Rules-Based Fiscal Policy in Emerging Markets

Rules-Based Fiscal Policy in

Emerging Markets
Background, Analysis, and
Prospects

Edited by

George Kopits
© International Monetary Fund 2004
Softcover reprint of the hardcover 1st edition 2004 978-1-4039-2085-0
All rights reserved. No reproduction, copy or transmission of this
publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or transmitted
save with written permission or in accordance with the provisions of the
Copyright, Designs and Patents Act 1988, or under the terms of any licence
permitting limited copying issued by the Copyright Licensing Agency, 90
Tottenham Court Road, London W1T 4LP.
Any person who does any unauthorized act in relation to this publication
may be liable to criminal prosecution and civil claims for damages.
The author has asserted his right to be identified
as the author of this work in accordance with the Copyright,
Designs and Patents Act 1988.
First published 2004 by
PALGRAVE MACMILLAN
Houndmills, Basingstoke, Hampshire RG21 6XS and
175 Fifth Avenue, New York, N. Y. 10010
Companies and representatives throughout the world
PALGRAVE MACMILLAN is the global academic imprint of the Palgrave
Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd.
Macmillan® is a registered trademark in the United States, United Kingdom
and other countries. Palgrave is a registered trademark in the European
Union and other countries.
ISBN 978-1-349-51512-7 ISBN 978-1-137-00157-3 (eBook)
DOI 10.1057/9781137001573
This book is printed on paper suitable for recycling and made from fully
managed and sustained forest sources.
A catalogue record for this book is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Rules-based fiscal policy in emerging markets: background, analysis, and
prospects / edited by George Kopits.
p. cm.
Includes bibliographical references and index.
ISBN 978-1-349-51512-7
1. Fiscal policy—Developing countries—Case studies. I. Kopits, George.

HJ1620.R85 2004
339.5
2
091724—dc22 2004046490

10 9 8 7 6 5 4 3 2 1
13 12 11 10 09 08 07 06 05 04
Contents

List of Tables vii

List of Figures and Box ix
List of Contributors x
Acknowledgments xi
Foreword xii

1 Overview of Fiscal Policy Rules in Emerging Markets 1

George Kopits

Part I Political Economy and Macroeconomic Setting 13

2 Fiscal Rules from a Political Economy Perspective 15
Allan Drazen
3 Good Credit Ratios, Bad Credit Ratings: The Role of Debt Structure 30
Ricardo Hausmann
4 Can Fiscal Rules Help Reduce Macroeconomic Volatility? 53
Guillermo Perry
5 Fiscal Policy and High Capital Mobility 66
George Kopits
6 Fiscal Institutions versus Political Will 81
Allen Schick

Part II Design Issues at the National Level 95

7 EMU Fiscal Rules: What Can and Cannot be Exported 97
Marco Buti and Gabriele Giudice
8 Fiscal Rules and Debt Sustainability in Brazil 114
Ilan Goldfajn and Eduardo Refinetti Guardia
9 Fiscal Rules in Mexico: Evolution and Prospects 131
Andrés Conesa, Moisés J. Schwartz, Alejandro Somuano,
and J. Alfredo Tijerina
10 Fiscal Rules on the Road to an Enlarged European Union 146
Fabrizio Coricelli and Valerio Ercolani

v
vi Contents

11 Fiscal Rules for Economies with Nonrenewable Resources:

Norway and Venezuela 164
Olav Bjerkholt and Irene Niculescu

Part III Design Issues at the Subnational Level 181

12 Subnational Fiscal Rules: A Game Theoretic Approach 183
Miguel Braun and Mariano Tommasi
13 Rules-Based Adjustment in a Highly Decentralized
Context: The Case of India 198
Kalpana Kochhar and Catriona Purfield
14 Fiscal Rules for Subnational Governments: Lessons
from the EMU 219
Fabrizio Balassone, Daniele Franco, and Stefania Zotteri
15 Rules for Stabilizing Intergovernmental Transfers in
Latin America 235
Christian Y. Gonzalez, David Rosenblatt, and Steven B. Webb

References 250
Author Index 268
Subject Index 271
List of Tables

1.1 Emerging markets: summary of fiscal policy rules 3

3.1 Selected regions: public debt (unweighted averages),
1993–2000 31
3.2 Cross-country estimates of credit rating equation, 1990–99 33
3.3 Selected regions: volatility of tax revenue, GDP, and terms
of trade, 1990–99 37
3.4 OECD and Latin America: impact of a shock in tax
revenue and GDP 38
3.5 Latin America and United States: volatility and cyclical
properties of exchange rates and interest rates, 1990–99 41
3.6 Latin America: annual decline in GDP growth 43
3.7 Estimates of cyclical comovement of the real exchange
rate and GDP 44
3.8 Correlation between the real exchange rate, the real
interest rate, and the import gap, 1990–99 45
3.9 OECD and Latin America: stress test on debt service capacity 46
4.1 Latin America and OECD: response of fiscal balance to
GDP growth, 1970–94 55
5.1 Crisis episodes: selected fiscal indicators of
vulnerability, 1994–99 68
7.1 Macroeconomic volatility in Latin America and the
euro area, 1970–2001 105
7.2 Public finances in Latin America and the euro area, 1997
and 2001 105
8.1 Brazil: public sector debt, December 2002 117
8.2 Brazil: general government debt, December 2002 120
8.3 Brazil: public sector debt, 1994–2002 121
8.4 Brazil: baseline scenario for public sector debt, 2002–11 126
8.5 Brazil: primary surplus required to stabilize the
debt–GDP ratio, 2002–12 127
9.1 Mexico: fiscal adjustment required under selected fiscal rules,
2002–30 143
10.1 Accession countries: fiscal elasticities, minimal benchmark,
and cyclically adjusted balance 151
10.2 Accession countries: output and fiscal indicators, 2000 152
11.1 Norway and Venezuela: selected indicators, 2000 171
12.1 Fiscal policy problems and their consequences 190
13.1 India: central and state government finances, 1998–2002 200

vii
viii List of Tables

13.2 India: maturity structure of central government securities,

1997–2002 202
13.3 Selected regions: volatility of tax revenue, GDP, and
terms of trade, 1990–99 204
13.4 India: macroeconomic assumptions, 2004–08 208
13.5 India: central government finances under the FRBM,
2003–08 209
13.6 India: general government finances under the FRBM,
2003–08 210
13.7 India: effect of elimination of growth-interest rate
differential on government debt under the FRBM, 2003–08 212
List of Figures and Box

Figures

3.1 Credit rating versus net debt–GDP ratio 31

3.2 Credit rating versus debt–revenue ratio 32
3.3 Volatility and VaR in debt service 36
3.4 Illustration of debt service dynamics 36
3.5 Illustration of the role of volatility 37
4.1 Selected regions: volatility of real GDP growth by decade 54
7.1 Public finance convergence in the euro area,
1993–2000 100
7.2 Composition of the fiscal adjustment in the euro area,
1993–2000 101
8.1 Brazil: public sector borrowing requirement and primary
deficit, 1998–2002 116
8.2 Brazil: net public debt, assuming constant exchange rate
and hidden liabilities, 1994–2002 118
8.3 Brazil: net public debt, assuming constant primary surplus
and interest rate, 1994–2002 118
9.1 Mexico: estimates of PSBR, 1991–2002 138
9.2 Mexico: fiscal adjustment required under structural
balance rule, 1991–2002 139
9.3 Mexico: public debt and financial absorption, 2002–30 141
9.4 Mexico: public debt under the balanced-budget rule, 2002–30 142
10.1 Accession countries: fiscal stance, 1990–2000 153
10.2 Volatility of government revenue, 1991–2000 154
10.3 Estimates of budget balance, 1990–2000 161
12.1 Outcomes under alternative policy choices 187
13.1 India: growth-interest rate differential, 1992–2003 199
13.2 India: nominal interest and inflation rates, 1991–2003 202
13.3 India: ratio of capital to current spending, 1991–2000 203
13.4 India: government deficit, investment, and growth, 1991–2002 205
13.5 India: government debt dynamics under the FRBM, 2003–08 211
14.1 EU member countries: internal stability pacts 229

Box

13.1 India: state-level fiscal responsibility legislation 206

ix
List of Contributors

Fabrizio Balassone Bank of Italy

Olav Bjerkholt University of Oslo
Miguel Braun CIPPEC and University of San Andrés
Marco Buti European Commission
Andrés Conesa Secretariat of Finance, Mexico
Fabrizio Coricelli University of Siena and CEPR
Allan Drazen Tel Aviv University, University of Maryland, NBER, and CEPR
Valerio Ercolani University of Siena
Daniele Franco Bank of Italy
Gabriele Giudice European Commission
Ilan Goldfajn Central Bank of Brazil
Christian Y. Gonzalez Georgetown University
Eduardo R. Guardia State of Sao Paulo, Brazil
Ricardo Hausmann Harvard University
Kalpana Kochhar International Monetary Fund
George Kopits National Bank of Hungary
Irene Niculescu Central University of Venezuela
Guillermo Perry The World Bank
Catriona Purfield International Monetary Fund
David Rosenblatt The World Bank
Allen Schick Brookings Institution and University of Maryland
Moisés J. Schwartz Secretariat of Finance, Mexico
Alejandro Somuano Secretariat of Finance, Mexico
J. Alfredo Tijerina Secretariat of Finance, Mexico
Mariano Tommasi University of San Andrés and CEDI
Steven B. Webb The World Bank
Stefania Zotteri Bank of Italy

x
Acknowledgments

The majority of the chapters contained in this volume were selected from
papers presented and discussed at a conference held in Oaxaca, Mexico,
February 14–16, 2002, under the auspices of the Government of Mexico, the
International Monetary Fund, and the World Bank. Several chapters were
commissioned subsequently for regional and thematic completeness. In
addition to the presentations, senior government officials from Brazil,
Colombia, Mexico, and Peru participated in a concluding panel discussion.
This volume has been enriched by the views of participants from academic
and research institutions, as well as policymakers from national and subna-
tional governments, and staff from international financial institutions, who
shared their considerable range of experience and analysis. The conference
benefited from inspiration and support by Eduardo Aninat, Francisco Gil-
Díaz, Vijay Kelkar, Guillermo Perry, and Teresa Ter-Minassian, and from
organizational support by David Colmenares, Moisés Schwartz, Augusto de
la Torre, and Jaime René Jiménez. The authors are grateful for comments
from assigned discussants and other participants mentioned in an initial
footnote in each chapter. Also, useful comments were provided by Max Alier,
James Daniel, Hugo Juan-Ramón, and Paolo Manasse, in their capacity as
anonymous referees on selected chapters. Editorial assistance by Carol Ann
Robertson and administrative assistance by Elizabeth Handal-Kocis are grate-
fully acknowledged. The views expressed are solely those of the authors and
should not be ascribed to their institutional affiliations.

xi
Foreword

Emerging market economies are exposed to considerable macroeconomic

volatility in tandem with a high level of capital mobility. The combination
of this inherent volatility and failed discretionary macroeconomic policy
management has often led to major capital account crises, with unfortunate
consequences for activity, employment and overall welfare. Against this
backdrop, in recent years, a number of emerging market economies have
turned to a rules-based framework for the conduct of monetary and fiscal
policies. In particular, given the endemic deficit bias as a major source of
instability in many of these countries, the introduction of constraints on the
budget balance, government expenditure growth or stock of public indebt-
edness has become rather appealing. These constraints are typically enacted
in the context of comprehensive fiscal responsibility legislation.
Increasingly, fiscal and monetary rules are seen as useful vehicles for
cementing policy credibility, as well as for preventing rapid shifts in investor
sentiment and attenuating vulnerability to financial crises. The International
Monetary Fund welcomes the recent interest in fiscal policy rules and
encourages member countries to consider their potential usefulness,
whether introduced as part of an adjustment program, or more important,
as part of a broader effort to strengthen key institutions and embrace inter-
nationally accepted good practices. However, experience shows that such
rules can contribute to stability and growth only if they are properly
designed and they take into account country-specific cultural, political, and
economic conditions. Furthermore, their success hinges on the informed
support of the authorities and the electorate.
The studies collected in the present volume, authored by academic scholars
and public officials from national and international institutions, represent a
valuable contribution to the ongoing debate on the merits and drawbacks of
fiscal policy rules in emerging market economies. They provide a thorough
analysis of the volatile environment faced by authorities in these economies,
taking into account a variety of key features that are present in some but not
all of these economies: openness to capital movements, nonrenewable
resource endowment, excessive fiscal decentralization, and other structural
and behavioral characteristics. Beyond the broader issue-oriented chapters,
each study focuses on a country or a group of countries in major regions: Latin
America, Asia, and Central and Eastern Europe. Thus a rich variety of circum-
stances is brought under scrutiny, drawing insofar as possible lessons from the
experience of some advanced economies, notably the European Union.
The chapters explore a range of design features, along with various statu-
tory provisions and operational arrangements. However, a major unifying

xii
 Foreword xiii

theme in the volume is the sober tone and balanced assessment which helps
counter the unrealistic view (popular in some quarters) that policy rules
automatically insure fiscal sustainability and macroeconomic stability. A
consensus emerges that fiscal policy rules can provide meaningful commit-
ment technology only if supported by a robust institutional infrastructure,
including a high degree of transparency in public finances. In all, this
comprehensive and timely volume encompassing a major policy area will be
useful for analysts and practitioners alike.

Agustín Carstens
Deputy Managing Director
International Monetary Fund
1
Overview of Fiscal Policy Rules
in Emerging Markets
George Kopits1

Introduction

In the past decade, several advanced economies have shifted from

discretion-based to rules-based fiscal policies. This shift has taken place in
countries such as New Zealand, Australia, and the United Kingdom, but
perhaps most visibly in the European Union (EU) in support of monetary
unification.2 Thus, experience with fiscal policy rules, in the context of
Europe’s economic and monetary union (EMU), has been the object of
extensive analysis.3 Also, there has been considerable research on the effects
of much older subnational rules in the United States.4
In emerging market economies, the adoption of fiscal policy rules has
been much more recent and limited mainly to Latin America. In some
instances, the rules were introduced following a financial crisis; in others
they were adopted to reduce vulnerability to a potential crisis. Often the
immediate motivation has been to reverse the buildup of public debt, to
restore fiscal sustainability and, more generally, to enhance the credibility of
macroeconomic management. In addition, in some regions, mainly Central
and Eastern Europe, rules are increasingly viewed as an anchor in the
convergence to a broader monetary union.
The growing interest in rules-based fiscal policies in emerging market
economies, stemming from an increasing realization that it is essential to
correct the fundamentals, has not been matched until now by sufficient
in-depth analysis.5 In the meantime, primary attention to the role of mon-
etary policy, especially regarding the choice of an appropriate exchange rate
regime, has gradually given way to the recognition of the importance of
sound fiscal policy and robust institutions in these economies.6 Moreover,
recent empirical evidence supports the view that discretionary fiscal policy
has been detrimental to overall macroeconomic performance.7
Accordingly, fiscal policy rules – if well designed and properly
implemented – are viewed as potentially useful commitment technology for
emerging market economies exposed to macroeconomic volatility and high

1
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
2 George Kopits

capital mobility. From a political economy perspective, fiscal rules can be

instrumental in avoiding the myopic policies that result from dynamic
inconsistency and/or political distortions, especially in highly decentralized
countries.8 Broadly speaking, fiscal rules can help depoliticize the macro-
economic policy framework.9
This volume of collected studies – intended primarily for policy analysts and
practitioners – is comprised of three parts. The first part reviews the macro-
economic setting and rationale for rules-based policies in emerging markets,
taking into account political economy aspects. The second and third parts are
devoted to design issues respectively at the national and subnational levels of
government, in the light of current country practices under a wide variety of
circumstances and institutional conditions. In addition to focusing on emerg-
ing market economies in Latin America, Asia, and Central Europe, the rele-
vant experience of advanced economies is also brought under scrutiny.

Current practices

With very few exceptions, the fiscal policy rules in emerging market
economies are of recent vintage, mainly since the late 1990s (Table 1.1);
thus actual experience has been relatively brief, unlike in some advanced
economies.10 Most of the existing rules are expressed as a permanent con-
straint on a broad performance indicator, such as the government balance
(in flow terms) or the public debt (in stock terms), usually as a proportion of
GDP. In practically all these countries, fiscal policy rules have been embed-
ded in a rules-based monetary framework. The latter includes an inflation-
targeting regime (Brazil, Chile, Colombia, Mexico, Peru, Poland), a currency
board arrangement (Estonia, and until recently Argentina), or a dollarized
regime (Ecuador). In this sense, fiscal rules can be viewed as means to reduce
or eliminate fiscal dominance in macroeconomic policy.
Rules on budget balance are subject to some variations: overall balance,
current balance, and in a few cases numerical limits on the overall deficit
(Argentina, Peru, India), or a floor for the overall surplus (1 percent of GDP
in Chile). Observance of a current-balance rule, also called the “golden rule”
(Brazil, Mexico, India, Venezuela), can prevent a crowding-out of much-
needed public investment. In some countries, the budget-balance rule is
accompanied by additional limits on total government expenditures
(Venezuela), primary (noninterest) outlays (Argentina, Ecuador, Peru), inter-
est payments (Colombia), and/or the wage bill (Brazil, Colombia), in order
to contain the fastest growing components of fiscal imbalance.
In some countries, under the budget-balance rule, escape clauses are pro-
vided in the form of a contingency fund and/or a multiyear definition of the
rule in order to accommodate shocks or cyclical fluctuations in activity. The
contingency fund (also called stabilization or countercyclical fund) is intended
to release resources to finance a cyclically induced deficit or to withdraw
Table 1.1 Emerging markets: summary of fiscal policy rules1

Rule/Country Effective Coverage2 Basic rule3 Escape Additional Statute4 Sanction5

date clause3 rule3

Budget rules
Argentina 2000 NG6 OB, DN CF EN L R
Brazil 2001 NG, SG CB WN C, L J
Chile 2000 NG SN MY, CF P R
Colombia 1997, 2001 SG CB WN, RN L J, F
Ecuador 2003 NG NB, DN CF EN L J
Estonia 1998 GG OB CF P R
India 2004 NG6 CB, DN L R
Indonesia 1967 GG DF P R
Mexico 1917 SG CB C R
Peru 2000 NG OB, DN CF EN L J
Venezuela 2004 NG CB MY, CF TN C, L R
Debt rules
Brazil 2001 NG, SG FN C, L J
Colombia 2004 NG, SG FN L R
Ecuador 2003 NG PN L J
Poland 1998 GG, SG PN C, L J

Notes
1
Table excludes standard restriction or prohibition on financing from the central bank.
2
General government (GG), national (central, federal) government (NG) or subnational governments (SG).
3
Budget rules consist of overall balance (OB), current balance (CB), or non-oil balance (NB); prohibition on domestic financing of deficits (DF);
and/or numerical limits on deficit (DN) or numerical threshold for surplus (SN), as ratio of GDP. All rules are applied on an annual basis, unless
specified on a multiyear (MY) basis; and/or a contingency fund (CF) is provided – besides the standard escape clause in event of war or natural
calamity. Additional rules consist of numerical limits on total expenditure (TN), primary expenditure (EN), wage bill (WN), and/or debt service (RN),
in terms annual growth or ratio of GDP. Numerical limits on debt stock targeted for a future year (FN) or permanently (PN) as ratio of GDP.
4
Constitution (C), legal provision (L), or policy guideline or target (P).
5
Sanctions for noncompliance: reputational (R), judicial (J ), or financial (F).

3
6
Adopted also by one or several subnational governments.
4 George Kopits

them from a cyclically generated surplus (Argentina, Estonia, Peru). More

directly, the rule may be defined in terms of structural or cyclically adjusted
balance (Chile). A multiyear or medium-term balanced-budget requirement
(Ecuador, Venezuela), which allows not only for the operation of automatic
stabilizers but also for some countercyclical discretionary action, performs a
similar function. To cushion the budget from output changes, some countries
(Ecuador) specify ex ante the real growth of primary expenditure in terms of
a constant rate, broadly in line with trend or potential GDP growth.
Either in conjunction with a budget-balance rule, or simply with the goal of
securing medium- to long-term fiscal sustainability, several countries have
established targets for the phased reduction of the debt–GDP or debt–revenue
ratio (Brazil), or limits on the debt–GDP ratio (Poland). The debt-ratio target
or ceiling usually presupposes, either implicitly or explicitly (Brazil), an annual
operational target in terms of a minimum primary surplus.
Generally, the institutional coverage of rules depends on the degree of fiscal
decentralization and autonomy of various levels of government. In a rela-
tively centralized fiscal system the rules are imposed only on the central gov-
ernment (Chile, Peru) without much loss of control. However, in federal
systems, rules must encompass the national and subnational levels of gov-
ernment. Depending on the degree of subnational autonomy, uniform rules
are imposed top-down on all subnational governments (Brazil, Colombia,
Mexico, Poland) or differentiated rules are voluntarily self-imposed from the
bottom-up by some subnational governments (Argentina, India).
Fiscal policy rules can be specified in a constitutional provision (Mexico,
Poland), high-level legislation (Brazil), or ordinary legislation (India) that
applies to governments over successive electoral cycles. The most frequent
statutory vehicle is a comprehensive Fiscal Responsibility Law (FRL) – named
after the legal framework introduced by New Zealand in 1994. Alternatively,
fiscal rules may merely consist of a policy guideline declared by a given
government and not necessarily binding on future governments (Chile,
Estonia, earlier in Indonesia).
In terms of contents, the statute may be very detailed (Brazil), specifying
not only the nature of the policy rules, but also detailed procedural rules
governing compliance. At the other end of the spectrum, it may define a
broad framework (India), to be accompanied by regulations issued by the
government in charge. Increasingly, countries enhance their credibility by
implementing policy rules subject to transparency standards – in contrast to
earlier opaque applications.
Whatever the statutory form, at present most policy rules are supported by
institutional arrangements encompassing the budget process (possibly in a
rolling medium-term budget framework), accounting conventions, periodic
reporting and projection requirements, and penalties for noncompliance.
Also, there is an assignment of responsibilities for implementation versus
monitoring and audit – the latter usually to be undertaken by an independent
 Overview of Fiscal Policy Rules 5

authority. Legal sanctions for noncompliance (treated as a criminal offense

in Brazil) may exist as well, though typically these remain untested in the
courts. Rarely, in the case of top-down subnational government rules
(Colombia), deviation from the rule is subject to financial penalties. However,
in most countries, noncompliance, especially by the national government, is
punished with loss of reputation toward the electorate or financial markets.

Political economy and macroeconomic setting

From a survey of the literature, Drazen shows how fiscal policy rules can help
address the problem of time inconsistency and the deficit bias. He indicates
how properly designed fiscal rules can be a useful means for building
reputation and serve as a disciplining device, as long as they are accompa-
nied by various procedural rules – including those that prevent creative
accounting practices. In particular, rules do not work when there are no real
costs attached to deviating from or changing them. Drazen further argues
that rules are more likely to be effective if they are enshrined in the consti-
tution or other high-level legislation.
Given that emerging market economies are exposed to considerable
macroeconomic volatility (in output, terms of trade, interest rates, and
exchange rate), Hausmann observes that they suffer from a much higher
risk premium than advanced economies, even with the same level of pub-
lic indebtedness expressed in terms of GDP. An additional explanation
for the risk differential lies in the excessive share of dollar-denominated,
short-term maturities in emerging market government liabilities. A major
implication is that emerging market economies would benefit from
fiscal rules that aim not only at eliminating deficits and reducing debt
ratios, but more important, also at containing the risk in the composition of
the debt.
High macroeconomic volatility, experienced especially in Latin America,
has been aggravated by the procyclical stance adopted under various fiscal
adjustment programs. Thus, to mitigate volatility and to avoid procyclical
adjustment, Perry argues, these countries ought to follow a rule that incor-
porates a countercyclical stance through a structural-balance target or a sta-
bilization fund. Admittedly, however, limited access or loss of access to
external financing precludes a countercyclical expansion during a recession
in a highly indebted economy.
The openness of emerging market economies to high capital mobility,
according to Kopits, underscores the case for predictable time-consistent
macroeconomic policies, and in particular for well-designed fiscal policy
rules. A review of crisis episodes points to the significant contribution of
public sector indebtedness to capital account crises. In view of serious rigidi-
ties that prevent rapid fiscal response to currency crises – often precipitated
by sudden shifts in market sentiment – it is important to signal early on a
6 George Kopits

commitment (as a precursor of the actual adjustment) to debt reduction.

Thus, fiscal policy rules can be a useful signaling device if accompanied by
strict transparency standards.
Fiscal policy rules do not guarantee sound fiscal management. Schick
emphasizes the critical role of political will for the success of any fiscal pol-
icy rule (or of prudent fiscal policymaking in general), when supported by
appropriate procedural rules. He notes that the recent literature on fiscal
institutions and budgetary process neglects political will and fails to distin-
guish between formal rules and informal practices. Schick identifies innova-
tions in budget procedures that are conducive to strengthening political will
and enforcement of rules.

Design issues at the national level

In an attempt to draw some lessons from the experience of advanced

economies, Buti and Giudice examine key elements of the EMU rules, as
applied under the Stability and Growth Pact (SGP), particularly its collegially
enforced mechanism of prevention of fiscal laxity and dissuasion from non-
compliance with the rules. The authors argue that EMU rules cannot be
exported unchanged to emerging market economies. Given the much
greater macroeconomic volatility and stronger procyclical deficit bias in
many of these economies, they should consider targeting a structural pri-
mary surplus and devising their own political economy incentives coupled
with steps to strengthen national budget procedures.
In Brazil, the FRL prescribes for each level of government detailed pol-
icy rules, budget procedures, transparency standards, and judicial penalties
for noncompliance. Following a review of past fiscal performance and the
initial experience under the FRL, Goldfajn and Guardia simulate the
government’s financial position over the medium term. The simulation
exercise suggests that, barring an unexpected deterioration of macroeco-
nomic conditions (e.g. declining growth rate, real interest rate increase, or
exchange rate depreciation), the FRL and associated innovations should
ensure public debt sustainability. Two key structural measures are critical to
this outcome: public pension reform and tax policy reform – both currently
under way.
Besides the golden rule for state governments, in recent years Mexico has
introduced procedural rules to enhance the effectiveness of budgetary con-
trol. At the federal level, contingent rules (measures specified ex ante to com-
pensate for any excess or shortfall, relative to the budget deficit target) in
operation since 1998 have contributed to fiscal discipline though at the
expense of a procyclical stance. Besides assessing the existing rules, Conesa,
Schwartz, Somuano, and Tijerina explore various options for a balanced-
budget rule (including a structural variant) to ensure fiscal sustainability
while allowing for revenue volatility, along the lines of a recent government
 Overview of Fiscal Policy Rules 7

proposal. The authors discuss simulation results in terms of their fiscal and
macroeconomic consequences.
Several emerging market economies in Central and Eastern Europe have
just become, or are prospective, EU members on the path of convergence to
the EMU fiscal reference values. Coricelli and Ercolani report that these
economies have experienced much higher output volatility than older EU
members, to which they have responded largely with a procyclical fiscal
stance. To correct this anomaly, the authors suggest a modification of the
existing EMU rules. The modified rule would call for a structural deficit tar-
get large enough to accommodate pending transition- or accession-related
expenditures. According to the authors, the rule would be relatively easy to
apply, based on an ex ante computation of the structural balance.
Emerging market economies endowed with nonrenewable resources face
very high volatility in the world price of the resource, which often translates
into extreme procyclicality of government expenditure. Bjerkholt and
Niculescu compare the adequacy of two alternative approaches to cope with
this situation: Norway’s policy rule, limiting the cyclically adjusted nonre-
source budget deficit (i.e. exclusive of resource-related revenue) to the return
on accumulated proceeds from resource extraction, and Venezuela’s combi-
nation of fiscal rules and an oil stabilization fund, which allows for a sizable
nonresource deficit. Major differences between the two countries – an aging
population in the former and a significant current need for public spending
in social programs and infrastructure in the latter – explain the differences
between the selected approaches.

Design issues at the subnational level

Attempts to contain subnational government indebtedness in Latin America

through the adoption of fiscal rules have been met with mixed success. In
particular, Argentina’s failure with subnational fiscal rules can be attributed
to large vertical imbalances, lack of incentives for local revenue raising, easy
access to and incentives for bailouts, and lack of enforcement of intergov-
ernmental agreements. In a game theoretic context, Braun and Tommasi
argue that lacking an outside enforcer, subnational rules can be effective only
if supported by institutional arrangements that stimulate cooperative behav-
ior by subnational governments. This requires mapping out a reform strategy
taking into account key country-specific economic and political features.
At present, India is among the most indebted emerging market economies;
also, it is running a very high fiscal deficit. Further external liberalization will
require a major fiscal adjustment effort that poses an extraordinary challenge
in the context of a highly decentralized system. Kochhar and Purfield explore
the adequacy of the recently enacted FRL in meeting this formidable adjust-
ment challenge over the medium term. On the basis of a set of quantitative
simulations, the authors conclude that the public debt–GDP ratio is bound
8 George Kopits

to rise even further in the future, jeopardizing India’s fiscal sustainability and
macroeconomic stability, and dampening the prospects for economic
growth, unless the federal FRL is matched by implementation of similar leg-
islation by all state governments and is supported by key structural reforms.
Several fiscally decentralized EU member countries have introduced vari-
ous forms of subnational rules, in order to support compliance with EMU
rules – as the latter are applied to the general government as a whole.
Balassone, Franco, and Zotteri examine practices in five EU countries (Austria,
Belgium, Germany, Italy, and Spain). They find that although in principle
explicit EMU-like subnational rules have distinct advantages over purely
cooperative arrangements, in practice the strengthening of consensus- based
institutions and procedures may be preferable not only in the EU economies,
but also in emerging market economies. While cooperation has been effec-
tive for general government deficit reduction, it may be suboptimal as regards
resource allocation and may not withstand stress situations in the future.
Traditionally, the central government is responsible for macroeconomic
stabilization; in addition, the central government usually has cheaper and
more stable access to financial markets than subnational governments do.
In the context of Latin America, Gonzalez, Rosenblatt, and Webb discuss
how stabilizing transfer rules (withholding transfers during upturns and
releasing them during downturns) can protect subnational governments
from the effects of economic cycles. They observe that, these rules can be
risky, as the protection provided in a downturn becomes a contingent lia-
bility for the central government. A major precondition for success lies in
the elimination of structural vertical imbalances.

Lessons for policy

To sum up, a number of lessons can be drawn from the analysis and experi-
ence presented in this volume. Although some of these lessons may be tenta-
tive, in general they affirm that emerging market economies can gain
credibility by adopting a rules-based fiscal framework. However, there is nearly
unanimous agreement that fiscal policy rules are not a magic wand that some-
how will immunize the economy against macroeoconomic volatility or finan-
cial crises and will sustain high economy growth. Clearly, the timing, design,
circumstances, and overall institutional basis are critical for the success of the
rules-based approach. The main lessons can be summarized as follows:

1. In emerging market economies, just as in advanced economies, fiscal

rules need the support of the electorate. Without such support translated
into political will by the authorities, fiscal rules are bound to weaken or
collapse altogether (Argentina, Peru). As a first step, this requires a broad
and deep consensus, built on a thorough and informed debate about the
benefits and responsibilities associated with the rules, taking into account
the legal and cultural specifics of the country (Brazil, Colombia, India).
 Overview of Fiscal Policy Rules 9

2. As a corollary, although in principle it is preferable to enshrine fiscal rules

in the constitution or in a high-level law or statute, informal rules might
be equally (or even more) effective as long as they are backed by broad
public consensus. Ironically, besides a promising case of formal rules
(Brazil), the most successful rules in emerging market economies simply
consist of a policy guideline (Chile and Estonia), applied with support
across the political spectrum. By implication, governments should refrain
from enacting a strong statutory obligation until after a period of exper-
imentation with a self-imposed policy guideline.
3. Macroeconomic policy rules – whether in the fiscal or the monetary area –
can be viable only if underpinned by strong procedural rules, including
good practices in transparency and accountability (Brazil); by contrast,
lack of transparency leads to the demise of any set of policy rules
(Indonesia). A clear and centralized process of decision-making at all
stages of budget formulation and execution, as well as an orderly legisla-
tive budget debate and oversight, are highly desirable. Finally, an inde-
pendent audit mechanism is essential. For emerging market economies,
this implies a major and relentless institution-building effort.
4. Markets have far lower tolerance for relatively high public debt–GDP
ratios in emerging market economies than in advanced economies.
Hence, to avert vulnerability to capital account crises, resulting in a pos-
sible loss of access to markets, it is essential for highly indebted govern-
ments to commit above all to a realistic reduction in the debt ratio
through a rule that targets a primary balance consistent with the envis-
aged debt reduction path (Brazil).
5. In emerging markets, fiscal rules must be designed to take into account
significant macroeconomic volatility (in output, terms of trade, interest
rates, and exchange rate). Indeed, it is essential that balanced-budget or
expenditure rules accommodate exogenous shocks and cyclical fluctua-
tions by allowing for the operation of automatic stabilizers and, if neces-
sary, their support with a mild discretionary countercyclical stance.
However, for such rules to be credible, they must operate symmetrically,
permitting budget deficits during downturns but requiring surpluses during
upswings in activity. In any event, a structural or cyclically-adjusted
balanced budget rule can be pursued only upon achieving a moderate
debt ratio (Chile, Estonia).
6. As an alternative, particularly for economies with nonrenewable resources,
a commodity stabilization fund that complements limits on the budget-
deficit and expenditure can cushion pressures stemming from wide fluc-
tuations in the terms of trade (Ecuador, Venezuela). In addition, in these
economies a saving fund might be useful to accumulate reserves to
finance infrastructure and social needs or to meet the eventual aging of
the population.
7. Fiscal decentralization requires considerable care in the design and
enforcement of rules. In a federal system, numerical policy rules need to
10 George Kopits

be enforced at the decision-making level, that is, at each level of

government, rather than for the consolidated public sector or general
government as a whole. In some countries, under a top-down approach
(Brazil), there is scope for setting uniform policy and procedural rules for
each level of government. It is far more difficult to establish consistent
fiscal rules in a highly decentralized system through a bottom-up
approach (Argentina, India), in which subnational governments voluntar-
ily adopt binding rules – particularly when needed to support an economy-
wide fiscal adjustment. This, of course, presupposes a credible no-bailout
clause toward subnational governments. In addition, in a decentralized
structure, rules for stabilizing intergovernmental transfers are necessary.
8. Finally, for fiscal policy rules to be credible, initiating key long-term
structural reforms early on is indispensable, especially in areas such as
taxation, social security, or intergovernmental relations (Argentina,
Brazil, India, Mexico, Peru). In this regard, rolling medium- to long-term
macroeconomic fiscal projections are useful for anticipating the need for
reforms that will facilitate compliance with the rules.

