Fiscal Decentralization and Local Finance in

Developing Countries
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
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

© 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 530 9

Contents
Preface

PART I WHY DECENTRALIZATION MATTERS

1 Fiscal decentralization 101

2 Has decentralization worked?

PART II DECENTRALIZING EXPENDITURE

3 Expenditure assignment and management

4 Decentralizing and financing infrastructure

PART III FINANCING LOCAL GOVERNMENT: THE KEY TO THE

PUZZLE

5 Financing local and regional government

6 Taxing land and property
7 Intergovernmental transfers

PART IV SUMMING UP

8 Financing metropolitan areas

9 Giving decentralization a chance
References
Index
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 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
 To our grandchildren
Roy – Margot, Thomas, Hadley and Carleigh
Richard – Austin, Spenser, Jack, James, and Rose.
PART I

Why Decentralization Matters

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 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
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 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 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 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 (Martinez-Vazquez 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 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 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 governments in its vision for Gross
National Happiness in its ninth five-year 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 governments 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 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 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 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
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
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
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 low-income 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
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 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 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 concludes 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 nontransparent fiscal systems
– probably the most they can or should 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 well-known 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.
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 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 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
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
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 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 governments, 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 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 –
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.

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), Martinez-Vazquez 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 varying 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 in the general context of federal
arrangements, see Bednar (2009), and for an interesting early treatment, see Cremer et al. (1995).
9. 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.
10. 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).
11. 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).
12. 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).
13. We take up the discussion of the relationship between decentralization and growth in Chapter 2.
14. 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).
15. 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.
16. 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.
17. 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.
18. India has 3,609 urban local bodies and, in rural areas, 474 zilla parishads, 5,906 panchayats
samithis and 227,698 gram panchayats.
19. 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).
20. 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).
21. http://www.citymayors.com/government/mayors-americas.html (consulted February 2, 2016).
22. 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.
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).
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.
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
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
measures are really needed for different purposes. There is no one-size-fits-
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.

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.
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.

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 α = the percentage share of subnational government expenditures over

which the subnational governments have discretion;
LEj = the direct expenditure of the subnational governments in country j
(including the expenditure of revenue from intergovernmental transfers);
and
CE j = the direct expenditure of the central government in country j
(excluding intergovernmental transfers to subnational governments).
 Then an index of expenditure decentralization (DE) for country j would be

where 0 ≤ α ≤ 1.3
Many empirical analyses (based on GFS data) implicitly assume that α = 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 α 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 α = 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:

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.

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

where
CT = total revenues raised from all central government taxes in country j;
Ti = revenues raised from subnational government tax sources in country j;
and
βj = 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 (βj) refers to the degree to which
the subnational government can control the level of revenue raised. If βj = 1,
then the subnational government has complete control, as is true for many
subnational government taxes in some OECD countries. As noted above,
where βj = 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 β 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, in which various
possibilities are arrayed broadly in descending order of tax autonomy.

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).

 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 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 ‘back-door’ 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).

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 ‘second-generation’ 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 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.

Table 2.1 Fiscal decentralization indicators (average values for data

available for 2000s)
Source: IMF, GFS data and author calculations.

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 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:

● 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
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 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.

Table 2.2 Determinants of fiscal decentralization

Dependent Subnational government expenditures as % Subnational government taxes as % of
variable of GDP GDP
Constant −2.71 −3.60
(4.83) (5.14)
Per capita GDP 0.39*** 0.39***
(6.67) (5.26)
Population size 0.27*** 0.17
(3.76) (1.91)
Federalism 0.74** 1.21**
dummy
(2.78) (3.40)
Transition 1.03*** 1.52***
country
dummy (4.70) (5.55)
R-squared 0.57 0.50
Sample size 71 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.

