ryneveld, Philip (2006) “Mobilizing Urban Infrastructure Finance within a Responsible Fiscal Framework: The South African Case,” Policy Research Working Paper No. 4042, World Bank, Washington DC. Van Ryneveld, Philip (2007) “Fiscal Decentralization and Financing of Urban Infrastructure in South Africa,” in George Peterson and Patricia Clarke Annez, eds, Financing Cities (Washington DC: World Bank and Sage), pp. 183–204. Varanyuwatana, Sakon (2004) “Property Tax in Thailand,” in Richard Bird and Enid Slack, eds, International Handbook of Land and Property Taxation (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 159–64. Vázquez-Caro, Jaime and Richard Bird (2011) “Benchmarking Tax Administrations in Developing Countries: A Systemic Approach,” eJournal of Tax Research, 9 (1): 5–37. Vázquez-Caro, Jaime and José Abelino Ospina (2006) “Evaluación sistémica de la ley impositiva colombiana,” in Juan José Echavarría, ed., Bases para una reforma tributaria estructural en Colombia (Bogotá: Fedesarrollo y Banco de la República), pp. 385–507. Vehorn, Charles and Ehtisham Ahmad (1997) “Tax Administration,” in Teresa Ter-Minassian, ed., Fiscal Federalism in Theory and Practice (Washington DC: International Monetary Fund), pp. 108–34. Vergara, Victor (1999) “Linking Local Integrity to Decentralization: An Overview to Making Local Government Credible and Effective,” paper prepared for Intergovernmental Fiscal Relations and Local Financial Management Core Course, World Bank Institute, Caracas, Venezuela, June.

M4439-BAHL_9781786435293_t.indd 482

01/03/2018 11:09



References ­483

Venables, Anthony (2005) “Spatial Disparities in Developing Countries: Cities, Regions, and International Trade,” Journal of Economic Geography, 5 (1): 3–21. Vetter, D.M. and M. Vetter (2011) “Could Land-Based Financing Help Create Fiscal Space for Investment by Brazil’s Municipalities?,” Lincoln Institute of Land Policy Working Paper WP11 DV1, Cambridge MA. Vetter, David, Kaizo Beltrao and Rosa Massena (2014) “Residential Wealth Distribution in Rio de Janeiro,” Land Lines, 26 (1): 14–21. Vickrey, William (1963) “General and Specific Financing of Urban Services,” in Howard Schaller, ed., Public Expenditure Decisions in the Urban Community (Washington DC: Resources for the Future). Vigneault, Marianne (2005) “Intergovernmental Fiscal Relations and the Soft Budget Constraint Problem,” available at www.queensu.ca/iigr/sites/ webpublish.queensu.ca.iigrwww/files/files/WorkingPapers/Archive/​20​0​5/​ 2005-2Vigneault.pdf. Vinuela, Lorena (2016) “Trends and Quality of Decentralized Public Investment,” in Jonas Frank and Jorge Martinez-Vazquez, eds, Decentralization and Infrastructure in the Global Economy (New York: Routledge), pp. 54–98. Von Braun, J. and U. Grote (2002) “Does Decentralization Serve the Poor?,” in Ehtisham Ahmad and Vito Tanzi, eds, Managing Fiscal Decentralization (London: Routledge), pp. 68–94. Wallich, Christine, Rosario Manasan and Saloua Sehili (2007) “Subsidiarity and Solidarity: Fiscal Decentralization in the Philippines,” in Richard Bird and Robert Ebel, eds, Fiscal Fragmentation in Decentralized Countries (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 363–95. Ward, Richard (2002) “Demonstration of Computer-Assisted Mass Appraisal: Three Pilot Studies,” in Michael Bell and John Bowman, eds, Property Taxes in South Africa: Challenges in the Post-Apartheid Era (Cambridge MA: Lincoln Institute of Land Policy), pp. 113–38. Wasylenko, Michael (1987) “Fiscal Decentralization and Economic Development,” Public Budgeting and Finance, 7 (4): 57–71. Webb, Steven (2004) “Fiscal Responsibility Laws for Subnational Discipline: The Latin American Experience,” Policy Research Working Paper 3309, World Bank, Washington DC. Weingast, Barry (2009) “Second-Generation Fiscal Federalism,” Journal of Urban Economics, 65: 279–93. Werner, Jan (2016) “Options to Support Urban Infrastructure Delivery in South Africa,” in Eric Sucky et al., eds, Mobility in a Globalised World 2015 (Bamberg, Germany: University of Bamberg Press), pp. 91–101.

M4439-BAHL_9781786435293_t.indd 483

01/03/2018 11:09

484

Fiscal decentralization and local finance in developing countries

West, Loraine and Christine Wong (1995) “Fiscal Decentralization and Growing Regional Disparities in Rural China: Some Evidence in the Provision of Social Services,” Finalization Conference on Public Finance Issues in the Transition Economies of Asia, December. Wetzel, Deborah (2013) “Metropolitan Governance and Finance in São Paulo,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of land Policy), pp. 309–38. Wibbels, Erik (2000) “Federalism and the Politics of Macroeconomic Policy and Performance,” American Journal of Political Science, 44 (4): 687–702. Wilensky, Harold (1975) The Welfare State and Equality (Berkeley: University of California Press). Winer, Stanley (1983) “Some Evidence on the Effect of the Separation of Spending and Taxing Decisions,” Journal of Political Economy, 91 (1): 126–40. Winer, Stanley (2016) “The Political Economy of Taxation: Power, Structure, Redistribution,” CESifo Working Paper No. 6252, December. Woller, Gary and Kerk Phillip (1998) “Fiscal Decentralization and IDC Economic Growth,” Journal of Development Studies, 34 (4): 139–48. Wong, Christine (1998) “Municipal Finance in China: The Development of Extrabudgetary Revenues,” paper presented at International Municipal Finance Forum, World Bank, Washington DC, April. Wong, Christine (2013) “Paying for Urbanization in China: Challenges of Municipal Finance in the Twenty-First Century,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 273–308. Wong, Christine and Richard Bird (2008) “China’s Fiscal System: A Work in Progress,” in Loren Brandt and Thomas G. Rawski, eds, China’s Great Transformation: Origins, Mechanisms, and Consequences of the Post-Reform Economic Boom (Cambridge, UK: Cambridge University Press), pp. 429–66. Wong, Christine, Christopher Heady and Wing Woo (1995) Fiscal Management and Economics Reform in the People’s Republic of China (Hong Kong: Oxford University Press). Wong, Susan and Scott Guggenheim (2005) “Community-Driven Development: Decentralization’s Accountability Challenge,” in Roland  White and Paul Smoke, eds, East Asia Decentralizes: Making Local Government Work (Washington DC: World Bank), pp. 253–67. World Bank (annual) Doing Business (Washington DC: World Bank).