Notes

1. Peter Heller, Guido Tabellini, and Julio Viñuela provided helpful comments.
However, responsibility for all views expressed rests with the author.
2. See, for instance, Kopits and Symansky (1998) and papers collected in Bank of Italy
(2001) with a primary focus on advanced economies.
3. For a recent set of articles on the experience with the EMU, including a review of
the rapidly growing literature, see Allsop and Artis (2003).
4. See the broad survey and empirical analysis on institutions in US states by Besley
and Case (2003).
5. Unlike the assessment of policy rules in advanced economies, in emerging market
economies analysis has been limited mainly to procedural rules. See, for example,
Poterba and von Hagen (1999) and Kopits and Craig (1998) on the effects of
various institutional arrangements, as well as of their transparency, on fiscal
performance.
6. Calvo and Mishkin (2003) and Kopits (2002) have stressed the importance of insti-
tutional arrangements and fiscal discipline as preconditions to viable exchange rate
systems in emerging market economies.
7. In a large cross-country empirical study, Fatas and Mihov (2003) found that
discretion-based fiscal policy contributes to significant output volatility and
reduced economic growth.
8. For a comprehensive analysis of dynamic-consistency and common-pool prob-
lems, see Persson and Tabellini (2000).
9. In this regard, there are two types of policy rules (as distinct from procedural rules):
direct constraints imposed on the elected policy makers and policy constraints del-
egated to nonelected authorities. Typically, in the monetary area, an example of
the latter consists of an inflation-targeting regime delegated to an independent
central bank. By contrast, the absence of delegated fiscal rules might be explained
 Overview of Fiscal Policy Rules 11

by the difficulty of entrusting policy making to an agent (an independent fiscal

authority) far removed from the principal (ultimately the electorate), as this
would in fact obviate the involvement of elected executive and legislative bodies
from budget formulation and debate. In fact, most of the fiscal literature,
including this volume, deals with the first type of policy rules.
10. Exceptions of older rules are Mexico’s current budget-balance requirement at the
subnational level and Indonesia’s prohibition on domestically financed deficits
for the general government. Both have been applied in a highly opaque fashion,
with mixed success. In Mexico, until recently the central authorities have
enforced the golden rule through considerable discretion and control of state gov-
ernment finances. In Indonesia, the prohibition did not prevent (and even
induced) excessive external borrowing by the public sector – including through
guarantees extended to the private sector.
Part I
Political Economy and
Macroeconomic Setting
2
Fiscal Rules from a Political
Economy Perspective
Allan Drazen1

Introduction

Persistent deficits, implying a secular growth of debt, have led many to argue
that fiscal rules may play an important role in helping reduce or eliminate
deficits and control the growth of government debt. There are two general
classes of fiscal rules. First, there are legislated quantitative constraints on
fiscal policy. These limits take a variety of forms: restrictions on deficit
financing, including balanced-budget laws; expenditure ceilings; numerical
targets for fiscal variables; borrowing rules; and restrictions on issuance of
debt. Much of the discussion of fiscal rules concerns such restrictions.2
Second, there are restrictions or rules on the procedure by which fiscal deci-
sions are made. These procedural rules may concern fiscal policy formula-
tion, as well as the actual execution of policy. Examples of policy formation
procedures include those that limit the extent to which the process is “hier-
archical,” list requirements for effective “transparency” in the budget docu-
ment, specify rules of amendment in both the formulation and approval of
the budget, and describe the nature of voting in the approval process.
Examples of policy execution procedures include those that restrict supple-
mentary budgets and open-ended appropriations in the budget implemen-
tation stage, provide automatic contingency rules, such as across-the-board
spending cuts, and allow sequestering of funds.
Perhaps the key question about fiscal rules is whether they can slow
the growth of deficits. Many observers argue that fiscal rules are ineffective,
because governments can so easily get around them. A related question is
whether a government’s commitment to fiscal restraint takes the place
of such rules. One may ask whether rules matter at all, in the sense that
governments truly committed to fiscal discipline build a reputation for
sound budget policy and hence make credible their announcements to
that effect, while governments not committed to fiscal discipline find ways
to get around fiscal rules (Kopits 2001a). In the latter case, a fiscal rule may
in fact be worse than useless, as it invites “creative fiscal accounting,” which

15
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
16 Allan Drazen

introduces fiscal distortions that would not be present in the absence

of rules.
The answer to the question of whether fiscal rules can play an important
role in reducing or eliminating deficits involves far more than the technical
design of rules, which has been a focus of much research. It concerns the
basic question of what problems a fiscal rule is meant to address and how a
rule can address these problems. This in turn raises general questions of
how formal rules or laws can be more effective than announced nonlegal
commitments to the same objectives. A political economy perspective on fis-
cal rules in particular, and the nature of rules in general, will help answer
these questions. It will also help us better understand the relation between
fiscal rules and the notion that governments that can build a reputation for
fiscal discipline do not need to rely on rules. As we shall see, rules, rather
than substituting for reputation, may help a government build reputation.

The rationale for fiscal rules: deficit bias

This discussion begins with a general perspective on why rules, fiscal or oth-
erwise, may be optimal. Basic economic theory suggests that a policymaker
can maximize social welfare when he is free to choose policy without arbi-
trary constraints, that is, to use his discretion. Because an unconstrained pol-
icymaker can always follow a specific policy guideline if he so chooses,
allowing him to deviate from a preset guideline should only increase welfare.
A standard argument for rules over discretion for a social welfare–
maximizing policymaker concerns cases where first, individual behavior
depends on expectations about future policy, and second, the policymaker
is limited in his choice of policy instruments. If he can change policies over
time, he will often have the incentive to announce one policy for the future
and then implement a different policy when it comes time to carry out his
policy announcement. This is the well-known problem of time inconsistency
in the choice of policy (Drazen 2000, chapter 4). Moreover, the policymaker
has the incentive to be time inconsistent in the choice of policies because his
objective is to maximize social welfare.
However, when individuals know the incentives for time inconsistency, a
time-inconsistent policy is not an economic equilibrium. People will form
their expectations of future policy on the assumption that the government
will deviate from announcements. The cost of having the discretion to
change is that the equilibrium resulting may imply low welfare, lower than
the case in which the government could credibly commit itself ex ante to a
specific policy. If rules can be made credible in the sense that the govern-
ment is expected to follow them, rules give higher welfare than discretion.
The literature on time inconsistency provides many examples. The best-
known macroeconomic example is probably the inflation bias result of
Barro and Gordon (1983b). In their example, the government maximizes
 Fiscal Rules from a Political Economy Perspective 17

the welfare of the representative individual, whose utility depends on the

fluctuations of unemployment and inflation around the target values.
Surprise inflation will lower actual unemployment relative to the natural
rate of unemployment, where the representative individual’s (and hence the
government’s) target unemployment rate is below the natural rate. Suppose
that the target inflation rate is zero. If people expected a zero inflation rate
(perhaps because of a government announcement), the government would
have the incentive to choose an inflation rate above zero in order to lower
unemployment below the natural rate and move it closer to the socially
optimal target. This incentive to inflate would be anticipated so that a zero
inflation announcement would not be believed. The equilibrium is one
where a positive rate of inflation is correctly anticipated, so that unemploy-
ment is at the natural rate, but inflation is suboptimally high. The problem
is that if the government can, it will use its discretionary policymaking
power to try to lower unemployment. This attempt will be unsuccessful in
equilibrium, but will have adverse implications for inflation. If the govern-
ment had a mechanism to commit itself credibly to the choice of zero infla-
tion, welfare would be higher. Hence, we have an illustration of the
argument for a (credible) rule over discretion.
Two points should be stressed concerning the basic time-inconsistency
result on inflation. First, there is a case for constraining the policymaker
even if he is a social welfare maximizer. Second, the argument for rules
revolves not around the unpredictability of inflation policy, but around its
known positive bias. Credible rules may also improve welfare by lowering
the unpredictability of policy, but that is not the main lesson.
In the case of fiscal policy, we begin from a similar perspective. Credible
rules may make fiscal policy more predictable, but the main argument for
fiscal rules is the positive bias toward government budget deficits we observe
in many countries. That is, as indicated, the attraction of rules is that by con-
straining policymakers, they will reduce or eliminate the tendency toward
budget deficits.
There is another major argument why there is an inherent bias toward
fiscal deficits in many countries, in addition to any time-inconsistency
reasons. Put simply, budgets are not chosen by politicians to maximize social
welfare, but are the result of a political process of budgeting that appears to
have a deficit bias. The term “political process” does not refer only to the
legislative budget process, the rules and institutions by which the budget is
made. It refers to the process of making and implementing the budget
combined with the political forces that determine the nature of the budget
emerging from that political process. Hence, one must ask two questions.
First, what is the nature of the legislative budget process that allows political
forces to bias the budget outcome relative to what is considered socially
optimal? Second, what aspect of these political forces encourages exces-
sive deficits? There is a significant body of literature discussing the first
18 Allan Drazen

question, how the details of the budget process may lead to a deficit bias in
general.3
This section considers the second question, briefly summarizing the
arguments on the political forces that explain a positive bias. Politicians may
use the budget process to increase expenditures in excess of taxes for their
own political aims. There are several arguments as to how this happens.
First, there is the electoral motive toward high spending in election years. In
many countries incumbents appear to increase government spending before
elections in order to improve their reelection prospects. Fiscal manipulation
before elections is especially strong in developing countries.4
More generally, it has been argued that voters suffer from fiscal illusion in
both considering the size of government and analyzing budget deficits. The
first argument is that voters can be led to underestimate the size of govern-
ment expenditures, thereby accepting a government larger than they would
if their perceptions were correct. Hence, fiscal illusion is not simply an
empirical statement about misperceptions about government size, but a
hypothesis about how policymakers may succeed in deceiving voters about
the true size of government. Voters tend to measure the size of government
by their tax bill and policymakers can disguise taxes so that voters underes-
timate the true tax bill.
Parallel to the discussion of fiscal illusion about government expenditures,
there is the argument about fiscal illusion with respect to government deficits.
One explanation of persistent deficits is in terms of misperceptions about
deficits. A classic argument is that individuals favor expenditures, but do not
want to pay for them. Wagner (1976) and Buchanan and Wagner (1977) have
formalized this point in the notion of a deficit illusion, whereby voters do not
understand the government’s intertemporal budget constraint. Faced with
deficit-financed expenditure, voters overestimate the value of the expenditure
and underestimate the future tax burden. Opportunistic incumbents take
advantage of this misperception, running deficits to win the favor of voters.
A third political argument concerns bureaucratic behavior. Niskanen
(1971) argues that the behavior of bureaucrats may be explained by budget
maximization. Bureaucrats try to maximize their budgets since a higher
budget translates into both higher salaries and more power. He views bureau-
cratic behavior as a principal–agent problem with asymmetric information.
In brief, the bureau receives a budget (say, from the legislature) as an increas-
ing, concave function of the output it is perceived as producing. The
bureau’s budget, but not its true output, is observed by the principal (here,
the legislature), which takes the budget itself as a measure of the benefits
from the bureau’s activities. The bureau’s costs are an increasing, convex
function of output, where the cost function is known only to the bureau.
Given the asymmetric information problems, the principal cannot monitor
whether the bureau is efficient in providing services. Nonobservability
allows the bureau to maximize its budget, subject only to the constraint of
 Fiscal Rules from a Political Economy Perspective 19

covering costs. The resultant maximization implies a higher level of output

and a higher budget than would be implied by maximizing net benefits, that
is, by setting marginal benefit equal to marginal cost.
A fourth political argument is that conflict of interest over who should pay
for reducing the deficit (through either tax increases or expenditure cuts for
specific groups), means that deficit reduction may be a long-delayed process.
The basic idea is that deficit reduction is a public good, so that the classic free-
rider problem is present. This argument was presented by Alesina and Drazen
(1991) in a dynamic context. Grilli et al. (1991), among others, offer evidence
that this problem may be especially severe for coalition governments.
A fifth argument stressing political factors is that deficits are used to con-
strain successor governments who may have different spending preferences
(Persson and Svensson 1989; Tabellini and Alesina 1990). To summarize the
argument, policymakers, though partisan, care about social welfare. Only dis-
tortionary taxes are available to finance public spending and to service the
debt, with the level of distortion rising with the tax burden. Hence, the
spending a government would find optimal would depend on the level of
debt (via debt service) existing when it began office: the higher the level of
debt, the lower the desired spending for given preferences. If a government
knew it would be retained in office, utility smoothing would imply no
issuance of debt in nonstochastic models. In contrast, under certain reason-
able parameter configurations, the probability of being replaced leads to debt
issuance in order to reduce the spending by a successor government, with the
higher the debt issuance, the higher the probability of being replaced.
The deficit bias is especially problematic in the use of countercyclical fis-
cal policy. In many OECD countries, there is evidence that fiscal policy has
a positive bias in the sense that expenditures are raised in a recession, but
are not lowered in an expansion to balance the budget over the cycle
(Hercowitz and Strawczynski 2001). The foregoing arguments on the politi-
cal pressures for high government spending, combined with the problems
of cutting spending when the government coffers are full, may help explain
the bias to government spending. Such a bias implies that deficits in reces-
sions, without compensatory surpluses in expansions, risk calling into ques-
tion the government’s long-term commitment to fiscal discipline. We will
return later in the chapter to the issue of the implications of spending when
the government’s commitment to fiscal discipline is unobserved.

On the effectiveness of rules: a general perspective

Much of the research on fiscal rules treats the technical question of how
fiscal rules should be best designed to constrain policymakers and reduce
deficit bias, given the political nature of policymaking and the loss of flexi-
bility that rules imply. This is an important question. Before considering –
in several chapters in this volume – the details of how fiscal rules may best
20 Allan Drazen

constrain policymakers, we need to consider a more general question. How

can legislated rules, especially on outcomes, be effective and make policy
more credible than simply an announced commitment to the same goal?
In order to achieve a specific outcome, in the face of political pressures,
legal restraints that attempt to bind the policymaker, enjoining him to
follow a specific course of action, may seem especially attractive. If a society
wants policy to achieve a specific goal, why not legally require the policy-
maker to achieve that goal? Statutory restrictions can take several general
forms. They can be embodied in the country’s basic set of laws, its constitu-
tion. For example, rules prohibiting issuance of debt to finance current
expenses are embodied in the constitutions of most states in the United
States. On the federal level, the last decade has seen continued debate on a
balanced-budget amendment to the Constitution. In addition, governmen-
tal units on all levels pass laws attempting to regulate their own economic
behavior.5 To take but a few examples, there are restrictions on the financial
or commercial agreements the government is allowed to enter, limits on tax
authorities with the effect of disallowing certain types of tax collection,
or – on a more macroeconomic level – mandates concerning the level of eco-
nomic aggregates, meant to reflect current policy goals or concerns. The fis-
cal authority may be enjoined to achieve full employment; the monetary
authority, to hit an inflation target. These injunctions can take the form of
specific laws (such as a full-employment act) or of directives that have the
force of law. Restrictions can also take the form of unwritten laws, such as
norms or social contracts that also have much of the force of law. Standards
of ethical behavior for public servants may be unwritten, but have a power-
ful effect on their actions if such standards are widely accepted.
Investing a policy with credibility by means of a law directing a policy-
maker to carry out the policy raises a basic question. Why do such laws have
any force? Put another way, what forces a policymaker to obey the law, espe-
cially in cases where breaking the law ex post improves citizens’ welfare? In
the absence of any explicit or implicit restrictions to the contrary, a politi-
cian will renege on a promise if it benefits his constituents to do so. If a law
is passed directing him to fulfill his promise, the benevolent (i.e. welfare-
maximizing) policymaker will similarly be tempted to break the law if so
doing will increase social welfare.6 By considering laws that direct the poli-
cymaker to adopt a rule or norm, are we just moving the problem of time
inconsistency one level higher? What makes an inflation target written into
law any more credible than an announcement of the same inflation rate as
a policy goal? Is there any real, as opposed to semantic, difference?
There are important differences between promises that have no legal back-
ing per se and laws. One primary difference is that laws have penalties
attached to them, so that there are explicit costs to breaking them.7
Similarly, social norms have recognized costs associated with not following
them (Elster 1989, chapter 3). In short, one key to understanding how laws
 Fiscal Rules from a Political Economy Perspective 21

can make policy credible is in understanding the specific mechanisms that

give laws their force, namely, the penalties or costs of breaking the law.8
(There is the obvious follow-up question of what prohibits the government
from changing a law, so that the old law is in fact “broken” legally; we con-
sider this point in detail later.) The key conclusion is that laws (and institu-
tions more generally) enhance credibility to the extent that they raise the cost
and lower the benefit from deviating from a given policy.
This argument may be obvious when laws regulate individual behavior and
the government is the body that punishes the lawbreaker, but it holds equally
well for laws that regulate the government itself, such as fiscal rules. What
exactly are the costs that can be imposed on the government for breaking its
own laws? Rather than addressing this question in the abstract, we can con-
sider the specific case of fiscal rules. First of all, many fiscal rules have both
explicit monitoring of the fiscal authority by some other agency as well as
explicit penalties. In the case of the Maastricht fiscal criteria, significant
divergences of budgetary positions from the medium-term budgetary objec-
tives trigger early warnings, and the Stability and Growth Pact (SGP) spells
out the type and scale of sanctions in the event of persistent excessive deficit,
with the Excessive Deficit Procedure (EDP), as Buti and Giudice mention in
Chapter 7. Australia, Canada, New Zealand, and the United Kingdom have
very similar monitoring and reporting requirements with external auditing
and ministerial responsibility for the results (Craig and Manoel 2002).
Another type of cost is that failure to meet a fiscal policy target triggers
an automatic expenditure cut of some sort. For example, the Gramm–
Rudman–Hollings Deficit Reduction Act of 1985 in the United States, which
set explicit deficit targets over the ensuing five years, legislated an equal cut
in defense and nondefense expenditures to meet the target in any year.
Many other fiscal rules are similar. (Though, in the case of “Gramm-
Rudman,” as it was commonly referred to, or other laws, one may ask what
happens if legislators simply change the law. We return to this later.) In
Chapter 3, Hausmann argues that fiscal rules must be enforced by an open
and politically independent review panel or court with significant sanctions
for violations. Rules without independent enforcement, he suggests, are sim-
ply not credible.
There are other costs as well. Formalizing a directive as a law may also
increase the probability that it is carried out because deviations become
more obvious. That is, fiscal rules cannot force legislators to be fiscally
responsible; however, the rules may significantly increase the public’s aware-
ness of deviations from fiscal responsibility by means of negative publicity.
Gramm–Rudman contained targets intended to reduce the deficit to zero
over five years. The act did not have its desired effect of eliminating the
deficit, in part because when the targets became binding, Congress passed
new legislation to modify the targets. Nonetheless, it is argued that
Gramm–Rudman did have an effect by negatively spotlighting lawmakers
22 Allan Drazen

who introduced “budget-busting” bills. This “negative-spotlight” effect is an

important possible effect of formal fiscal rules.
A skeptic may point to the problem of the creative accounting that fiscal
rules often engender. In many countries, fiscal policy rules dictating specific
numerical targets induce policymakers to use fiscal accounting tricks that
make it appear as if the targets are being met, when in practice they are not
being met at all. Such practices may result in significant distortions, so that
poorly designed rules are likely to backfire, perhaps even causing a greater
distortion than they were meant to address. This effect is an important con-
sideration for the use of fiscal rules and their design. But the use of creative
accounting is not an argument against the existence of a “negative-spotlight”
effect. Just the opposite is true. Legislators or governments use creative
accounting because they perceive that costs are attached to the failure to
meet the fiscal targets. The greater the cost, the more a government may
engage in such creative accounting. That is, the problem of creative account-
ing arises from lack of transparency, not from the political cost of the failure
to meet targets. The visibility that a fiscal rule affords may also reduce
the costs of monitoring compliance. The formalization of a fiscal rule may
create new mechanisms to monitor compliance. These may exist within the
political system, but may also be created outside of the narrowly defined
political system, for example, in the press.

Policy rules versus procedural rules: some observations

There are several reasons to prefer procedural rules over policy rules, in line
with authors who argue that the focus on fiscal rules should shift away from
numerical policy rules to emphasize institutions and the budget process (von
Hagen 1992, von Hagen and Harden 1995, Alesina and Perotti 1996, and
Milesi-Ferretti 1997). The first broad reason is, quite obviously, that a fiscal
rule is more likely to be effective the more it addresses the specific cause of
the problem. This is a key argument in favor of procedural instead of numer-
ical policy rules and is in line with earlier discussion on the political causes
of deficit bias. Procedural rules should be tightened to address the aspect of
the budget procedure responsible for the deficit bias, to the extent that one
can pinpoint this aspect. Of course, pinpointing the cause of the problem is
far easier said than done.
In this respect, numerical policy rules seem like rather clumsy tools to
reduce deficit bias. There is a difficult trade-off in the design of a numerical
policy rule. Too simple or rigid a rule (one with no stated contingencies or
escape clauses) lacks the flexibility that may be necessary in the face of eco-
nomic developments. Hence, over the long term, the rule will be impossible
to satisfy and hence not credible. By the same token, creative accounting fur-
ther reduces its credibility. Thus, it will be ignored or changed. However,
allowing too much flexibility also reduces the credibility of the commitment
 Fiscal Rules from a Political Economy Perspective 23

to fiscal discipline itself. Though state-contingent quantitative rules are

generally preferable from a theoretical point of view, they are not always
workable. First of all, it is impossible to specify all possible contingencies ex
ante. Second, when the information is private, it is often difficult to verify
whether or not the government has reneged on a state-contingent rule. The
difficulty in verification suggests that to be credible, numerical fiscal policy
rules must be simple, though we come back to the problems listed earlier.
The requirement that compliance with a fiscal rule be easily verifiable is
usually labeled transparency of fiscal rules. The problems of too simple rules
notwithstanding, transparency is generally thought to be central for quan-
titative fiscal restrictions to be effective in controlling the growth of deficits.
Among the methods used to thwart the effectiveness of balanced-budget
rules are overoptimistic predictions of key macroeconomic variables, the
strategic use of what is kept on- or off-budget, the measurement of fiscal
adjustment relative to an inflated baseline, and multiyear budgeting, in
which difficult changes are postponed, with the budget’s being revised
before the day of reckoning. Quantitative restrictions may increase the
incentives for creative accounting and hence can actually lead to worse out-
comes in terms of welfare. Another aspect of transparency is that fiscal vari-
ables are even more difficult to measure than inflation. Often there is
disagreement about which measure of the deficit is “correct,” and also, even
when there is agreement about which measure to use, the ease with which
deficit measures may be manipulated makes verification especially difficult.
Milesi-Ferretti (1997) suggests that numerical policy rules can play a posi-
tive role if they are adopted as part of a budget process reform that addresses
procedural problems, especially in the budget formulation stage. Otherwise,
he argues, such rules are not only likely to be ineffective, but also, as argued
earlier, they create incentives for creative accounting and reduce the trans-
parency of the budget process. Therefore, rather than relying on numerical
targets, one may want to concentrate on procedural rules that modify the
structure of the budget process so that it is more difficult to adopt fiscally
irresponsible behavior. Specific reforms include strengthening the position
of the finance minister, limiting the scope for amendments to the budget at
the parliamentary level, and enforcing hard budget constraints at the imple-
mentation stage. The adoption of numerical fiscal rules alone may be coun-
terproductive, diverting attention from the need to change the fiscal
policymaking process itself.
Why then is there often a preference for simple numerical policy rules? A
key reason may be that both policymakers and the public are not convinced
that procedural reforms will yield as unambiguous a discipline as simple
numerical targets. Unless there is a clear link between an aspect of the bud-
geting process and a high deficit outcome (sometimes the case), the argu-
ment that process reform will reduce the deficit even partially may not be
persuasive. Corsetti and Roubini (1996) argue that this is the case in the
24 Allan Drazen

European Union (EU), with Germany, for example, concerned about excessive
deficits of other EU members, having expressed a preference for clear numer-
ical constraints on public debt and deficits.9

Reputation and fiscal rules

There is another way in which a fiscal rule can make the commitment to
fiscal discipline more credible than simply an announced commitment to
deficit reduction. This has to do with the adoption of a rule as a signal in
itself. To understand this, we need to backtrack a bit and consider the notion
of the information content of government actions in the absence of a for-
mal commitment to fiscal discipline. This also helps highlight the relation
between a reputation for fiscal discipline and fiscal rules.
We begin with a brief review of reputation under incomplete information.
Generally speaking, the argument that observing an increase in government
spending reduces our belief in the government’s commitment to fiscal dis-
cipline concerns unobserved characteristics. It is assumed that there is
incomplete information about each player’s “type,” with different types
expected to play in different ways. We then make an inference about what
is unobserved (the type) on the basis of observed past actions, so that the
expectation of future behavior is based on what has been observed in the
past. “Reputation” is summarized by the public’s belief about the govern-
ment’s type, where reputation depends on observed past actions.
We can thus identify two types of policymakers: those committed and those
less committed to fiscal discipline. If commitment is not mandated by law, the
public uses observations of government spending and deficits to form infer-
ences on type, with lower deficits leading the public to update the probabil-
ity it assigns to the government’s being committed to fiscal discipline.
Therefore, actions showing a commitment to fiscal discipline may substitute
for fiscal rules in making credible the commitment to deficit reduction.
There are, however, several ways in which this basic argument must be
amended. First, in applying the notion of reputation to fiscal policymaking,
the key point is that there is no one policymaker, but a whole political process
that generates outcomes. Hence, observing the fiscal outcome gives infor-
mation on the nature of the political budgetary process as described earlier.
“Type” thus refers to the characteristics of the fiscal process, including the
strength of the political forces in pushing for higher spending and higher
deficits.
Second, once we realize that the budget process reflects the interactions
of a large number of political actors, a signal of commitment to fiscal disci-
pline may be important not so much to an outside observer, but to partici-
pants in the process itself. To take a homespun analogy, when I buy or
refrain from buying something at a store, the important signal may not be
of my own commitment to fiscal discipline. It may be an indication to my
 Fiscal Rules from a Political Economy Perspective 25

two-and-a-half-year-old daughter of the unlikelihood that she can succeed

in inducing me to buy something for her. Anything that signals fiscal disci-
pline in the process may thus be self-reinforcing.
What then is the role of fiscal rules for reputation? A fiscal rule may serve
not as an alternative to actions that build reputation, but as one of those
actions in itself. When a single policymaker chooses a policy, his willingness
to adopt a rule conveys information about the preferences of the policymaker
himself. Hence, adoption of a rule can make the commitment to fiscal disci-
pline more credible not because it imposes constraints on a policymaker with
a known incentive to fiscal profligacy, but because it signals commitment.10
When fiscal outcomes reflect a political process with many actors, a fiscal
rule that signals the willingness of those involved in policymaking to limit
their own flexibility may similarly convey information about their prefer-
ences. However, unlike the case of a single policymaker, it may also convey
information about the budget process itself. The ability of politicians them-
selves to legislate the fiscal restriction may provide information consistent
with the expectation that future fiscal behavior will be more disciplined.
This leads again to the issue in the background. The fact that the legisla-
tors who are the forces pushing for increased spending are also the ones who
must credibly commit themselves may weaken the expectation that the law
will actually be carried out. Even if a fiscal discipline law is passed, what pro-
hibits the government from changing a law when it perceives that it is opti-
mal to do so? We now turn to this issue.

Credible commitment to unchanging fiscal rules

As indicated earlier, in the United States the Gramm–Rudman–Hollings

Deficit Reduction Act failed to eliminate the budget deficit because when the
targets became binding, the US Congress passed new legislation to modify
the targets. In Israel, the Deficit Reduction Law of 1991 has similarly seen
the target modified numerous times since its adoption. Every new govern-
ment that has been elected since the law’s adoption has, in fact, changed the
targets, even in the very first budget it submitted. Such experiences should
perhaps not be surprising, since one way to break a law legally is simply to
change the law, when the government perceives it is advantageous to do so.
If a fiscal rule is continually being changed, however, it cannot really be con-
sidered a rule.
On the basis of experiences such as these, Kopits and Symansky (1998,
p. 2) argue that to be considered a fiscal rule, a restriction, by definition,
must be “intended for application on a permanent basis by successive
governments.” If a rule is followed by successive governments, then it
certainly may be seen as a credible commitment to fiscal discipline,11 while
one that is regularly revised is not really a rule. However, this is only a part
of the story, for one may ask what makes credible an announcement that
26 Allan Drazen

successor governments will be bound by the rule? More generally, how will
the public believe that existing laws will not be changed whenever current
circumstances make it convenient to do so? To answer this, we detour into
what are known as “constitutional laws” and, more generally, the issue of
constitutionalism.
Constitutional laws have at least one of four characteristics: they concern
restrictions on the government’s use of authority; they concern the process
of policymaking; they have more stringent amendment procedures than
other laws; they often treat issues that are seen as fundamental in a deeper
conceptual sense, that is, basic rights or liberties. Fiscal rules may certainly
be seen as constitutional in all of the first three senses. As we shall see,
constitutional legislation to achieve fiscal discipline may satisfy the fourth
characteristic as well.
The third and fourth characteristics of constitutional laws are key to
making it credible that the law will not be changed whenever circumstances
make it tempting. The third characteristic, the use of more stringent amend-
ment procedures, may be thought of as a concrete rather than conceptual
approach to engendering such an expectation. Procedures that make it
difficult to amend the constitution include supermajorities, waiting periods,
confirmation, referendums, and (in federal systems) ratification by subna-
tional governments. Such restrictions are seen as protecting the electorate
against itself, in that a majority may act under the influence of a “momen-
tary passion” (Elster 1995). In short, a promise to follow through on a cer-
tain action may be made credible by enacting it into a law which itself is
difficult to undo, that is, by “constitutionalizing” it. Effective commitment
follows from the extreme difficulty in changing a law once it is given constitutional
status. This is analogous to the key result in the earlier discussion on what
gives force to a legal restriction. In both cases, raising the cost of deviating
from the law (or changing it) makes it more credible that the law will be fol-
lowed (or not changed).
The fourth characteristic, the fundamental nature of the issue the law
addresses, is a more conceptual approach to engendering the expectation that
the law will not be changed whenever it is deemed convenient. A funda-
mental right, by definition, is one that is seen as having more permanence
than an ordinary piece of legislation. Hence, giving a provision constitutional
status is meant to give it a permanence it would not otherwise have. In fact,
constitutional fiscal rules may send the signal that fiscal discipline is a pre-
eminent goal of society, hence relating to the fourth constitutional charac-
teristic. While zero budget deficits do not confer a fundamental right on a par
with freedom of speech, a balanced-budget restriction in the constitution
sends the signal that a society attaches fundamental importance to it.12
However, since constitutional laws are meant as an extreme form of com-
mitment (and hence loss of flexibility), this solution should be used for fis-
cal restraint when other solutions have been tried, but have repeatedly failed.
 Fiscal Rules from a Political Economy Perspective 27

This discussion explains why in practice numerical policy rules are often
changed. To the extent that lawmakers see little cost to changing a fiscal
target, it will be changed. Perhaps somewhat paradoxically, in the cases of
the legislation in the United States and Israel, lawmakers apparently placed
a lower cost on changing the law than on breaking it. In part, this may be
because the cost of breaking the law was only the resulting negative public-
ity (the “negative-spotlight” effect). It may be that not meeting the target at
a well-specified time would have been more damaging than changing the
target as part of the budget proposal, but this is a conjecture. Furthermore,
in the United States, commentators have pointed out that Gramm–Rudman
may have been unconstitutional, in that read literally, the law allowed one
congress to restrict the lawmaking power of a successive congress. In that
sense, for each congress to choose its deficit target may have been viewed as
more acceptable. Clearly, a balanced-budget amendment to the Constitution,
for all its drawbacks (and there are many), would not suffer from the prob-
lem of being so easily changed.
The ease with which laws with numerical targets were changed in these
cases seems to reflect unenthusiastic acceptance of a numerical target law to
begin with. The lack of flexibility inherent in a zero-deficit law in the face
of changing economic circumstances may mean that the public fully expects
the law to be changed as circumstances change, and accepts this procedure.
Conversely, allowing for contingencies in the rule makes it less necessary to
be changed, but as argued earlier, this introduces the problems of credibility
and effectiveness of the rule.
The Maastricht fiscal reference values remain unchanged even though
many EU member countries have not found them easy to comply with. One
reason is that contingencies are allowed for.13 Perhaps more important, these
policy rules are seen as a crucial part of the process of integration and as a
condition for successful monetary union; hence, the importance attached to
the larger goals that fiscal rules are meant to advance ensures that they will
not be easily changed. In addition, the multicountry nature of the policy
authority (invested in the European Union Council of Economy and Finance
Ministers (ECOFIN) ) provides an enforcement mechanism across govern-
ments, backed by a treaty obligation (Maastricht) not present in the case of
fiscal rules for a single government.