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 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, β = 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 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-to-GDP 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 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.
Recent research has reduced some of these problems. For example, more
sophisticated estimation techniques allow fiscal decentralization to 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 because the impact is often indirect and difficult
to sort out.27 To take a topic of current interest: suppose, for 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:
where G = economic growth and C = corruption.
Theory suggests that fiscal decentralization may increase income equality;
but whether it does so or not may depend on the interaction between the
(presumed, for illustrative purposes) negative impact of decentralization on
corruption, the negative impact of 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 problems arise with most impact
analysis.
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 been heard. Here we simply review briefly what
has been learned so far about four important potential impacts of fiscal
decentralization: on economic growth, the size of government, the quality of
governance, and equality.

Economic Growth

Low-income countries want to move up the economic growth ladder and

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
may, for instance, be effects of decentralization on corruption, income
inequality, technical 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

where EG = economic growth, FD = fiscal decentralization and C =

corruption.
If fiscal decentralization leads to lower levels of corruption, and less
corruption is growth-enhancing, then the marginal effect of fiscal
decentralization on economic growth would be positive. This approach seems
a 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 growth be
positive or negative? The answer depends on the relative size of 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 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 Lago-Peñas 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 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 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
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 country-specific 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 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
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
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
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 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 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 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.

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 accounts in
the US (Ebel and Wang, 2017), but they usually are in Canada (Slack and Tassonyi, 2017).
14. 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.
15. 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.
16. 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.
17. 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.
18. 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).
19. A good review of a wide range of empirical studies focusing on the three basic hypotheses may be
found in Letelier (2005).
20. This hypothesis has been much explored in the historical literature: for an interesting application to
Latin American development, see Centeno (2002).
21. For an alternative explanation, see note 15 above.
22. 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).
23. 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 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.
24. Vertical tax base competition is a form of vertical fiscal externality, as discussed in e.g. Dahlby
(1996).
25. 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).
26. For surveys of the literature on this subject, see Martinez-Vazquez (2011), Lago-Peñas et al.
(2016), Mascagni (2016) and Treisman (2007).
27. The general argument is developed in Martinez-Vazquez and McNab (2001) and demonstrated
theoretically in Brueckner (2006).
28. 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.
29. 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).
30. 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.
31. 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.
32. 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.
33. To mention only one recent study providing some supporting evidence, see Tassonyi et al. (2015).
34. Alternatively, they may perhaps, as in China, appeal over the head of local officials to those in the
official (or party) hierarchy for relief.
35. 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.
36. 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).
37. See http://info.worldbank.org/governance/wgi/index.aspx#home for a full description of these
indicators.
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/oecdfiscaldecentralisationdatabase.htm. See also such webpages as
www.uclg.org.
PART II

Decentralizing Expenditure
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 decentralization 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.
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 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.

Table 3.1 Assigning subnational expenditure responsibilities

Notes:
N = national; R = regional; L = local; P = 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.

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.

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 sub-
function. 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 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
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 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.

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 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 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 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 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 cost-minimizing
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 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 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 higher-income 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.

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 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 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 (Martinez-
Vazquez 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).
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 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 inevitably
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 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 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 (Martinez-
Vazquez 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
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 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 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). 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 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.

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 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.

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 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
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 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
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 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 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
transfer of funds to democratic local government, ever, anywhere in the
world” – was launched in 2005 and is discussed briefly in Box 3.4.

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).

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 obscurity is seldom
conducive to good decision-making at any level. Subsidy policy may often be
justified on distributive or externality grounds, but whenever 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.

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 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.

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)