M4439-BAHL_9781786435293_t.indd 484

01/03/2018 11:09



References ­485

World Bank (1996) Colombia: Reforming the Decentralization Law: Incentives for an Effective Delivery of Services, 2 vols. (Washington DC: World Bank). World Bank (1996a) Argentina: Provincial Finances Study: Selected Issues in Fiscal Federalism, 2 vols. (Washington DC: World Bank). World Bank (1998) Public Expenditure Management Handbook (Washington DC: World Bank). World Bank (2000) World Development Report 1999/2000. Entering the 21st Century (Washington DC: World Bank). World Bank (2003) Chile: Subnational Government Finance (Washington DC: World Bank). World Bank (2004) World Development Report 2004: Making Services Work for Poor People, (Washington DC: World Bank). World Bank (2005) The Effectiveness of World Bank Support for CommunityBased and Driven Development (Washington DC: World Bank). World Bank (2006) Where is the Wealth of Nations? Measuring Capital for the 21st Century (Washington, DC: World Bank). World Bank (2007) India: Rural Governments and Service Delivery (Washington DC: World Bank). World Bank (2007a) Brazil – São Paulo: Inputs for a Sustainable Competitive City Strategy, 2 vols. (Washington DC: World Bank). World Bank (2008) Brazil: Topics in Fiscal Federalism (Washington DC: World Bank). World Bank (2009) Decentralization, Democracy, and Development: Recent Experience from Sierra Leone (Washington DC: World Bank). World Bank (2009a) Colombia Decentralization: Options and Incentives for Efficiency (Washington DC: World Bank). World Bank (2009b) World Development Report 2009: Reshaping Economic Geography, (Washington: World Bank). World Bank (2009c) Improving Municipal Management for Cities to Succeed: An IEG Special Study (Washington DC: World Bank). World Bank (2010) Peru: The Decentralization Process and Its Links with Public Expenditure Efficiency (Washington DC: World Bank). World Bank (2012) “China: Land Policies for a Modern and Harmonious Society,” Sustainable Development Department, East Asia and Pacific Region Discussion Paper, World Bank, Washington DC, June. World Bank (2013) Beyond the Annual Budget: Global Experience with Medium Term Expenditure Frameworks (Washington DC: World Bank). World Bank (2014) Public–Private Partnerships: Reference Guide Version 2.0 (Washington DC: World Bank). World Bank and Development Research Center of the People’s Republic

M4439-BAHL_9781786435293_t.indd 485

01/03/2018 11:09

486

Fiscal decentralization and local finance in developing countries

of China (2013) China 2030: Building a Modern, Harmonious, and Creative Society (Washington DC: World Bank). World Bank and Development Research Centre of the People’s Republic of China (2014) Urban China: Toward Efficient, Inclusive, and Sustainable Urbanization (Washington DC: World Bank). World Bank Independent Evaluation Group (2008) Decentralization in Client Countries: An Evaluation of World Bank Support, 1990–2007 (Washington DC: World Bank). World Water Council (2006) 4th World Water Forum Synthesis (Mexico: World Water Council). Yao, Guevera Assamoi (2007) “Fiscal Decentralization and Poverty Reduction Outcomes: Theory and Evidence,” PhD dissertation, Georgia State University, available at http://digitalarchive.gsu.edu/econ_diss/26. Yatta, François and François Vaillancourt (2011) “Africa,” in Jorge Martinez-Vazquez and Paul Smoke, eds, Local Government Finance: The Challenges of the 21st Century (Cheltenham, UK: Edward Elgar), pp. 23–60. Yilmaz, Serdar, Yakup Beris and Rodrigo Serrano-Berthet (2010) “Linking Local Government Discretion and Accountability in Decentralisation” Development Policy Review, 28 (3): 259–93. Yilmaz, Serdar and Varsha Venugopal (2011) “Obstacles to Decentralization in Ethiopia: Political Controls versus Discretion and Accountability,” in Jorge Martinez-Vazquez and François Vaillancourt, eds, Decentralization in Developing Countries (Cheltenham, UK and Northampton MA, USA: Edward Elgar Publishing), pp. 431–64. Youngman, Joan (2005) “Taxing and Untaxing Land: Current Use Assessment of Farmland,” State Tax Notes, 37 (10): 727–38. Youngman, Joan (2016) A Good Tax: Legal and Policy Issues for the Property Tax in the United States (Cambridge MA: Lincoln Institute of Land Policy). Youngman, Joan and Jane Malme, eds (1994) An International Survey of Taxes on Land and Buildings (Amsterdam: Kluwer). Yusuf, Shahid (2013) “Metropolitan Cities: Their Rise, Role and Future,” in Roy Bahl, Johannes Linn and Deborah Wetzel, eds, Financing Metropolitan Governments in Developing Countries (Cambridge MA: Lincoln Institute of Land Policy), pp. 31–56. Zhang, H. and X. Chen (2007) “Fiscal Competition and the Structure of Local Public Expenditure in China,” Frontiers of Economics in China, 2 (2): 237–49. Zhang, Tao and Heng-fu Zou (1998) “Fiscal Decentralization, Public Spending and Economic Growth in China,” Journal of Public Economics, 67 (2): 221–40.

M4439-BAHL_9781786435293_t.indd 486

01/03/2018 11:09



References ­487

Zhang, Zhihua and Jorge Martinez-Vazquez (2006) “The System of Equalization Transfers in China,” International Studies Program, Georgia State University, Working Paper 03-12, July. Zhong, H. (2010) “The Impact of Decentralization of Health Care Administration on Equity in Health and Health Care in Canada,” International Journal of Health Care Finance and Economics, 10 (3): 219–37. Zodrow, George (2006) “Reflections on the New View and Benefit View of the Property Tax,” in Wallace Oates, ed., Property Taxation and Local Government Finance (Cambridge MA: Lincoln Institute of Land Policy), pp. 79–111.