Conclusions

This chapter highlights a number of conceptual and practical issues in the

political economy of fiscal rules. It is not meant as a prescription for what
kind of rules are optimal (though some conclusions can nonetheless
be drawn), nor as a user’s guide to fiscal rules. It is meant to help us think
more clearly about some basic issues involved in the use of fiscal rules. Here
we find insight into why rules often do not work: they do not address the
28 Allan Drazen

cause of deficit bias, they attach no real costs to deviating from the rules, or
to changing the rules. It has also suggested some different kinds of costs that
may help enforce rules. Both perspectives should be helpful in thinking how
rules should be designed.
Although much of the existing literature on the rationale for fiscal rules
refers to the experience of advanced economies, most of the arguments
presented in this chapter are just as valid for emerging markets. In fact,
several of the political economy issues arise even more forcefully in emerg-
ing market economies, where the political process and the budget institu-
tions are relatively more fragile and still evolving. Furthermore, it is well
documented that much of Latin America is vulnerable to significant eco-
nomic and financial volatility, in part associated with sharp fluctuations in
the terms of trade – discussed in Chapter 3 by Hausmann. The confluence
of these forces – against the background of strong demand for public goods
and services and a relatively narrow tax base – has exposed many of these
economies to pressures for time-inconsistent fiscal behavior, reflected in
both a procyclical fiscal stance and a deficit bias. For some of these
economies, the buildup of public debt has contributed to significant
increases in the cost of financing and even jeopardized market access.
Therefore, in recent years, fiscal policy rules have become particularly
appealing for these countries in their effort to withstand pressures and to
reverse the rise in public indebtedness.

Notes

1. I wish to thank George Kopits, Andrés Conesa, and conference participants for
useful comments. Financial support from the Yael Chair in Comparative
Economics, Tel Aviv University, is gratefully acknowledged.
2. Kopits and Symansky (1998) provide an excellent in-depth discussion of fiscal
policy rules. These are sometimes termed “numerical rules,” but as Kopits and
Symansky point out, restrictions on fiscal procedure may have a distinct numeri-
cal character as well.
3. See Drazen (2000, chapter 14), or Alesina and Perotti (1996), for a summary of the
literature on the budget process from a political economy perspective.
4. See Drazen (2000, chapter 7), for a discussion of opportunistic political business
cycles, in which incumbents manipulate economic variables before elections in
order to influence election outcomes. A summary of the evidence and references
can be found in Drazen (2001). On the other hand, there is also the argument that
voters penalize fiscal excess at the polls. Niskanen (1975) analyzes US presidential
elections from 1896 to 1972 and finds that, holding macroeconomic performance
constant, increases in federal spending imply lower vote totals for the incumbent
party. Similarly, Peltzman (1992) presents econometric evidence for this effect on
several levels of government.
5. A problem that arises is how to interpret the constraining power of laws that give
only qualitative targets, or whose escape clauses imply that the law imposes few if
any effective constraints.
 Fiscal Rules from a Political Economy Perspective 29

6. As discussed, a utilitarian policymaker is equally tempted to break the law to

enhance his reelection prospects.
7. To the extent that there is significant leeway in the application of the penalty, as
is often the case, the law clearly loses much or all of its force. Hence, in discussing
the role of penalties in enforcing good behavior, we assume penalties that are
unambiguous, and unambiguously applied.
8. It is beside the point to argue that with a harsh enough penalty for disobedience,
the law will certainly be obeyed. If the penalty for disobedience is so harsh that
it implies that the law will be obeyed no matter what, then it is not credible that
the penalty will be applied in all circumstances. Hence, harsh penalties simply
shift the credibility problem from whether the law as written will be obeyed, to
whether it will actually be enforced. That is, since individuals know that such a
harsh reaction is not optimal ex post, the punishment itself is not credible. If laws
make policies credible only to the extent that the penalties that enforce the laws
are themselves credible, then enhancing credibility depends on choosing the opti-
mum structure of penalties to do this. McCallum (1995) makes a similar point in
criticizing some of the work on institutional solutions to the time-inconsistency
problem in monetary policy. He argues that some of the proposed solutions do
not solve the problem: they merely relocate it, in that it is not clear why the insti-
tutions are themselves credible.
9. They also suggest that in the EU, the need to balance the economic integration,
which was crucial for monetary unification on the one hand and the principle of
political sovereignty of member countries on the other, made numerical targets
more politically acceptable than directives for similar procedural rules across
countries.
10. This distinction mirrors the discussion of reputation under incomplete informa-
tion versus complete information in Drazen (2000, chapter 6).
11. One can also argue that even a restriction meant to bind only a single govern-
ment can be thought of as a fiscal rule if it is credibly believed to really bind the
government and not be changed over the period of time it is meant to be effec-
tive. For example, governments and organizations facing large budget shortfalls
put in place severe expenditure restrictions meant to remain in force until the
budget situation improves. These are often rigidly enforced and credible. They are
fiscal rules, their relatively short horizon notwithstanding.
12. The third and fourth characteristics are connected. If a law is seen as fundamen-
tal, it is natural that it cannot be changed through the ordinary legislative process,
but requires a more stringent procedure. Many argue (e.g. Elster 1995) that this
connection is the heart of constitutionalism.
13. Another is the recourse to procedural loopholes, such as the recent delays in con-
vergence for certain members, approved by ECOFIN, acting as the ultimate
authority for enforcing the SGP.
3
Good Credit Ratios, Bad Credit
Ratings: The Role of Debt Structure
Ricardo Hausmann1

Introduction

Many emerging market economies suffer from bad credit ratings, limited
access to finance, and large and unstable risk premia. Policymakers usually
attribute this result to weak fiscal policies: countries run large deficits and
accumulate debt that puts them on the brink of insolvency. The policy rec-
ommendations that emerge in these circumstances are clear. Countries must
tighten fiscal policies so as to reduce the debt–GDP ratio gradually. Budget
institutions must be transformed in order to meet credibly the intertempo-
ral budget constraint. In due course, as the debt ratio falls and as the mar-
kets see a consistently improving pattern, credit ratings will improve and
access to capital markets will become less expensive and more secure.
This conventional view overlooks an important fact: what distinguishes
highly rated countries from those with low credit ratings is not necessarily
their public indebtedness. Emerging markets often exhibit public debt ratios
lower than those in industrial countries. Yet their credit ratings are drasti-
cally worse. As shown in Figure 3.1, the debt–GDP ratio is a poor guide to
perceptions of creditworthiness as measured by credit ratings across coun-
tries.2 For example, the public debt ratios in Belgium, Italy, and Canada are
similar to the ratios in India, Pakistan, and Jordan. Also Japan, the United
States, the United Kingdom, and Spain exhibit debt ratios similar to those in
Argentina, Brazil, Mexico, and Turkey. However, the industrial and the
developing countries received radically different ratings.
One potential explanation for the lack of systematic correlation between
credit ratings and debt–GDP ratios is that GDP is not a reliable measure of a
government’s capacity to service its debt. Debt service is paid out of the por-
tion of GDP the government can appropriate from the effective tax base. In
developing countries, the average tax–GDP ratio is 26 percent, while indus-
trial countries average 43 percent (Table 3.1). Since developing countries
have a smaller tax base, they can support a smaller public debt. Hence, a bet-
ter metric for debt service capacity is the debt–tax revenue ratio, which

30
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Good Credit Ratios, Bad Credit Ratings 31

19 NOR JPN GBR AUT, DEU, USA

SWE DNK CAN BEL

FIN AUS ESP
ITA
PRT

CYP
ISL
Credit rating (foreign currency)

SVN
CZE
ISR

EST
CHN GRC
LVA
TUN HUN
POL
TTO
PAN
IND
MEX
CRI ARG MAR

DOM BRA JOR

PRY TUR
5 PAK
0
–0.291965 1.13803
Debt /GDP

Figure 3.1 Credit rating versus net debt–GDP ratio

Table 3.1 Selected regions: public debt (unweighted averages), 1993–2000

Public debt Tax ratio

(% of GDP)
(% of GDP) (% of tax
revenue)

OECD 53 128 43
Developing and
transition countries 45 177 26
Latin America 35 143 25
Eastern Europe 95 301 34
Middle East
and North Africa 66 200 32
South Asia 78 388 22

Sources: IMF and author’s calculations.

stands at 177 percent in developing countries, as compared to 128 percent

in industrial countries. However, this difference provides only a partial
explanation for the large difference in credit ratings, as revealed by a scatter
plot of the debt–tax revenue ratio against the ratings of 51 countries aver-
aged over the 1990s (Figure 3.2). For example, the debt–revenue ratio is
32 Ricardo Hausmann

19 NOR AUT, GBR, DEU, JPN USA

SWE DNK CAN BEL

FIN AUS ESP
ITA
PRT

CYP
ISL
Credit rating (foreign currency)

SVN
CZE
ISR

EST
CHN GRC
LVA
HUN TUN
POL
TTO
PAN
IND

MEX
CRI, ARG MAR

DOM BRA JOR

PRY TUR
5 PAK
0
–0.558143 5.41445
Debt /revenue

Figure 3.2 Credit rating versus debt–revenue ratio

similar in Paraguay and Denmark, in Brazil and Spain, and in Mexico and
the United States, but their credit ratings are very different.
A regression of credit rating on measures of fiscal fundamentals, control-
ling for the level of development, as measured by the log of GDP per capita,
shows that neither the debt–GDP ratio nor the debt–revenue ratio is signif-
icantly correlated with the rating (Table 3.2). In fact, the coefficient on the
debt–GDP ratio has the wrong sign while that of the debt–revenue ratio is
very close to zero.
These results suggest that something else determines the perception of
risk, beyond the debt ratio and the level of development. This chapter con-
tends that debt structure plays an important role. In particular, it argues that
a crucial question is whether countries can borrow abroad in domestic cur-
rency. If they cannot, external debt service will be impacted by movements
in the real exchange rate, which are substantially more volatile than real out-
put and move in such a way as to make debt service even harder in bad
times. In addition, many countries are unable to borrow at long maturities
and fixed interest rates in the domestic market. This makes domestic debt
service sensitive to movements in the real short-term interest rate, which is
also volatile and countercyclical, making debt service even more difficult.
The key point is that debt structure affects the volatility of the debt ser-
vice burden relative to the capacity to pay. The greater the volatility is, the
Table 3.2 Cross-country estimates of credit rating equation, 1990–99

Independent variable (1) (2) (3) (4) (5) (6)

Log GDP per capita 3.380 3.352 3.329 2.981 3.231 3.193
(13.79)** (12.07)** (12.41)** (10.46)** (10.07)** (8.40)**
Debt–GDP
0.987 — — —
0.113 —
(1.06) — — — (0.10) —
Debt–revenue — 0.042 — — —
0.077
— (0.13) — — — (0.21)
SD revenue growth — —
14.241 — — —
— — (2.51)* — — —
SD terms of trade — — —
17.445 — —
— — — (3.04)** — —
SD GDP growth — — —
38.290 — —
— — — (2.11)* — —
Original sin — — — —
2.923
2.918
— — — — (2.70)* (2.71)*
Constant
17.083
17.097
16.286
11.527
14.015
13.613
(7.60)** (6.32)** (6.21)** (3.78)** (4.09)** (3.34)**
Observations 59 51 57 55 35 35
R2 0.79 0.80 0.82 0.85 0.83 0.83

Note: Dependent variable is the credit rating. Absolute value of t statistics is shown in parentheses (* significant at 5 percent; ** signifi-
cant at 1 percent).
Sources: IMF and author’s calculations.

33
34 Ricardo Hausmann

greater the value at risk, and consequently the higher the risk premium. This
creates a positive feedback between the interest rate and the value at risk.
One implication is that a country’s debt-carrying capacity depends on the
structure of its debt and the variances and covariances of the real exchange
rate, real interest rate, and real income. This may explain why some coun-
tries get into trouble with the same debt ratio at which other countries
are rated very favorably.
From a policy perspective, managing the debt structure may reduce risk
premia and allow rapid fiscal consolidation through a self-reinforcing reduc-
tion in interest rates. This implies that governments should be concerned
not only with the debt stock, but also with its riskiness.
Our results are in line with other works in the recent literature. Reinhart
(2002) finds that 84 percent of the debt crises were preceded by currency
crises. Catão and Sutton (2002) report that measures of terms of trade and
policy volatility can predict the likelihood of debt crises. Manasse et al.
(2002) find that measures of illiquidity, such as short-term debt or interna-
tional reserves as indicators of vulnerability to debt crises, help predict fiscal
crises. However, Detragiache and Spilimbergo (2001) report that the rela-
tionship between liquidity measures and crises is plagued with problems of
reverse causality, as countries that get into trouble lose the capacity to bor-
row, especially at longer maturities, and deplete their reserves before a crisis.
To begin this chapter, a simple model of fiscal risk is constructed to derive
implications for country risk. The next section examines the effect of GDP
and tax revenue volatility on credit ratings, and finds that while the effects
are significant, they are far from the whole story. This is followed by the
implications of debt denomination as a determinant of what types of volatil-
ity are relevant. Dollar-denominated debt is discussed in light of the real
exchange rate, its volatility and cyclical properties as relevant sources of risk.
Then we assess the role of short-term debt in making debt service contingent
on the volatility and cyclical properties of the short-term real interest rate.
The following section simulates the joint impact of tax revenue, real
exchange rate, and real interest rate volatility on an average OECD and an
average Latin American country, to highlight the role played by debt denom-
ination in affecting the risks associated with debt service. Finally, we derive
some policy implications.

A simple model of fiscal risk

To determine the relationship between risk and interest rates, it is useful to

construct a simple model of fiscal risk. Suppose a country has a public debt
service burden as a share of government revenue, labeled x:

iD
x= , (3.1)


 Good Credit Ratios, Bad Credit Ratings 35

where i and D are respectively the interest rate and the stock of government
debt, Y is GDP and
is the effective tax rate – subject to the risk that the debt
will not be repaid in full. Suppose for simplicity that the government will
repay its debt in full provided that the debt service ratio x is not larger than
some maximum value x. If it were larger, then the government would sim-
ply default on the total amount forever.3 This assumption is consistent with
standard sovereign risk models, in which the government wants to maintain
the net present value of government spending, and avoid taxes and the con-
sequences of default, in terms of the loss of access to credit or other costs.4
If the benefits of default, in terms of reduced debt service, rise faster with
debt than the costs, then there will be an optimal point to default, namely,
when x  x.
Investors are risk neutral and hence require the expected income from
holding government debt to equal the risk-free rate . However, government
debt pays according to the following schedule:

iD if x  x, and 0 if x  x .

In order for investors to earn the risk-free rate , the contractual interest rate
must be


i  . (3.2)
Prob (x < x )

What are the determinants of the probability that x  x? Figure 3.3 shows
two stylized probability distributions of x that differ in volatility, but have
the same expected value. The risk premium must cover the value at risk, that
is, in situations where x x. The narrow distribution has low volatility and
a negligible or no value at risk (VaR). The fatter distribution has a significant
part of the distribution in which x x. The contractual interest rate has to
be higher for the fatter distribution, even though the expected value of debt
service is the same as for the narrow distribution. If two economies have
the same expected debt service ratio x, but are exposed to different volatili-
ties, then the more volatile economy must have a lower debt stock and a
smaller debt–revenue ratio to compensate for the higher interest rate.
However, as the contractual interest rate increases, x rises.
Equations (3.1) and (3.2) are depicted in Figure 3.4. The model is solved
in the x versus i space. The vertical line represents the locus of points in
which the interest rate i is equal to the risk-free rate . The horizontal line
expresses the points at which x  x. The ray from the origin is equation (3.1);
it traces x as a function of i, with a slope equal to the debt–revenue ratio
D/
Y. We draw three such rays at different D/
Y ratios.
The hyperbola represents equation (3.2). For low expected values of x, the
probability that x  x is essentially zero and hence i is very close to the
36 Ricardo Hausmann

Expected value x–

Prob.

Figure 3.3 Volatility and VaR in debt service

Bankrupt
–
x

Fragile i=
Pr (x < x–)

x
D
x= i

Y

Sound
i* i

Figure 3.4 Illustration of debt service dynamics

risk-free rate . For high expected values of x (drawn from a probability dis-
tribution) the probability that x  x increases and hence the interest rate
must be higher. Obviously, at no point can x  x, since in that case the gov-
ernment pays nothing.
Equilibrium is determined where the ray crosses the hyperbola. As shown,
at a low D/
Y ratio, the equilibrium is very close to the riskless rate. At higher
ratios the ray crosses the hyperbola twice. This does not mean that there are
multiple equilibria, as the second intersection is unstable: small increases in
i cause a rise in x that is larger than what would be consistent with equa-
tion (3.2), causing thus an even larger increase in i until the government
becomes insolvent. By contrast, in the first intersection, small increases in
i lead to increments in x that are lower than would be consistent with equa-
tion (3.2), causing the interest rate to fall back. Finally, there is an even
higher D/
Y ratio in which the ray does not cross the hyperbola: there is no
interest rate at which the expected return is : the country is bankrupt.5
 Good Credit Ratios, Bad Credit Ratings 37

x–
Low variance

High variance

Figure 3.5 Illustration of the role of volatility

Table 3.3 Selected regions: volatility of tax revenue, GDP, and terms of trade (in
percent), 1990–99

Tax revenue GDP Tax revenue GDP Change in

growth growth growth growth terms of
trade
(in local currency) (in US dollars)

Industrial countries (20) 3.6 2.0 12.6 12.1 4.4

Developing countries (56) 12.6 4.8 18.4 14.0 11.6
Latin America (20) 11.8 4.6 17.9 13.8 10.7
Other countries (36) 13.0 4.9 18.7 14.1 12.0

Note: Standard deviation for period 1980–99. Data in local currency are shown at constant prices,
and data in US dollars at market prices.
Sources: IMF and author’s calculations.

Figure 3.5 analyzes the impact of volatility with two alternative hyperbo-
las. An increase in volatility implies a southeastern shift in the hyperbola.
For a given D/
Y ratio, higher volatility leads to bankruptcy, while at lower
volatility the country has market access at a reasonable interest rate.
Therefore, volatility may be a major factor in explaining why countries with
the same debt–revenue ratio have very different risk profiles.

Volatility of tax revenue and fiscal risk

A major source of volatility lies in the revenue base, which in turn depends
on overall economic stability. In the 1990s, the volatility in real growth of
government revenue was four times higher in developing than in industrial
countries, explained largely by the relatively high volatilities in GDP growth
and in the terms of trade (Table 3.3).6
38 Ricardo Hausmann

Table 3.4 OECD and Latin America: impact of a shock in tax revenue and GDP
(in percent)

OECD Latin America

Initial value Shock New value Initial value Shock New value

Real interest rate

(in US$) 4.0 — 4.0 8.0 — 8.0
Tax revenue–
GDP ratio 40.0
3.6 38.6 25.0
11.8 22.1
Debt–GDP ratio 50.0 — 51.9 35.0 — 39.7
Debt service
revenue ratio 5.0 0.4 5.4 11.2 3.2 14.4

Note: Impact of one standard deviation change in tax revenue and GDP for the average economy
in each region.
Sources: IMF and author’s calculations.

Can higher volatility in tax revenues fully explain the differential in credit
ratings? This question can be answered with a set of regressions of ratings
on either revenue volatility or its more fundamental determinants, such as
terms of trade and output volatility. In both equations we control GDP per
capita as a proxy for institutional and economic conditions for develop-
ment. In both cases, the estimates show a statistically significant relation-
ship between measures of volatility and ratings, even though the effects are
not very large (columns 3 and 4, Table 3.2). For example, while the differ-
ence in ratings between industrial and developing countries in our sample
is about nine increments, the elasticity estimate implies that the difference
in revenue volatility can account for slightly less than one increment (col-
umn 3), and the difference in GDP growth and terms of trade volatility
together can account for about 1.8 increments (column 4).
To illustrate why the effects might be limited, let us simulate a one
standard deviation shock to tax revenue and GDP in an average OECD coun-
try and in an average Latin American country (Table 3.4), differentiated in
terms of their debt–GDP ratio, tax–GDP ratio and real interest rate. The
greater tax–GDP shock, the higher interest rate, and the smaller tax base in
Latin America than in the OECD generate an impact on x that is eight times
larger in the former than in the latter. However, this effect only increases the
debt service burden by 3.2 percent of the government budget in Latin
America – not an impressive amount. Therefore, revenue volatility is a small
part of the explanation for the difference in credit ratings.

Original sin, liability dollarization, and real

exchange rate volatility

When emerging market economies borrow abroad, they have no choice but
to do so in foreign currency. This phenomenon, called original sin by
 Good Credit Ratios, Bad Credit Ratings 39

Eichengreen and Hausmann (1999), has important implications for macro-

economic stability and financial fragility in these economies. If a country
with a sizable net foreign debt exposure suffers from this sin, then real
exchange rate movements will have aggregate wealth effects. This will tend
to render depreciations contractionary and monetary policy less effective. In
these circumstances, monetary policy is usually subject to “fear of floating,”
and the attempt to avoid exchange rate volatility makes domestic interest
rates more volatile.
There are several hypotheses advanced to explain this phenomenon.
Jeanne (2003) emphasizes the role of monetary policy credibility. Tirole
(2002) discusses commitment problems in protecting the property rights of
foreigners. Corsetti and Mancowizk (2002) study the role of weak fiscal sol-
vency. Chamon (2001) and Aghion et al. (2001) stress that a positive corre-
lation between depreciation and default risk creates incentives for borrowers
to increase their foreign currency debt after they have secured a given
amount of domestic debt. In anticipation of this process, the domestic debt
market may disappear. Chamon and Hausmann (2002) analyze the strategic
interaction between debt denomination and monetary policy: risk-averse
borrowers have an incentive to issue debt in a variable that the central bank
is trying to stabilize.
However, Hausmann and Panizza (2003) find little empirical support for
any of these hypotheses. In fact, they show that 98.5 percent of cross-border
lending takes place in seven currencies. Original sin seems to be robustly
related to the relative size of a country in the world economy and to acci-
dents that may have converted it into an international financial center
(e.g. Switzerland and Luxembourg), consistent with the role of liquidity in
generating economies of scale in currency denomination. While countries
with weak institutions and poor policy credibility suffer from original sin, so
do other, well-behaved small countries. Thus, we can view original sin not
as one more consequence of loose fiscal performance, but as a relatively
exogenous determinant.
To understand the implications of original sin for the volatility of x it is
useful to expand equation (3.1) in order to split the debt into three compo-
nents: long-term fixed-rate domestic currency debt (l); short-term domestic
currency debt (s); and long-term fixed-rate foreign currency debt (f ),7 so that

ilt
1 Dlt
1  ist Dst
1  ift
1 et Dft
1

xt  . (3.3)

Yt

At time t
1, while the cost of servicing long-term fixed-rate domestic cur-

rency debt is fully known, the interest rate on the short-term domestic cur-
rency debt is uncertain and the exchange rate on long-term, foreign currency
debt is unknown. Uncertainty about is and e will affect the volatility of x.
Consider first the impact of dollar-denominated liabilities. If debt is
denominated in US dollars, then what matters is not the volatility of GDP
40 Ricardo Hausmann

or tax revenue growth measured in constant domestic prices, but instead,

the volatility of these variables measured in US dollars. The volatilities of the
growth of tax revenue or of GDP measured in US dollars at market prices
are much larger than in local currency at constant prices (Table 3.3). These
volatilities are higher in developing countries than in industrial countries.
However, the critical point is that, while for industrial countries that can
borrow in local currency the relevant volatility is measured in local currency,
for most Latin American countries where the bulk of the debt is denomi-
nated in foreign currency, the latter is the relevant volatility measure.
In this regard, the yearly volatility of the real exchange rate in Latin
America during the 1990s averaged over 21 percent, more than double than
in the United States (Table 3.5). But in the United States, the debt is denom-
inated in domestic currency, so changes in the real exchange rate do not
affect debt service; by contrast, in Latin America, external public debt aver-
aged 30 percent of GDP.8 This means that in Latin America a real exchange
rate depreciation of one standard deviation would, on average, increase the
debt–GDP ratio by 6 percentage points and the interest burden on that debt
by 21 percent.
Moreover, if the currency tends to depreciate in bad times, then the capac-
ity to service dollar-denominated debts will decline at the worst possible
moment. A list of years since 1972 in which the US dollar value of the GDP
fell by more than 25 percent in a Latin American region – which not sur-
prisingly coincides with major financial crises in the region – shows that the
average decline in US dollars is nearly 40 percent, though the fall in local cur-
rency is only about 2 percent (Table 3.6). Clearly, the decline in the capacity
to service dollar-denominated debts is severely impaired by the volatility in
real exchange rates, in real GDP and by their comovement. This point is also
illustrated with a fixed-effects panel regression separately for a sample of 20
Latin American and 20 industrial countries, which indicate that on average
the real exchange rate appreciates by 0.37 percent when output increases by
1 percent; the pattern is not observed in industrial countries (Table 3.7).
Hence, liability dollarization makes the real exchange rate a relevant variable
in the determination of the debt service burden in Latin America. Its high
volatility and its comovement with the business cycle make the burden of
debt service more uncertain and harder to bear in bad times.
Whether original sin is robustly correlated with measures of country risk
constitutes a valid question. This can be answered with estimates of a credit
rating equation including a measure of original sin.9 The latter is calculated
as the ratio of the debt placed abroad in domestic currency to total debt
issued abroad by the country’s residents. Zero implies that there is no inter-
nationally traded debt in domestic currency, that is, full original sin. The
estimates show that while the debt ratios have no power to explain the cross-
country variation in ratings, original sin is strongly correlated with ratings,
after controlling for the level of development. The coefficient implies that
Table 3.5 Latin America and United States: volatility and cyclical properties of exchange rates and interest rates, 1990–99

Country Real exchange rate Real interest rate External public debt
(% GDP)
Elasticity to imports Standard deviation Elasticity to imports Standard deviation

United States
0.03 9.1
3.25 0.9 —

(
0.9) (
4.1)
Latin America 0.18 21.4
126.26 10.5 31
(8.9) (
10.9)
Argentina 0.02 21.1
221.87 4.0 30
(1.2) (
10.3)
Brazil 0.42 18.8
451.64 17.2 16
(10.4) (
3.4)
Chile 0.32 16.0
8.76 5.4 8
(4.5) (
1.0)
Colombia 0.26 19.3
16.63 7.8 25
(6.2) (
2.3)
Costa Rica — —
19.72 5.0 —
(
5.0)

41
 42
Table 3.5 Continued

Country Real exchange rate Real interest rate External public debt
(% GDP)
Elasticity to imports Standard deviation Elasticity to imports Standard deviation

Ecuador 0.27 25.7
2.35 12.2 81

(3.7) (
0.5)
Mexico 0.61 18.3
73.31 23.0 15
(10.7) (
13.2)
Panama — —
0.43 0.06 —
(
0.6)
Peru 0.15 28.4
151.40 11.2 43
(3.1) (
1.7)
Uruguay — — 2.59 11.8 —
(0.4)
Venezuela
1.04 23.6 0.08 17.6 28
(
7.7) (0.0)

Note: Based on monthly data, excluding periods in which inflation rate exceeded 40 percent. Absolute value of t statistics is shown in parentheses.
For the real exchange rate, annual data; for the real interest rate, monthly data: For external public debt, end-2000 data.
Sources: IMF and author’s calculations.
 Good Credit Ratios, Bad Credit Ratings 43

Table 3.6 Latin America: annual decline in GDP growth (percent)

Country Year Tax revenue GDP GDP

(in local
(in US dollars) currency)

Average
27.7
39.6
2.3

Bolivia 1986
23.1
28.2
2.6
Chile 1975 —
58.2
11.4
1982 26.4
28.8
10.3
Costa Rica 1981
14.0
50.8
2.3
Dominican Rep. 1985
36.8
52.4 1.0
Ecuador 1986
44.3
31.2 3.1
1999
20.7
31.8
7.3
El Salvador 1986
30.6
30.8 0.2
Guatemala 1986
4.4
26.6 0.1
Guyana 1987
48.1
35.5 0.9
Honduras 1990
35.4
43.0 0.1
Jamaica 1984
26.0
37.1
0.9
1976
39.6
40.6 4.4
Mexico 1982
45.4
49.3
0.6
1986
32.6
30.7
3.8
1995
31.9
33.3
6.2
Paraguay 1989
1.7
30.3 5.8
Suriname 1994 —
95.3
2.3
Trinidad and Tobago 1986
45.5
36.7
3.3
Uruguay 1983 —
56.0
10.3
1984 —
25.8
1.1
Venezuela 1984 —
29.3 1.4
1989
9.6
30.2
8.6

Note: Decline of more than 25 percent in GDP, measured in US dollars. Data in local currency are
shown at constant prices, and data in US dollars at market prices.
Sources: IMF and author’s calculations.

the presence of original sin can account for up to three increments in rat-
ings (columns 5 and 6, Table 3.2).

Effects of short-term domestic currency debt

Many countries that cannot borrow in local currency abroad also cannot bor-
row at home in domestic currency at long maturities and fixed rates. Thus
domestic public debt is typically either short term or long term at floating
rates. For example, Brazil’s domestic currency debt is indexed to the overnight
SELIC rate. In Venezuela, it is indexed to the average lending rate of the six
largest banks. This implies that movements in short-term real interest rates
44 Ricardo Hausmann

Table 3.7 Estimates of cyclical comovement of the real exchange

rate and GDP

Independent variable Latin America Industrial countries

Log real GDP 0.374 0.058

(3.77)* (0.91)
Year
0.024
0.002
(6.94)** (1.19)
Constant 50.812 8.571
(7.74)** (2.62)**
Observations 440 402
Countries 20 20
R2 0.14 0.00

Note: Dependent variable is the log of the real exchange rate. Absolute value
of t statistics is shown in parentheses (* significant at 5 percent; ** significant
at 1 percent).
Sources: IMF and author’s calculations.

will affect debt service, as governments roll over their short-term debt or as
the floating-rate long-term debt is repriced. Moreover, if the country is unable
to borrow abroad in local currency, the central bank is likely to exhibit “fear
of floating” and domestic interest rates are more volatile.10
Consequently, in countries with these characteristics, not only is debt ser-
vice more sensitive to short-term interest rates, but these rates are also more
volatile. The monthly volatility of 12-month real interest rates averaged
10.5 percent in Latin American countries, while it was less than 1 percent in
the United States (Table 3.5). The real interest rate volatility in Latin America
is not only large, but it is inversely related to the business cycle, which is
unfavorable from the point of view of risk diversification: in good times,
when the economy is buoyant, real interest rates are low; in bad times, they
are high. The average elasticity of Latin America implies that a 1 percent
decline in real imports relative to trend is associated with an increase in the
real interest rate of 1.26 percentage points. This elasticity estimate is about
40 times larger in Latin America (particularly large in Brazil, Peru, Argentina,
and Mexico) than in the United States. The real interest rate also displays a
high negative correlation with the real exchange rate for Latin America
(except Argentina and Colombia), relative to the United States (Table 3.8).
In fact, the three shock variables that we have focused on – the real interest
rate, the real exchange rate, and the import gap (as a measure of income) –
have, on average, an unfavorable correlation in Latin America in the sense
that their comovement amplifies the fiscal consequences of fluctuations. Bad
times – defined as periods in which the import gap is negative – are also peri-
ods in which real interest rates are high and the real exchange rate is weak.
 Good Credit Ratios, Bad Credit Ratings 45

Table 3.8 Correlation between the real exchange rate, the real interest rate, and the
import gap, 1990–99

Real exchange rate Real exchange rate Real interest rate

versus import gap versus interest rate versus import gap

United States
0.09
0.17
0.37

Latin America 0.21
0.43
0.33
Argentina 0.12
0.09
0.72
Brazil 0.71
0.67
0.43
Chile 0.41
0.59
0.10
Colombia
0.47
0.06
0.35
Ecuador 0.34
0.46
0.05
Mexico 0.72
0.85
0.79
Peru 0.48
0.34
0.19
Venezuela 0.61
0.39 0.0

Note: Based on monthly data, excluding periods in which inflation rate exceeded 40 percent.
Sources: IMF and author’s calculations.

Bringing it all together

To answer the central question as to why many emerging market economies

are perceived as being so much riskier than advanced economies in spite of
the fact that debt–GDP ratios are low and declining, we have examined the
roles of the lower tax revenue ratio, the higher interest rate, the foreign cur-
rency denomination, and the high volatilities and comovements of the rev-
enue base, the real exchange rate, and the real interest rate. The roles of all
these factors can be further illustrated with a stylized stress test with the
ratios used earlier (Table 3.5) to mimic the situation of the average OECD
and Latin American countries (Table 3.9). While the debt–GDP ratio is
50 percent and 35 percent, respectively, the debt composition is widely dif-
ferent. In the OECD, it is all in domestic currency and mainly long-term lia-
bilities; in Latin America, it consists primarily of foreign and short-term
liabilities (or with floating rates).
Let us assume that the economies receive simultaneously a shock to tax
revenues, GDP, real interest rate, and the real exchange rate, equal to about
one standard deviation – in line with the volatilities estimated above. This
means that the short-term rate increases by 1 percent in the OECD and by
10 percent for Latin America. The real exchange rate shock is 9 percent in
the OECD, though without fiscal consequences as the debt is denominated
in domestic currency, and it is 20 percent in Latin America. The revenue
shock is 3.6 percent and the GDP shock amounts to 2 percent in the OECD
and the revenue shock and GDP shock are respectively 11.8 percent and 4.5
percent in Latin America.
 46
Table 3.9 OECD and Latin America: stress test on debt service capacity (percent)

OECD Latin America

Initial value Shock New value Initial value Shock New value

Real interest rate (US$) 4.0 4.0 8.0 8.0

Real interest rate
(local currency) 4.0 1.0 5.0 8.0 10.0 18.0
Real exchange rate 100.0
9.0 91.7 100.0
20.0 83.3
Tax revenue–GDP ratio 40.0
3.6 38.6 25.0
11.8 22.1
GDP 100.0
2.0 98.0 100.0
4.5 95.5
Debt–GDP ratio 50.0 51.0 40.0 47.1
foreign currency 0.0 0.0 25.0 31.4
domestic long term 40.0 40.8 0.0 0.0
domestic short term 10.0 10.2 15.0 15.7
Debt–revenue ratio 125.0 132.3 160.0 213.7
Debt service–GDP ratio 2.0 2.1 3.2 5.3
Debt service–revenue ratio 5.0 5.5 12.8 24.2

Note: Impact of one standard deviation in the real interest rate, real exchange rate, tax revenue, and GDP, for the average economy in
each region.
Sources: IMF and author’s calculations.
 Good Credit Ratios, Bad Credit Ratings 47

In the OECD, these shocks would cause an increase in the debt service
burden of 0.1 percent of GDP and 0.5 percent of the budget. By contrast, in
Latin America the shock increases debt service by 2.1 percent of GDP and by
11.4 percent of government revenue, almost doubling x. These effects are
more than 22 times as large as those in the OECD (Table 3.9). In all, there is
a very large difference in the variance of x between a typical OECD and
a Latin American country, owing mainly to the interaction between the
debt structure and the variances and covariances of GDP, tax revenues, inter-
est rates, and real exchange rates, with the latter two sources playing a quan-
titatively significant role.