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 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;
● 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 excessive,
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). 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 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 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 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 revenues,
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 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.
8. 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/85-002-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.
9. 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.
10. The main direct responsibility of local municipalities in Colombia is to look after school
maintenance.
11. Two useful embellishments and updates are Oates (1999) and Oates (2008).
12. 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.
13. 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.”
14. There may also be ‘fiscal externalities’ arising from how expenditures are financed (Dahlby, 1996),
but we leave these for discussion in later chapters.
15. 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).
16. The special problems of metropolitan governance and finance are discussed further in Chapter 8.
17. 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.
18. 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.
19. Economies of density may also come into play (Bel, 2013). For example, a study of 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).
20. 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.
21. 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.
22. After implementing the 2001 commission recommendations, for instance, the aggregate state
government deficit rose significantly (see Andrew Young School of Policy Studies, 2005).
23. As Box 3.1 indicates, however, even if the pay arrives on time, there is no guarantee that any
education will ensue.
24. 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.
25. How much spending is increased depends on many factors, as we discuss further in Chapter 7.
26. 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-
institute-of-land-policy-launches-campaign-to-promote-municipal-fiscal-health). The situation is
no better elsewhere in the developed world (Bird and Slack, 2015).
27. 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.
28. 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).
29. 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.
30. 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.
31. Similar issues arise with respect to public–private arrangements concerning infrastructure, as
discussed further in the next chapter.
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.
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 decision-makers 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.
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
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.

Table 4.1 Public investment and decentralization in Latin America

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).

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, 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 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 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).

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).

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
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 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
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
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 set
of funding sources also makes it exceptionally difficult either to invest
sensibly or to maintain investments properly once made (Bird, 2012a).

Table 4.2 The economics of decentralizing infrastructure

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).

Externalities

When the benefits (or costs) of an infrastructure project spill over to

households that reside outside the boundaries of the jurisdiction, lower-tier
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 sub-
functions, 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

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 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 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
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
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.
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).

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.
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 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 higher-level
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.

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.
Subnational governments in developing countries have access to three
sources for financing capital expenditures: local government tax and non-tax
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 own-source
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 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.
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 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.

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 – especially 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 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 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.

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
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. 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 externality-compensating 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; 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 over-borrowing
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.

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).

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 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.

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.

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 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 (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 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.
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).
 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 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).
7. 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.
8. For a more broad-ranging review of the provision and maintenance of rural roads, see Ellis and
Menendez (2016).
9. 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.’
10. 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).
11. 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.
12. For an interesting example of the importance of monitoring O&M see Hu and Ebel (2016) on the
water sector in Albania.
13. 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.
14. 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.
15. 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.
16. 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 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.
17. Procurement changes may have other beneficial effects also, as noted in Lewis-Faupel et al.
(2016).
18. As we discuss later, such earmarking is essential when infrastructure (including debt service) is
financed by user charges.
19. 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).
20. 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.
21. 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).
22. 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).
23. 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.
24. This subject is discussed further in Chapters 5 and 6.
25. An interesting discussion of this issue is in World Bank (2012).
26. 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.
27. 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.
28. Often, however, some in-period borrowing is permitted in order to smooth cash flow over the
budget year.
29. When central transfers are used as collateral, as in Mexico, this has been seen as a problem
(Sutherland et al., 2005).
30. 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).
31. For a study emphasizing the importance of appropriate incentives, fiscal and otherwise, with
respect to establishing and sustaining sound decentralized governance, see Faguet (2011).
32. IMF (2014) estimates that less than 25 percent of infrastructure investment has been added by
private investment.
33. 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).
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.
PART III

Financing Local Government: The Key to the

Puzzle
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 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) 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 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 non-resident
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 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.
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 decision-makers 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
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 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 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 non-residents.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
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.

Table 5.1 Appropriate local revenue source by category of expenditure

Notes:
P = primary source; S = secondary source.
A = Borrowing appropriate for major capital expenditures; (A) = borrowing is appropriate for capital
expenditures but likely to account for a small share of spending.
* = 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 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 decision-makers 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

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):

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:

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 = 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 = TR –
TRC). The amount of local borrowing (B) required to finance the amount of
investment spending in the current period (GK) is thus:

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 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).