M4439-BAHL_9781786435293_t.indd 487

01/03/2018 11:09

M4439-BAHL_9781786435293_t.indd 488

01/03/2018 11:09

Index accountability 6, 12–15, 42, 75, 78–9, 88, 89, 95–7, 107–8, 140, 169, 213, 224, 259–61, 299, 396, 399, 401–3 dual 6, 42, 44, 96, 140, 224 link to autonomy 89 see also metropolitan areas accrual budgeting, see budgeting Addis Ababa 250, 380 administration, local 362–3; see also local government; tax administration administrative costs 212, 225, 258, 288, 321 Afonso, José Roberto 122 Africa 285, 344 agglomeration economies 222, 285, 341, 344, 361 agricultural land 193, 259–61, 389 Ahluwalia, Isher 361 Ahmad, Ehtisham 60 Ahmedabad 237, 362 Albania 100, 161 Alegre, Juan 124 Alexeev, Michael 320 Alm, James 319, 325 Andres, Luis Alberto 145 Angola 26 annexation, see mergers Annez, Patricia 158 Antioquia 250 appeals 256, 275 Araujo, M. Caridad 59 area-based competition 126 area-based property tax, see property tax Argentina 38, 105, 147, 162, 195, 196, 204, 223, 234, 376, 395, 396, 409 borrowing 24, 221 tax sharing 289, 294, 298, 299 transfers 313, 314, 320

turnover tax 207–8, 369 see also Buenos Aires Armenia 213 Artana, Daniel 320, 369 assignment, see expenditure assignment; tax assignment asymmetry 20, 26–8, 35, 129, 141, 204, 274, 314, 348–9, 406–8, 417 audit 112–13, 316, 326, 363 Australia 194, 203, 239, 268, 291, 295, 304, 323, 324, 328 Austria 200 authority, need for 140; see also accountability autonomy 40, 89–90, 142, 188, 191, 192, 401–3 limits on 69–90, 93, 191 see also autonomy; conditional grants; mandates; taxing power Bach, Steffen 40 bailouts 135, 153, 155, 178, 403 banded property tax 240 Bangalore (Bengalaru) 218, 237, 239, 249, 362 Bangkok 295, 360, 380, 381, 387 Bangladesh 63, 105, 143, 289, 320, 407 bankruptcy, municipal 186, 351 barangays, see Philippines Barankay, Iwan 60 Barco Vargas, Virgilio 349 Bardhan, Pranab 39 Barenstein, Matias 59 basic needs 323 Baskaran, Tushyanthan 56 Basurto, Pia 63 Beijing 387 Belo Horizonte 255 benchmarking 45 benefit model 142, 168–77, 221 489

M4439-BAHL_9781786435293_t.indd 489

01/03/2018 11:09

490

Fiscal decentralization and local finance in developing countries

benefit taxes 129, 147, 188, 205, 211, 222, 259–60, 378 Besley, Timothy 126 betterment 267; see also value capture Bhattacharyya, Sambit 309 Bhutan 10 ‘big bang’ reform 65, 89, 395 vs. gradualism 65, 396, 404–6 Blöchliger, Hansjörg 67, 190 block grants 313 Boadway, Robin 55, 86, 328, 394 Boex, Jameson 319, 325 Bogotá 249, 250, 254, 256, 267, 348, 357, 359, 369, 372, 380 Bolivia 18, 126, 149, 158, 194, 303, 311, 391, 413 decentralization 410–11, 421 borrowing, subnational 24–5, 122, 143, 152–7, 167, 176, 221, 383, 395 limiting 155–6 packaging 153–6, 392 see also debt Bosnia and Herzegovina 18 Bossert, Thomas 61 Botswana 252–3 brain drain 345 Brazil 8, 12, 13, 16, 24, 25, 28, 42, 52, 91, 105, 122, 147, 187, 193, 204, 234, 238, 276, 303, 316, 375, 384, 386, 395, 396, 407, 409 property tax 242, 250, 252, 255, 259 service taxes 206, 369 state VAT 195, 202, 206 tax sharing 291, 299, 326 transfers 23, 306, 312, 314, 320, 325 value capture 267, 268, 269 see also São Paulo Brennan, Geoffrey 188 Breton, Albert 14, 172 Breuss, Fritz 55 Brodjonegoro, Bambang 319 Buchanan, James 188 Bucharest 293 Budapest 292, 356 budget constraint 24, 185–7; see also hard budget constraint; soft budget constraint budgeting 113–14, 118, 179, 223, 363 capital 179–80, 402 cash and accrual 110–11, 223

M4439-BAHL_9781786435293_t.indd 490

transparency 96, 111 see also fiscal gap; PFM Buenos Aires 254, 345, 357, 361, 369, 372 Bulgaria 356 business property tax 169, 174, 210, 222, 237, 246; see also property tax Byrnes, Joel 83 cadaster 242, 248, 249–50 Cairo 380 CAMA (computer assisted mass appraisal) 253, 263; see also tax technology Cambodia 7, 12, 129 Cameroon 294 Canada 35, 52, 77, 83, 116, 117, 129, 192, 211, 262, 268, 275, 279, 334, 335, 347, 352, 357, 359, 390 equalization 186, 295, 304, 331 piggybacked taxes 215–16 VAT 297, 335 see also Toronto Canberra 268 capacity, see administration; fiscal capacity; local governments Cape Town 254, 352, 360, 387 capital account 178–80; see also budgeting; deficit capital gains tax 263 capital grants 179, 180; see also infrastructure; transfers capital value, see property tax Capuno, Joseph 325 cascading, see tax cascading Case, Anne 126 case study approach 31, 41, 45, 56 cash transfer schemes 105, 127 Catalonia 17, 347 CDD, see community-driven development Central African Republic 237 central government, key role 101, 110, 112, 133, 135, 214, 274, 382, 385, 392, 408, 418 CEPACS 269 champions of decentralization 387, 410–13 of grants 295, 308

01/03/2018 11:09



Index ­491

Chelliah, Raja 184, 320 Chen, Xian 126 Chile 105, 136, 258, 359, 382 property tax 231, 239, 242, 246, 250 China 6, 10, 13, 15, 42, 52, 69, 78, 84, 88, 102, 105, 126, 128, 154, 159, 192, 197, 204, 279, 283, 315, 341, 362, 376, 380, 387, 392, 396 decentralization 53, 224, 37, 398, 404 earmarking 94, 307 expenditure assignment 100, 304 extrabudgetary levies 184, 203, 216 fiscal disparities 285, 299, 324, 325 land sales 148–9, 223, 266, 268 local business tax 44, 217, 300, 367, 370 local expenditures 87, 91, 314 tax sharing 291, 293, 296, 320, 381 VAT 292, 300, 370 cities, see metropolitan areas citizen report cards 96, 118; see also community score cards; surveys; transparency collection-driven reform 255, 271–2 collection rate 255, 271 Cnossen, Sijbren 264 Collier, Paul 364 Colombia 7, 10, 12, 16, 18, 20, 32, 33, 58, 77, 89, 90, 92, 104, 126, 131–2, 149, 159, 161, 196, 204, 206, 213, 234, 250, 254, 267, 316, 356, 367, 375 benefit financing 129, 147, 222 decentralization process 103, 138, 315, 396, 406, 416 monitoring grants 327, 333, 409 property tax 246, 256, 277, 373 subnational expenditures 87, 279 transfers 23, 294, 295, 308, 312, 320, 325, 392 turnover (industry and commerce) tax 195, 208, 218, 369 see also Bogotá Commonwealth Grants Commission 328 communal ownership 243 community-driven development (CDD) 6, 98–9 community score cards 96, 118;