Conclusions of the analysis

The analysis sheds some light on three key issues: why country risk is so high
and volatile in Latin America, why countries borrow in dollars, and why
fiscal policy is procyclical.

Why interest rates are so high and volatile in Latin America. Risk premia are
related to the probability that x  x. If this probability is high, interest rates
are high. The probability that x  x is related to the debt–tax ratio and to the
volatility of x. Beyond the difference in debt–tax ratio between Latin
American and industrial countries, there is a significant difference in the
volatility of x. This volatility is affected by the volatility of the real exchange
rate and of the real interest rate, given the debt structure associated with the
so-called original sin. This explains why interest rates are high and credit rat-
ings low, in spite of relatively moderate debt ratios.

Why Latin American fiscal prudence has not been rewarded with more secure market
access, and why interest rates are so volatile and market access so uncertain. The
model presented has a multiplier effect: increases in interest rates cause a rise
in x which increases interest rates further; hence, shocks are amplified.
Moreover, increased volatility can lead to insolvency even if fiscal ratios do
not change. This implies that while the debt–GDP ratio or the debt–tax ratio
may have fallen in some countries, the probability that x  x may have
increased if perceptions of real exchange rate and interest rate volatility have
gone up. Spikes in interest rates after the Russian crisis and the large exchange
rate movements in 1998–2002 suggest that this may well be the case.

Why governments borrow in dollars. Critics often speculate that governments

do so because of irresponsible moral-hazard behavior. Borrowing in dollars
exposes the public sector balance sheet to real exchange rate volatility.
However, if the alternative is to borrow at short maturities in local currency,
interest rate volatility may expose the fiscal accounts to even greater risk.
The evidence presented suggests that the volatility of the real interest rate is
48 Ricardo Hausmann

higher than that of the real exchange rate. Since the two are not perfectly
correlated, a prudent government will choose an optimal portfolio that
includes both. However, given the high volatility in the interest rate, a large
share of foreign currency debt may make prudent sense.11
Moreover, governments should pay attention not only to the fiscal con-
sequences of relative price volatility. The interest rate is mainly a monetary
policy instrument. A government that wants to defend monetary indepen-
dence from fiscal considerations tends to value the fact that the cost of ser-
vicing dollar-denominated public debt does not rise when the central bank
decides to respond to an acceleration of inflation by raising interest rates. In
fact, to the extent that interest rate increases cause the exchange rate to
appreciate, this lowers the debt-service cost of the foreign debt. By contrast,
a large stock of short-term domestic currency debt results in an increase in
debt service when the central bank is fighting inflation, thus creating so-
called unpleasant fiscal arithmetic problems. A government that wants to
maintain a strong monetary stance to mount an interest rate defense may
be fearful of accumulating short-term domestic debt.
Hence, both for prudential reasons and for the maintenance of monetary
policy independence, governments may opt for foreign currency debt.
However, liability dollarization is likely to give the central bank a stronger
preference for stabilizing the exchange rate by allowing a more volatile inter-
est rate. This will prevent the development of a long-term domestic currency
market and will induce the private sector to borrow also in dollars to avoid
the interest rate risk.12

Why fiscal policy is procyclical in Latin America. Analysts often ask why
governments do not let automatic stabilizers take effect and instead raise
taxes and cut spending in bad times. As shown earlier, in bad times, debt
service increases because of the interaction between debt denomination and
the pattern of movements in output, real exchange rates, and real interest
rates. This has the effect of raising debt service in bad times, which further
increases the borrowing requirement. A country without original sin would
only need to finance the worsening of the primary balance during the reces-
sion. A country burdened by original sin would need, in addition, to finance
the increased debt service. Moreover, as x increases, interest rates increase,
making further borrowing even more difficult. This limits the scope for the
primary balance to adjust countercyclically.

Policy implications

The preceding analysis suggests that fiscal problems in Latin America and
other emerging markets are related not so much to the accumulation of large
debts or deficits, but instead to the risks inherent in the composition of the
 Good Credit Ratios, Bad Credit Ratings 49

accumulated stock of debt. These risks make reductions in debt ratios less
effective in terms of restoring solvency and securing access to markets.
Therefore, policies geared simply to limiting the flow of deficits may not be
adequate. An alternative and potentially more successful policy involves act-
ing on debt denomination. This means that policies should internalize the
costs associated with debt service risks.
Public debt takes on risky denominations because of the original sin.
Although the obvious solution would be to get rid of the problem, the causes
of original sin are not well understood, and it is not easy to learn from the
few countries that have overcome it. It is too common a phenomenon glob-
ally to be attributed to Latin American-style fiscal or macroeconomic pecca-
dilloes.13 Unfavorable borrowing affects virtuous Chile and prudent East Asia
as much as it does countries with poorer credit ratings. A discussion of inter-
national policy options lies beyond the scope of this chapter.14 Instead, let
us focus on the implications of the analysis presented here for fiscal rules.
Targeting an overall balance rule is not a good policy approach. In a Barro-
style “first-best world,”15 the overall balance should be volatile in the short
run. In the Latin American context, given the volatility of government rev-
enue and of debt service, a stable overall deficit would imply a very unsta-
ble level of primary spending, which is not desirable.
Targeting a cyclically adjusted overall balance is preferable, since it would
ensure a balanced budget over the medium term while providing for short-
term stability in primary expenditure. However, doing so in an emerging
market context is rather complex. In the standard IMF approach (Hagemann
1999), cyclical adjustments incorporate only the effect of the output gap on
the budget. The volatilities and elasticities shown in this chapter suggest that
while the volatility of the output gap in Latin America is larger than in
industrial countries, much larger adjustments are required to account for the
fiscal impact of the difference between current and equilibrium values
of the real exchange rate, the real interest rate, and the terms of trade.
Making these corrections requires interpretation, estimation, and ultimately,
discretion. If these adjustments are made by an interested party such as a
ministry of finance, they will hardly be credible. Hence, an independent
scorekeeper is needed to constrain discretion.
An alternative is to target a variable that is more clearly within the con-
trol of the authorities: primary spending, and to a lesser extent the primary
balance, are good candidates. Movements in the exchange rate or the real
interest rate do not directly affect these variables, although the primary
balance is affected by the output gap and the terms of trade. For example,
between 1999 and 2002 Brazil was able to meet its primary surplus targets
consistently and build some policy credibility, at a time when the overall
deficit was ballooning due to movements in the real interest rate and the real
exchange rate. Nevertheless, this credibility was not large enough to avoid
50 Ricardo Hausmann

very high risk premia and rising debt ratios. Hence, the fact that a country
may choose to control what it can does not mean that it controls what
matters. As long as debt service swells in bad times, the space for counter-
cyclical policy is very limited: in bad times, governments must make sure
that they can find additional financing at a reasonable cost to cover the ris-
ing debt service burden before they can contemplate a deterioration of the
primary balance.
The more substantive conclusion of this chapter is that deficits and debt
ratios are just two of the ingredients for determining fiscal risk. The proba-
bility that x  x is affected strongly by debt denomination and its interaction
with the structure of variances and covariances of the real exchange rate, the
real interest rate, the output gap, and the terms of trade. These factors can
trap relatively prudent policies in a path to high interest rates at which the
debt ratio becomes unsustainable. Hence incorporating the debt structure
and taking account of the monetary and exchange rate regime (which affects
the structure of variances and covariances) is key. The fact that major
improvements in the overall deficit and in debt ratios over the last decade
in Latin America have not delivered reduced risk premia and secure market
access, suggests that working on the other determinants may increase the
payoff.
In an ideal world, fiscal rules should help internalize the risks associated
with the public debt structure. Governments should be able to judge
whether a strategy that substitutes relatively cheap short-term domestic cur-
rency debt for more expensive long-term fixed-rate domestic currency debt
is convenient. A target on the overall balance discourages such a strategic
approach. One way of incorporating this logic into a rule is to create a sys-
tem of risk weights for public debt. The idea is to have a synthetic way to
value the implied risks of each obligation by borrowing a page from current
banking regulations that focus on a risk-adjusted level of capital. However,
in our context, the risk weights should reflect the variance and covariance
structure of obligations. Long-term fixed-income domestic currency debt
would get a relatively low weight, as it would protect the budget from shocks
to the short-term real interest rate or the real exchange rate. Commodity-
linked debt would also receive a lower weight as it would protect from
shocks to the term of trade. Debt indexed to the short-term interest rate or
debt denominated in dollars would receive a higher weight to reflect the esti-
mated structure of variances and covariances. Liquid assets, hedging opera-
tions, or contingent credit lines should be given a negative weight as they
reduce the overall fiscal risk.16
Under this system, the goal of fiscal policy would be to target the risk-
adjusted level of debt. To incorporate cyclical considerations, it could target
the risk-adjusted debt relative to trend tax revenues. Governments could
then internalize the trade-offs between targeting the budget balance vis-à-vis
managing the debt structure. For countries with a floating exchange rate, the
 Good Credit Ratios, Bad Credit Ratings 51

aggressive creation of a long-term fixed-income debt market could radically

reduce the implied fiscal risks and move countries into a virtuous circle of
declining risk premia and improving overall fiscal performance. Mexico has
moved in this direction. However, if countries find this too difficult or
expensive, indexing the debt to the price level may facilitate the extension
of maturities. Chile is a good example of this. For countries that dollarize in
a sustainable manner, in the sense that the equilibrium real exchange rate is
relatively stable,17 the debt can be converted more easily into long-term dol-
lar debt, thus limiting the fiscal risks.
Debt denomination has powerful effects on creditworthiness. Fiscal rules
that focus only on limiting the flow of deficits and the accumulation of
debts are not efficient. They leave aside important avenues for the reduction
of fiscal risks. Fiscal rules should facilitate the internalization of the costs
associated with these risks. A system that targets a risk-weighted level of debt
is one way to do it.

Notes

1. I am indebted to the discussant, Mario Teijeiro, as well as Ugo Panizza, Dani

Rodrik, Roberto Rigobón, George Kopits, Andrés Velasco, conference participants
and an anonymous referee for useful comments. I have made liberal use of the
database developed jointly with Ugo Panizza and Alejandro Riaño. I would like to
thank Luisa Palacios and Alejandro Riaño for able research assistance.
2. Standard and Poor’s credit ratings have been converted into a numerical scale
where each increment represents a unit increase in rating, with AAA assigned a
value of 19.
3. In the real world, the government would not default on the entire debt, nor would
it stop paying forever. However, a more realistic default rule would complicate the
algebra without adding any new insights.
4. See Eaton and Gersovitz (1981) for a seminal treatment.
5. If the distribution of x’s were uniform, the hyperbola would be substituted by a
straight line that would complete a triangle with the i   line and the
x  x line.
6. These estimates are in line with, though somewhat lower than those reported in
Gavin et al. (1996) calculated for the period 1970–94.
7. We are abstracting from short-term foreign currency debt and from long-
term floating-rate domestic currency debt because they do not contribute to the
analysis.
8. There is also a significant proportion of domestic debt denominated in foreign
currency. Absence of comparable data on this variable understates the importance
of the effects shown in this section.
9. See Hausmann et al. (2001) and recalculated and expanded in Hausmann and
Panizza (2003).
10. See Calvo and Reinhart (2002).
11. Jeanne (2003) argues similarly as regards corporate borrowing: domestic currency
debt is risky, especially if a government has low credibility so that it may suffer
from a significant peso problem. Chamon and Hausmann (2002) argue that there
52 Ricardo Hausmann

is a strategic interaction between the public’s choice of debt denomination and

the central bank’s choice of stabilization instrument. They argue that if people
borrow in dollars (pesos), the central bank will favor exchange rate (interest rate)
stability, thus justifying the initial choice.
12. Hausmann et al. (2001) show that countries with original sin that float allow
much more interest rate volatility compared to exchange rate volatility than
countries that do not suffer from original sin. Chamon and Hausmann (2002)
have determined endogenously both the currency denomination of debt and mon-
etary policy, and show multiple equilibria in debt denomination and monetary
policy.
13. Hausmann and Panizza (2003) show that the inflation history of a country, the
independence of its monetary authority, the strength of its fiscal stance, and the
quality of its institutions have little effect on the presence of original sin.
14. Eichengreen et al. (2003a) propose an international initiative to create liquidity in
baskets of emerging market currencies.
15. See Barro (1977).
16. A basis for implementing this approach can be found in the application of the
value-at-risk methodology to Ecuador’s public sector, in Barnhill and Kopits
(2004).
17. This requires the country to form part of an optimal currency area with a hard
currency country (such as the United States), a condition that was obviously not
met in the case of Argentina.
4
Can Fiscal Rules Help Reduce
Macroeconomic Volatility?
Guillermo Perry1

Introduction

The debate on fiscal policy rules in Europe2 centers mainly on how to

facilitate the workings of automatic stabilizers while achieving fiscal consol-
idation. In particular, the discussion focuses on two issues. The first is
whether the goal of medium-term equilibrium established in the Stability
and Growth Pact should be interpreted as a cyclically adjusted balance. The
second is to what extent the deficit limit of 3 percent of GDP will be enough
to conduct countercyclical policy for countries affected by strong asymmet-
ric shocks. There is, however, significant agreement within the EU on the
importance of using fiscal policy as a countercyclical instrument, as mone-
tary policy can no longer play this role.
In contrast, most of the discussion on fiscal policy in Latin America and the
Caribbean (LAC) and other emerging market economies deals only with long-
term sustainability issues, largely ignoring the effects of the economic cycle.3
This is rather surprising as these economies are much more volatile than their
European counterparts and have generally been applying procyclical fiscal
policies that exacerbate volatility. Some analysts and policymakers maintain
that countercyclical fiscal policy is a luxury that only developed countries can
indulge in, or at least that LAC countries (with the exception of Chile) need
to deal first with pressing adjustment and solvency issues before they attempt
to reduce the highly procyclical character of their fiscal policy.
This chapter argues that ignoring the economic cycle in emerging markets
is a major mistake for several reasons, including the costs of procyclical fis-
cal policy in terms of growth and welfare, especially for the poor, and the
implications of such policy for a deficit bias. It also discusses how well-
designed fiscal rules may help to deal with the political economy and cred-
ibility factors behind procyclicality. A survey of the experience with different
fiscal rules and institutions in LAC in this chapter leads to a policy proposal
that should interest policymakers, market participants, and international
financial institutions (IFIs).

53
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
54 Guillermo Perry

Effects of procyclical fiscal policy

Effects on growth
Volatility has long been a trademark of economic performance in emerging
markets. By whatever measure, LAC economies have been more volatile than
those of most other developed regions. Although their volatility decreased
during the nineties, after significant increase in the seventies and eighties,
these economies remain twice as volatile as OECD countries and significantly
more volatile than South Asia, in terms of real GDP growth (Figure 4.1). The
picture is roughly the same whether macroeconomic volatility is measured
in terms of aggregate consumption or spending.
Economic instability can affect growth through different channels. Most
of the existing empirical evidence shows that this impact is generally nega-
tive. Servén (1998) finds that five alternative definitions of instability –
volatility of inflation, terms of trade, real exchange rate, growth rate, and
price of capital goods – have a strong negative correlation with investment
ratios. High volatility also tends to skew investment towards short-run gains
in a nonoptimal way. The destruction of informational and organizational
capital of firms and banks during deep recessions has long-lasting effects.
There can also be irreversible human capital losses. Such effects are corrob-
orated by statistical evidence of a significantly negative effect of macroeco-
nomic volatility not only on growth, but also on poverty, education, income

% 4

0
Industrialized East Asia 7 Latin America Middle East South Asia Sub-Saharan Other East
economies and the and North Africa Africa Asia and Pacific
Caribbean

1960s 1970s 1980s 1990s

Figure 4.1 Selected regions: volatility of real GDP growth by decade (regional medians)
Source: De Ferranti et al. (2000).
 Macroeconomic Volatility 55

Table 4.1 Latin America and OECD: response of fiscal balance to GDP growth, 1970–94

High growth Low growth Average

periods periods growth

OECD LAC OECD LAC OECD LAC

Total surplus
(% of GDP) 0.17 0.15 0.61
0.02 0.25 0.08
Primary surplus
(% of GDP) 0.14 0.12 0.58
0.05 0.22 0.05
Total revenue
(% change) NA NA NA NA 0.84 1.32
Total expenditure
(% change) NA NA NA NA 0.09 0.61

Note: Typical impact of a one percentage point increase in real GDP on the budget category. A
positive number for the surplus means that it moves in a stabilizing direction.
Source: Gavin et al. (1996).

distribution, and financial deepening (Inter-American Development Bank

1995a; World Bank 2000b).

Effects on macroeconomic volatility

Estimates of the causes of excess volatility in LAC, taking volatility in OECD
and East Asian countries as a benchmark (De Ferranti et al. 2000) indicate
that nearly one-third of the excess volatility is due to exogenous shocks
(terms of trade and capital flows), one-third to insufficient financial inte-
gration and development of domestic financial markets, and one-third to
volatility in fiscal and monetary policies. The importance of volatile mone-
tary policy has been reduced over time, especially since most countries
adopted floating exchange rate regimes, but that is not the case with fiscal
policy. In fact, a likely explanatory determinant is procyclical fiscal policy.
Fiscal performance indeed remains highly procyclical (Gavin et al. 1996).
It deepens the cycle, particularly during recessions with a tendency to gen-
erate surpluses, usually after a major financing crisis (Table 4.1). On average,
during recessions LAC countries tend to generate a small surplus of 2 cents
for every dollar of real output fall. In contrast, OECD countries tend to gen-
erate a significant countercyclical deficit of 61 cents. Furthermore, GDP
growth volatility tends to increase as the degree of procyclicality of fiscal
policy increases (Suescún 2003).

Effects on poverty
High macroeconomic volatility is harmful for the poor. The poor have less
human capital to adapt to downturns in the labor market. They have fewer
assets and less access to credit to facilitate consumption smoothing.
Irreversible losses in nutrition and educational levels may occur if there are
56 Guillermo Perry

no appropriate safety nets, usually the case in many emerging market

economies. It should not come as a surprise, then, that we find an asym-
metric behavior of poverty levels during deep cycles: poverty levels increase
sharply in deep recessions and do not come back to previous levels as out-
put recovers (World Bank 2000b). De Ferranti et al. (2000) found that social
expenditures are at best kept constant in proportion to GDP during down-
turns, and targeted transfers tend to fall in relation to GDP, instead of
expanding as the number of poor increases. As a consequence, in a typical
downturn, social expenditures per poor are reduced by 2 percent for each
1 percent of reduction in output. By contrast, social expenditures usually
grow as a percentage of GDP in upturns, when they are less needed. That is,
the procyclicality of safety nets adds substantial income risk for the poor.
They thus suffer both from higher consumption losses and from higher cuts
in social transfers during deep recessions.

Effects on the deficit bias

Procyclical fiscal policy often leads to a deficit bias. Indeed, whenever coun-
tries do not generate surpluses in booms, as they should, they will be forced
by markets to compensate for the workings of automatic stabilizers and to
reduce the deficit through expenditure cuts during downturns. However,
there are political and legal limits to what can be done in such cases.
Procyclical policy generates as a consequence unsustainable results over the
cycle, as in the case of some European countries (e.g. Italy and Spain) prior
to fiscal consolidations or major macroeconomic crises. This has been the
experience – though more pronounced – in many LAC countries, more
recently in Argentina, Ecuador, and Colombia, where fiscal expansion during
booms has been followed by a deep fiscal crisis during economic downturns.

Causes of procyclical fiscal policy

As mentioned, fiscal policy has been strongly procyclical in LAC, in sharp

contrast with what has happened in many OECD countries (Table 4.1).
During the recent downturn, only Chile and Venezuela (the latter due to
high oil prices) were able to apply mild countercyclical fiscal policy. All oth-
ers had to adopt a strong procyclical stance. The case of Argentina was the
most dramatic example. The reason for this behavior is a mutually reinforc-
ing vicious circle comprised of the volatility of macroeconomic outcomes,
the procyclicality of the fiscal response and the procyclicality of capital
inflows. Fiscal procyclicality can in turn be explained by the combination of
faulty policy stance and weak budget institutions, coupled with asymmetric
information problems in international financial markets.
First, there is a serious credibility problem for most LAC and other emerg-
ing market economies that attempt to pursue a fiscal expansion in a bust.
Most of them arrive at the end of a boom with a fragile fiscal position (high
 Macroeconomic Volatility 57

or at least moderate deficit and relatively high debt stock), and face a time-
consistency problem. Governments may borrow today and choose not to
pay back in good times; that is, they may continue to increase indebtedness
in good times. Indeed, most of them have not reduced their debt in past
booms, so there is no reason for an investor to expect that they will do so
next time. No country in LAC but Chile has been able to run a surplus in
booms, and very few have reduced public debt indicators in such episodes.
Recent serious fiscal crises in both Colombia and Argentina can be traced to
excessive expenditure increases during the booms, that is, to the incapacity
to achieve or maintain surpluses in good times.4 In the event, a fiscal expan-
sion during bad times can be expected to lead to an intertemporally unsus-
tainable deficit bias, and can be rightly interpreted as leading to an increase
in default risk. In these conditions, the proper response of creditors should
be to refuse to finance an increased deficit during busts, or to do it at sig-
nificantly higher spreads.
But why is fiscal policy procyclical in booms? The main reasons lie in the
political economy of fiscal policy and the lack of strong budget institutions.
It is hard enough for a responsible minister of finance to avoid a deficit, espe-
cially in periods in which financing is readily available, as is usually the case
in booms. It is politically much harder to maintain a visible surplus through
discretionary decisions. Political incentives are aligned with spending any
potential surplus in a boom.5
Similarly, political pressures may inhibit a responsible minister of finance
from indulging in an explicit discretionary countercyclical policy (or just
from permitting automatic stabilizers to operate) during downturns, as once
the Pandora’s Box of a deliberate increase in expenditure or in the deficit
is open, it may not be easy to close it again because of these pressures.
Moreover, as mentioned, markets will punish any such move, as they can-
not distinguish easily between a responsible countercyclical policy and out-
right fiscal laxity.
In addition, financial market failures also contribute to fiscal procyclical-
ity. Markets often finance outright boom deficits that exacerbate the trend
towards deficit bias. Spreads are found to be procyclical and may be traced
to underlying problems of asymmetric information and herd behavior.

Why fiscal rules may help

Authorities in emerging market economies may well want to adopt rules that
at least allow automatic stabilizers to work during the cycle, and eventually
follow a full-fledged rules-based countercyclical fiscal policy. Such a frame-
work would help keep any surplus in good times out of sight and especially
out of the reach of the political process involving normal discretionary bud-
get decisions. The question remains whether policymakers would actually
adhere to fiscal rules during difficult times when they might be tempted to
58 Guillermo Perry

ignore them, or cheat through creative accounting. More precisely are rules
capable of changing the incentive to spend potential surpluses in good
times?
Fiscal rules may work if they impose high enough exit costs to compen-
sate for the incentives to spend during the normal budget process. The
authorities will have strong incentives to comply with a rule if there are
enforceable penalties (as with Brazil’s Fiscal Crimes Law) and if the rule is so
clear and simple that it does not leave much room for cheating. More impor-
tant, breaking an explicit rule may indeed be more costly than just indulging
in silent discretionary expansionary policy when there are no rules, due to
the visibility of such a decision. This is particularly true if breaking the rule
requires changing the law, especially if it is a constitutional law that requires
a qualified majority. Opposition parties and the press will find in such cases
a golden opportunity to criticize the lack of responsibility of the governing
political party or coalition. For a responsible finance minister, such a rule is
a gift from heaven, as it will facilitate resisting the pressure from peers and
politicians to spend in booms, shifting to politicians the cost of an explicit
breaking of the rule. The experience in Chile at the beginning of the demo-
cratic period during a major boom in copper prices shows clearly the use-
fulness of a tight legal rule for responsible economic authorities.
In downturns, a well-designed rule may facilitate the operation of auto-
matic stabilizers and enhance credibility of a countercyclical fiscal stance.
This will be the case especially when the rule induced a surplus in the pre-
vious boom. If such surpluses were saved in a stabilization fund for bad
times, the government would have resources at its disposal fully or partially
to finance the deficit in bad times, reducing the need to resort to market
finance. The fact that the deficit in the downturn is predictable and allowed
by the rule – and that the same rule will credibly limit spending in future
booms – will give a clear signal that the government is not indulging in
unsustainable lax policies, but merely following an established rule, and
thus help convince markets to cover any remaining financing requirement.
In other words, the market will face clear signals that distinguish a respon-
sible and sustainable (i.e. limited and predictable in scope) countercyclical,
rules-based fiscal policy from outright indulgence in intertemporally unsus-
tainable policies. A good rule should help reduce asymmetric information
and facilitate the financing of deficits in bad times. Such a rule should also
help a responsible minister of finance to avoid pressures promoting exces-
sive expansionary policy in downturns.

Flexibility and credibility of fiscal rules

Any rule may entail a dilemma between flexibility and credibility. Too
strict a rule in the pursuit of credibility may lead to rigidity. Moreover, an
excessively rigid rule may become altogether nonviable. If this is the case,
 Macroeconomic Volatility 59

economic actors may anticipate the nonsustainability of the rule, weakening

credibility. In other words, an excessively rigid rule may limit flexibility and
not enhance credibility; it may entail only costs and few benefits, if any at
all. It would thus be counterproductive.
This is the case with fiscal rules that attempt to reduce the deficit bias and
enhance solvency without correcting for the potential effects of shocks and
the economic cycle, as has happened with some of the fiscal responsibility
laws (FRLs) and stabilizing transfer schemes recently enacted in LAC. If there
is a positive commodity price shock and/or a boom in economic activity,
such a rule will not be binding: it will be too easy to comply with. It will not
improve the underlying fiscal balance and will permit a procyclical fiscal
stance to accentuate the boom. If there is a negative shock and/or a down-
turn, the rule may become excessively tight, exacerbating the downturn. It
may thus turn out to be too difficult to comply with and thus be abandoned
altogether – as happened in both Argentina and Peru with their FRLs during
the economic downturn from 1999 onwards. In other words, such rigid rules
may not help avoid a deficit bias. On the other hand, a rule that attempts to
support countercyclical fiscal policy but is not designed to achieve long-term
debt sustainability will also be unsustainable and noncredible, and will not
serve as an effective policy framework.
A well-designed rule must, as a consequence, attempt both to facilitate the
operation of automatic stabilizers (or even permit limited discretionary coun-
tercyclical fiscal policy) and to avoid a deficit bias. This will by necessity make
the rule somewhat more complex, but also more realistic and credible. In
light of these difficulties with rule design, let us review some fiscal rules intro-
duced in LAC with the purpose of coping either with procyclicality or with
the deficit bias or both, and attempt to draw some practical lessons.

Experience with fiscal rules in Latin America

Commodity stabilization funds

From a review of the copper stabilization fund in Chile, the coffee and oil sta-
bilization funds in Colombia, and the oil saving and stabilization funds in
Ecuador and Venezuela – all of them typically capturing government revenue
from nonrenewable natural resources – we can draw some general lessons.
Stabilization funds are easier to introduce before the fact (that is, given
low commodity prices or before an expected increase in volume) and, in the
case of shared revenue, when national and subnational governments
are treated symmetrically. This is highlighted by the experience of the
Colombia’s oil stabilization fund; it was a major consideration in the design
of Ecuador’s oil fund.
Stabilization funds may indeed play a useful role in insuring some savings
from fiscal revenue associated with commodity export booms, especially
where there are automatic saving rules and the accumulated net surplus
60 Guillermo Perry

remains “out of sight and out of reach” of the normal discretionary budget
process. This is clear in the cases of both the Chilean and Colombian funds.6
However, insofar as they cover a limited portion of government revenue,
the funds cannot by themselves assure aggregate expenditure restraint dur-
ing booms in economic activity (as happened in Colombia during the 1990s
or more recently in Venezuela with the macroeconomic stabilization fund).
In such cases some savings may be generated in the funds, but along with
an unsustainable deterioration of the non-oil fiscal accounts. To avoid such
a problem one would need a complementary rule that would require non-
oil deficits to be limited to the interest yield of the oil fund (as in Norway)
or would otherwise restrict overall expenditures (as envisaged in the recent
FRL in Ecuador).7
Even when effective in helping to restrain expenditure during booms (in
the case of Chile), if commodity stabilization funds do not include an auto-
matic rule for divestitures, they may play too limited a role in encouraging
countercyclical policy in downturns (as experienced in 1999), because there
is no way for markets to distinguish between responsible countercyclical pol-
icy and the beginning of fiscal laxity. These reasons underpin the adoption
of a structural balance rule (Chile).
All these facts conform with the previous conceptual discussion, and have
led these and other countries in the region to adopt or to plan to adopt addi-
tional rules with broader scope and more automatic overall expenditure
rules. Some of these attempts have eventually dealt with the difficult issue
of distinguishing between temporary and permanent shocks in commodity
prices by adopting as a reference a moving average of past prices (Colombia)
or a long-term projection by a panel of independent experts (Chile). This
helps both smooth the effect of cycles and adapt gradually the level of
expenditures to permanent shocks (which are welfare enhancing, given the
inefficiencies associated with sharp increases or cuts in expenditures). For
the same reason, some countries (e.g. Colombia and Ecuador through funds)
have attempted to smooth out (or gradually adapt to a new level of ) revenue
from oil exports, using a moving average of past dollar revenue as a bench-
mark. In both cases there was an explicit objective of gradual adjustment to
a sharp expected increase in export volume.
However, the moving-average approach might be problematic when
commodity prices tend not to revert toward a constant mean, but rather to
experience a random walk.8 This means that when shocks bring prices down,
an expenditure rule based on a moving average might lead to the exhaus-
tion of the resources accumulated in the fund. If the shocks to prices are pos-
itive and long-lasting, the fund might tend to accumulate large savings and
entail costs in terms of forgone investment. Nevertheless, even a moving-
average approach might be better than no rule, given that political pressures
tend to lead to deficit bias and the quality of public investment tends to dete-
riorate remarkably during a rise in commodity prices. Moreover, establishing
 Macroeconomic Volatility 61

rules that set ceilings and floors on the total savings of the fund can ame-
liorate these risks. More generally, fiscal rules should aim to smooth overall
government expenditure and eliminate the deficit bias, rather than just sta-
bilize the portion related to commodity exports.

Laws on fiscal responsibility and stabilizing transfers

The overarching objectives of recently adopted FRLs in Brazil, Argentina,
Ecuador, Peru, and Colombia and proposed in other emerging market
economies have been to eliminate the deficit bias and to achieve and main-
tain fiscal solvency. Without elaborating on their potential virtue in terms
of these objectives, we should note that in contrast to the evident failure in
the Argentine and Peruvian cases, the Brazilian law has been more effective,
as it is a product of broad political consensus, is less rigid, and has been
accompanied by a law that can effectively punish deviations.
A particularly relevant issue involves the likely effects of these laws on the
fiscal policy stance. To the extent that they only set rigid, specific targets for
deficits or debt levels, they may become nonviable or actually reinforce the
procyclical policy stance. If the country in question faces an unexpected
negative shock to its public finances – a fall in a major revenue source or a
sharp downturn in overall activity – adherence to the rule may force author-
ities to cut public spending, preventing the operation of automatic stabiliz-
ers and deepening the downturn. Otherwise, the authorities may not be able
to comply and the rule is effectively abandoned, as happened in both
Argentina and Peru. A rule that does not take into account such outcomes
may end up trading long-term solvency benefits for short-term costs, becom-
ing eventually unsustainable and thus not credible.
Some of these rules and supporting legislation (especially in Argentina and
Colombia) included provisions intended to achieve some stabilization of sub-
national government expenditure during the cycle, as analyzed in Chapters
12 and 16 by González et al. and by Braun and Tommasi respectively.
Unfortunately, more often than not, these provisions have merely offered
guaranteed minimum transfers, creating sizable contingent liabilities for the
national government. This has led to incapacity to observe the commitment
in the case of Argentina in 2001 and may well lead to a problematic outcome
during the excessively long transition period established in the recent con-
stitutional reform in Colombia as well. However, these laws contain some
more promising norms that would limit transfers or expenditures to a mov-
ing average of past transfers or expenditures, following a transition period.
Applied to both national and subnational expenditures, such provisions
could go a long way in helping to avoid procyclical policy and facilitate
at least the operation of built-in stabilizers, while at the same time helping
to meet solvency goals. Even better, the law may encompass convergence to
structural balance, taking explicit account of cycle and commodity price
fluctuations.
62 Guillermo Perry

Structural balance rule

Chile recently adopted a rule facilitating credible countercyclical policy-
making and ensuring fiscal solvency. It is elegant and simple: an explicit
commitment to keep a structural surplus equivalent to 1 percent of GDP. The
structural balance is estimated by removing the effects of variations in cop-
per price (relying on expert opinion about long-term price trend) and the
economic cycle on revenues (based on revenue-elasticity estimates and a
measure of potential GDP). Accordingly, for example, the actual budget
deficit in 2001 and 2002, as well as the more recent modest surplus, has been
consistent with the rule.9
The fact that the size of both the surplus in good times and the deficit in
bad times is constrained and predictable facilitates curbing political pressures
either to use the surpluses in booms or to run excessively large deficits in busts,
as these outcomes would imply breaching the surplus rule. It also enhances
the credibility of fiscal policy in the markets, provided, of course, that results
are close enough to the target, as deficits in bad times can be clearly antici-
pated and will not be interpreted as a relaxation of the fiscal stance.
This rule helped Chile to conduct a credible countercyclical fiscal policy
during the recent downturn.10 Credibility gains have largely been immedi-
ate, given a solid track record and the fact that the authorities are using cred-
ible estimates of potential output (which have long been used for the
conduct of monetary policy) and revenue elasticities.11 Chile also may find
it useful to institutionalize the rule, as at present it is just a self-imposed pol-
icy guideline that binds the current government.
Other countries willing to follow this approach would have to establish a
sound analytical and statistical basis to be able to predict or assess potential
output, and to include an adjustment for interest payments,12 or define the
structural balance in terms of the primary accounts.13 They should probably
adopt such a rule in good times, so as to build some track record before the
bad times come, and establish it by law, setting adequate penalties for non-
compliance.