Table 5.2 How to soften the budget constraint

Budget line Comments
1. Recurrent revenues Revenues received regularly each year
1a. Tax and non-tax Includes user charges and delinquent tax payments
revenues
1b. Recurrent transfers Intergovernmental transfers regularly available to local governments for
current expenditure purposes, but not including year-end gap-filling grants
1c. Cash balances Accumulated (unrestricted) balance from previous years
2. Current expenditures All expenditures regularly made each year, excluding any capital
expenditures
2a. Operating Wages and salaries, plus normal operating and maintenance expenditures
expenditures
2b. Subsidies and Includes transfer payments to individuals, transfers to other governments
other for current services, transfers to government enterprises
2c. Debt service Repayment of principal and interest due
2d. Contributions Regular contributions to pension funds, health care funds, etc.
3. Total current Row 1 minus Row 2
surplus/deficit
4. Capital expenditures Long-lived capital assets acquired; excludes any current expenditures
5. Capital budget: Revenues available to fund capital expenditures
sources of funds
5a. Capital grants Conditional grants that must be spent for capital purposes
from higher-level
governments
5b. Cash balances Accumulated from current account surpluses
5c. Sales or lease of Buildings, land, businesses, etc.
assets
5d. Borrowing Long-term debt increases
6. Capital budget Row 4 minus Row 5
surplus/deficit
7. How to get around the
 hard budget constraint
7a. Deferred payments 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
7b. Unfunded Delinquent or inadequate contribution to pension fund, health care funds,
contributions etc. or ‘raids’ on balances in pension funds, etc.
7c. Reclassify Shift recurrent expenditures to capital account
expenditures
7d. Bailout by higher- Debt repayments forgiven or reduced
level government
7e. Debt rollovers Debt service paid by issuing new debt
7f. Deficit covered by Bank overdraft to cover current deficit
short-term loans
7g. Backdoor Create extra-budgetary accounts and/or levy unauthorized taxes or charges
approaches
8. Exhibit: contingent Other claims on budget revenues
liabilities
8a. Underfunded
pensions, etc.
8b. Loan guarantees
8c. Underfunded debt
repayment
schedules

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; deferring maintenance of
capital assets; failing to raise taxes or user charges rates to 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.

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 governments – ‘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).

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.

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 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 revenue-productive 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 (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
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
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
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.
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 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 broad-based
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, 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 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
governments 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 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 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
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 self-
employed 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
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.

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
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 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 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

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 non-resident 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.

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 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 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 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 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

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

VAT BVT
Base Consumption Income
Principle Destination Origin
Application Invoice-credit Accounts
Timing Monthly 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 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.
 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 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 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 credit-invoice
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 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 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
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
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 ‘tax-for-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 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
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 decentralizing taxing
power, many countries have instead continued to reform their system of
intergovernmental transfers. As we discuss in Chapter 7, they 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 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 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).
5. For a good discussion of the Tiebout model and the research it has generated, see Fischel (2006).
6. Obvious distributional questions may be raised about this argument, but for the moment we leave
them aside.
7. 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).
8. 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.
9. 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.
10. 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.
11. 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).
12. 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).
13. The pros and cons of earmarking and the many varieties found in practice are outlined in Bird and
Jun (2007).
14. 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).
15. 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.
16. 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).
17. 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.
18. 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 by money rather than
votes, they are probably less likely to impose differentially heavy taxes on business.
19. 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.
20. For useful reviews and discussion of this topic, see Fischer and Easterly (1990), Rodden et al.
(2003), Vigneault (2005), and Liu and Waibel (2010).
21. For a broader discussion of how one may assess the fiscal health of subnational governments, see
Bird and Slack (2015) and references cited there.
22. 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.
23. 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.
24. To simplify, we assume that revenues from the sale or lease of assets are earmarked for capital
spending.
25. 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.
26. 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).
27. 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.
28. 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).
29. 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).
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).
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.
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.
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.
 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; well-structured 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.
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 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 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).

Table 6.1 Property tax as share of GDP (%)

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).

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 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.

Table 6.2 Property taxation in selected countries

 An important barrier to more use of property taxes is that the tax is 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 two-
thirds 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
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 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.

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).

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 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 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 area-based
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:

● 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 area-based
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.

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 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).

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).

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
owner-occupied 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
owner-occupied 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 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 = tax
liability and TAV = 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 = (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 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, 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 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).

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 self-declaration, 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 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. 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.

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 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 sales-
assessment ratios – the usual ‘standard’ for appraising assessment practices –
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.