M4439-BAHL_9781786435293_t.indd 491

see also citizen report cards; transparency comparative approach 31, 45, 66 compliance cost 212, 225, 256, 322 concurrency 125–6; see also expenditure assignment conditional grants 40, 81, 87, 89, 94–5, 134, 150, 151, 287–8, 281, 306–7, 312–16, 401; see also externalities; transfers connection fees 145, 379 contingent liabilities 182 contracting 85, 102, 116, 130, 132, 136–7, 157–8, 160, 161, 173, 253 intergovernmental 5, 85, 102, 129, 296, 352, 359, 379 see also fiscal contracting; privatization; public–private partnerships Copenhagen 354 corporate income tax 202, 211, 297–8 correspondence principle 169, 201, 205; see also benefit model; linkage; Wicksellian connection corruption 29–30, 58, 59, 60, 66, 92, 107–8, 118, 138, 161, 173, 215, 220, 266, 310, 395–6, 419 cost reimbursement grant, see conditional grant cost–revenue ratio 258; see also administrative costs Côte d’Ivoire 237, 294 Craig, Jon 156 credit market 150, 153, 154, 383; see also borrowing; debt Croatia 251, 256 Crook, Richard 63 crowding out 200, 219, 235, 348, 385; see also fiscal competition Cuba 12 current use value 260 Cyan, Musharraf 52, 200 Danegeld 34 Das, Jishnu 78 data, need for better 28, 31, 34, 47, 54, 66, 118, 305, 315, 324, 415; see also IMF; ICTD; OECD; World Bank DAU (revenue-sharing grant) 311, 336

01/03/2018 11:09

492

Fiscal decentralization and local finance in developing countries

Davoodi, Hamid 29 De, Indrandil 59 De Cesare, Claudia 231, 243, 253 De Mello, Luis 59, 202 De Ree, Joppe 78 De Silva, Migara 385 debt 18, 68, 134–5, 153, 395; see also borrowing decentralization 30, 119, 140, 191, 220 dangers of 395–7 and infrastructure 121–2, 162 key role of central government 14–15, 25 measurement of 30–44 reasons for 7–22 reform, comprehensive 140 sequence 416–17 see also gradualism see also asymmetry; ‘big bang’ reform; champions; deconcentration; delegation; fiscal decentralization decentralization effect 280 decentralization theorem 89, 149, 189–90, 224 deconcentration 4–5, 27, 89, 102, 142, 191, 290, 395, 494 deficit 178, 180–83 delegation 5–7; see also contracting; principal–agent relationship Delhi 237, 250, 367, 373, 380 democracy 12–15, 86, 169, 171, 350; see also accountability; elections Denmark 93, 132, 238, 354 density, economies of 117 derivation basis 149; see also horizontal sharing; tax sharing; transfers Detroit 186 development charges 268 Devkota, K.L. 55 devolution 4, 8, 60, 286, 394 Dillinger, William 209 direct expenditure controls 90–91; see also mandates disaster relief 308, 403 discount rate 178 distributable pool, see transfers distributional goals 82, 103–8, 145, 217, 278; see also inequality; poverty alleviation; redistribution

M4439-BAHL_9781786435293_t.indd 492

Dollery, Brian 83 donors, see foreign aid dual accountability (dual subordination), see accountability Dunn, Jonathan 102 Duranton, Gilles 341 earmarking 94, 134, 173, 306, 307, 313, 338, 354, 370, 378, 391 for infrastructure 112, 116, 148, 265 see also conditional grant; intercept; special district East Africa 195 East Timor 20 economies of scale 26, 82–5, 94–5, 128, 129–32, 345, 350, 351, 361 in tax administration 212, 214–15 Ecuador 59, 251, 257 education 77–9, 84, 86–7, 88, 89, 90–91, 101, 136, 225, 307, 389 Egypt 13, 16, 24–5, 87, 100, 254, 283 El Salvador 257 elections 12, 33, 95–6, 191, 399, 421; see also accountability; democracy; malapportionment electricity tax 367 elite capture 6, 13, 59, 60, 80, 128, 149, 161 Eller, Markus 55 empowerment 4–7, 14, 39, 58, 67, 73, 107, 234, 394, 399 enforcement 256–7, 377 entitlement 180, 292, 300, 379 equalization 63, 66, 204, 285–7, 322–5, 336, 381, 388, 407; see also inequality; regional disparities; transfers Estache, Antonio 29 Estonia 238 Ethiopia 18, 250, 303, 306, 325, 380 European Union 117, 121, 149, 297 ex ante and ex post control 110–11, 155–6 exactions 267–8 excises 193, 203, 206–7, 298 exclusion principle 176 exemptions, see property tax; threshold expenditure assignment 74–7, 97–103, 115, 118, 125–6, 139–41, 304

01/03/2018 11:09



Index ­493

expenditure needs 284, 302, 303–4, 313, 323, 324, 328 expert capture 128, 226, 304, 420; see also foreign aid externalities 25, 80–82, 90, 126, 132, 287–8, 352, 358, 374; see also conditional grants; fiscal externalities; spillovers; transfers extra-budgetary levies 184, 203, 216 Faguet, Jean-Paul 18, 59, 126 Fan, C. 60 federations 15, 32, 52, 61, 75, 102, 116, 118, 155, 221, 224, 347, 421 finance commission 327, 328–31, 409 finance follows function 51, 52, 71, 197, 353, 416 Finland 83–4, 132 fiscal appeasement 20 fiscal capacity 150, 283, 301, 304, 320 fiscal competition 6, 57–8, 69, 104, 126, 188, 201–3, 213, 351, 358, 386; see also crowding out; yardstick competition fiscal contracting 344; see also contracting fiscal decentralization 51, 413 costs of 22–30 definition 4–7, 37 determinants of 10–11, 36, 45–52, 65, 68 effects of 9, 53–64 level of 45–52, 87–8, 394 as a system 397, 398–401 theory of 9, 11, 45, 75, 80–87, 139 see also decentralization theorem; second generation see also champions; decentralization; measurement fiscal dentistry, see fiscal gap fiscal discipline 109, 112–13, 184, 397; see also budget constraint fiscal disparities, see regional disparities fiscal externalities 190 fiscal flows 346–8 fiscal gap 184, 320, 344, 403; see also vertical balance fiscal health 180