Conclusion and policy proposal

The foregoing discussion highlights several considerations. First, it is as impor-

tant to reduce procyclical behavior as it is to eliminate the deficit bias in Latin
America and other emerging market regions. Procyclical policy accentuates
macroeconomic volatility, with harmful effects on growth, and hurts especially
the poor, by forcing sharp reductions in social expenditures precisely at the
time when they are most needed, thus compounding the income risk for the
poor. Second, whether based on rules or not, fiscal policy must attempt to deal
with both problems at the same time. Obviously, countercyclical policy that
contributes to a deficit bias will not be sustainable. But also, policies or rules
that attempt to reduce the deficit bias and achieve fiscal solvency while increas-
ing procyclicality are not likely to prove sustainable over the medium run.
 Macroeconomic Volatility 63

Properly designed fiscal rules may indeed help cope with the political
economy and credibility problems that underlie procyclical policy and the
deficit bias. They may help contain pressures to dissipate potential surpluses
in good times by tying the hands of authorities and restrain the political
process associated with normal discretionary budget practices; this can be
accomplished as long as exit costs from the rule are perceived to be high.
Rules can give credibility to the sustainability of deficits in downturns if in
fact they have generated surpluses in upturns: such deficits will be entirely
predictable and limited by the rule, thus increasing the likelihood that they
will be financeable (partly or wholly from the savings accumulated in good
times). The fact that deficits will be limited by the rule may also help to keep
pressures for excessive deficits at bay during bad times.
Of the different rules examined, the most convenient seems to be to set a
goal of structural balance, or a modest structural surplus, such as the one
recently adopted by Chile, which has worked remarkably well so far. Such a
rule would permit the full operation of automatic stabilizers during the eco-
nomic cycle and avoid sharp changes in public expenditure associated with
changes in fiscal revenue from commodity export receipts.
Emerging market economies would benefit from adopting a structural
balance framework for the presentation and discussion of their fiscal policy,
even if they do not adopt a structural balance rule. In addition to continu-
ing to improve fiscal accounting practices, they will need to prepare reliable
estimates of potential output and revenue elasticities. They will also have to
develop ways to adjust for the cyclical components in interest rates or, alter-
natively, to base their policy goals on adjusted primary balances. For this
purpose, assistance from the IMF could be valuable, applying a well-developed
methodology regularly used to analyze and discuss the fiscal stance of OECD
countries. The IMF might also consider requiring a structural balance frame-
work as part of the Code of Good Practices in Fiscal Transparency.
Structural balance rules should ideally form the basis of future attempts to
establish FRLs and stabilizing transfers for subnational governments, instead
of relying excessively on rigid ceilings that do not take into account the
effects of shocks or the economic cycle. The latter are likely to accentuate
the procyclicality of fiscal policy and prove nonsustainable over time, as
happened in Argentina and Peru. Such a limitation may reduce ex ante their
credibility, severely reducing their usefulness. The same can be said of recent
attempts to formulate rules for stabilizing intergovernmental transfers in
Argentina and Colombia: by introducing a long transition period of rigid
quantitative targets, they can become minimum guarantees that add to fiscal
risks for the national government.
Governments that do not yet have the capacity to adopt credible struc-
tural balance frameworks or rules may benefit from considering a more
simple rule that would limit real expenditure growth to a moving average of
past real revenue increase. Such a rule, included in the oil stabilization fund,
64 Guillermo Perry

as well as in the stabilizing transfer provisions, in Colombia and Argentina

(though in some cases with excessively short periods for estimating averages)
would significantly reduce procyclicality and avoid the deficit bias.
Structural goals need to be set according to fiscal consolidation needs.
Thus, a country that starts below a sustainable structural balance should set
goals that permit a gradual convergence to the target level. Alternatively,
again, it could limit real expenditure growth with a gradual approximation
to a moving average of real revenue increases. There is no need in principle
to wait until fiscal consolidation is completed to introduce rules that make
the achievement of solvency goals compatible with removal of the procycli-
cal stance. The best time to introduce such rules is of course during good
times, as they may gain enough credibility in the boom to facilitate market
financing of deficits in bad times. Countries affected by a large debt burden,
and unable to reduce or stabilize rapidly the debt–GDP ratio, may find it
impossible to credibly introduce a structural balance rule in bad times
because the country may remain a prisoner of market confidence.14
Most of all, it would be extremely useful if the IMF and private sector ana-
lysts were to decide to use systematically a structural balance framework
when examining and discussing the fiscal stance of all countries and set the
goals for adjustment programs accordingly. We should engage in extending
the best practices already applied to industrial countries to the analysis and
design of policies in emerging markets. In this regard, it was most unfortu-
nate that not only local authorities and analysts, but also international mar-
kets and the IFIs indulged in accepting and financing highly expansionary
procyclical fiscal policies in many Latin American countries, most notably
in Argentina (Mussa 2002; Perry and Servén 2003) during the good times in
the early 1990s, planting the seeds of major fiscal crises, or at least signifi-
cant fiscal stress, during the bad times of the late 1990s. These mistakes
should not be repeated in the future.

Notes

1. I would like to thank Nicholas Stern, Luis Servén, Augusto de la Torre, Rodrigo
Suescún, Mauricio Carrizosa, Sergio Schmukler, and George Kopits for useful com-
ments on an earlier version.
2. According to several authors, these rules have permitted some countries to elimi-
nate the deficit bias and improve fiscal stance to such an extent that there was a
significant reduction and convergence of interest rates before the start of the cur-
rency union. But whatever the merits of these fiscal rules in reducing the deficit
bias and facilitating the decline of interest rates (discussed in Chapter 7 by Buti and
Giudice), there is still a debate over to what extent they will permit recourse to
countercyclical fiscal policy. Many argue that the deficit limit is too rigid, though
it can be waived under exceptional circumstances. See Eichengreen and Wyplosz
(1998) and Hagemann (1999).
3. As exceptions, see Gavin and Perotti (1996) and Gavin et al. (1996).
 Macroeconomic Volatility 65

4. As noted by Gavin et al. (1996), “It is during booms that the seeds of crisis often
are sown, although the crisis becomes evident only when the boom subsides.”
5. As discussed in Chapter 6 by Schick, there are strong pressures in any normal dis-
cretionary budget process to increase expenditures.
6. The Copper Stabilization Law required keeping revenue generated by prices higher
than a benchmark price out of the current budget.
7. For an analysis of the Norwegian case, in contrast to the Venezuelan approach,
see Chapter 11 by Bjerkholt and Niculescu.
8. See Cashin et al. (2000) and Cuddington (1992).
9. This was noted by Jaime Crispi in comments on Chapter 5 by Kopits in this vol-
ume. See also Marcel et al. (2001).
10. In the words of Anne Krueger, Chile is using “virtue with a purpose” as fiscal con-
servatism is not a purpose in itself but a means to keep good access to markets,
low interest rates, and the capacity to conduct expansionary policies in bad times.
11. Technical problems are, however, nonnegligible. Simulation exercises indicate
that (by just using present output gap models and revenue elasticity estimates) the
potential for countercyclical smoothing of the present rule would be fairly lim-
ited, as effects of the estimated cyclical adjustment would be just a small fraction
of expected corrections related to the volatility in copper prices. In other words,
as presently applied the rule would not do much more than what the copper sta-
bilization fund has accomplished. See Fiess (2002).
12. Chapter 3 by Hausmann shows that interest payments have been the most
volatile component of public expenditures in LAC, with the possible exception of
Chile.
13. The Brazilian authorities have been following primary surplus targets since 1999,
without taking into account the revenue response to the economic cycle. The new
government has recently announced that it intends to set primary surplus goals
in the future adjusted by the level of economic activity in order to allow auto-
matic stabilizers fully to operate.
14. Amaury Bier, former Undersecretary of Finance in Brazil, holds this point of view
with respect to his country’s fiscal position. Present authorities, as mentioned,
have announced that they intend to adjust primary fiscal goals to the level of eco-
nomic activity in the future, but only after a couple of years of reduction in the
debt–GDP ratio, that is, after some further consolidation of credibility in present
fiscal policies.
5
Fiscal Policy and
High Capital Mobility
George Kopits1

Introduction

The conduct of fiscal policy in the context of an open capital account

deserves particular attention in emerging market economies that adhere to
an exchange rate peg either explicitly or implicitly, in view of their poten-
tial vulnerability to currency crises.2 Indeed, fiscal policy must be examined
as both a possible source of and a remedy for capital account crises – broadly
defined in terms of a sudden and sizable loss of foreign exchange reserves, a
large devaluation, or both.3
Recent capital account crises can help distill lessons for fiscal policymak-
ing in an economic environment increasingly characterized by high capital
mobility, low tolerance for information asymmetries, rigidities in the fiscal
system, and sensitivity to adverse social impact. This chapter focuses on fis-
cal adjustment issues associated with capital account crises, rather than with
current account imbalances. Thus it addresses the need for policy credibility
and explores the appropriate fiscal stance and the measures to achieve it,
including through the adoption of fiscal policy rules.

Role of fiscal policy

Fiscal contribution to capital account crises

The role of fiscal policy in balance of payments crises has been subject to
various explanations, depending on whether policy impacts primarily the
current account or the capital account. Traditionally, public dissaving
accommodated by domestic credit creation has been a major cause of cur-
rent account crises. With the opening of the capital account, persistent
budget deficits in combination with an exchange rate peg have often led to
large and sudden capital outflows. However, the fiscal contribution was less
obvious in capital account crises in the 1990s in countries where the budget
appeared to be broadly in balance. The emphasis on the fiscal imbalance as
the root of the capital account crisis in first-generation models has given way

66
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Fiscal Policy and High Capital Mobility 67

to a less deterministic interpretation in second-generation multiple-equilib-

rium models.
First-generation models highlight the inconsistency in fundamentals,
namely the inconsistency between fiscal policy and the exchange rate
peg: large monetized budget deficits are accompanied by a gradual fall in
foreign exchange reserves that collapse at the time of an attack on the
currency.4 With some qualifications, this approach seems to fit not only
earlier currency crises, but also recent ones, for example those in Turkey
and Ecuador (Table 5.1). The approach can be extended beyond recorded
budget deficits to include quasi-fiscal imbalances or prospective imbal-
ances, in particular when expected to be monetized in the future. In this
context, the net assets of the public sector can be regarded as a key deter-
minant of speculative attacks: as the net asset position declines (on the
strength of the accumulation of recorded and unrecorded deficits) and
becomes highly negative, the process inevitably results in abandonment of
the peg.5
In second-generation models, large-scale capital inflows (often attracted
by a speculative bubble) turn into precrisis outflows not necessarily predi-
cated on an obvious policy inconsistency. The actual attack on the currency
is prompted by a shift in investor sentiment – triggered by perceptions of a
sharp rise in the cost of defending the peg – from a good equilibrium to a
crisis equilibrium.6 The unfavorable shift in expectations can be influenced
by new information on the possible magnitude of future budget deficits, in
some cases in connection with implicit liabilities associated with a likely res-
cue of the collapsing banking sector. This explanation seems relevant for the
Asian, Mexican, Czech, and possibly the Russian crises.
More generally, asymmetric information on the extent of net public
debt exposure, or the official commitment to taking corrective fiscal
action, can translate into doubts about the sustainability of the peg and thus
render it vulnerable to a speculative attack. In these circumstances, the
actual attack takes place when investors feel that net government liabilities –
especially in the form of short-term obligations – exceed a certain threshold,
or when the authorities decide to extract seignorage to meet the intertem-
poral budget constraint, instead of embarking on a fiscal adjustment. In
the event, the immediate cause of the crisis is a signal that the government
can resolve the policy inconsistency only through a devaluation.
In the second half of the 1990s, major currency crises erupted in emerg-
ing market economies with various types of exchange rate pegs, degrees of
openness,7 and apparent differences in fiscal performance (Table 5.1).
Observed overall deficits, on the whole, were understated relative to the
cyclically adjusted position, as most countries (except Brazil and Ecuador)
were operating at full capacity prior to the crisis; also, in some countries the
effective coverage of general government accounts was incomplete.
Although these crisis episodes cannot be interpreted as statistical evidence
 68
Table 5.1 Crisis episodes: selected fiscal indicators of vulnerability (in percent of GDP, unless otherwise indicated)1 1994–99

Country/Date of crisis Capital Precrisis Gross public

account overall fiscal balance debt
restrictions
(index)2 Recorded Other3 Precrisis End–2000

Large recorded deficit (
6% of GDP)

Turkey (1994 Q1) 0.26 13 — 36 68
Russia (1998 Q3) 0.56 8 — 46 65
Brazil (1999 Q1) 0.47 8 — 42 47
Ecuador (1999 Q1) 0.13 6 — 83 1235
Significant (including unrecorded) deficit
Mexico (1994 Q4) 0.21 — 4 41 55
Argentina (1995 Q1) 0.11 2 1 36 50
Czech Republic (1997 Q2) 0.19 2 4 13 354
Small deficit or surplus ( 2% of GDP)
Thailand (1997 Q3) 0.40 2 — 5 57
Indonesia (1997 Q3) 0.34 1 — 24 102
The Philippines (1997 Q3) 0.32 1 — 57 1125
Korea (1997 Q4) 0.40 2 — 12 575

Notes
1
Precrisis indicators for general government (or consolidated public sector) refer to the period of, or prior to, the abandonment of the fixed or pre-
announced crawling peg, or (for Argentina, the Philippines, and Turkey) of sudden and substantial private capital outflows – date of crisis shown
in parentheses.
2
Index value ranges from 0 (lowest) for absence of controls, to 1 (highest) for most restrictive exchange and capital controls.
3
These are estimates (official for Mexico, Teijeiro (2001) for Argentina, and World Bank for the Czech Republic) of unrecorded balance encom-
passing mainly quasi-fiscal activities by state-owned banks.
4
Includes World Bank estimate of unrecorded public sector liabilities mainly involving state-owned banks.
5
Data for end–1999.
Sources: National authorities, World Bank, and IMF staff estimates.
 Fiscal Policy and High Capital Mobility 69

of linkages between fiscal indicators and external vulnerability, they suggest

certain patterns that escape large-scale empirical studies.8
Besides large recorded government deficits (Turkey, Russia, Brazil, and
Ecuador), fiscally induced vulnerability may take less obvious forms: a
buildup of public indebtedness, especially in short-term securities denomi-
nated in foreign currency (Mexico), or an accumulation of quasi-fiscal
liabilities by extrabudgetary funds or public financial institutions (Argentina,
the Czech Republic, and Mexico). While in itself not a source of vulnerabil-
ity, a large present value of net contingent liabilities of a defined-benefit
public pension system can also affect investor sentiment (Brazil). In other
cases (Asia), vulnerability is attributable to maturity and currency transfor-
mation from short-term foreign liabilities to long-term domestic assets by a
weakly regulated banking system, where the government assumed an indi-
rect role through the provision of formal or informal guarantees on corpo-
rate and bank liabilities.
These episodes illustrate three points. First, in the environment of a very
open capital account coupled with a hard peg (Argentina under the currency
board arrangement), even a modest fiscal imbalance can contribute to exter-
nal vulnerability; with less capital mobility and a softer peg (Russia), it may
take a larger imbalance to precipitate a crisis.9 Second, concealed or quasi-
fiscal deficits (the Czech Republic, Mexico) can also contribute to vulnera-
bility under such conditions. And third, though a precrisis fiscal imbalance
is absent, large short-term exposure of unregulated financial institutions
backed by government guarantees (Asia) is likely to lead to a significant
buildup of postcrisis public indebtedness. Even where the stock of explicit
plus implicit public debt may not seem to indicate vulnerability, it is cer-
tainly a determinant of the fiscal debt burden following an eventual crisis.
Summing up, the contribution of fiscal policy to capital account crises
in emerging markets with explicit or implicit pegs has been the object of
differing interpretations. Under a narrow interpretation, only a misalign-
ment of fundamentals reflected in a large flow of recorded budget deficits
renders the economy vulnerable to an attack. A broader interpretation,
borne out by most country episodes, suggests that external vulnerability is
associated with a large stock of explicit or implicit, actual or prospective,
public sector indebtedness10 – including its interaction with high interest
rates at a time when the government is unable to generate a sufficiently large
primary surplus.

Prevention and remedy

The standard prescription for correcting an unsustainable current account
disequilibrium includes a reduction in public dissaving commensurate with
the available level of foreign savings; that is, the goal is to contain aggregate
domestic demand by limiting the flow of fiscal imbalance.11 Under a
straightforward application of this approach, the prevention or cure of a
70 George Kopits

currency crisis calls for a tough, front-loaded fiscal adjustment to correct the
underlying policy inconsistency.12
By contrast, perceptions of an unsustainable stock of (mainly short-term)
public or publicly guaranteed private indebtedness may cause investors to
lose confidence, leading to a sudden capital outflow. The restoration of
confidence requires a commitment to addressing the underlying lack of sus-
tainability. As part of this commitment it is necessary to initiate correction
of the policy inconsistency with convincing upfront measures. However, the
central task is to influence perceptions, rather than to eliminate the nega-
tive net worth position of the public sector which is not likely to be either
feasible or credible in the short run.
Even if politically feasible, a sizable immediate adjustment could be coun-
terproductive; in particular, the case for fiscal consolidation is obviated
when the crisis cannot be directly traced to fiscal conditions. For one thing,
rapid adjustment with excessive recourse to quick-yielding distortionary
taxes or a cut in productive investment expenditures usually involves unde-
sirable allocative costs. Or worse, reduction of the cash deficit through a
buildup in payment arrears or across-the-board wage cuts can just as easily
be reversed in a subsequent period, and may preempt much-needed struc-
tural reforms. Moreover, the skewed composition and speed of such an
adjustment can result in a procyclical stance that will tend to aggravate the
adverse real impact of the capital flight.
On balance, in most of the above capital account crises, IMF-supported pro-
grams called for a moderate initial withdrawal of fiscal stimulus.13 Following
some easing of the adjustment – especially in the Asian countries where fis-
cal imbalances were not seen as the main source of vulnerability – the actual
fiscal impulse turned out to be positive or far less negative than initially
envisaged. Thus, except for Turkey, the output contraction was attributable
mainly to supply conditions rather than to cuts in public consumption.
Clearly, a fundamental question concerns what fiscal policy stance is an
appropriate response to a sudden capital outflow. If the crisis has fiscal roots,
some fiscal contraction may be necessary to help meet the external con-
straint and to restore confidence in macroeconomic management. However,
fiscal retrenchment in the event of a pronounced downturn in private con-
sumption and investment demand may aggravate the output loss associated
with the crisis.
In principle, during a currency crisis a neutral stance can help mitigate the
adverse impact of the fall in private demand through the operation of auto-
matic stabilizers, given an effective unemployment compensation scheme
and a progressive income tax system. In practice, in fiscally induced crises
an accommodating neutral stance could undermine the credibility of a
coherent adjustment program. This is particularly true if the precrisis period
is characterized by near, or at, potential output growth, justifying some fiscal
retrenchment. Nonetheless, formulation of the fiscal stance in an emerging
 Fiscal Policy and High Capital Mobility 71

market economy, especially during a turbulent period, is a complex and

uncertain exercise.14
From a rather qualitative perspective, in a multiple-equilibrium framework,
investor sentiment toward fundamentals can shift on the basis of new infor-
mation, shocks, or an announcement about changes in policy course. By the
same token, credible signaling of a future fiscal adjustment can influence
expectations favorably, as long as the signaling is followed up with imple-
mentation of appropriate structural reform measures over time. Broadly
speaking, it is useful to distinguish between two types of signaling accord-
ing to the time consistency of the adopted policy stance. The first type,
excessive initial toughness followed by laxity, lacks credibility. In the second
type, while sufficiently tough, the policy stance involves persistence and
credibility.15 The latter, of course, calls for high-quality, high-yield measures.

Major adjustment issues

Quality of adjustment
In earlier stabilization programs, especially when the task was to correct a cur-
rent account deficit, the primary focus was on the short-term macroeconomic
impact of the fiscal adjustment, with little regard for the composition of the
adjustment.16 Over time, however, it has been recognized that the quality of
the corrective measures is essential for the success of an adjustment program –
with success being defined in terms of durability of the adjustment and,
above all, in terms of its contribution to sustained growth. This view is cor-
roborated in a number of country studies, as well as broader cross-country
estimates, that suggest clear differences in effects depending on the compo-
sition of the adjustment.17 While fiscal adjustment involving cuts in pro-
ductive expenditures (mainly on education and investment) and increments
in income or payroll tax rates is likely to be contractionary, there is growing
evidence that an adjustment consisting of cuts in price subsidies and in the
wage bill, and of widening the effective tax base, can be expansionary.18
The quality of the adjustment is particularly critical when addressing a
capital account imbalance, which in essence involves an eventual stock
adjustment that helps reduce vulnerability to sudden capital outflows. In
fact, measures that are credible and durable, and that ultimately contribute
to fiscal sustainability, have a far greater chance of reversing a capital out-
flow than stop-gap measures (e.g. nominal cuts in salaries and pensions) that
are quickly reversible and have little or no long-term impact on the stock
imbalance.
The possible tradeoff between the speed and the quality of adjustment can
be eased by credible policy signaling. This should help boost investor
confidence, thereby mitigating the financing constraint, and afford the
breathing room needed to design and implement higher-quality adjustment
measures in the medium term – with increased scope for creating a broad
72 George Kopits

consensus for structural reform. Instead of relying only on short-run (and by

their nature, often unsustainable) demand-restraining measures, the fis-
cal adjustment should encompass reform in the tax system, intergovern-
mental fiscal relations, social security, civil service, public enterprises, or
official financial institutions. More immediately, if necessary, the adjustment
should envisage upfront restructuring of the banking system.
A prominent issue in capital account crises, with possible implications for
the quality of subsequent adjustment programs, is the alleviation of their
social impact. There is growing consensus that, instead of adopting an
accommodating fiscal expansion, the potential conflict between stabiliza-
tion and social-protection goals should be resolved by activating as soon as
feasible a social safety net targeted to the most vulnerable households.19

Constraints on adjustment
Often the authorities are hampered in undertaking a fiscal adjustment
because of rigidities in the structure of the public finances, which may pose
a major problem when pressures from high capital mobility demand a rapid
policy response. The combination of these constraints can seriously limit the
scope for introducing quality revenue or expenditure measures in the near
term. Relaxation of these impediments presupposes structural reform imple-
mented over an extended period.
The trend toward fiscal decentralization in the past decades may result in a
major rigidity and thus challenges the stabilization function of fiscal pol-
icy.20 Automatic transfer of revenue without an unambiguous, concomitant
assignment of expenditure responsibility to lower levels of government has
created significant vertical imbalances. The lack of incentives to levy taxes,
to exercise expenditure control, and to limit borrowing at subnational
levels shifts the burden of adjustment to the central government. These dif-
ficulties have been particularly acute in Argentina, Colombia, and Russia.
Another source of rigidity is the automatic revenue earmarking for specific
expenditure categories or decentralized government agencies or subnational
governments. In a well-defined application of the benefit principle, revenue
earmarking may be useful to gain public support or to secure stable financ-
ing for government programs that otherwise might not be financed through
the ordinary budget process. Good examples are the earmarking of payroll
contributions for social security programs or of user charges to the benefi-
ciaries of specific government services. However, earmarking of one-half or
more of government revenue (excluding social security contributions) in
countries such as Colombia and Ecuador cannot be justified on these
grounds. In essence, revenue earmarking implies a leakage in the fiscal cor-
rection, requiring a far greater effort to realize the envisaged adjustment.
Present or prospective social security imbalances may drain government
finances as they are not usually amenable to near-term correction. The
difficulty of trimming benefits protected under so-called acquired rights,
 Fiscal Policy and High Capital Mobility 73

of raising contribution rates at the risk of damaging competitiveness, or

of broadening the effective contribution base through time-consuming
administrative improvements, limits the room for adjustment. In some
countries, these impediments are reflected not only in a medium- to long-
term sustainability problem, but also in short-term financing pressures.21
With a large stock of public debt, substantial interest payments tie up a
major portion of government expenditures. High real interest rates, usually
owing to a significant rise in the risk premium before and during a specula-
tive attack, especially if the government is biased toward the external financ-
ing of deficits,22 compound the problem. A devaluation exacerbates the
debt-servicing burden to the extent the government carries (Mexico) or
assumes from private debtors (Asian countries) a large volume of short-term
liabilities denominated in foreign currency. Capital outflows prompted by
deterioration in investor sentiment lead to higher interest rates and thus to
a rise in the interest bill. This imposes a major burden that can only be
reduced over time if interest spreads, along with domestic interest rates,
decline as investor confidence is restored (Brazil).
An additional impediment to rapid adjustment is weak administrative
capacity in taxation and budget management. This is a common character-
istic of many emerging market economies. Administrative shortcomings can
limit the feasibility of improvements in tax policy design23 or in expenditure
structure. Efforts at correcting these weaknesses are resource intensive and
for the most part cannot be put into effect in a short time.
Given their openness, emerging market economies are exposed to tax base
erosion under high capital mobility. Obviously, competitive pressures limit the
scope for taxing income from capital. In fact, countries often forego with-
holding taxation of interest income and grant generous tax incentives on
income from direct investment – in some cases refraining from the applica-
tion of thin capitalization rules. In addition, the need to maintain compet-
itiveness in the current account inhibits increases in payroll tax rates.

Fiscal costs of structural reform

Any adjustment effort must contend with the fiscal costs of restructuring in
key areas, such as banking and public pensions, as part of the effort to reduce
external vulnerability.24 In principle, these costs are tantamount to making
transparent preexisting contingent public liabilities; in this sense, they need
not be regarded as an added fiscal burden, but rather as an explicit recogni-
tion of an already existing one. In practice, however, full recognition of these
liabilities imposes a concrete financing cost that may reflect, through
increased interest rates, an additional currency risk or default risk. Thus, the
question remains as to whether, and in what form, these costs need to be
incurred by the public sector.
A currency crisis often arises more or less simultaneously with a banking
crisis. The ensuing bank restructuring, of course, usually entails significant
74 George Kopits

fiscal costs manifest in the government interest bill over an extended period.
The experience of many countries (including Mexico, Korea, and Thailand)
suggests that the cost of the resolution of nonperforming portfolios of
commercial banks can be minimized if the restructuring is speedy and
comprehensive, accompanied by the strengthening of banking supervision
and regulation (Brazil).25 By contrast, a protracted and uncertain process,
coupled with open-ended liquidity injections, can lead to moral hazard at a
very high fiscal cost (Indonesia, Ecuador).
Likewise, reforming public pensions may carry a sizable fiscal cost.
Specifically, the shift from a defined-benefit to a defined-contribution (or to
a multipillar) system can be viewed as making explicit a contingent public
liability.26 The recognition of all benefits as acquired rights under the present
system, while requiring the application of the new system only to new
enrollees, involves sizable and uncertain transition costs to be accommo-
dated over an extended period (Argentina, Mexico).27 Consequently, there is
a strong argument first for tightening parameters for all current and future
beneficiaries (as partly achieved in Brazil), and only then shifting to the
new system, so as to ease the adjustment task and limit the future public
debt burden.28

Transitory measures
In view of the institutional impediments to adjustment and the costs of
essential structural reform tasks, it may be necessary to resort upfront to
measures that temporarily help ease the public sector borrowing require-
ment. These measures, of course, cannot stand as a substitute for more
permanent corrective action, since financial markets tend to discount such
measures as only short-term (in some cases, inefficient) solutions to the
underlying fiscal imbalance. Therefore, they should be phased out accord-
ing to a preannounced schedule.
Given the limitations to introducing quick, high-quality, revenue-raising
measures, in recent years some Latin American countries have resorted to
the taxation of financial transactions,29 which is a convenient device to gen-
erate immediate revenue due to the fairly inelastic transaction demand for
(noncash) money balances. However, there is evidence that, beyond a low
threshold, the revenue yield declines with rising statutory tax rates.30 Also,
at high statutory rates, this type of tax tends to be distortionary and con-
tributes to capital flight and to financial disintermediation.31
Another transitory measure consists of a uniform import surcharge. Some
economies have relied temporarily on a surcharge as a revenue-raising mea-
sure when an undeveloped tax system and other rigidities prevented a rapid
reduction in the budget deficit.32 Although at a low rate it might be less
distortionary than a tax on domestic financial transactions, an import sur-
charge is objectionable to the extent it raises the effective rate of protection
and induces trading-partner countries to retaliate in kind. For this reason the
 Fiscal Policy and High Capital Mobility 75

surcharge is subject to approval by the WTO under certain criteria of threshold

rates and of balance of payments need.
Proceeds from privatization – though neither permanent nor predictable –
can also help finance the public sector deficit (unless they are accompanied
by an equivalent cut in tax revenue or an increase in expenditure) at a time
when the economy is facing fiscal stress under the impact of capital out-
flows. Large-scale sale of state enterprises may actually dampen capital flight
or even contribute to its reversal.33
Beyond the fiscal adjustment task, supported by some of these transitory
measures, appropriate public debt management can help reduce vulnerability
to a crisis. In the first place, the strategy should be aimed at limiting the
issuance of short-term government liabilities denominated in foreign cur-
rency. Although, admittedly, it is difficult to shift toward long-term domes-
tic-currency maturities in the midst of a capital account crisis, some recent
experience (Colombia, Turkey) indicates that such conversion can provide
temporary relief from immediate financing pressures, if conducted in a mar-
ket-friendly manner.

Creating policy credibility

Transparency and institutional infrastructure

Policy credibility is key in an environment of high capital mobility. In an open
economy, a weak and opaque fiscal policy can undermine credibility and thus
contribute to a speculative attack. Conversely, public finances underpinned by
transparency and strong institutional infrastructure can foster credibility.
Accrual-based accounting, economic and functional classification of expendi-
tures, wide institutional coverage, and an explicit medium-term macroeco-
nomic framework enhance the clarity of public finances.34
The lack of satisfactory information on the stock of implicit (including
contingent) liabilities of the public sector have contributed to currency crises
in Argentina, Mexico, Indonesia, Korea, Russia, and Thailand.35 In the after-
math of the Mexican and Asian crises, there have been major strides toward
making publicly available clear, comparable, and frequent fiscal data. The
adoption of the Code of Good Practices in Fiscal Transparency as the primary
international guideline of good practices for IMF member countries has
been a major milestone. Compliance with the Code is documented in ROSCs
(Reports on the Observance of Standards and Codes).36 Although many
emerging market countries have moved toward the clear delineation of
the roles and responsibilities of (and within) the public sector, transparent
budget management practices, open procurement procedures, and improved
quality of fiscal data, relatively few of these countries regularly provide
information on contingent liabilities, quasi-fiscal activities, and tax
expenditures.37
76 George Kopits

Policy signaling and fiscal rules

Appropriate policy signaling can reassure financial markets of the govern-
ment’s commitment to fiscal prudence in a timely manner. This process
entails a politically realistic, sufficiently ambitious fiscal policy stance that is
underpinned by a high-quality structural content and conducive to the
restoration or maintenance of a public net worth position in line with vari-
ous standards of sustainability.38 The effectiveness of signaling in enhancing
reputation and reducing vulnerability will depend, of course, on public
perception of the authorities’ readiness to match policy announcements
with commensurate action.39
Adoption of well-designed fiscal policy rules is a potentially powerful
vehicle for signaling commitment.40 If accompanied by strict transparency
standards, such rules can ensure a superior macroeconomic outcome as com-
pared with time-inconsistent discretionary policies pursued by a rational
government to enhance the prospect of reelection.41 At a practical level, per-
manent balanced-budget requirements or limits on public debt, imple-
mented convincingly with the aim of reducing the public debt ratio to a
sustainable path, over time can confer considerable benefits, including a
likely decline in risk premia and thus in interest rates. In turn, the falling
cost of capital paves the way for increased private investment and growth.
The recent experience of Brazil, following the change of government in
2002, confirms the usefulness of adhering to fiscal policy rules. In fact, after
a sharp spike in risk premia, fueled by electoral uncertainties, the new gov-
ernment raised the primary surplus target and embarked on a long-awaited
reform of the public pension and tax systems. As a result, Brazil regained
market confidence, reflected in a significant decline in its sovereign spread,
with improved prospects for restoring public debt sustainability – along the
lines of the simulation presented in Chapter 8 by Goldfajn and Guardia.
These benefits, particularly in emerging market economies, can far out-
weigh the apparent loss of flexibility of discretionary fiscal policy that, in
fact, often results in a procyclical stance.42 Moreover, in countries with estab-
lished reputation in financial markets – as illustrated by the cases of Chile
and Estonia – a balanced-budget rule can be designed to allow for the oper-
ation of automatic stabilizers.43 For economies endowed with oil or other
nonrenewable resources, which usually are most exposed to exogenous
shocks, such flexibility can be complemented with an expenditure rule,
requiring maintenance of a constant ratio of primary expenditure to trend
non-oil GDP (within preset margins) over time.44 In addition, within a fed-
eral context, there is a strong case for mitigating asymmetric shocks through
an appropriate mechanism of intergovernmental transfers.