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 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 difference.
Some experts have therefore argued that property tax reform 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 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 is to benefit
the rural sector, providing rural local governments with some source of own
revenue is an important part of the equation.

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 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).

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). 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 (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 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 higher-
income 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.
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 under-declaration
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 third-party 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 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 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 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, 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

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 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
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.

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 = (TC/TL) (TL/TAV) (TAV/AV) (AV/MV) (MV/Y)

where TC = property tax collections, TL = property tax liability, TAV =

taxable assessed value, AV = total assessed value (including exempt
properties), MV = market value and Y = 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, 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

Table 6.3 Simulation of impact of alternative reforms (%)

Note: See text for details and designations.

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 – starting 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 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
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.

Table 6.4 Problems blocking property tax reform, with possible solutions
Issues and problems Promising approaches Problematic
approaches
Salience: property tax is more Couple tax reform with improvements in local Assessment
visible than other taxes servicess limits
Property tax
capping
Phase-in gradually
Withhold tax at source and other payment
options
Liquidity constraints: imperfect Means testing, Tax deferrals for seniors Assessment
association between taxpayers’ More payment options limits
incomes and property taxes, Phase-in Property tax
especially for seniors capping,
exemptions
based on age
Perceived regressivity: taxes higher Property tax credits Classified tax
as a percent of income for low- Tax deferrals rates
income taxpayers Bundle with other tax reforms Progressive tax
Package with expenditure changes rates
Low-income housing exemptions Assessment
limits
Property tax
capping,
owner-
occupied
exemptions
Volatility: potentially large swings More frequent reassessment Assessment
in taxes for some taxpayers Index base limits
Taxpayer education and communication in Property tax
 understandable form capping,
Phase-in frequent
statutory rate
changes
Presumptive tax: tax base inherently Taxpayer education Self-
arbitrary Accessible appeal process assessment
Phase-in Classified
property tax
rates
Assessment
limits
Property tax
capping,
preferential
treatment
Inelasticity (a problem for local Frequent reassessments Frequent
governments, not for taxpayers): Index base, eliminate tax preferences, keep statutory rate
taxes do not increase with growth property tax roll up-to-date, modernize increases
collection procedures and introduce more Tax amnesties
stringent penalties

Source: Based on Slack and Bird (2015).

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).
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 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.
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 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:

where Tr/SE = intergovernmental transfers as a percent of subnational

government expenditures (the transfer dependency effect); SE/Y =
subnational government expenditures as a percent of GDP (the
decentralization effect); and Tx/Y = 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, 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.

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.

Table 7.1 The importance of intergovernmental transfers

Note: Numbers in parentheses indicate number of countries included in each sample.

Source: Bahl and Wallace (2007).

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.
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 for
subnational governments, doing so is unavoidably a subjective exercise that
may easily turn into a political adventure.3 Even when we try for a more
formal approach to defining fiscal balance, this subjectivity gets in the way,
For instance, the vertical gap (or vertical fiscal imbalance) can be defined as
the difference between (a) the amount subnational governments can raise
from own-revenue sources if they exert a ‘normal’ revenue effort and (b) the
amount they must spend to provide a ‘minimum’ level of the government
services that have been assigned to them. Assuming the two terms are clearly
defined, the gap for any particular level of subnational government is then:

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

where α is the percent of the financing gap that the central government
intends to cover with the transfer system, and CR is the total amount of
revenues raised by the central government from current sources.4
 Few countries make such calculations, of course, or even attempt to
formalize exactly what they mean by such key terms as normal tax effort 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.
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 revenue-
raising 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 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 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 data-rich 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
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
imperatives 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.
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)
Origin of collection (derivation)
Formula
Cost reimbursement (matching)
Ad hoc
Specified share of national or Annual budgetary
state government tax decision
A Not applicable
B E
C F
D G

Note: For discussion of type A–G transfers, see text.

Source: Adapted from Bahl and Linn (1992).