M4439-BAHL_9781786435293_t.indd 493

fiscal laziness, see tax effort fiscal mischief 178, 181–4; see also soft budget constraint fiscal responsibility law 24, 152 Fiszbein, Ariel 138 floating debt 18 floor area ratio (FAR) 266, 267, 268 foreclosure 256 foreign aid (assistance) 227, 315, 327, 332–3, 401, 409, 411, 413, 414–15; see also expert capture formula grants, see transfers Fox, William 84 fractionalization 40, 49, 68, 350 fragmentation 353–7 France 209, 218, 254, 354, 357 Frank, Jonas 136, 176–7 Franzen, Riel 251, 252, 372 free riding 86 Freire, Maria 361, 364 Friedman, Jonathan 158 fuel subsidy 375 fuel tax 206, 374–5 fungibility of money 287–8, 307 Gabon 294 Gaebler, Ted 25 gambling, taxation of 367–8 gaming, intergovernmental 292, 294, 301 Germany 20, 38, 41, 52, 186, 192, 265, 269, 291 GFS data 30, 37, 38–9, 67, 281; see also IMF Ghana 253, 254, 294, 345 GIS (geographic information systems) 247, 249, 392 Glaeser, Edward 104, 121, 341, 345, 384 Goel, Rajeev 58 Gomez-Reino, Juan 49 governance 97–9; see also informality; metropolitan areas gradualism 139, 401; see also ‘big bang’ reform graduated personal tax 195, 205 grants, see transfers grants commission, see finance commission Gravelle, Jennifer 229

01/03/2018 11:09

494

Fiscal decentralization and local finance in developing countries

gross receipts tax 207–8, 366; see also turnover tax Grote, U. 63 GST, see VAT Guatemala 250 Guayaquil 251 Guerra, Susana 285, 324 Gurley, Tami 84 Guyana 12 hard budget constraint 24, 95, 102, 167, 177–80, 374, 385, 388, 395, 400, 403–4; see also budget constraint hard-to-tax sector 57–8, 121, 260; see also informality headquarters problem 207, 298, 300, 366 health care 119, 313 Henderson, Vernon 50 Hicks, Ursula 167 high-powered money effect 319; see also tax effort Hirschman, Albert 139 Hofman, Bert 285, 324 ‘hold-harmless’ provision 305, 325 homeowner relief 244, 251 Hong Kong 20, 237, 252 horizontal sharing 290, 296–308, 337, 382, 388, 392; see also equalization; transfers hotel tax 367 Human Development Index (HDI) 63, 303 Hungary 134–5, 291, 292, 320, 356, 391 Huther, Jeff 50 ICTD (International Centre for Tax and Development) data 39 IMF (International Monetary Fund) 30, 37, 38–9, 66, 110, 230, 327, 395, 411; see also GFS data incidence, fiscal 347, 367, 368, 371, 378–9; see also inequality; progressivity income tax 193, 217, 218, 301, 368, 371; see also corporate income tax; personal income tax increment, land value, see value capture incrementalism, see gradualism

M4439-BAHL_9781786435293_t.indd 494

indexing 254, 260, 275 India 10, 14, 28, 59, 64, 78, 87, 91, 96, 101, 106, 109, 124, 127, 159, 184, 190, 193, 254, 268, 377, 406–7, 416 Finance Commission 305, 314, 321, 329, 332–3, 409 informal governance 98–9 infrastructure 151, 364 local taxes 147–8, 205, 218, 367, 369–70 metropolitan areas 348, 351, 362, 386 octroi 147–8, 206, 217, 289, 369 panchayats 98, 105, 132, 259, 329 property tax 231, 235, 237, 242, 244, 245, 253 rural sector 105, 107–8, 137, 136, 198, 259–60, 307, 344, 345 state finance commissions 314, 329–31 taxing power 44, 184, 192 transfers 134, 295, 306, 324 water supply 84, 101 see also Delhi; JNNURM; Kolkata; Mumbai Indonesia 10, 16, 19, 20, 45, 60–61, 78, 79, 100, 126, 184, 285, 303, 314, 333, 348, 396 deconcentration 5, 27, 89, 102, 142, 290, 395, 404 property tax 234, 235, 259, 373 tax sharing 291, 293, 310–11 transfers 23, 160, 305, 306, 319, 324 see also Jakarta inequality 19, 49, 53, 62–4, 203–4, 285, 341; see also distributional goals; equalization; poverty alleviation; redistribution; regional disparities inertia, institutional 145–6 informality 57, 58, 98–9, 205, 341, 371 settlements 81, 242, 243 taxation 199–200 see also CDD; hard-to-tax; slums infrastructure 25, 124–5, 343, 345, 350 and decentralization 121–2, 128, 139–41 financing 105, 142–60, 364 see also capital grants; maintenance expenditure Ingram, Gregory 133, 158, 364

01/03/2018 11:09



Index ­495

Inman, Robert 187 innovation 119, 212, 344, 360, 394 Inter-American Development Bank 327 intercept (of transfers) 150, 153 interest charges 256, 277 IRAP (regional business tax) 217–18 Iregui, Ana 61 Israel 269 Istanbul 345, 360, 367, 380 Italy 209, 219, 257 Jacobs, Jane 345 Jakarta 10, 295, 367, 381 Jamaica 222, 238, 250, 251, 292, 391 Japan 186, 209, 218, 269, 291, 304, 356, 381 Jibao, Samuel 27 JNNURM (Jawaharlal Nehru National Renewal Mission) 106, 159, 381 Johannesburg 380 Jordan 237, 239, 244, 262 Joshi-Ghani, Abha 121, 384 Juul, K. 59 Kelly, Roy 248, 253 Kenya 79, 231, 245, 254, 257, 289, 292, 365 land value tax 238–9, 253 local business tax 208–9, 218 Kharas, Homi 414 King, David 190 KIS (keep it simple) principle 171 Klein, Michael 157 Kolkata 237, 355, 380 Korea 265, 269, 285, 304, 320, 347, 381 Kravchuk, Robert 12 Kurlyandskaya, Galina 320, 324 Kyrgyzstan 213 Kyriacou, Andreas 60 Lago-Peñas, Santiago 9, 56, 65 Lagos 215, 365 land adjustment 268–70 land leases 266, 268 land sales 148–9, 223, 266, 268, 342 land tax, see property tax land titles 230; see also cadaster land use, tax effects on 236, 237, 241, 272, 372, 379