Conclusions

In spite of some interpretative disagreements in the literature about the fiscal

roots of capital account crises, major crisis episodes of the 1990s suggest that
 Fiscal Policy and High Capital Mobility 77

actual or perceived public debt sustainability problems can contribute to

such crises. Emerging market economies that are saddled with a large present
value of public sector liabilities (explicit and implicit, including contingent
liabilities) tend to be more vulnerable to shifts in market sentiment and to
the onset of a crisis.
The key to the prevention of and remedy to a capital account crisis, in the
face of an unsustainable fiscal position, lies in creating policy credibility,
rather than enforcing a tough and rapid fiscal adjustment. To this end, it is
necessary to launch an ambitious yet credible fiscal adjustment, encom-
passing upfront policy action followed by high-quality structural measures
that may require phased implementation. Reliance on cuts in productive
expenditures or on distortionary tax hikes should be avoided. Rapid adjust-
ment to prevent or remedy a crisis is hampered in many countries by struc-
tural rigidities, most of which can be relaxed only through reform over the
medium term. In the short term, transition measures can provide temporary
relief from immediate financing pressures.
Formulation of an appropriate fiscal stance must be guided by objectives of
macroeconomic stability and of policy credibility, while minimizing collateral
output loss. The adverse social impact should be alleviated through the imple-
mentation of a targeted social safety net that can be activated rapidly in the
event of a crisis, rather than through the adoption of an expansionary stance.
Policy credibility should be an overarching goal of fiscal policy in an envi-
ronment of high capital mobility, based mainly on the observance of strict
transparency standards and a sound institutional infrastructure for public
finances, including a disciplined yet flexible budget process set within a
rolling multiyear macrofiscal framework. In this context, as implemented in
a number of emerging market economies, fiscal policy rules can be a power-
ful signal of the government’s commitment to fiscal prudence. However, fis-
cal rules alone, particularly if introduced in turbulent times or under duress,
are totally ineffective in bestowing credibility on fiscal policy. Overall, both
policy and procedural rules can only be successful if they are well designed
and supported by political will – as shown in Chapter 6 by Schick.

Notes

1. The author is grateful for useful comments from Jaime Crispi, Timothy Lane and
other IMF colleagues. However, he alone is responsible for the views expressed.
2. See Calvo and Reinhart (2002), for evidence on fear of floating.
3. Thus the definition of a capital account crisis used here is broader than of a cur-
rency crisis, normally defined as the abandonment of the pegged exchange rate
regime coupled with a significant devaluation.
4. See Krugman (1979). The basic model is open to qualifications as to sterilization of
the decline in reserves, money demand behavior, and debt-financed budget
deficits; see Calvo (1997).
5. For this purpose, net assets can be defined as incorporating reserves plus some
contingent assets less public debt, including insured bank liabilities and external
78 George Kopits

obligations. An analysis of the crisis as caused by the stock of public debt (or net
public assets) is provided in Calvo (1994).
6. In comparison to the linearity assumed in first-generation models, which leads
to the inevitability of the crisis, second-generation models allow for policy
nonlinearities to explain such shifts. See a derivation of the model in Obstfeld
(1994). For an analysis of various types of state contingent policies and reconcili-
ation between first- and second-generation models, see Flood and Marion (1998).
7. Four countries are more open than the average (a mean index value of 0.26)
and all, except Brazil and Russia, are located within plus one standard error
from the mean (an index value 0.46). For the construction of the index, see
Tamirisa (1999).
8. For example, Aziz et al. (2000) exclude explicit fiscal determinants from the expla-
nation of currency crises; instead, fiscal policy is subsumed under monetary vari-
ables. On the other hand, Hemming et al. (2003) test statistically the contribution
of a range of fiscal variables, though without explicitly accounting for unrecorded
components.
9. Although fiscal policy is not central to the ongoing debate on hard versus soft
pegs, reviewed in S. Fischer (2001a), a fiscal sustainability problem in combina-
tion with high capital mobility tends to undermine exchange rate stability. At the
end of the spectrum, under a very hard peg (e.g. dollarization), the influence of
fiscal policy on the nominal exchange rate vanishes.
10. For a narrow interpretation of the fiscal role in the Mexican and Asian crisis, see
Sachs et al. (1996), Radelet and Sachs (1998) and Summers (2000). For a broad
interpretation, see Dooley (1998), Corsetti et al. (1999) and Burnside et al. (1999).
11. See the discussion of the traditional approach, for example, in Tanzi and Blejer
(1984).
12. For example, according to Begg (1998), the Czech crisis could have been prevented
with a substantial budget surplus. A general argument for preemptive tightening
of the fiscal stance, as the capital account is opened, can be found in Heller (1997).
13. See Ghosh et al. (2002). Initially, most programs provided for a fiscal impulse
between  0.5 and  1.5 percent of GDP. For Brazil and Thailand, the programmed
impulse was about  3 percent, while for Turkey it was roughly  8 percent of GDP.
It should be noted that the fiscal stance measures the actual budget deficit against
a counterfactual cyclically adjusted deficit; the fiscal impulse is defined as the
annual change in the fiscal stance.
14. This exercise is subject to data limitations, as well as to the difficulty of separat-
ing changes in economic environment (exchange rate, GDP growth, commodity
prices, and other exogenous shocks) and discretionary fiscal measures; see IMF
(1998, Box 2.5).
15. See an application of this approach to monetary policy in Drazen and Masson
(1994). Moreover, as shown in Krugman (2000) in a multiple-equilibrium context,
a fiscal contraction may actually ensure a crisis equilibrium.
16. For a critique of the traditional approach followed in IMF-supported programs
and a discussion of the need to focus on structural measures, see Tanzi (1989).
17. See, for example, Alesina and Ardagna (1998) and Kneller et al. (1999).
18. See, for example, the analysis of fiscal consolidation episodes in Denmark and
Ireland in Giavazzi and Pagano (1990), and Bertola and Drazen (1993). Similar
experiences can be found in developing and transition economies, including in
the context of Fund-supported stabilization programs: Ghana and Turkey in
the early 1980s, Chile since the mid-1980s and Poland in the early 1990s. For
example, the Turkish program, undertaken in the context of a wide-ranging
external liberalization, is documented in Kopits (1987).
 Fiscal Policy and High Capital Mobility 79

19. This view has been expressed in the recent statement submitted to the APEC
Finance Ministers by ADB, IADB, IMF, and World Bank (2001), which contains a
survey of social safety nets and their application in Asia and Latin America.
20. See, for example, Tanzi (1995).
21. In Brazil, the target for the primary surplus for 2004–06 is over 4 percent of GDP
for the public sector as a whole, while the public pension system totals a primary
deficit of close to 4 percentage points. In other words, the rest of the public sector
must generate a primary surplus of some 8 percent of GDP to accommodate the
imbalance of the pension system. Hence, the proposed extension of an increased
contribution obligation to retired civil servants, as part of a comprehensive pen-
sion reform package, is seen as a key measure to ease the public sector borrowing
requirement in the short term.
22. This bias may arise in the presence of an overvalued currency, compounded under
a currency-board arrangement which precludes government borrowing from
domestic sources – at the cost of fully crowding out domestic private activity.
This explains the rapid buildup of external public indebtedness in Argentina
during the 1990s. For estimates of the government deficit during that period, see
Teijeiro (2001).
23. For example, in Ecuador, until recently the absence of withholding taxation has
been a major obstacle to income tax collection or to a redesign of the income tax
system.
24. There are other areas that entail upfront fiscal costs including those associated
with environmental cleanup, healthcare reform, and civil service reform – espe-
cially in transition economies. While they are similar in several respects (i.e, as
implicit contingent liabilities) to bank restructuring and pension reform, their
connection with currency crises is far less direct and the magnitude in terms of
contribution to public sector liabilities is significantly smaller.
25. Whereas under rapid restructuring programs total costs ranged from less than
1 percent to 15 percent of GDP, under slower programs they varied between 6 per-
cent and 45 percent of GDP; see Dziobek and Pazarbasioglu (1998). For a review
of alternative restructuring and recapitalization schemes, and their fiscal impact,
see Daniel and Saal (1998).
26. See Mackenzie et al. (1999).
27. For example, in Mexico, the present value of the fiscal cost (not yet absorbed by
the budget) of the transition to the new system is estimated between 45 percent
and 80 percent of GDP. The yearly cost is envisaged to rise from over 1 percent of
GDP at present to nearly 4 percent of GDP by 2040.
28. Parameter adjustments usually include the increase and unification of the mini-
mum retirement age, redefinition of the pension base to reflect average lifetime
earnings, and broadening of the contribution base.
29. Tobin-type taxes on international capital movements or compulsory deposits on
capital inflows (as imposed until recently on short-term capital inflows in Chile)
are intended to serve as market-friendly capital controls, rather than as a revenue
function. Thus, they fall outside the scope of this discussion.
30. In its most common form, this tax has been in effect, for periods of up to nearly
six years, since the early 1980s in Argentina, Brazil, Colombia, Ecuador, Peru, and
Venezuela. Statutory tax rates ranged from 0.2 to 1.4 percent, with an annual yield
of as much as 3.5 percent of GDP; see Coelho et al. (2001). For an analysis of the
Brazilian experience, see Albuquerque (2001).
31. In Ecuador, the tax was introduced at a 1 percent rate along with a freeze on bank
deposits, thus aggravating the banking crisis with an added incentive to capital
outflow.
80 George Kopits

32. Since the early 1990s, a uniform import surcharge has been introduced in
Argentina, Bulgaria, Hungary, Poland, and Slovakia, with initial statutory rates
between 3 and 8 percent and an annual yield of up to 1 percent of GDP. In all
cases, the rate was gradually reduced until its elimination over a maximum period
of four years. Argentina and Hungary adopted the surcharge as part of a stabiliza-
tion program in the wake of the Tequila crisis, whereas Slovakia introduced it in
response to the Asian crisis. The other countries applied it mainly for revenue rea-
sons, absent sufficient tax handles.
33. In principle, however, the short-run macroeconomic impact of privatization is
equivalent to bond financing of the fiscal deficit, as shown in Mackenzie (1998).
While the short-run impact on the fiscal balance is negligible, if present at all, the
medium- to long-run fiscal consequences are beneficial on account of improve-
ments in allocative efficiency and X-efficiency.
34. See Kopits and Craig (1998).
35. In Argentina, unrealistic cash-based measurement of the government accounts
masked significant budget deficits in 1991–99 (estimated in Teijeiro 2001), which
were inconsistent with the currency board arrangement and eventually led to its
abandonment.
36. See the review of these reports (posted on the web) in Allan and Parry (2003).
37. Chile stands out as having one of the most advanced institutional infrastructures
since the mid-1980s, after having experienced a series of financial crises. Building
on sound constitutional and legal foundations, Chile’s fiscal management has
established a disciplined and transparent decision-making process – admittedly
made easier by the centralized unitary structure – supported by a stable and sim-
ple tax system. See Vial (2001).
38. According to the most generally accepted definition, sustainability obtains when
the debt–GDP ratio does not rise over time; see Buiter (1985). Alternatively,
Blanchard et al. (1990) define as sustainable a fiscal policy whereby the debt ratio
eventually converges to its initial level.
39. Everything else being equal, policy signaling is likely to be more effective at the
beginning of the electoral cycle, with greater scope for the authorities to carry out
their promises, than at the end of it.
40. For a definition and classification of fiscal policy rules, see Kopits and Symansky
(1998). Fiscal rules resemble monetary rules and exchange rules in the sense that
they are established to gain credibility in the eyes of financial markets and the
electorate.
41. For a formal demonstration, see Cukierman and Meltzer (1986), along the lines of
Kydland and Prescott (1977).
42. See the procyclical discretionary fiscal stance in Latin America, documented by
Gavin et al. (1996).
43. The rules enshrined in New Zealand’s Fiscal Responsibility Act or the European
Union’s Stability and Growth Pact go further, since the requirement of maintain-
ing balance or surplus over the medium term provides, in principle, some latitude
for discretionary countercyclical fiscal policy.
44. This approach is a simplified approximation of the rule proposed in Hausmann
et al. (1993). As a companion to the expenditure rule, surpluses (shortfalls) in oil
revenue above (below) a certain threshold would be deposited in (withdrawn
from) an escrow account managed by the monetary authorities following con-
ventional criteria of foreign exchange reserve management.
6
Fiscal Institutions versus
Political Will
Allen Schick

Introduction

Numerous contemporary studies have found that fiscal institutions strongly

influence budget outcomes, explaining why some countries maintain fiscal
discipline while others do not.1 The studies do not agree on the factors that
account for differences in budget results, but they do agree that rules matter.
These findings have spurred international organizations and some national
governments to seek new fiscal institutions to strengthen budget discipline
and keep public finance on a sustainable course.
Although recent institutional innovations2 have drawn the attention of
reformers, most fiscal studies focus on traditional procedures of preparing,
adopting, and implementing the annual budget. The most important
procedural rules pertain to the role and power of the finance minister
vis-a-vis sectoral ministers and other claimants on the budget, and to the
capacity of the legislature to amend the budget proposal submitted by the
government. The findings are generally in line with conventional wisdom:
a strong finance minister bolsters fiscal discipline, as does a weak legislature.
The studies also delve into basic governing arrangements, and generally
conclude that majoritarian regimes have more success than coalition gov-
ernments in balancing public finances. Some recent research, however, dis-
tinguishes between governments which enforce fiscal discipline through
contract-like commitments and those that delegate the task to budget con-
trollers (Hallerberg and von Hagen 1999). This distinction helps explain
why, contrary to expectations, coalition regimes are sometimes more fiscally
disciplined than majoritarian ones.
Inasmuch as the studies conducted thus far have focused on conventional
budget practices, evidence on the extent to which rule changes affect budget
outcomes is limited. To understand how fiscal rules interact with budget
results, it is necessary first to look at standard rather than best practices.
This chapter seeks to explain the anomaly that seemingly sound budget
processes often produce adverse or unwanted fiscal outcomes. It argues that

81
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
82 Allen Schick

political commitments play a large role in determining the effectiveness of

fiscal institutions, drawing on the experience of advanced economies that
have taken different fiscal paths. To conclude, the chapter addresses the
question of whether fiscal rules or political will is more important, particu-
larly in the circumstances of emerging market economies.

Limitations of conventional practices

Budgeting is the process of allocating government expenditure according to

prescribed rules. However, adverse outcomes such as unplanned deficits
make it appear that there are no rules in place or that they are ineffective.
In almost all countries, detailed procedural rules dictate the way spending
units bid for resources, the review of these bids, the forms used in compil-
ing the annual budget, the timetable for legislative action, and the proce-
dures for spending appropriated funds and reporting on financial stocks and
flows.
Before reforming these practices, it is important to understand why the
existing rules do not work. It makes no sense to reengineer budget practices
unless there is reasonable assurance that the changes will promote fiscal dis-
cipline. In addition, the apparent failure of existing rules calls into question
the premise that fiscal institutions drive fiscal outcomes. It may be that the
effectiveness of these rules depends not only on their design, but also on the
political and economic conditions under which they are implemented.

Due process in budgeting

As a process, budgeting is governed by distinctive procedural rules that cover
every step in the expenditure chain. Over time, the procedures have become
routines repeated yearly with little change. The routines are important
because they ease the task of allocating public money and reduce the conflict
inherent in the competition for scarce resources. Each government has its
particular language and forms, its distinctive procedures, and its own orga-
nizational chart. Early in the development of budgeting, the basic routines
were codified into good- or best-practice principles to be followed by
governments, regardless of their political-administrative culture.3 The basic
principles have been refined over time, but most have had remarkable
staying power. The most important principles of good budgeting are
comprehensiveness (inclusion of all revenue and expenditure), accuracy
(accurate recording of transactions), annuality (coverage of a single fiscal
year), authoritativeness (expenditure as authorized by law), and trans-
parency (timely publication of information on projected receipts and
expenditures).
These principles are enforced through detailed procedures specifying the
scope and format of the budget, the types of information processed through
it, the timetable for the budget cycle, the roles and responsibilities of
 Fiscal Institutions versus Political Will 83

participants in the process, the authorization required and controls maintained

in spending public funds, and so on. Compliance is maintained by budget
overseers at the center of government and in spending entities.
The rules and conventions constitute due process in budgeting. The term
“due process” implies that if procedures are properly applied, the budget
outcomes must then be accepted as correct. That is, the procedures used, not
substantive objectives or criteria, determine the legitimacy and propriety of
the results. Arguably, therefore, whatever results ensue from a well-run bud-
get process should be accepted as appropriate for the country.
Due process is procedurally indifferent to outcomes. It has no preference
for more or less spending, a balanced or unbalanced budget, rising or stable
public debt burdens, frozen budget priorities, or significant reallocations.
What matters is that the procedures are followed. In this sense, due process
in budgeting is analogous to due process in litigation: if proper judicial
procedure is followed, the resulting verdict must be accepted. Also, due
process is politically neutral. It can accommodate both left- and right-of-
center governments, as well as politicians who want to cut back or enlarge
the size of government. Commonly, an incoming government with a
markedly different political agenda than the government it has replaced
continues the inherited budget process.
A due process approach has important advantages. It establishes the basis
for financial regularity and accountability and stabilizes expectations as to
the tasks that budget participants perform. With the routines repeated year
after year, participants know what they have to do and when. These essen-
tial building blocks of public expenditure management are adhered to in the
conduct of substantive fiscal policy. No government can manage its finances
effectively if procedural due process is materially impaired.

Good process and bad outcomes

Although it is essential, due process is an inadequate basis for managing pub-
lic finances because it tolerates or generates unplanned, adverse outcomes.
The dismal fiscal performance of many developed and developing countries
impels one to conclude that good budget practices often produce outcomes
at variance with those sought by the government or deemed appropriate by
independent experts. For decades, international organizations have assisted
developing countries in installing sound budget systems, but in many cases,
the results have been consistently subpar. Most developing countries today
have formal budget systems that pass muster by international standards.
What they do not have are disciplined budgets, effective programs, or effi-
cient operation: improving the forms of budgeting is not likely significantly
to alleviate their deeply embedded pathologies.
Budget results generally appear to be more favorable in affluent
than poorer countries, but the differences may be more in perception than in
performance, and they usually owe more to the abundance of resources than
84 Allen Schick

to the quality of the budget process. Because few developed countries have
experienced the capital flight and economic destabilization that have beset
emerging market countries, they have been able to finance budget shortfalls
without facing pressure to overhaul their fiscal institutions. Rich countries
often credit their good fortune to fiscal discipline, but the truer explanation
lies in economic plenitude, not in fiscal rules and procedures.
Part of the explanation for why standard budget practices do not assure
disciplined fiscal results lies in the machinery of budgeting. Routines may
bias budgeting in favor of higher public spending. Although budgeting is a
means of rationing resources among competing uses, it invites spending
units to campaign for more money each year. It is a rare agency that requests
only as much as or less than it obtained for the previous year. The common
practice is for spending agencies to request increases, to have a portion of
the request denied by the finance ministry and in the end receive more
money than the previous year. Central budget officials receive credit for cut-
ting the budget and spenders gain funds to finance expansion.
It is not only that spenders want more; government leaders want to give
them more. Little opprobrium attaches to a government that tables a bud-
get with spending increases. This is a normal occurrence, built into the
expectations of budgeting and the behavior of participants in the process. In
fact, governments often point to spending increases as prima facie evidence
of the good they are doing. It is the budget cuts which stir political unrest
and analytic curiosity, not the increases over the previous years.
Budgeting is an incremental process that extends the past into the future
by focusing on year-to-year changes (Wildavsky 1964). It is near universal
for governments to format the budget in ways that facilitate interyear com-
parisons. In most governments, the budget and supporting documents show
spending for one or more past fiscal years, the year in progress, and the next
fiscal year. This structure formalizes incrementalism by focusing attention
on the amounts by which each budget varies from previous years.
Efforts by reformers to uproot incrementalism through zero-based bud-
geting and other innovations have been notoriously unsuccessful.
Incrementalism thrives because it simplifies the process by significantly
reducing the number and scope of decisions that have to be made within
the short time available for compiling and reviewing the estimates, and cur-
tails conflict by protecting agencies and interest groups from cutbacks in
existing programs. Yet, at the same time, incrementalism undermines bud-
get discipline because it impels government to accommodate fresh demands
by spending more, not by substituting new expenditures for old. So, too,
does the stickiness of public expenditures. In most developed countries,
more than half of central government expenditure is governed by perma-
nent laws that establish legal rights to benefits based on eligibility criteria
and payment formulas. These entitlements must be paid regardless of the
condition of the budget or other demands for public funds.
 Fiscal Institutions versus Political Will 85

The spread of entitlements has weakened budgetary due process. Typically,

entitlements have been enacted with little regard to downstream budget
impacts, and often without adequate information on their prospective cost.
In contrast to standard budget estimates and appropriations that are for
fixed amounts, entitlement programs usually are open ended; this tends to
transform budgeting into an accounting device for the past obligations of
government. By comparison, mandatory payments generally are less promi-
nent in the budgets of less developed economies where the typical fiscal
problem is a lack of resources, not an overabundance of commitments.
However, quite a few emerging market economies that have enacted bene-
fits in good times have been stuck with the costs when economic conditions
deteriorated.
The discussion of standard fiscal rules offers two broad conclusions. First,
due process in budgeting does not shield a government against adverse
financial outcomes. Second, budgetary incrementalism and sticky expendi-
tures bias budgeting toward higher spending and chronic deficits.
Contemporary efforts to devise robust fiscal institutions must deal with
these realities.

Do institutions matter?

The foregoing analysis places all countries in the same fiscal starting point.
Patently, however, countries differ in their fiscal performance. Efforts to
account for these differences have generated numerous studies, many of
which recently focused on the experience of the European Union (EU),
including under the Stability and Growth Pact (SGP).
Studies of fiscal institutions follow two main paths. One is to examine
electoral systems and the governing arrangements that ensue from them; the
other is to look at fiscal rules, particularly those pertaining to the distribu-
tion of budgetary power within government. A few recent studies focus on
the interaction of political regimes and fiscal rules. Most of them assign
weights to different elements of political regimes or fiscal rules, aggregate the
weights for all the elements, and correlate the score with fiscal outcomes.
This method reduces politics and finance to a few elements and ignores
much of the inherent complexity of political regimes and fiscal rules, slight-
ing informal or behavioral characteristics.
The template for the regime-oriented studies is a paper by Roubini and
Sachs (1989), who found a predictable relationship between governing
arrangements and changes in the ratio of public debt to GDP among OECD
countries. The authors concluded that “differing institutional arrangements
in the political process in the various OECD economies help to explain the
markedly different patterns of budget deficits in the different countries.”
The key argument is that weak, fragmented governments have difficulty
mobilizing support for fiscal austerity.
86 Allen Schick

In an alternative approach, focusing on budget institutions in EU member

countries, von Hagen (1992) differentiated between fragmented and central-
ized institutions at each of the three main stages of budgeting: compilation,
legislation, and implementation. Application of this approach to differences
in fiscal performance suggests that a budget process giving the prime minister
or finance minister strategic dominance over spending ministers, limiting the
amendment power of parliament, and leaving little room for change during
the executive phase is strongly conducive to fiscal discipline. Reviewing sub-
sequent research, Poterba and von Hagen (1999) reached similar conclusions.
Procedural design thus emerges as an important alternative to policy rules
restricting the outcome of the budget process, such as balanced-budget laws.
As sound and appealing as these conclusions may be, they are based on
research that is heavily skewed to formal elements and on an overly sim-
plistic notion of how budgets are decided and implemented. Each stage of
budgeting has subtleties and behavioral nuances, which institutional stud-
ies do not take into account. During budget preparation, regardless of the
formal power structure, the finance minister’s actual influence varies from
one budget cycle to another and has much to do with personality, relations
with other ministers, political cycles, and the budget situation. Fiscal targets
prepared by the finance ministry may serve as hard constraints one year and
platforms for demanding more money another year.
Institutionalists take the view that legislative independence and activity
on the budget is a prescription for fiscal irresponsibility. But the plain fact is
that the explosive growth in government spending and deficits during the
twentieth century occurred for the most part under legislative subservience.
With the rise of hegemonic political parties, legislatures lost their dominant
role in public finance. In some countries, they lost the power of amendment;
in others, their amendments nibbled at the margins of the budget, provid-
ing targeted, usually inexpensive, benefits to particular constituencies.
Furthermore, the main influence of parliaments has been through substan-
tive legislation that entitles broad swaths of the population to payments
from the government, not through annual appropriations.
Implementation of the budget cannot be analyzed independently of the
spending plan prepared by the government and voted by the parliament. In
some emerging market countries, the government prepares a budget that
authorizes more spending than can be accommodated with available
resources, and it is understood that the “real” budget will emerge during
implementation. In other countries (such as Japan), the original budget is
purposely kept small in expectation that significant supplementals will be
voted during the year. Rather than measuring the ease of modifying the bud-
get, it would be better to analyze the impact of supplementals within the
overall fiscal framework.
Fiscal rules are political instruments; they are made by politicians and
enforced or breached by them. Rules are not necessarily self-enforcing, and
 Fiscal Institutions versus Political Will 87

courts rarely intervene to stop violations. The effectiveness of budget

constraints depends on the willingness of political leaders to abide by them.
When the rules work, it may be because voters and politicians have a pref-
erence to be fiscally responsible. In fact, the difference between countries
that have strong fiscal rules and those that do not may lie in political pref-
erences rather than in the rules themselves (Poterba and von Hagen 1999).
It can be argued that rules matter when politicians are predisposed to act in
a fiscally disciplined manner, by making it easier for them to resist spending
demands. Rules strengthen politicians who want to be fiscally prudent,
but they do not stand in the way of those who are determined to spend more
than the rules allow. Political will spells the difference between rules that are
effective and those that are not. Rules have to be willed into existence, and
have to be sustained by political commitment. The institutional approach rec-
ognizes the importance of political commitment in regulating budget out-
comes, but it defines commitment as a key feature of fiscal rules rather than as
an enabling condition that gives the rules effect. In so doing, their reasoning
comes close to being tautological: commitments are fiscal rules that constrain
spending; when spending is not constrained, it is because commitment is lack-
ing. This construct leads to the conclusion that rules are always effective.
Interest in commitment has been spiked by evidence that, contrary to con-
clusions in the early literature on fiscal institutions, coalition regimes are
sometimes more disciplined than majoritarian governments. To explain this
anomaly, Hallerberg and von Hagen (1999) distinguish between delegation
and commitment. Governments that rely on delegation to enforce bud-
getary discipline empower the finance minister to set targets and decide the
estimates. This arrangement is not suitable for coalition governments where
the finance minister may come from one party and sectoral ministers from
other parties. When the coalition parties have conflicting views on govern-
ment policies and priorities, they will not entrust budget making to the
finance minister. Instead, if they are determined to act in a disciplined man-
ner, they may negotiate a coalition agreement that sets out the boundaries
of the budget and the main funding priorities for the life of the government.
During the 1990s, coalition agreements became more detailed in some coun-
tries (e.g. the Netherlands, as discussed below) and accounted for their suc-
cess at reining in public spending and curtailing the deficit. Coalition
agreements are effective when they are “credible commitments”; the parties
to the agreement have a strong incentive to abide by the terms, because oth-
erwise the government may collapse.
A budget commitment is worth no more than the willingness to comply
with its terms. Although the parties to an agreement may pay a political
price for violating it – new elections and a backlash from voters – politicians
may judge that it is better to break the constraints than to live with them.
Even when politicians are predisposed to abide by the agreements, one year’s
understandings may turn into next year’s misunderstandings. All budget
88 Allen Schick

commitments are at risk of being overtaken by changing conditions, such as

a weakening economy, rising revenues, international obligations, changes in
public sentiment, and so on. Therefore, commitments that cover only a sin-
gle fiscal year or the term of a government may have a better chance of being
honored than those for a longer time frame.

Budget cycles
The relatively short life of commitments may help explain why countries
may exercise fiscal discipline at some times and not at others. Commitment
depends as much on willingness to comply as on formal enforcement. When
the commitment wavers, budget discipline weakens. In fact, the natural fate
of budget commitments may be to erode over time as pent-up pressure for
more money overwhelms the rules. It may be that all fiscal rules are inher-
ently weak and made to be broken, either directly or by accounting tricks,
and that governments therefore have difficulty maintaining a disciplined fis-
cal posture over time. In some countries, the rules may have an effective life
of only a few years; in others, they may last a decade or longer. Sooner or
later, however, just about every country has a need to reinvigorate budget
procedures.
Case studies of zigzagging budget fortunes indicate the fragility of rules, as
illustrated by an in-depth study (US General Accounting Office 1994) of five
countries moved from large deficits to balance in the early 1990s. These
countries (Australia, Germany, Japan, Mexico, and the United Kingdom)
restructured their procedures, set top-down, multiyear limits on aggregate
spending, took steps to reduce the public-sector wage bill, and curtailed
some social benefits. A few also reduced payments to subnational govern-
ments or trimmed capital spending. In all the countries, political leaders
took an active role in promoting fiscal discipline, and persuaded voters to
accept fiscal austerity. These successes indicate that eliminating deficits is
possible in modern democracies and prompt action can be taken to avert a
crisis. The study also concluded that sustaining fiscal balance over the longer
term, however, is difficult. In fact, by the mid-1990s, four of the five coun-
tries had reverted to budget deficits. The common-pool condition that gen-
erates deficits does not go away when fiscal rules are tightened or when
surpluses occur. Formal rules may remain the same; budget behavior and
outcomes change.
It is useful to examine the experience of two advanced countries that have
had structural reversals in their fiscal condition. The Netherlands, once
deemed to be helplessly mired in deficits, has pursued fiscal discipline for
more than a decade and now has one of the strongest fiscal positions in
Europe. Germany, by contrast, once renowned for its sound fiscal policy, has
lapsed into chronic deficits. Nevertheless, both countries have some impor-
tant similarities. Both are governed by coalitions; both are high-benefit wel-
fare states; both made a strong currency the cornerstone of economic policy;
 Fiscal Institutions versus Political Will 89

both suffered fiscal distress in the early 1980s; and both installed
governments that promised to curtail spending and deficits, and had suffi-
cient time in office to stay the course. There are important differences,
however, especially Germany’s federal structure, which compels the national
government to cooperate with the länder (states) on most major budget
matters. Also, Germany has not perfected the coalition agreement as an
instrument of budget control, but relies instead on the finance minister to
steer fiscal policy.

From Dutch disease to Dutch model

As recently as the early 1980s, the Netherlands was regarded as the nesting
ground of the “Dutch disease,” a term that referred to a country living well
beyond its means and lacking fiscal discipline. Barely a decade later, it was
being celebrated as the “Dutch model,” a reform program that corrected
many of the structural imbalances in the economy and stabilized the gov-
ernment’s finances.4 Although the transformation occurred within a seem-
ingly short time, it actually unfolded in a gradual and sustained manner
through a handful of election cycles.
What lesson does this model hold for budgetary discipline? The
Netherlands had several distinctive characteristics hard to replicate else-
where, but sustained political commitment was essential for lasting fiscal
discipline. The aim has been to produce structural change in the budget
without stirring social unrest. The key features of Dutch-style fiscal disci-
pline can be summarized as follows: two decades of uninterrupted growth,
supported by fiscal adjustment and structural reform; political stability and
continuity, conducive to responsible budgeting; government consultation
with “social partners” to forge consensus in support of policy change; con-
sistency of budget implementation with its fiscal targets; and sustained fis-
cal discipline during economic upswings.
All told, the Netherlands’ success story can be characterized as due either
to new fiscal institutions or to political resolve. One can label commitment
as a fiscal rule, as institutional economists do. But it should be clear from the
Dutch experience that without political backing, carried forward by five suc-
cessive governments over a span of two decades, coalition agreements would
have been mere scraps of paper, stringent fiscal rules would have been
evaded, and the outcome would have been much different from what actu-
ally occurred.

From fiscal prudence to imbalance

Germany exercised strong fiscal discipline during the postwar period, took
effective steps to liquidate deficits when they emerged, and had a relatively
low public debt. But over time, public finance was driven by generous
social transfers and elevated government spending. During the 1990s,
these embedded expenditures combined with costly unification and slow
90 Allen Schick

economic growth, boosted public deficits and debt to the point that the
government, which had pushed for the establishment of the SGP, now
has difficulty (along with France and other countries) meeting the obliga-
tions under the pact. The deterioration in the fiscal position may be
explained simply by the fact that political commitment to discipline has
evaporated under the pressure of unification and aging.
Institutionalists, by contrast, attribute the emergence of deficits to a fun-
damental decay of budget institutions (von Hagen and Strauch 2001). Key
unification decisions, including costly decisions to give parity to East
German currency and to equalize social benefits, were made hastily, outside
regular budget channels. The chancellor concentrated policy work in his
office, and often relied on his own staff, task forces, and roundtables to make
budget-impacting decisions. Some of the costs of unification were hidden in
special funds that were outside the scope of the budget. The government
resorted to supplemental budgets when the amounts provided in the regu-
lar budget did not cover the year’s full costs.
Budget institutionalism is often an exercise in circular, self-fulfilling logic.
If the outcome is favorable, it must be because the institutions are strong; if
it is not, it is because the institutions have been bypassed or weakened. This
circularity hinges on defining political commitment as the most critical
institution; this is one of the ephemera of budgeting, ever changing from
one budget season to another. Budget institutions, as the label “institution”
denotes, should be seen as regular, ongoing practices such as the basic
budget routines discussed earlier, that change only slowly. In contrast to
institutions, substantive budget results do fluctuate from one year to the
next because political and other conditions change.