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 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 (Canavire-Bacarreza
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. 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 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
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.
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 time.
As we note below, the issue has been a particularly contentious one when it
comes to sharing natural resource revenues.

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 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.

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 Martinez-Vazquez (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
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 (Martinez-
Vazquez 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 transactions 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 extra-budgetary 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
higher-level 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 affect the
final allocation. Once the 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 γ1 and γ2,
respectively, in the formula, where γ1 + γ2 = 1. This formula will then allocate
to jurisdiction i an amount (Ri) determined as follows:

If one knows the right numbers for X1 and X2 for region i and the weights
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.
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 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.

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).

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 above-
average 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 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
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 freedom
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
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. 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 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 revenue-
sharing 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 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.
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.
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 governments 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.

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 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.

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, 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 decision-makers 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, 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 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.

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 Martinez-Vazquez (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, 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 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
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:

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; αi 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 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 α 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.
 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 αi(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.
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
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 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.

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

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 (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 (third-tier) local governments in
addition to bringing together the provincial and local governments to negotiate
intergovernmental fiscal relations more generally. In 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.

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 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 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.
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.
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).
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).
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).
PART IV

Summing Up
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 lower-income
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 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
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 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
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
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 non-metropolitan
areas.
The inter-regional flow of funds may sometimes play a role in arguments
about potential secession (Vaillancourt and Bird, 2016). Richer regions 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 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 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 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, 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, 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 metro-wide 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.
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 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 financial base often also provides
greater access to capital markets than most general-purpose local
governments have.

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).

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 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 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).

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 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; upgrade the quality of the local
government staff; and develop a viable plan for resource mobilization.
 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 area-wide 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.

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).
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.
● 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.
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.

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 low-income
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 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 middle-
income 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.
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.

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 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),
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).

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 own-source
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 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.

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
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
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 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 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.
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 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 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 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 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 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 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 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.
Such sentiments are not new: World Bank (2009b), Glaeser and Joshi-
Ghani (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 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 provisions for
improved information, transparency and accountability at 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 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 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
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 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.
8. 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).
9. On the policy relevance of different approaches to incidence analysis, see especially Break (1974).
10. For a relatively sophisticated (though still crude) attempt along these lines for Canada, see
Vaillancourt and Bird (2007).
11. See e.g. the studies in Bosch et al. (2010); see also Bird (2006).
12. Such fears are not always well founded, as discussed in Chapter 2.
13. For an interesting review of the effects of legislative malapportionment on tax policy in general,
see Ardanaz and Scartascini (2011).
14. 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).
15. 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.
16. 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).
17. 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.
18. 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).
19. 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).
20. Other countries also have parastatals: for a general discussion, as well as a somewhat out-of-date
country example, see Bird (1984).
21. 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 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).
22. For a more recent discussion and analysis, see Henriquez Diaz et al. (2011).
23. For an interesting discussion of cities that have taken greatest advantage of the urbanization era,
and those that have lagged, see Yusuf (2013).
24. 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.
25. These estimates are derived from an income-driven model based on a sample of 30 low-income
countries.
26. 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).
27. 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).
28. 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).
29. 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).
30. 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.
31. 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.
32. 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).
33. On the Singapore experience, see Chin (2010).
34. 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.
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.
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.
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 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 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 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.

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.

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 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 means and local officials to be
accountable for the hard choices they must make.

Table 9.1 Components of a successful system of fiscal decentralization

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.
● 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 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.

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
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.
Table 9.2 Contrasting approaches to intergovernmental fiscal reform

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.

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 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, 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.
● 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.
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 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 support, and outside
comments are usually quickly forgotten when the money is gone.