M4439-BAHL_9781786435293_t.indd 495

land value tax 236, 238–9; see also property tax; value capture Lao PDR 10 Latin America 231, 233, 253, 314, 376 Latvia 291, 356 Lausanne 356 learning curve 213, 219, 220, 314, 322, 344 Letelier, Leonardo 49, 51 Levtchenkova, Sophia 328 Lewis, Arthur 11 Leviathan argument 57 Li, Bingqin 126 Liberia 253, 291 Libya 20 licenses 47, 148, 199, 208, 215, 218, 376; see also user charges lifeline tariff 145, 171 Lin, Justin 55 Lindblom, Charles 139 linkage, rural areas 175–6, 261; see also benefit principle; Wicksellian connection Linn, Johannes 175, 290, 353, 414 Lithuania 250 Litvack, Jenny 21 Liu, Zhiqiang 55 loans, subsidized 281; see also borrowing local business taxes 44, 200, 203, 208–9, 218, 335, 367; see also business property tax; VAT local economic development (LED) approach 69 local governments 32, 84, 186 budgets 182–3 capacity 11, 24–5, 80, 97, 104, 129, 137–8, 142, 213, 413 informational advantage 104–5, 126, 127, 130–31, 132, 199, 212 see also administration; budgeting; local taxes local investment corporations 154, 159 local knowledge, see local governments local taxes 184, 190–91, 204–9; see also property tax Lockwood, Ben 60 lotteries 367 Lotz, Jorgen 200 Loyalka, Prashant 78

01/03/2018 11:09

496

Fiscal decentralization and local finance in developing countries

Lucknow 237 Lustig, Nora 347 Macedonia 256 Madrid 357, 390 maintenance expenditure 125, 130, 133–7, 141, 142, 173, 315 malapportionment 349, 390; see also elections Malawi 245 Malaysia 44, 237, 379 Manasan, Rosario 319, 325 mandates, expenditure 6, 87, 91–4, 134, 139, 226, 316 unfunded 6, 93, 115, 133, 312, 385, 401 Manila 355, 358, 370, 372, 380, 381 Manor, James 73, 97, 105, 107–8 Maputo 250 Marcesse, Thibaud 108 marginal cost pricing 169, 378 market-preserving federalism 6; see also fiscal competition Martinez-Vazquez, Jorge 49, 52, 55, 63, 100, 136, 147, 176–7, 191, 231, 299, 304, 319 Mascagni, Giulia 9, 65 matching grant, see conditional grants Mathur, Om 244, 253, 280, 289, 319, 325 Mauritius 277 Mauro, Paolo 29 Mbiti, Isaac 78–9 McCluskey, William 235, 251, 252, 255, 372 McLure, Charles 297, 310 McNab, Robert 55 measurement 37, 97 economies of scale 83–5 expenditure 39–41 externalities 80–81, 94 level of decentralization 45–52 revenue 42–4, 257, 258–9 Medellín 267 media, role of 34, 349 median voter 189 ‘mega-cities’ 344 mergers, municipal 83–4, 387 Merk, Olav 158

M4439-BAHL_9781786435293_t.indd 496

metropolitan areas 28, 81, 85, 102, 176, 188–9, 342–9, 395 accountability 353, 363, 386 competition in 58, 104 government 82, 130, 134, 349–51, 358–9, 366, 402, 407, 415 one-tier 358, 387 reform 383–9 self-financing 146, 195, 197 taxes 204–9, 366–78 transfers 379–83 two-tier 351, 380 see also regional governments; special districts Mexico 12, 20, 38, 88, 91, 116, 162, 201, 224, 250, 267, 307, 346, 355, 357, 359, 373, 380 payroll tax 202, 205, 371 tax sharing 291, 293, 298, 305 transfers 23, 89, 92, 134, 150, 162, 289, 295, 306, 312, 321 middle class 231, 233 migration, rural 103, 249, 314, 323, 337, 341, 342–3, 345, 361, 389 Mohmand, Shandana 98 monitoring 110, 150, 245, 249, 258–60, 263, 286, 316, 322, 325–33, 408–10, 417, 419 Montenegro 256 Montevideo 250 Montreal 390 Mookherjee, Dilip 29 moral hazard 125, 153, 155 motor vehicle taxes, see vehicle taxes Mozambique 250 MTEF (medium-term expenditure framework) 111–12 Mumbai 109, 136, 148, 206, 234, 237, 243, 245, 357, 358, 365, 369, 370 municipal development fund 151, 154 Musgrave, Richard 57, 170, 172, 175, 187 Myanmar 20 Nag, Tirthankar 59 Nagpur 237 Nairobi 254, 365 Namibia 238, 245, 262 nation-building 17–22, 314 national priorities 91, 126, 173

01/03/2018 11:09



Index ­497

national standards 82, 104, 312 natural resources 19, 20, 334, 337 revenues 67, 149, 194, 292, 299, 305, 308–12 Nelson, M. 66 Nepal 10, 53, 197, 319 Netherlands 294, 354, 390 Netzer, Dick 173, 227 New York 186 New Zealand 238, 285 Neyapti, Bilin 55 Nigeria 19, 20, 215, 309, 314, 319, 325, 365 Nilekani, Nandan 96 non-residents, tax on 175 Nordic countries 192, 193, 200, 202, 371 norms, see expenditure needs; national standards Norway 67, 149, 337, 356 Oates, Wallace 3, 9, 45, 57, 72, 188, 189 octroi, see India OECD (Organisation for Economic Co-operation and Development) data 38–9, 41, 66, 67 OECD countries 40, 42, 57, 60, 121, 188, 191, 194, 195, 230, 253, 258, 285, 286, 345, 361 Olken, Benjamin 60, 61 option demand 175 Osborne, David 25 Oslo 356 Pakistan 43, 77, 205, 206, 237, 289, 293, 303, 319, 369, 377 expenditure assignment 102–3 finance commission 330–31, 409 property tax 244, 254 transfers 23, 289 Panama 161, 244, 246, 250, 253 panchayat, see India Panizza, Ugo 49 Papua New Guinea 10, 310 parastatals 85, 132, 357; see also stateowned enterprises; utilities Paris 354, 357 parking tax 377 participatory budgeting 118