Interaction of political will and fiscal institutions

Budget institutions make a difference in budget outcomes. Ignoring the

importance of institutions is as mistaken as disregarding political will. But it
is necessary to disentangle the two, and to examine each separately, as well
as the ways they interact with each another. Rules may make it easier or
harder for politicians to maintain a disciplined budget policy. The concept
of “credible commitments” describes rules that affect political behavior by
changing expectations about the future (Williamson 1983, 1996). Trust
funds are an example of credible commitments in the budget, for though
governments can terminate or siphon money from trust funds, the estab-
lishment of these funds generates commitments as to how the accumulated
reserves will be used (Patashnik 2000).
There are two pending issues pertaining to the interaction of fiscal rules
and political will. First, what do we know about the types of rules and con-
temporary innovations that fortify political will and improve budget results?
The answer is: not very much, but it is worth setting out what we do know.
 Fiscal Institutions versus Political Will 91

Second, do rules and political will differ in highly developed countries

compared to emerging market economies? As indicated earlier, the facile
assumption that all countries are in the same institutional boat, regardless
of their political and economic development, is not tenable.

Rules that constrain political will

Any rule that bars certain actions, requires that actions be reviewed by inde-
pendent entities, or raises the political cost of acting, relative to the benefits,
impacts on political will. For example, politicians may be barred by a bal-
anced-budget rule from submitting a budget in which outlays exceed
revenue. The rule is more powerful if it has constitutional rather than leg-
islative status, but it may be even more powerful if it is an embedded norm
that has been upheld for generations. Some governments bar politicians
from making the official revenue forecasts, and entrust this responsibility to
an independent board or to nonpolitical experts. Not all action-barring rules
are equally effective. A rule that requires actual budget balance may induce
more discipline than one that merely requires a planned balance. Even when
these rules are effective on paper, they may be imperfectly enforced. A strict
fiscal rule may produce both more discipline and more evasion. When rules
are lax, there is no need to evade them; when they are stringent, however,
politicians may seek to spend more than they are supposed to.
Requiring that the actions of politicians be reviewed by independent
actors may strengthen enforcement. Auditors review financial statements to
assure that generally accepted accounting principles have been adhered to.
These reviews are based on a common presumption that left to their own
will opportunistic politicians will seek to evade the rules.
The final means of influencing fiscal behavior is to raise the cost or lower
the benefits of evading budget discipline. When budget commitments are
credible, as may be the case with certain trust funds, politicians may be penal-
ized by voters for failing to abide by promises. Many contemporary budget
innovations seek to alter the cost–benefit ratio of budgetary politics. For exam-
ple, a requirement that the medium-term cost of policy initiatives be esti-
mated may discourage politicians from shifting expenditures into future years.
However, these rule changes are effective only when budgeting is transparent,
the media and interest groups are attentive, and citizens feel they can influ-
ence public policy. Many emerging market economies lack such conditions.

Contemporary budget innovations

Two of the most prominent contemporary reforms – fiscal targets and
medium-term expenditure frameworks (MTEFs) – demonstrate the difficulty
of changing budget behavior when opportunistic politicians are in control.
In some cases, targets are substitutes for genuine fiscal discipline when the
government is incapable of living within its means, but nevertheless
wants to convey the notion that it is fiscally prudent. In some countries, the
92 Allen Schick

targets are so accommodating that they are not real constraints; in others,
the targets are so beyond reach that politicians give up without trying.
Fiscal targets need an array of supporting institutions to be effective. They
need realistic, implementable budgets, honest and timely accounts, and
comprehensive budgets that do not permit spending to be hidden in special
accounts. They also need independent enforcers, self-enforcing restrictions
on political decisions, or incentives that induce compliance. These condi-
tions are most likely to be present in countries that need targets the least,
and least likely to be present in countries that need them the most.
Multiyear frameworks are supposed to ensure that the medium-term
implications of budget decisions are accurately accounted for and that policy
initiatives comply with constraints on budget aggregates. The clear intent is
that the MTEF should be a ceiling, but in the hands of wily politicians, it is
sometimes treated as a floor. Despite the active promotion of MTEFs, there
have not been many success stories. When the MTEF does work, it is likely
because of a political decision to constrain spending, not because of the
mechanics of medium-term frameworks.
Some contemporary innovations strengthen fiscal discipline by constrict-
ing political opportunism. One such reform is conversion of accounting
standards to the accrual basis. It should be noted, however, that few coun-
tries budget on an accrual basis.5 Moreover, if a country has shoddy account-
ing practices and accumulates large, unreported arrears, as some poor
countries do, accrual-based financial statements will be no more reliable
than cash-based ones.
It makes sense for governments to budget for contingent liabilities. These
off-budget transactions are rarely recognized in the budget until they come
due, at which time it is too late to regulate them. At present, few govern-
ments systematically estimate the risks they are holding or set aside funds
for such contingencies. This is an area where new systems may have to be
devised to supplement conventional budget practices.

Emerging market economies

Budget innovations are like migratory birds; they move from one place to
another. They typically germinate in highly developed economies, then
spread to emerging market economies, and finally over time are transported
by reformers, consultants, and international organizations to less developed
economies. The driving assumption is that what works in one setting should
also work in other settings because the basics of fiscal discipline and budget
management are pretty much the same in all countries. Furthermore, by
adopting the latest and most advanced practices, countries can accelerate
their development and improve government performance.
There are examples, though admittedly few, of emerging market
economies that have successfully adopted contemporary institutional
 Fiscal Institutions versus Political Will 93

innovations – including procedural and policy rules – backed by unflagging

political will and broad-based consensus. In this respect, Chile and Estonia
stand out among these economies as cases of solid macroeconomic and fis-
cal performance.6 Although of a more recent vintage, Brazil has so far
applied successfully its fiscal responsibility legislation, in part facilitated by
major reforms in social security and taxation. This case is a vivid illustration
of the interplay between recently introduced institutions and political will at
all levels of government. It is often noted that, at present, Brazilian politicians
across the ideological spectrum abide by the new rules-based constraints,
which in turn enjoy popular support, but in the recent past Brazil has passed
through boom and bust periods, during which corrective fiscal policy has been
followed by fiscal profligacy. It is much too early to discern whether fiscal dis-
cipline in Brazil will survive political change and economic cycles.
To the extent, however, that political and economic conditions differ,
reform may have difficulty taking root in the less hospitable conditions of
many developing countries. In these countries, basic practices should be
given priority over avant garde innovations. For example, reliable cash
accounting should be in place before a government tries to introduce accrual
accounting. It should have a realistic annual budget before it budgets on a
multiyear basis. It should have sturdy financial and management controls in
place before it adopts a devolved system of management. It should have a
good measure of outputs before it reaches to measure outcomes.
Emerging market economies face fiscal issues that differentiate them from
both affluent and poor countries. Because they are not yet fully developed,
they are vulnerable to destabilizing fiscal shocks. Because they are not poor,
they have substantial fiscal capacity, generate significant tax revenue and
have the organizational resources and professional skill to manage public
finances. Nevertheless, these countries may encounter enormous difficulties
as they progress toward developed status.
As an emerging market economy develops, demand for public services
tends to rise and spending increases as a share of GDP. Some of the increase
is used to improving public services, but much is spent on transfer pay-
ments. Development boosts expectations and triggers a “Wagner’s Law”
expenditure boom. Moreover, development itself intensifies pressure on gov-
ernment to cope with urbanization, population shifts, environmental dete-
rioration, and other costs. The new spending does not displace the old, and
these countries often use tax collections for subsidies, state-owned enter-
prises, and a bloated civil service. These countries also are likely to drag their
feet in uprooting grand and petty corruption. As a consequence, fiscal stress
is likely to rise as they edge toward the developed ranks.
Caught in a vicious circle, not all of their own making, emerging market
economies are often blamed for being fiscally imprudent even when they
have behaved responsibly. As indicated in Chapter 3 by Hausmann, good fis-
cal behavior is not always recognized or rewarded. Clearly there must be
94 Allen Schick

market imperfections that explain why good fiscal behavior is not requited.
Thus a fundamental task is to counter these imperfections by building robust
capital markets and by implementing strong fiscal rules, during good times,
to put these economies on a sustainable course and buffer them against
shocks.

Notes

1. See Poterba and von Hagen (1999), Alesina and Perotti (1995), Alesina et al. (1999),
Roubini and Sachs (1989) and von Hagen (1992).
2. According to the US General Accounting Office (2000), about half a dozen coun-
tries have placed their budgets on an accrual basis, though a larger number now
prepare accrual-based financial statements. Many countries have expanded the
time horizon of budgeting, but few have formal MTEFs which explicitly limit total
spending for each of the next several years. A growing number of countries list con-
tingent liabilities in financial statements or other documents, but few actually set
aside funds in the budget for calls on guarantees or other contingent liabilities.
3. The classical statement of budget principles was compiled by Rene Stourm in the
late nineteenth century. See Stourm (1917) and Sundelson (1935).
4. See OECD (1988) for a discussion of the “Dutch model.” Major changes in gov-
ernment policy are discussed in Schick (1993).
5. The US General Accounting Office (2000) notes that even countries that budget on
an accrual basis do not include pension liabilities in their budgets.
6. For an overview of recent experience in Latin America and Central Europe, see
Kopits (2002).
Part II
Design Issues at the
National Level
7
EMU Fiscal Rules:
What Can and Cannot be Exported
Marco Buti and Gabriele Giudice1

Introduction

A basic feature of Europe’s Economic and Monetary Union (EMU) is strong

fiscal discipline. The Maastricht Treaty sets tight public finance requirements
for joining the euro area while the Stability and Growth Pact (SGP) makes
fiscal prudence a permanent feature within the area.
The rationale for the EMU fiscal rules can be found in the fiscal policy fail-
ures in Europe during the 1970s and 1980s: high and persistent budget deficits
feeding a rising stock of public debt; tendency to run a procyclical policy
which, instead of smoothing the business cycle, has accentuated its swings;
and finally, high share of public sector in the economy, hand in hand with a
rising tax burden that has hampered efficiency, growth, and job creation.
In the run-up to EMU, the Maastricht-cum-SGP framework has been
widely debated. Some, especially in the academic community, have pointed
to its excessive rigidity: the loss of national monetary independence may
hamper cyclical stabilization. On the opposite side of the spectrum, others
have pointed to the weakness of sanctions in the event of fiscal misbehav-
ior, which may eventually take Europe back to its pre-Maastricht years.
While the EMU is still in its infancy and some of its institutional features
are not yet fully consolidated, several lessons can nonetheless be drawn from
the design and implementation of its fiscal rules. The aim of this chapter is
to review the EMU’s fiscal policy framework with a view to identifying its
strengths and weaknesses and assessing its potential exportability outside
the European Union (EU) to emerging market economies in general, and
Latin America in particular.

Maastricht fiscal rules

Why fiscal rules in the EMU?

In principle, fiscal rules can be justified either to internalize spillovers or to
protect national interest. The spillover argument is particularly relevant in a

97
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
98 Marco Buti and Gabriele Giudice

currency area formed by sovereign countries. Spillovers occur either directly

between fiscal authorities or indirectly via the impact of national fiscal
policies on the single monetary authority (Buti 2003). The perception of a
less steep EMU interest rate schedule may lead to an overly expansionary
fiscal policy and an excessive accumulation of public debt with conse-
quences on the overall cost of borrowing and default risk. Even though the
commitment of the European Central Bank (ECB) to price stability is crucial
in preventing such an outcome, the commitment is itself a function of fiscal
behavior. To protect stability, the EMU framework was set up as a screening
device to ensure that only countries with a sufficiently good fiscal track
record could enter the euro area.
Rules justified by national interest are intended to tie governments’ hands
in a binding supranational agreement. They help counter the factors that have
led to fiscal profligacy and a deficit bias in the domestic political game. Given
the relentless increase in the stock of debt, the need to regain a sustainable fis-
cal position was increasingly recognized even in countries traditionally char-
acterized by weak fiscal performance. The argument of having to make painful
budgetary retrenchments “for the sake of Europe” was used to win support of
reluctant public opinion (McKinnon 1997; Buti and Sapir 1998).
In order to achieve and sustain fiscal prudence, fiscal rules can take the form
of numerical targets or procedural norms. Numerical targets impose a perma-
nent constraint on fiscal policy by setting specific targets or an upper ceiling
on key performance variables (spending, borrowing, debt). Their severity
depends on the degree of coverage of the government sector, the fiscal indi-
cator chosen, and the threshold being targeted. Procedural norms involve insti-
tutional arrangements according to which government budgets are presented,
adopted, and executed. “Hierarchical” procedures are more conducive to fis-
cal discipline than “collegial” procedures. At the national level, hierarchical
rules attribute strong authority to the finance minister to overrule spending
ministers during the intragovernmental preparation of the budget, and limit
the ability of the parliament to amend the government’s budget proposals.
Numerical and procedural rules have both proven effective to achieve and
sustain fiscal discipline. Eichengreen (1994) finds that the statutory and con-
stitutional deficit restrictions exert a significant restraining influence on fis-
cal behavior of U.S. states. Von Hagen (1992) and von Hagen and Harden
(1994) provide empirical evidence for Europe which suggests that procedural
rules help avoid excessive government spending and deficits.
A drawback of numerical rules is the incentive for creative accounting,
which entails a loss of information about the government’s true budgetary
situation and may reduce the credibility of the commitment to fiscal disci-
pline. Still, as empirical evidence for U.S. states shows, accounting devices
do not appear to be the primary source of deficit reduction in the longer run
(Poterba 1996). Anyway, to prevent circumvention, the targets and, more
broadly, the accounting framework, need to be simple and transparent.
 EMU Fiscal Rules 99

The choice between numerical and procedural rules depends on several

factors. Von Hagen and Harden (1994) find a clear correlation between the
size of a country and the nature of its commitment to fiscal discipline: the
larger EU member states, such as Germany and France, which were relatively
successful in maintaining fiscal discipline during the 1980s, relied on pro-
cedural rules, while the smaller countries opted for numerical targets. While
numerical targets and procedural reforms can be alternative options to bud-
getary prudence, they are not mutually exclusive in practice, and are fre-
quently implemented in parallel. In the case of the EMU, while numerical
targets have a clear primacy, procedural rules are also called upon to ensure
compliance with the budget constraints.

Rules on budget deficit and debt

The Maastricht Treaty requires a high degree of sustainable convergence for
admitting a member state to the monetary union, in terms of two criteria:
the government deficit should not exceed the reference value of 3 percent
of GDP (unless justified by exceptional circumstances), and the government
debt stock should not exceed the reference value of 60 percent of GDP (or at
least the debt ratio should be on a decreasing trend at a satisfactory pace).
In addition, the treaty rules out monetary financing and privileged access to
credit by public authorities.
When a country is subject to a decision of the EU Council of Economy and
Finance Ministers (ECOFIN) on the existence of an excessive deficit, a pro-
cedure aimed at correcting this situation is initiated. This procedure includes
several steps designed to increase pressure on the member state to take effec-
tive measures to curb the deficit. If such corrective measures are not imple-
mented, sanctions may be applied.
While the Treaty recognizes the importance of effective national bud-
getary procedures, their design and application is left to the member states
themselves. The combination of a harmonized accounting framework and
the need to meet the numerical targets has led to significant reforms in
domestic procedures conducive to budgetary discipline (European
Commission 2001a; J. Fischer 2001; Fischer and Giudice 2001).

Have the Maastricht rules been effective?

Size of the adjustment

The imposition of the Maastricht fiscal criteria set off a genuine consolida-
tion process in the euro area. Budget deficits have declined substan-
tially since 1993, during which the euro area registered the historically
high deficit ratio of 6 percent of GDP (Figure 7.1). Aided by lower interest
rates thanks to reduced risk premia, the deficit fell by 3.5 percentage
points to below the 3 percent of GDP threshold in 1997. However, the pace
of consolidation has slowed down considerably and structural balances
100 Marco Buti and Gabriele Giudice

15
Budget
10 balance + 3%

5
Public debt – 60%
0
D A
E Euro NL IRL
–5 F P area
FIN B
I
–10
EL
–15
–40 –20 0 20 40 60 80 100

Figure 7.1 Public finance convergence in the euro area (percent of GDP), 1993–2000
Sources: European Commission and authors’ estimates.

stopped improving as of 1999 in several member states. The public debt–GDP

ratio also declined in most countries, though in some of them very slowly; in
Germany and France, the debt ratio actually increased in recent years.
Von Hagen et al. (2001) examine whether the consolidation represents a
Maastricht effect, that is, whether the convergence process created its own
political dynamics of fiscal adjustment. The authors find that most of the
consolidations that began before 1995 in the euro area could not have been
predicted by a model estimated over past data, confirming that the
Maastricht process did create some political pressure to undertake fiscal con-
solidations, mainly in the first half of the 1990s.
The sheer size of the adjustment may have induced favorable non-
Keynesian effects in some countries, thereby helping to sustain the retrench-
ment effort. As argued first by Giavazzi and Pagano (1990), there is a
nonlinearity between fiscal adjustment and economic activity: although tra-
ditional Keynesian effects dominate in the event of small cuts, confidence
and crowding-in effects may help in offsetting the direct reduction in
demand in response to larger adjustment packages.2

Composition of the adjustment

The composition of the consolidation appears to play an important role in
determining its success. There is increasing evidence that deficit reductions
that take place through expenditure cuts, rather than tax increases, have a
much higher probability of reducing the stock of debt and permanently
reduce the deficit.3
In order to capture the composition of the adjustment, discretionary
policy changes since 1993 are separated into changes in revenue and
primary expenditure (Figure 7.2).4 In practically all countries the cycli-
cally adjusted primary balance improved. Only Portugal and Greece – starting
 EMU Fiscal Rules 101

15
Expenditure-induced Deterioration
deterioration
Tax rises cum
deterioration
Change in cyclically adjusted

Revenue-induced EL
P
primary expenditure

deterioration
Consolidation
D
0 IRL
A F
I B Euro area
Revenue-based
NL E consolidation

FIN
Tax cuts cum Expenditure-based
consolidation consolidation
–15
–15 –10 –5 0 5 10 15
Change in cyclically-adjusted total revenue

Figure 7.2 Composition of the fiscal adjustment in the euro area (percent of GDP),
1993–2000
Sources: European Commission and authors’ estimates.

from a low level of total revenue – pursued a revenue-based retrenchment,

while several countries combined discretionary cuts in spending with
a reduction in tax revenue, thus reducing the overall size of the public sector.
In sum, the fiscal adjustment appeared to be of good quality. Moreover,
during the consolidation process, the composition of the adjustment tended
to improve, as in a number of countries where initially the adjustment was
revenue based, it later became expenditure based.

Stability and Growth Pact

While the Treaty establishes the requirements for admission to the euro area,
the SGP creates the conditions to make fiscal discipline a permanent feature
of EMU.5 The SGP consists of a preventive arm, which aims to strengthen
the surveillance of budgetary positions and the surveillance and coordina-
tion of economic policies, and a dissuasive arm, which aims to accelerate
and clarify the Excessive Deficit Procedure (EDP) under the Treaty. It also
includes political guidelines to implement the SGP in a strict and timely
manner, spelling out the responsibilities of each institutional actor (ECOFIN,
Commission, and member states).6
102 Marco Buti and Gabriele Giudice

Prevention
The SGP prescribes that the medium-term budgetary position must be “close
to balance or in surplus.” This allows the full operation of automatic stabi-
lizers in recessions without exceeding the 3 percent of GDP reference value
for the deficit – which is to be seen as a ceiling, not a target.
To facilitate monitoring national budgetary developments and make an
early identification and assessment of possible risks, member states having
adopted the single currency submit “stability programs” while the other sub-
mit “convergence programs.” These programs, which are made public, indi-
cate the medium-term objective for the budget balance, the adjustment
path, and policy measures to attain the objective. In the latest updates, long-
term sustainability issues are also covered. ECOFIN is committed to carrying
out the examination and may deliver an opinion on the programs and their
updates – on a recommendation from the Commission – within at most two
months of their submission.
These programs represent the key element of the enhanced surveillance
introduced with the SGP (Fischer and Giudice 2001). They provide a trans-
parent frame of reference for fiscal monitoring at the EU level and as such
allow for a consistent cross-country assessment of budgetary developments
and policies. In this context, particular attention is given to possible “sig-
nificant divergences” of budget outcomes from the medium-term objective.
Should significant slippage from the targets set in the programs be identi-
fied, ECOFIN can issue an “early warning” recommendation urging the
member state concerned to take adjustment measures.
Dissuasion
The SGP contains provisions to speed up and clarify the EDP, in order to dis-
courage excessive deficits and, if they occur, to further their prompt correc-
tion. Above all, it specifies when a deficit above 3 percent of GDP is not
considered excessive, as well as the extent of the sanctions in case of persis-
tent excessive deficits.
A deficit is deemed excessive if it is higher than the reference value “unless
the excess over the reference value is only exceptional and temporary and
the ratio remains close to the reference value.” Exceptionality can be invoked
when the excess results from an unusual event outside the control of the
member state in question or a “severe” economic downturn. An excess is
considered temporary if budgetary forecasts provided by the Commission
indicate that the deficit will fall below the reference value following the end
of the unusual event or the severe economic downturn. Closeness to the ref-
erence value was not defined in the SGP, the reason presumably being that
no member state wanted to prejudge the level of the deficit acceptable for
qualifying for euro-area membership.
The EDP is subject to a tight timetable so as to arrive at a speedy decision
on the existence of an excessive deficit.7 Finally, the SGP spells out the type
 EMU Fiscal Rules 103

and scale of pecuniary sanctions in the event of a persistent excessive deficit

in a euro-area member. So far the implementation of the EDP has never
arrived at the sanctions stage.

Will the pact work?

While the Treaty has been effective in the run-up to EMU, the question remains
about whether the SGP will work too. In order to provide an answer, Buti and
Giudice (2002) and Buti et al. (2003) provide an assessment of the Maastricht
convergence criteria and the SGP against a number of desirable features
identified by Kopits and Symansky (1998) and Inman (1996). Both papers
conclude that the SGP put flesh onto the bones of the Treaty, leading to better
rules and procedures, although at the cost of somewhat more complexity.
From political economy perspective, the main ingredients of the Treaty’s
success were its public visibility, clear structure of incentives, strong politi-
cal ownership, constraining calendar, effective monitoring, and collegial cul-
ture.8 The SGP clearly strengthened the monitoring of fiscal policies and the
collegial culture within the multilateral surveillance of economic policies.
However, the political ownership of the SGP shifted toward countries with
structural surpluses, which, although numerous, have a relatively small
weight in the euro area. Relative to a simple deficit ceiling, the close-to-
balance rule enjoys lower political visibility and its specification is more con-
troversial. Similarly, and probably more important, the structure of
incentives has changed with the move to a single currency: market incen-
tives have been reduced with the convergence of interest rates; the carrot of
entry has been eaten while the stick of exclusion has been replaced by the
threat of uncertain and delayed pecuniary sanctions.
While the jury is still out on the effectiveness of the SGP in securing fiscal
discipline, the gloomy predictions that the consolidation of the 1990s was
simply an opportunistic move for countries to be admitted to the euro area,
and thus large deficits would reappear, have not materialized. The SGP,
however, faces the challenge of governing fiscal policy in a currency union –
necessarily a complex task in a regime of decentralized fiscal responsibility.
The outcome will depend on how the open issues in implementation are
tackled.9 Among these issues, paramount is the definition of a medium-term
position of “close to balance or in surplus.” Other issues are the asymmetric
working of the pact, the different degrees of fiscal stabilization across coun-
tries, the necessity to secure sustained discipline at all levels of government,
the quality and sustainability of public finances, and the implications for the
new EU members from Central and Eastern Europe. 10
In order to illustrate the relevance of the definition of the medium-term
target for output stability, Buti and Giudice (2002) laid out a simple model
which proxies the basic features of EMU rules on the behavior of individual
countries where output is subject to transitory shocks and can deviate
104 Marco Buti and Gabriele Giudice

temporarily from a fixed potential level. Given the simplicity of the analytical
framework, the results obviously should not be overstated. Nevertheless, a
powerful policy message emerges on the implementation of the SGP frame-
work: if fiscal authorities want to achieve output stabilization while main-
taining fiscal discipline, they have to focus on structural rather than actual
deficits; as a corollary, they must create the necessary room for maneuver,
especially if they have a preference for active fiscal management.11 Such an
interpretation would reinforce the role of the pact as commitment technol-
ogy to free national fiscal policies from the burden of high deficits and debt
which have in the past hampered their use for stabilization purposes (Buti et
al. 1998). The European Commission (2002c) has adopted this approach in
the evaluation of fiscal policies of EU member states.
Emerging market economies, and in particular Latin American countries,
are affected by large cyclical fluctuations, low budgetary stabilization, and a
tendency to run procyclical policies. This implies that the message under-
pinning the SGP becomes even more crucial for countries in Latin America:
stabilizing high output while ensuring fiscal discipline might be achieved
only if they focus on attaining and sustaining structural surpluses so as to
have a sufficient safety margin for fiscal stabilization in bad times.

Fiscal performance in Latin America

High macroeconomic volatility and limited government creditworthiness

influence the conduct of fiscal policy in Latin American countries. While
output volatility in the euro area has decreased over time, in Latin America
it has remained twice as high (Table 7.1). In addition, the average dispersion
of growth rates across countries in a given year has been much higher in
Latin America than in the euro area since the 1970s, although to a lesser
extent in the 1990s.
Several explanations have been offered to explain this high volatility:12
strong dependence on the export of one or a few commodities, with quite
volatile prices; exposure to lending booms which make these countries vul-
nerable to financial and currency crises; and policy-induced shocks, given
inadequate central bank mandate and fiscal institutions, or inappropriate
exchange-rate regimes.
High macroeconomic volatility and precarious creditworthiness tend to
combine with some intrinsic features of fiscal policies in Latin America in a
“mutually, self-reinforcing, vicious circle” (Gavin et al. 1996). Typical charac-
teristics of public finances in Latin American countries are high budgetary
imbalances, volatile tax bases, a procyclical bias, and weak budget institu-
tions. Some of these characteristics are more evident when comparing key
fiscal variables in Latin America and the euro area in 1997 and 2001 (Table 7.2).
Although the deficit–GDP ratio in Latin America does not show a stark dif-
ference from that of the euro area, the fiscal imbalances are substantially
 EMU Fiscal Rules 105

Table 7.1 Macroeconomic volatility in Latin America and the euro area, 1970–2001

1970–01 1970–80 1981–90 1991–01

Weighted average of standard

deviation of GDP growth over
the period of each country within
the area
Euro area 2.4 2.7 1.9 1.7
Latin America 4.4 3.5 4.6 3.2
Period average of the yearly
standard deviation of GDP growth
across countries in each area
Euro area 2.0 2.2 1.6 2.0
Latin America 3.6 3.9 4.4 2.8

Note: Latin America includes Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Mexico, Peru,
and Venezuela.
Sources: European Commission (2001b) and IMF (2001e).

Table 7.2 Public finances in Latin America and the euro area, 1997 and 2001

Latin America
Euro Area

1997 2001 1997 2001

Budget balance*
percent of GDP 2.4 2.4** 2.6 1.1
percent of revenue 8.8 8.7** 5.5 2.4
Public debt
percent of GDP 32.5 40.0 75.3 68.8
percent of revenue* 125.5 151.5** 158.2 147.0
Revenue*
percent of GDP 25.9 26.4** 47.6 46.8
Interest payments
percent of revenue* 20.6 23.9** 10.7 8.3
percent of exports of goods 25.5 23.5 15.9 10.3

Notes


Consolidated government.



General government.
* Excluding Bolivia, Uruguay and Venezuela.
* * Data from 1999 for Brazil and from 2000 for Ecuador.
Sources: European Commission (2001b), Global Insight (2001), and authors’ calculations.
106 Marco Buti and Gabriele Giudice

higher when measured as a share of total revenue. As pointed out by Gavin

et al. (1996), the same pattern would emerge if the deficit were expressed as
a share of M2 or other financial market variables. This indicates that the
capacity of Latin American countries to carry out fiscal retrenchment via tax
increases or to finance budget deficits on the domestic financial market is
limited compared to that in the euro area.
Large deficits in Latin America coupled with high interest rates due to bad
credit ratings have led to large debt levels relative to the capacity of the coun-
tries to service them. While the average ratio of debt–GDP is smaller than that
of the euro area, relative to total revenue the debt stock is of a similar magni-
tude. However, given the higher interest rates, the resulting interest burden is
much higher. In addition, because of the inability of these countries to borrow
in domestic currency at long maturities, the debt either is dollar denominated
or has a short-term maturity. This debt structure exposes Latin American coun-
tries to shocks to short-term interest rates and exchange rates.13
Compared to the euro area, a larger portion of revenue in Latin America
is made up of trade taxes and nontax revenues, which have very volatile
bases. A relatively large share of spending consists of investment, interest
payments, and wages, the latter two being difficult to reduce in the short
run. Highly volatile tax bases and rigid spending make public finances vul-
nerable to domestic and foreign shocks and often require a large discre-
tionary adjustment. Unlike in preceding decades, since the 1990s inflation
is no longer the main adjustment mechanism.
Fiscal policies tend to behave procyclically in Latin America, especially
during recessions. By comparison, the procyclical behavior observed in the
euro area, especially in countries with large public finance imbalances and
high debt ratios, has been considerably milder. Talvi and Végh (1996)
explain the procyclical bias in Latin America in terms of the high share of
public consumption. An important role is played by the access to interna-
tional financial markets which, already scant in normal times, virtually dis-
appears in face of adverse shocks (Gavin et al. 1996); thus, fiscal contractions
become inevitable when the economy is affected by a negative shock. In
periods of financial distress, however, a procyclical fiscal contraction may
even be desirable, as it tilts the distribution of the limited credit in favor of
the private sector (Caballero 2000).
Other explanations of the procyclical bias pertain to the realm of political
economy. Lane and Tornell (1998) interpret the rise in consumption com-
pared to output in response to a positive shock as the outcome of strategies
of powerful groups, creating a sort of “voracity effect.” Talvi and Végh (1996)
note that in response to these lobbying pressures in good times, govern-
ments optimally restrain the balance in bad times, thereby accentuating the
procyclical bias. Also, procyclical behavior can be the result of a commit-
ment to sound finances demonstrated in policies that are more restrictive
than necessary (Saint-Paul 1994).
 EMU Fiscal Rules 107

Empirical evidence shows that procyclicality and other negative features

of fiscal policy behavior are more accentuated in countries with weak bud-
get institutions (Alesina et al. 1999 and Gavin et al. 1996). Latin American
countries with less transparent and hierarchical procedures have higher
deficits and debt (Stein et al. 1999). In particular, these procedures have not
provided a satisfactory solution of the “commons” problem in the allocation
of resources and have not constrained the strategic, politically induced use
of fiscal policy (Eichengreen et al. 1999).

Can EMU rules be exported to Latin America?

Exporting EMU fiscal rules

Well-designed numerical rules ensuring discipline and flexibility, coupled
with judicious multilateral surveillance and peer pressure, are key in the suc-
cess of the EMU process. Although such elements may also play a role in
Latin America, their application needs to take into account the specific fea-
tures of this region.14 Moreover, any step toward an EU-like fiscal architec-
ture would need to be supplemented by national reforms underpinning a
commitment to fiscal prudence.
In broad terms, the European approach combining budgetary discipline
and flexibility appears well suited to Latin America’s needs: limited credit-
worthiness argues in favor of maintaining low levels of public debt and not
letting the deficit deteriorate sharply in downturns. Sound public finances
in normal times help ensure solvency and create the room to use fiscal pol-
icy for cyclical stabilization in bad times.
However, the EMU debt limit, currently favored by most Latin American
countries, is likely to be fiscally imprudent,15 as stressed by S. Fischer
(2001b). Hausmann in Chapter 3 suggests that the debt limit should be risk
adjusted, to account for the higher risk intrinsic in the short-term, foreign-
denominated structure of public debt in these countries. 16
Similarly, the EMU deficit limit may prove not viable in Latin American
countries. The high volatility of economic activity, tax rates, and interest
rates removes credibility from a limit on the nominal, interest-inclusive bud-
get deficit. A realistic alternative would be to set a target in terms of the struc-
tural primary surplus. Although this enjoys considerably less visibility than
the overall balance, it is much more controllable by fiscal authorities. Such
a target should be set so as to allow a continuing decline in public debt in
most economic downturns, thereby internalizing the constraint of limited
creditworthiness. In order to prevent an excessive deficit deterioration in
downturns, the adoption of the target for the structural primary surplus
should be accompanied by measures that prevent excessive deterioration of
the budget balance in bad times.
The target for the structural primary surplus should be consistent with an
overall budget surplus under “normal” economic conditions. This would
108 Marco Buti and Gabriele Giudice

permit an expansionary fiscal policy in response to a negative shock without

the need to borrow and hence reduce the risks carried by limited creditwor-
thiness.17 It would also create a safety margin in the budget balance for
unexpected fluctuations unrelated to the business cycle. For instance in
2000, Chile’s government committed itself to achieving by 2001 and main-
taining thereafter a structural surplus of 1 percent of GDP.
A crucial choice in setting a target for the budget balance is how to define
“normal” economic conditions. In the EU, this is done with reference to the
estimated gap between actual and potential output; this corresponds to a
policy rule defined in cyclically adjusted terms. In Latin American countries,
however, output fluctuations may not be the overwhelming factor auto-
matically affecting the budget balance. For instance, changes in a key com-
modity price may be as important as output swings. Such factors vary from
country to country; hence there is a need to incorporate a country-specific
notion of normal economic conditions. As suggested by Latin American
authors,18 these could be defined in terms of the consumption cycle – poten-
tially appropriate for Argentina and Uruguay – or in terms of the oil price –
which could suit Venezuela. Chile’s structural surplus target reflects the
estimated effects on revenue of the output gap and deviations of copper
export prices from a notional reference price. Also, the terms of trade and the
real exchange rate could be included in the adjustment of the actual balance.
By inducing countries to create a safety margin under the deficit limit, the
SGP allows the use of fiscal policy for cyclical stabilization. As earlier
stressed, its underlying philosophy is that automatic stabilizers should be
allowed to play freely and symmetrically, in order to cushion economic
shocks. Instead, due to the widespread skepticism of demand management,
discretionary policy is regarded as the exception rather than the rule.
In the case of Latin America, relying on traditional automatic stabilizers
only would be problematic. Given the limited size of tax and welfare sys-
tems, automatic stabilizers are very small in Latin American countries: 0.1,
as estimated by Gavin et al. (1996), compared to 0.5 in the euro area
(European Commission, 2000). In order to avoid the well-known shortcom-
ings of active fiscal policies, Gavin and Hausmann (1999) suggest develop-
ing “quasi-stabilizers” in the form of stabilization funds that would link their
operation to specific features of the economy. For instance, building on the
example of some existing funds,19 they could be triggered by the collapse of
key export prices while resources would be accumulated when revenue from
a price boom is higher than a given threshold. Accordingly, the budget
should include provisions to secure surpluses in good times.