Table 9.3 Champions of fiscal decentralization

Comments
Potentially
strong
supporters
The people and Demand for more control over local services and for participation in governance at
their elected the local level.
representatives
President 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.
Parliament or Decentralization in general may be popular, but politicians like be able to claim
 Congress credit for specific local projects, and hence are likely to favor less transparent and
less structured decentralization.
Urban local Bigger cities are most likely to be concerned with achieving more spending
governments autonomy and more access to the tax base.
External donors 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 In part for stabilization purposes, likely to support strict limits to decentralization,
Finance particularly on the taxation and borrowing side.
Ministry of Concerned about controlling public investment (both what and where) and about
Economy focusing on programs with benefits that are wider than the local area.
Line ministries 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 Will support a bigger (and preferably guaranteed) share for local governments, but
Local would also like a major voice in controlling the distribution of those resources.
Government
Weaker local Would like a guaranteed transfer of resources from the urban and wealthier local
governments governments to the rest. More interested in a transfer system than a local taxing
system.

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.

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 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 outline a normative
approach to sequencing decentralization that might serve as a useful baseline
against which to compare real-world practice.

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
Step 2 Do the policy design and develop the White Paper Pass the decentralization law
Step 3 Pass the decentralization law
Step 4 Develop the implementation regulations
Step 5 Implement the decentralization program
Step 6 Monitor, evaluate and retrofit

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

 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 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.
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 never-ending 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 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.

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 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.
4. 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.
5. 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.
6. Ideally, such reviews should not just be at the central government level, but also by
intergovernmental as well as independent agencies.
7. 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.
8. 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.
9. 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.
10. This section draws heavily on Bahl and Martinez-Vazquez (2006a).
11. 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.’
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 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
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
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-WP-282,
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 Sub-
Central Governments: A Pilot Study,” OECD Economics Department
Working Paper No. 705.
Backhaus, Jurgen (2000) “Germany,” in Robert Andelson, ed., Land-Value
Taxation around the World, 3rd edn (Malden MA: Blackwell), pp. 221–38.
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 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. 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
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).
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).
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?key=1290.
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.
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?
doi=10.1.1.399.1915&rep=rep1&type=pdf.
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 Capital-
Intensive Services: What Evidence Is There of Economies of Scale,
Density and Scope?,” 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. 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: Vote-Seeking,
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).
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.
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 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: McGill-
Queen’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).
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.
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.
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
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.
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 Allingham-
Sandmo,” 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 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,
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 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).
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).
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 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:
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 Martinez-
Vazquez, 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 Hard-
to-Tax (Amsterdam: Elsevier), pp. 275–311.
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.
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 Sub-
National 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 Martinez-Vazquez,
eds, Decentralization and Infrastructure in the Global Economy (New
York: Routledge), pp. 305–23.
Fox, William and Enid Slack (2011) “North America,” in Jorge 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. 257–82.
Fox, William and Christine Wallich (2007) “Fiscal Federalism in Bosnia-
Herzegovina: 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, 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: Inter-
American 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
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
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/wp-
content/uploads/PU5-HABITAT-III-POLICY-PAPER-
FRAMEWORK.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 Murillo-
Garcia (2011) “The Challenge of Designing Capital Equalization
Transfers: An Application to Spanish Regions,” Publius: The Journal of
Federalism, 42 (2): 260–88.
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 Inter-Regional
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).
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 Martinez-Vazquez, 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.
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/making-
public-private-partnerships-more-effective.
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 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.
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
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. Cuadrado-Roura
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.
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 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. 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 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/abstract=2775461.
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_id=2802655.
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 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. 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
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 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. 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
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.
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 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,” 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/AHMEDABAD%20BRTS.pdf.
Natural Resource Governance Institute (2016) “Natural Resource Revenue
Sharing,” available at https://resourcegovernance.org/analysis-
tools/publications/natural-resource-revenue-sharing.
Nauges, Céline and Caroline van den Berg (2008) “Spatial Heterogeneity in
the Cost Structure of Water and Sanitation Services: A Cross-Country
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.
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).
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., TheTaxation 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.
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: 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.
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.
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.
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 Martinez-Vazquez, 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.
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,
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.
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
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 Decentralization
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.
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.
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.
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.
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.
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/A
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.
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).
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 Community-
Based 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 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.
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.
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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 state-owned enterprises; utilities
Paris 354, 357
parking tax 377
participatory budgeting 118
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
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

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, 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
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
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
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
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
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