M4439-BAHL_9781786435293_t.indd 497

patente 208; see also local business taxes paternalism 91, 92, 197, 315, 394, 414 Patna 218, 237 Paul, Samuel 96 payroll tax 175, 202, 203, 204, 205, 298, 368, 371 penalties 256; see also interest charges Pennsylvania 236 performance grants 321–2 personal income tax 202, 293, 298 Peru 14, 101, 122, 149, 214, 250, 292, 310 tax sharing 291, 311 Petchey, Jeffrey 328 Pethe, Abhay 160 Philippines 12, 16, 19, 20, 32, 64, 88, 91, 126, 129, 186, 205, 208, 225, 303, 319, 348, 351, 370 property tax 43, 248, 256 tax sharing 291, 293–4 transfers 289, 314, 324, 325 see also Manila piggybacking 192, 215, 218, 226, 299 pilot projects 114, 127, 161, 384–5 Pirenne, Henri 345 Poland 231, 291 policing, see public security political decentralization, see devolution Pommerehne, Werner 49 Porto Alegre 252 Portugal 41 Pöschl, Caroline 18, 59, 126 poverty alleviation 62–4, 115, 362 presumptive taxes 215, 217, 227, 252, 260, 275, 298; see also property tax Prichard, Wilson 27 principal–agent relationship 5, 171, 385 private schools 78, 225 privatization 157, 257, 294; see also contracting; public–private partnerships procurement 141 professional tax 205 progressivity 229, 236, 240, 262, 301, 372; see also incidence; inequality; redistribution Prohl, Silke 57

01/03/2018 11:09

498

Fiscal decentralization and local finance in developing countries

project cycle 136, 141 property tax 43, 147, 194, 244, 262, 275, 371–4 administration 147, 215, 234, 247–57, 373 administrative cost 238 area-based 45, 215, 218, 239–40 assessment 229, 237–9, 252, 253, 271 as benefit tax 228, 229 capping 275, 276 collection 255–7 effective rate 245–6 effort 52, 25 as excise tax 240 exemptions 229, 242, 244–5, 271 government as income tax 241 as presumptive tax 227, 238 progressivity 229, 236, 240, 262, 273 property 242, 245 rates 245–7, 273 reform 228, 234, 247, 248, 255, 270–75 revaluation 238, 254–5 revenue potential 229, 230–35 rural areas 241, 257, 259–61 self-assessment 248–9, 275 tax base 235–45 tax roll 251, 254, 269 technology 247–8 threshold 242, 275 unpopularity of 227–8, 367 as user charge 173, 174 as wealth tax 218, 227, 241, 270 see also business property tax; cadaster; homeowner relief; valuation property transfer tax 215, 228, 255, 261–4, 373 Prud’homme, Remy 15, 29, 254, 395 public enterprises, see parastatals; state-owned enterprises public financial management (PFM) 108–14, 362–3 public goods 168 public–private partnerships (PPPs) 135, 139, 157–60, 161, 163 public security 77, 83, 116, 341, 391 Puga, Diego 341 Punjab 254

M4439-BAHL_9781786435293_t.indd 498

Qian, Yinyi 13, 300 ‘race to the bottom’ 58, 202 rainy day fund 185, 187, 194 Rajaraman, Indira 260 random control trials 61, 69 Rao, Govinda 184, 268, 279, 320, 324 Razafimahefa, Ivohasina 61 redistribution 170–71, 220; see also distributional goals; equalization; inequality; poverty alleviation regional disparities 149, 203, 282, 285, 299, 301, 307, 310, 322–5, 352, 354, 411 regional governments 222, 228; see also metropolitan areas regional redistribution, see equalization Regional Services Council levy 195, 295, 216, 370 registration tax 376 regressivity, see progressivity regulation 214, 367, 376, 387 rent control 237 rental value, see property tax representative tax system (RTS) 304 reserve funds 223 results-based grants, see performance grants retaliation 201, 225 revaluation 253, 254–5 revenue effort, see tax effort revenue mobilization 16–17, 197–200, 219, 321, 323 revenue mobilization effect 280 revenue sharing 192, 288, 301, 308–12; see also tax sharing Rhoads, William 267 Ricardo, David 227 Riga 356 Rio de Janeiro 267, 276 risk bearing 157–8, 310 road fund 148 roads 131–2, 146, 374, 376 Roca-Sagalés, Oriol 60 Rodriguez, Edgard 62, 64 Rodrik, Dani 36 Romania 356, 371 Rome 219 rural local governments 28–9, 85,

01/03/2018 11:09



Index ­499

197–8, 130, 137, 141, 142, 195, 198, 314–15, 344, 407 taxation 241, 257, 259–61 Russia 8, 10, 18, 20, 44, 102, 105, 192, 217, 285, 290, 291, 294, 304, 320, 325, 396 Rwanda 205, 208 sales-assessment ratio 253, 271 sales tax 175, 369–71; see also gross receipts tax; turnover tax salience 198, 275; see also tax rate; visibility Salmon, Pierre 56 Sánchez de Lozada, Gonzalo 411, 413 Sánchez Torres, Fabio 235, 254, 320 Santiago (Chile) 382 São Paulo 242, 268, 269, 299, 345, 355, 361, 380, 382 Saunoris, James 58 scale economies, see economies of scale Schick, Alan 111 Schneider, Friedrich 57 secession 8, 20–21, 48, 290, 309, 345–7 second-generation theory 189–90, 224 Self, Peter 355 self-assessment (self-declaration) 248, 249, 275 Senegal 59, 231, 237, 262, 325 separatism, see secession Sepulveda, Cristian 49, 63 Serbia 244 services, taxes on 206, 367–8, 369 shadow economy 57–8; see also informality Shah, Anwar 50, 55, 59, 86, 328, 380 Shanghai 380 Shoup, Carl 84 Siemiatycki, Matti 158 Sierra Leone 27, 66, 219, 224, 253, 278, 392 Singapore 223, 376 Singh, Nirvakar 279 Slack, Enid 236, 274 Slovenia 244 slums 151, 260, 350, 361, 364–5, 384; see also informality small business 200; see also licenses; local business tax Smoke, Paul 99, 341

M4439-BAHL_9781786435293_t.indd 499

Smolka, Martim 243 social capital 61, 190, 257; see also trust social investment fund 154 Sofia 356 soft budget constraint 24, 153, 155, 177, 182–4; see also budget constraint; fiscal mischief Sokoloff, Kenneth 51, 78 Somalia 20, 238 Sorens, Jason 20 South Africa 8, 12, 17, 27, 154, 204, 206, 209, 218, 290, 303, 373, 396, 406, 407–8, 417 equitable share 295, 325 finance commission 331, 332 metropolitan areas 345, 348, 359, 380, 386 property tax 253, 254, 257 Regional Services Council levy 195, 205, 216, 370 see also Cape Town Soviet bloc, see transition countries Sow, Moussé 61 Spain 52, 193, 222, 347, 357, 390 special district 82, 129, 132, 139, 142, 160, 180, 354, 355, 385, 387, 402 spillovers 81–2, 137, 176, 188, 222, 285, 352; see also externalities Spratt, Stephen 145 Sri Lanka 17 stabilization policy 23–4, 193–4, 220, 221, 310, 411 stamp duty 251, 261 state finance commissions, see India state-owned enterprises 132, 182, 350, 355, 388; see also parastatals; utilities Stein, Ernesto 57 Stockholm 354, 356 Stren, Richard 87 sub-municipalities 129 subnational taxes 204–9; see also local taxes subsidiarity principle 117 subsidy policies 106, 109, 379; see also lifeline tariff; user charges Sudan 20, 917 ‘sunset’ rule 93, 245, 273, 332, 419 ‘sunshine’ rule 93; see also transparency

01/03/2018 11:09

500

Fiscal decentralization and local finance in developing countries

surveys 129, 213 Sverrisson, Alan 63 Sweden 218, 354, 356 Switzerland 60, 193, 200, 202, 225, 356, 378 Syria 20 Tanzania 95, 205, 208, 242, 289, 295 Tanzi, Vito 29, 395 tax administration 27, 211–19, 335; see also tax technology tax assignment 172, 175, 187–95 tax base sharing, see crowding out; fiscal competition; fiscal externalities; piggybacking tax cascading 207, 366 tax clearance certificate 256 tax competition, see fiscal competition; fiscal externalities tax decentralization 51–2, 196–8, 209, 211–19, 286 tax effort 203, 235, 282, 284, 285, 287, 319, 320, 323 tax exporting 169, 173, 175, 190, 201, 366 tax incentives 292, 300, 384 tax price 97, 284, 299, 306, 388 tax rates 191–2, 245–7, 273, 277 limits on 201, 211, 259 see also salience; visibility tax roll, see property tax tax room 213 tax segmentation 226 tax sharing 149, 218, 289, 291–4, 295, 296–8, 320, 395; see also revenue sharing tax technology 211, 214–15, 227, 247, 253, 263, 376 taxing power 16, 21, 44, 52, 97, 144, 180, 184, 187, 192 limits on 6, 201, 319, 397 Taylor, Zachary 97 technical assistance, see expert capture; foreign aid Ter-Minassian, Teresa 156 Thailand 87, 231, 295, 360, 380, 387 third-party information 252, 263 Thirsk, Wayne 304 threshold, tax 200 Tiebout, Charles 58, 168

M4439-BAHL_9781786435293_t.indd 500

Timofeev, Andrey 100 tolls, road 373 Toronto 352, 359, 390 Tosun, Mehmet 55 trade taxes 202, 217 transfers, fiscal 23, 92, 112, 134, 150–52, 162, 176, 223, 279, 281, 312, 314, 349, 379–83, 403 dependence 195–6 derivation approach 296, 298–301 distributable pool 301, 322, 329 distribution formula 289, 290, 295 effects 318–25 equalization 144, 285–7, 322–5 externalities 283, 287–8 formula grants 23, 150, 301–6 importance 280–82 rationales 282–90, 302 and revenue 319–22 see also equalization; expenditure need; fiscal capacity transit, public 109, 124, 127, 158, 159, 352, 356 transitional countries 12, 19, 20, 185, 192, 201, 239–40, 289, 299, 371 transparency 93, 96, 112–13, 118, 191, 255, 296, 301–2, 335; see also accountability Treisman, Daniel 9, 32, 60, 64 Trinidad 237 trust 92, 111, 134, 160, 190, 246, 363; see also social capital Turkey 20, 53, 148, 345, 360, 367, 380 turnover tax 195, 202, 298; see also gross receipts tax; sales tax two-tier government, see metropolitan areas Uganda 17, 78, 95, 96 decentralization in 27, 92, 396, 409 graduated personal tax 194, 205 Ukraine 93, 304 UN Millennium goals 364 unbundling 101, 132, 135–6; see also expenditure assignment; PPPs UNCDF 321 unconditional grants 150, 337; see also transfers underutilized land, see land use

01/03/2018 11:09



Index ­501

unearned increment 236, 265, 266; see also value capture unfunded liabilities 182, 184 unfunded mandates, see mandates United States 42, 52, 79, 186, 192, 193, 201, 202, 222, 236, 238, 260, 354 urban fringe 260 Uruguay 250, 260 USAID 411 user charges (fees) 29, 46–7, 67, 81, 106, 108, 141, 143, 144–6, 169, 334, 356, 378–9, 402; see also licenses; utilities utilities 47, 67, 106, 160, 175–6; see also parastatals; state-owned enterprises Uttar Pradesh 331 vacant land tax, see land use Vaillancourt, François 294 valorization 265 valuation, of property 252–5 valuation-driven reform 271–2 value capture 147, 174, 228, 264–9, 342, 372 Vancouver 357 VAT (value added tax) 193, 195, 200, 213, 216, 218, 254, 263–4, 291, 292, 296, 297, 300–301, 335, 369–70 as local business tax 209, 210–11, 367, 291 vehicle taxes 148, 206–7, 374–8; see also fuel tax Venables, Anthony 364 Venezuela 256 vertical balance (imbalance) 51, 184, 187, 282, 283–4; see also fiscal gap vertical programs 51, 141, 313–16 vertical share 280, 288, 290, 291–6, 329 Vickrey, William 173, 377

M4439-BAHL_9781786435293_t.indd 501

Vietnam 20, 91, 161, 197, 285, 319, 324 Vinuela, Lorena 121 visibility of taxes 191, 227, 234, 402; see also salience; tax rate Von Braun, J. 63 Wallace, Sally 49, 229 water 82, 86, 94–5, 101, 129–30, 145, 146, 315 subsidies 109, 379 wealth 230, 262, 270; see also property tax Weingast, Barry 13, 45, 300 West Bengal 259, 314 Wetzel, Deborah 102 Wicksellian connection 172–7; see also benefit model; linkage Widodo, Joko 349 willingness to pay 197, 199 Winer, Stanley 152 World Bank 7, 17, 18, 39, 59, 66, 67, 97, 99, 115, 158, 161, 327, 384, 395, 411 World Values Survey 61 Yao, Guevara 63 Yao, Ming-Hung 57 yardstick competition 45, 58, 126; see also benchmarking Yatta, François 294 Yeltsin, Boris 349 Yilmaz, Serdar 55 Zagreb 356, 371 Zambia 78 Zhang, Henglan 126 Zhang, Tao 55 Zhang, Yongmei 126 Zhang, Zhihua 299 Zolt, Eric 51, 78 zoning 264, 267 Zou, Heng-fu 55

01/03/2018 11:09

M4439-BAHL_9781786435293_t.indd 502

01/03/2018 11:09