Fiscal convergence and reforms

The empirical literature shows that budget procedures, supported by appro-
priate reform measures, are conducive to fiscal discipline and help curb
the deficit bias resulting from politically induced behavior.20 The potential
 EMU Fiscal Rules 109

application of EMU rules to Latin America, particularly in designing

appropriate budget institutions and procedures at both national and supra-
national levels, deserves closer attention.
Peer pressure and multilateral surveillance are distinctive features of the EMU
rules. At a technical level, this process aims at improving transparency and
predictability. At a political level, it involves rewards for compliance or penal-
ties for noncompliance with the limits. EMU fiscal rules address the exter-
nalities generated by sharing monetary control among sovereign countries.
Nevertheless, even countries not aiming at creating a monetary union may
have an interest in ensuring orderly fiscal behavior in neighboring countries.
Exchange rate instability threatening free trade and risks of contagion from
financial crises are obvious examples of externalities brought about by reck-
less fiscal behavior – a scenario that applies in particular to Latin American
countries. In the event, a process of convergence and macroeconomic coop-
eration may help internalize these spillovers (Ghymers 2001).
An embryonic system of multilateral surveillance, largely inspired by the
EMU’s fiscal architecture, is taking shape at the subregional level in Latin
America. The agreements of Mercosur and the Andean Community repre-
sent major initial steps toward regional convergence and coordination.21
Not surprisingly, the EMU framework has been taken as point of reference
in the efforts to overcome Latin America’s fiscal failures. These efforts are
clearly inspired by the EU convergence process, notably with regard to final
objectives, intermediate targets, and procedures.22
In the Treaty of Ouro Preto in 1994, Mercosur (Argentina, Brazil, Paraguay,
and Uruguay, with association agreements with Chile and Bolivia) estab-
lished a coordination structure with objectives very similar to those of the
EU, that is, the creation of a common market, and eventually an economic
and monetary union. In December 2000, the presidents of Mercosur mem-
bers as well as associate members stated that “economic policy coordination
represents an essential element for the process of integration” and reaf-
firmed “their commitment to fiscal solvency and monetary stability”
(Mercosur 2000). The agreement outlines numerical and procedural rules for
macroeconomic convergence, including to a deficit limit of 3 percent of GDP
by 2002, and to a ceiling on net debt of the consolidated public sector of
40 percent of GDP by 2010. A High Level Macroeconomic Monitoring
Group, made up of senior officials from the ministers of finance and central
banks, is responsible for monitoring the implementation of agreements, fol-
lowing macroeconomic developments in member countries and initiating
proposals for stepped-up policy coordination.
A similar process of convergence is being designed by the Andean
Community (Bolivia, Colombia, Ecuador, Peru, and Venezuela). The embry-
onic institutional architecture supporting the integration is, at least on
paper, much in line with that of the EU. At the political level, the Andean
Advisory Committee sets targets and deadlines, and deliberates on the
110 Marco Buti and Gabriele Giudice

fulfillment of the convergence criteria. The General Secretariat and the

Permanent Technical Group oversee the inflation and fiscal targets and
examine the progress toward common goals. Fiscal convergence envisages
that the budget deficit should not exceed 3 percent of GDP as of 2002
(though allowing for deficits up to 4 percent of GDP in 2002–04), and public
debt should not be in excess of 50 percent of GDP, at the latest by 2015.
However, during years of recession, member countries may exceed the deficit
limit, if they present a plan for regularizing their positions within one year.
In order to achieve these objectives, each country must submit a yearly con-
vergence action program.
Although far-reaching on paper, these agreements are not very forceful in
practice and their effectiveness in constraining national policies remains, so
far, untested. In terms of design, a ceiling for the overall deficit is unlikely
to prove viable. Indeed, the selected limits may be too high in normal times
and too low in times of crisis. Also it remains to be seen whether the limits
set for the public debt are prudent for all the members. However, if deficit
and debt targets are preferably country specific (in both level and defini-
tion), centralized monitoring may become more difficult. More important,
given the lack of discernible penalties and enforcement mechanisms, it is
unclear how the supranational dimension could enhance the incentives to
abide by the agreed-upon rules.
Effective fiscal rules and budget reforms at the national level are a precondition
for any progress toward multilateral surveillance. Several Latin American
countries have recently made significant steps by adopting reforms of insti-
tutions and procedures, introducing hierarchical elements and higher trans-
parency. In particular, fiscal responsibility laws encompassing numerical
fiscal rules and more transparent accounting standards have been adopted
in several countries.23
Probably the most comprehensive fiscal responsibility law is that intro-
duced in Brazil in 2000. It requires all levels of government to observe cur-
rent balance and limits on debt–current revenue ratio and payroll spending.
The law also sets a general framework for budget planning, execution, and
reporting. In case of noncompliance there are mechanisms to induce com-
pensation and correction, with institutional sanctions and individual penal-
ties. Transparency is ensured by the preparation of a fiscal policy document
attached to the government’s multiyear plan.24
Given their recent adoption, the effectiveness of fiscal responsibility laws
cannot be evaluated yet. Nevertheless, a number of conditions to enhance
their effectiveness can be identified (Ter-Minassian 2002). First, the laws
need to cover all levels of government. Unsustainable fiscal behavior at the
subnational level might create difficulties for macroeconomic management,
as illustrated by the differing experiences of Argentina and Brazil. Second, it
is necessary that these laws require transparency – so that their implemen-
tation can be effectively assessed by the markets and the public at large – and
enforceable sanctions for noncompliance. Third, to be credible, these laws
 EMU Fiscal Rules 111

should be enacted by higher-level legislation, so that it is more difficult to

change them. Finally, the specificity of the targets should be assessed on a
case-by-case basis, depending on the political and economic circumstances
of the country in question.
While the current examples of fiscal responsibility laws build on existing
budget institutions, more radical reforms have been proposed such as the
establishment of an independent national fiscal council (NFC) empowered
with the decision over the maximum permissible change in public debt.25
Extension of the NFC to various Latin American countries might prove a use-
ful step toward multilateral surveillance.

Conclusions

An examination of EMU fiscal rules leads to several conclusions. First, while

numerical and procedural rules are complementary, numerical targets appear
more adequate to jumpstart the process of fiscal retrenchment. Second, the
Maastricht criteria contributed to a genuine fiscal consolidation which, given
its size and composition, is unlikely to be reversed in future years. Third, polit-
ical economy ingredients (visibility, incentive structure, ownership, con-
straining calendar, central monitoring, and collegial culture) have played key
roles in its success. And fourth, several open issues need to be addressed if the
pact is to become an effective framework for conducting fiscal policy in the
single-currency area: correction of its asymmetric operation, improved defin-
ition of medium-term targets, more coherent institutional reforms at the
national level, and improved quality and sustainability of public finances in
the multilateral surveillance.
Rooted in the history of European integration, the EMU fiscal architecture
is inevitably EU specific. While this reduces its exportability, a number of
lessons can be drawn for nonmember countries intent on embarking on
budgetary reforms. In the case of emerging market economies, the EU expe-
rience may prove useful in designing rules to pursue greater fiscal discipline
and flexibility. However, were fiscal rules inspired by the EMU architecture
to be adopted by Latin American countries, a number of elements would
have to be tailored to the typical features of these economies. All in all, more
ambitious fiscal targets (compared to those of the EMU), coupled with com-
mitments on fiscal variables that governments can control, appear to be nec-
essary. National institutional reforms are needed to buttress the numerical
targets and represent a precondition for any step toward convergence and
multilateral surveillance.
While a number of technical features of EMU fiscal rules can be exported,
the European process also shows that the political economy dimension of
the rules is key for their success. Emerging market economies will have to
devise their own systems of political incentives and arrangements under-
pinning the implementation of fiscal rules.
112 Marco Buti and Gabriele Giudice

Notes

1. We would like to thank our discussant Pablo Guidotti, as well as Teresa

Ter-Minassian and George Kopits, for their views. We are also indebted to our
Commission colleagues Giuseppe Carone, Declan Costello, Heliodoro Temprano
Arroyo, and Christian Ghymers, for useful comments and suggestions. The opin-
ions expressed are the authors’ alone and should not be attributed to the European
Commission.
2. Giudice et al. (2003) find evidence of non-Keynesian effects in fiscal consolida-
tions in the EU over the last thirty years.
3. See, for example, Alesina and Perotti (1996), Perotti (1996;1999).
4. The diagonal from top right to bottom left indicates the direction of the budgetary
adjustment: the area above it marks deterioration in the cyclically adjusted pri-
mary balance, while the area below it indicates a structural consolidation. The
diagonal from top left to bottom right marks the composition of the adjustment,
indicating the combinations where revenue changes or expenditure changes
dominate.
5. For a historical perspective, see Costello (2001) and Stark (2001).
6. For a detailed account of the legal aspects of the SGP, see Cabral (2001). For a thor-
ough review of the debate on the SGP, see Brunila et al. (2001), Buti et al. (2003),
and Giudice and Montanino (2003).
7. This decision was made concerning Portugal in 2002 and concerning France and
Germany, in 2003 for excessive deficits that occurred in 2001 and 2002 respec-
tively. Further procedures have been launched in 2004 with regard to Greece, the
Netherlands and the United Kingdom. The procedure on Portugal has been closed.
8. See Buti and Giudice (2002) for a detailed analysis of the main ingredients of the
success of the Treaty in influencing fiscal policy behavior in the run-up to EMU.
9. For proposals for improvement in the implementation of the pact, see Buti and
Giudice (2002), Giudice and Montanino (2002; 2003), Buti et al. (2002), Buti and
Sapir (2002) and Buti et al. (2003).
10. On the specific issues of application at subnational levels of government, see
Chapter 14 by Balassone et al. and of implications for accession countries, see
Chapter 10 by Coricelli and Ercolani, in this volume.
11. The model also highlights the importance of avoiding the use of fiscal policy
artificially to boost output beyond potential.
12. See, for example, Inter-American Development Bank (1995b), Caballero (2000),
and Gourinchas et al. (2001).
13. See Chapter 3 by Hausmann.
14. According to Wildavsky (1986, p. 147), “Before prescribing the budgeting proce-
dures of rich countries for poor ones, it might be well to bear in mind the obsti-
nate waywardness often displayed by institutions when transplanted from their
native habitat.”
15. Kletzer (1997) shows that, in the face of standard interest rate and oil price shocks,
current debt levels might not be sustainable.
16. Pablo Guidotti, as discussant of this chapter, supported the application of more
stringent criteria to Latin America than those embedded in the Maastricht Treaty.
For example, when liquidity considerations are important, the targets should be
more prudent the lower the average maturity and the higher the share of foreign
currency debt.
17. See Gavin et al. (1995) and Gavin and Hausmann (1999).
 EMU Fiscal Rules 113

18. See Gavin et al. (1996), Gavin and Hausmann (1999), and Talvi and Végh (1998).
The IADB research network (1998) has carried out extensive work in this direction,
identifying new indicators capable of more accurately capturing the effects of
changes in macroeconomic variables and public finances.
19. Acting as automatic spending rules, stabilization funds for copper and for oil and
coffee were introduced in Chile and Colombia, respectively (Gavin et al. 1995).
20. See, for example, Alesina and Perotti (1996) and Poterba (1996), implying that
proper rules may prevent governments from using fiscal policy to push output
beyond potential.
21. A similar agreement, established in 1993, involves seven Central American
countries.
22. The European Commission actively supports coordination attempts by Latin
American countries. With the cooperation of the Commission, the Economic
Commission for Latin America and the Caribbean (ECLAC) is implementing an
informal network for macroeconomic dialogue for major Latin American subre-
gions. The spirit of this project is to prepare technical grounds for monitoring
national policies at the subregional level.
23. Argentina and Peru, in 1999; Colombia, in 2000; and Ecuador, in 2002. On these
and other reforms, see Stein et al. (1999), Ter-Minassian and Schwartz (1997),
Fischer (2000), Kopits (2000) and Ter-Minassian (2002).
24. For an analysis of Brazil’s fiscal rules and their implications for sustainability, see
Chapter 8 by Goldfajn and Guardia.
25. See Eichengreen et al. (1999). Wren-Lewis (2000, 2003) and Wyplosz (2002)
suggest that a similar reform may also be helpful in the euro area.
8
Fiscal Rules and
Debt Sustainability in Brazil
Ilan Goldfajn and Eduardo Refinetti Guardia1

Introduction

A fundamental policy issue faced by the Brazilian authorities centers on the

medium- to long-term sustainability of the public sector debt. The prospects
for fiscal sustainability can be regarded as an important argument in the for-
mulation of the macroeconomic policy stance and of structural reforms.
Assessing fiscal sustainability in a real economy involves, however, a degree
of subjectivity. Possible future outcomes for relevant variables – real growth,
real interest rates, and real exchange rates – may lead to different assess-
ments. Debt sustainability exercises should be based on medium- and long-
run scenarios, but transitory adverse market swings commonly result in
biased assessments. In general, objective assessments are more feasible in
tranquil times.
Fiscal sustainability calculations depend on probabilities associated with
underlying variables. In the case of Brazil, for example, what is the proba-
bility of a further real depreciation of the exchange rate over the next five
to ten years? What are the chances that equilibrium real interest rates will
remain as high as they are currently? Both questions are relevant, given the
sensitivity of the public debt to these variables. We argue in this chapter that
both probabilities are small in Brazil for a time frame of up to a decade. At
present, the real exchange rate is probably undervalued and real interest
rates are very high and on a declining trend.
Some analysts tend to extrapolate the past trend of the public debt–GDP
ratio in Brazil into the future.2 However, most factors that have contributed
to the increase in the debt ratio are nonrecurrent. These explanatory factors
included a weak institutional framework for controlling public finances at all
levels of government, the recognition of hidden outstanding liabilities (the
so-called skeletons) totaling about 10 percent of GDP, insufficient public sec-
tor primary surplus until 1998, high real interest rates, and significant real
depreciation since 1999. A forward-looking analysis of fiscal sustainability in
Brazil should exclude or correct all of these factors. The fiscal stance has

114
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Fiscal Rules in Brazil 115

improved considerably since the phase-in of a rules-based fiscal policy begun

in 1998, and there is reason to expect that this policy will continue; the real
exchange rate has adjusted under the present floating exchange rate regime
and most of the hidden liabilities have already been identified.
At the core of assessing fiscal sustainability lies an analysis of the current
institutional framework of fiscal policy. Such an analysis reveals whether the
present level of primary surplus is compatible with a sustainable debt posi-
tion and whether there is scope for further adjustment, if conditions so
require. In this respect, a number of important points are worth considering.
First, although a tax reform would increase efficiency, there are no structural
difficulties in generating sufficient revenue;3 the overall tax yield amounts
to about 35 percent of GDP, among the highest in Latin America. Second,
fiscal discipline has been achieved at all levels of government – as evidenced
by their primary surplus position – partly due to successful debt-restructuring
agreements between the federal government and subnational governments.
Third, the Fiscal Responsibility Law of 2000 (LRF) provides a sound, perma-
nent fiscal regime; borrowing limits prevent governments, at all levels, from
spending beyond their means, and the closure of most state banks further
constrains their borrowing capacity. Fourth, a constitutional ban on any law
that modifies existing financial contracts or that can be interpreted as forced
restructuring guarantees institutional continuity and stability – lacking in
many other countries at a similar stage of development.4 Finally, though
the present imbalance of the social security system is manageable over the
medium term, it imposes an increasing adjustment burden on the rest of the
public sector in order to remain within the primary surplus target.5 Indeed,
it is widely recognized that further reforms are needed to increase flexibility
in government spending, to enhance the efficiency of taxation, and to
reduce the social security deficit.
The importance of institutions in explaining the differences in fiscal out-
comes among countries has received growing interest in the economic liter-
ature. In spite of sparse empirical evidence, several studies have emphasized
that the characteristics and implementation of the fiscal structure, the bud-
get process, and the political system, as well as policy rules and procedural
rules, are likely to affect fiscal performance.6 Although there is a consensus
that institutions matter, the usefulness of fiscal rules as instruments of fiscal
discipline remains controversial. This debate becomes even more complex
considering that a fiscal rule can be designed and implemented in various
different ways, and thus can have quite different effects.
A fiscal policy rule can be defined broadly as a permanent constraint on
fiscal performance. The experience of developed and emerging market
economies points to three major types of fiscal policy rules: limits on bor-
rowing or on debt stock; targets or limits for selected fiscal summary indica-
tors, such as the overall, current, or primary balance; and limits on payroll,
interest expenditure, or primary expenditure.7 These rules are cast in a
116 Ilan Goldfajn and Eduardo Refinetti Guardia

broader statute on public finances in a number of countries, often under the

heading of fiscal responsibility legislation – following the lead of New
Zealand. Analogously, in the European Union, the Stability and Growth Pact
is a treaty-based statute applicable to all member countries.
In this chapter, we evaluate the contribution of recent innovations in
Brazil’s fiscal system, particularly the adoption of fiscal rules in the context
of the LRF, to public debt sustainability. A basic argument is that adherence
to the rules and their implied performance targets at all levels of government
reinforces the credibility of monetary policy under the inflation-targeting
regime adopted in 1999, in the promotion of stability and growth.

Recent trends

The fiscal results in Brazil have improved significantly in the recent past, as
reflected in the main measures of public sector balance. The public sector bor-
rowing requirement (PSBR) had reached about 7 percent of GDP at end-1995,
peaked at 8 percent in the first quarter of 1999, and then declined steadily to
4.6 percent of GDP by end-2002. The improvement in the primary deficit (i.e.
noninterest deficit) of the public sector was remarkable: after fluctuating near
balance 1995–98, it turned into a surplus that reached nearly 4 percent of
GDP by mid-2000, and remained close to that level through 2002 (Figure 8.1).
The operational balance (i.e. net of real interest rate payments) has registered
an even more impressive improvement of 7.5 percentage points since 1995,
reflecting the full extent of the adjustment effort.
In spite of this adjustment, the outstanding gross debt of the general gov-
ernment stood at 72.7 percent of GDP at the end of 2002, three-fourths of

10
8
6
4
2
0
–2
–4
–6
Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec
1998 1999 2000 2001 2002

Primary deficit PSBR

Figure 8.1 Brazil: public sector borrowing requirement and primary deficit (in percent
of GDP), 1998–2002
Note: Cumulative totals over 12 months.
Sources: Central Bank of Brazil and authors’ estimates.
 Fiscal Rules in Brazil 117

Table 8.1 Brazil: public sector debt, December 2002

Domestic External Total Total

(R$ billion) (percent of GDP)

Net consolidated liabilities, 654 226 881 56.5

total (A
B
C
D)
General government (A
B) 581 278 859 55.2
Federal government (A) 310 256 566 36.4
States and local 271 21 292 18.8
governments (B)
Central Bank (C) 53 59 5 0.4
Public enterprises (D) 19 7 27 1.8

Source: Central Bank of Brazil.

which was in domestic obligations and the remaining one-fourth in external

liabilities. However, by subtracting relatively liquid asset holdings (mainly in
the form of various government deposits, including certain autonomous
funds),8 we can derive a robust measure of net debt of the general govern-
ment totaling 55.2 percent of GDP (Tables 8.1 and 8.2).9 The federal govern-
ment (including the social security system) accounts for nearly two-thirds of
this amount, while one-third is the share of the state and local governments.
Upon consolidation with the balance sheet of the central bank and the pub-
lic enterprises, the estimated net debt of the public sector is at 56.5 percent
of GDP.10
The near doubling of the net debt ratio over the past decade (more than
26 percentage points’ increase since 1994) reflects far more than the accu-
mulated fiscal outcome during this period. In fact, substantial reforms were
implemented during this period, resulting in a sharp deceleration of infla-
tion, increased transparency and debt recognition, and strong exchange rate
depreciation in the last three years. These developments significantly influ-
enced the rise in the debt ratio.
A hypothetical decomposition of the present debt ratio indicates that
the depreciation accounts for 19 percentage points, and the recognition of
hidden liabilities, 12 points.11 In other words, if the exchange rate had
been kept constant since the end of 1994, the debt ratio would have reached
37.8 percent, instead of 56.5 percent, in 2002 (Figure 8.2). Since the effect of
the depreciation is calculated on an accrual basis, part of this effect might
have reversed itself if the exchange rate had appreciated toward its earlier
level. Alternatively, without recognizing any hidden liabilities, the debt ratio
would have reached 44.7 percent of GDP. These developments are of course
nonrecurrent: the depreciation has occurred (thus the real exchange rate is
now probably undervalued) and a large share of hidden liabilities has been
recognized.
118 Ilan Goldfajn and Eduardo Refinetti Guardia

60
56.5
55
50
44.7
45
40 37.8
35
30.0
30
25
20
1994 1995 1996 1997 1998 1999 2000 2001 2002

No debt recognition (skeletons) Constant exchange rate Actual

Figure 8.2 Brazil: net public debt, assuming constant exchange rate and hidden lia-
bilities (in percent of GDP), 1994–2002

60
50 56.5

40
30.0 30.7
30
20 26.8

10
0
1994 1995 1996 1997 1998 1999 2000 2001 2002

Actual 3.5 percent average primary surplus (1995–98)

3.5 percent primary surplus + decline of 5 percent SELIC (1995–98)

Figure 8.3 Brazil: net public debt, assuming constant primary surplus and interest rate
(in percent of GDP), 1994–2002
Sources: Central Bank of Brazil and authors’ estimates.

As mentioned earlier, since 1999 the general government has achieved a

significant surplus in the primary balance. In this respect, the question arises
as to whether the adoption of this policy stance earlier would have pre-
vented the sharp increase in the debt ratio. For this purpose, we simulated
the effect of a constant primary surplus of 3.5 percent of GDP (i.e. below the
present target of 4.25 percent of GDP) from 1995 onward, all else being kept
at actual levels. In these circumstances the debt stock would have declined
and then increased to the initial ratio, barely above 30 percent of GDP in
2002 (Figure 8.3). Under this hypothetical situation, a virtuous cycle would
have ensued, resulting in lower interest rates and a higher growth path than
actually observed. The combination of the assumed primary surplus and a
5 percent reduction in the basic interest rate (SELIC) since1995 would have
resulted in a pronounced fall in debt ratio, before a rebound to less than
27 percent of GDP in 2002.
These simulations illustrate that despite some rather conservative
assumptions and even under the adverse economic conditions that charac-
terized this period, a stronger fiscal stance would have stabilized the debt
 Fiscal Rules in Brazil 119

ratio. Moreover, the improved primary surplus position would have pro-
duced an initial reduction in the debt stock; this debt reduction would have
become steeper with smaller interest payments. Further, the endogenous
feedback on GDP growth would have depressed the debt ratio. Such favor-
able dynamics would have allowed a considerable decline in the debt ratio
over the period.
If the past is any indication of the future, these exercises suggest that the
debt ratio is likely to decline in the future. Specifically, in the absence of
additional hidden liabilities or further depreciation in the real exchange rate,
the present targeting of a sizable primary surplus should help pave the way
to public debt sustainability.

Institutional innovations

The institutional reforms implemented in recent years – culminating in the

adoption of fiscal rules – have set the stage for achieving a primary surplus
target consistent with a sustainable path for the public debt ratio. To appre-
ciate the importance of the reforms, it is necessary to take a comprehensive
view of public finances, including the role of subnational governments.
After a brief review of the federal structure and the resolution of intergov-
ernmental financial problems, we discuss key features of the fiscal adjust-
ment implemented in recent years and present an outline of the fiscal
responsibility legislation.
The federal constitution guarantees financial and administrative auton-
omy for subnational governments, assigns spending responsibilities to
them, and clearly defines their tax base and legal transfers from the federal
government. Under this high degree of fiscal decentralization, 27 state and
over 5 500 local governments are responsible for approximately one-half of
public expenditure, concentrated mainly in the provision of basic education,
health, and public security. However, subnational government borrowing
has been subject to direct control by the Senate, exercised through resolu-
tions. In addition, the Central Bank of Brazil sets limits on domestic bank
credit to subnational governments.12 As in most other countries, the Central
Bank is forbidden to finance the nonfinancial public sector; it is not autho-
rized to extend loans to any public sector entity or to purchase primary
issues of government securities.13
Despite Brazil’s extensive and complex legislation to control subnational
government borrowing, during the 1990s state and local government
indebtedness grew apace with that of the federal government, so that their
share remained about one-third of total public sector debt (Table 8.3).14 Two
major developments explain the debt growth and the failure of the existing
system. First, the regulations were extremely permissive in terms of debt
rollover. On several occasions, Senate resolutions authorized the states to roll
over up to 100 percent of the debt service (interest plus principal). Given the
120 Ilan Goldfajn and Eduardo Refinetti Guardia

Table 8.2 Brazil: general government debt, December 2002

R$ billion Percent
of GDP

Net liabilities, total 859 55.2

Gross liabilities 1 132 72.7
Gross assets 272 17.5
Bank deposits1 102 6.8
Investments of financial
funds and programs 38 2.5
Labor Assistance Fund (FAT) 67 4.3
Credit to public enterprises 32 2.1
Other government credit 23 1.9

Note: 1 Includes deposits of all government levels and the social secu-
rity system, plus tax revenue in transit (collected but not yet deposited).
Sources: Central Bank of Brazil and authors’ estimates.

prevailing high real interest rates, this led to the rapid growth in debt due to
the capitalization of interest, even without new borrowing. Second, the fed-
eral government had become accustomed to bailing out insolvent state and
local governments.15 Furthermore, the federal government was forced to
exchange state bonds for federal bonds to facilitate the rollover of state debt,
as states had no access to financial markets. These procedures artificially
reduced the subnational government cost of borrowing, created incentives
for indebtedness, and introduced dangerous moral hazard in intergovern-
mental relations.
State-owned banks and public enterprises were also a major source of fis-
cal imbalance through the early 1990s. State banks routinely provided
financing to state governments, and nonfinancial public enterprises bor-
rowed to support quasi-fiscal operations. Although few states borrowed
directly from their commercial banks, the banks facilitated state borrowing by
underwriting state bond issues. In addition, a significant stock of liabilities
of state enterprises was assumed by the respective states and renegotiated
with the federal government at a lower cost.
At the federal level, fiscal performance deteriorated after the introduction
of the Real Plan in 1994. The combination of high real interest rates and a
weak primary surplus position resulted in a significant rise of the net federal
debt ratio. This situation worsened in the wake of the Asian and Russian
crises, and was met with an active monetary policy response, reversing the
interest rate reduction initiated in 1995. Inevitably, these developments also
contributed to the increase in the subnational debt ratio.
More generally, until 1998 the conduct of fiscal policy was plagued with an
inefficient budget process at all levels of government. Prior to 1994, this inef-
ficiency was compounded by the impact of inflation on budget execution.
 Fiscal Rules in Brazil 121

Table 8.3 Brazil: public sector debt (in percent of GDP), 1994–2002

1994 1995 1996 1997 1998 1999 2000 2001 2002

Total, net 30.4 30.8 33.4 34.4 41.7 49.2 48.8 52.6 56.5
Federal
government1 13.1 13.4 15.9 18.7 25.0 30.1 30.6 32.8 36.0
State and local
governments 10.1 10.7 11.6 12.9 14.1 16.3 16.1 18.3 18.8
Public enterprises 7.2 6.7 5.9 2.8 2.6 2.8 2.2 1.6 1.8

1
Note: Consolidated with Central Bank accounts.
Source: Central Bank of Brazil.

Most revenue was subject to some sort of formal or informal indexation, but
without a comparable inflation adjustment on the expenditure side. In this
context, cash management affected the allocation and level of real expendi-
ture.16 In practice, the budget had become a fictitious instrument, with very
little influence over expenditure allocation and fiscal performance.
In view of these weaknesses, any serious attempt to improve fiscal policy
in Brazil in the late 1990s had to deal with the following issues: the stabi-
lization of the debt ratio, the need for effective control of state and local gov-
ernment debt, the creation of conditions for avoiding future bailouts of state
and local governments, the overhaul of the budget process, the introduction
of a medium-term macroeconomic budgetary framework (MBF), and the
increase in transparency in fiscal reporting.
Accordingly, in 1998, the approach to fiscal policymaking changed dra-
matically with the announcement of the Fiscal Stabilization Program. The
program comprised four interrelated initiatives: a front-loaded adjustment
aimed at increasing the primary surplus of the public sector, steps toward
social security and administrative reform, comprehensive restructuring of
subnational government debt, and enactment of the LRF, including reform
of the budget process and introduction of fiscal rules.
The recent improvement in the fiscal position already reflects the effec-
tiveness of some of the reform initiatives. As mentioned, between 1998 and
2000, the consolidated public sector has shown an increase in the primary
surplus from 0 to 3.5 percent of GDP and remained above that level there-
after (Figure 8.1). This sustained adjustment can be explained above all in
terms of the subnational debt-restructuring agreements and the introduc-
tion of a rules-based fiscal policy – the most important innovations in
Brazil’s public finances since the 1988 Constitution.
Most states signed debt-restructuring agreements with the federal govern-
ment.17 According to the agreements, state debts were refinanced (with
securities issued by the federal government) over a maturity of 30 years at a
122 Ilan Goldfajn and Eduardo Refinetti Guardia

6 percent fixed real interest rate. The agreements required a minimum debt
service equivalent to 13 percent of the state (or municipal) net revenue. The
federal government assumed the cost (approximately US$22 billion by July
2001) of the refinancing, reflected in the differential between the market
interest rate paid by the federal government and the fixed interest rate paid
by state governments. Following state models, the federal government also
restructured 95 percent of the local government debt incurred before May
2000. By December 2001, total debt restructured amounted to more than
US$100 billion, resulting in an annual flow of debt service (interest plus
principal) of US$6 billion.
Under their debt-restructuring agreements, states were obliged to imple-
ment an adjustment program, including an agreed-upon debt-reduction
path. The adjustment program, approved by the Senate on a case-by-case
basis, set targets for revenue and expenditure and required that privatization
proceeds be used to redeem public debt. Perhaps the most important ele-
ment of the adjustment was the reduction of payroll expenditures from
more than 70 percent of their net revenue in 1987 to less than 60 percent in
2001. In addition, states and municipalities had to offer their own revenue
and federal transfers (including state transfers, for municipalities) as collat-
eral. In the case of default, the federal government was authorized to retain
the legal transfers or, if this was insufficient, to withdraw the amount due
from the state’s bank account. This zero-default guarantee proved to be very
effective, especially in combination with the sanction that states failing to
comply are charged interest penalties on the rescheduled debt and can be
denied federal guarantee on new state borrowing.
The debt-restructuring agreements with subnational governments in
Brazil can be characterized as a coordinated approach.18 The goal was to estab-
lish collective credibility for macroeconomic policy by creating conditions
for sound fiscal management at the subnational level. The agreements, in
fact, provided the basis for the improvement in subnational fiscal perfor-
mance after 1998, which was strengthened after the approval of the LRF in
May 2000. In the fact, the LRF prohibits the federal government from financ-
ing state and local governments. This restriction not only prevents future
bailouts, but it also preserves the existing contracts, including the obliga-
tions under subnational debt-restructuring agreements.
More generally, the LRF sets a framework for the conduct of fiscal policy,
including budget planning, execution, and reporting requirements at each
level of government.19 The law is intended to sustain the structural adjust-
ment of public finances and constrain public indebtedness. To this end, it
established policy rules consisting of limits and targets for selected fiscal
indicators, procedural rules (including transparency requirements), and cor-
rective steps and legal sanctions for noncompliance.
All levels of government are subject to fiscal policy rules: to maintain cur-
rent balance (the so-called golden rule that allows borrowing only to finance
 Fiscal Rules in Brazil 123

investment projects), to limit indebtedness, and to limit payroll expendi-

ture. In addition, the government is required to meet the operational target
for the primary balance, set in the Annual Budget Guidelines Law (LDO) in
accordance with the debt limit. Each government must observe all the
policy rules, which in effect means compliance with the most stringent or
binding requirement, at present the primary balance target – except for a
few low-indebted subnational governments that are thus bound by the
golden rule.
In this regard, the most important innovation introduced by the LRF is the
formulation of the debt ceiling for each level of government. These ceilings,
proposed by the executive branch for approval by the Senate,20 are defined
as a percentage of the net current revenue of each government, consistent
with a declining debt–GDP ratio. In broad terms, the objective was to reduce
the general government net debt ratio to about 40 percent over a 15-year
period. Senate Resolution No. 40, approved December 2001, required an
annual adjustment equivalent to 1/15 of the difference between the actual
(if in excess) and the limit debt–revenue ratio set for the corresponding gov-
ernment level. In other words, although the federal, state, and local
governments will all have to achieve the limit set for the corresponding
level, each subnational government has its own path, depending on the debt–
revenue ratio on December 2001.21
The limit on payroll expenditure, a binding fiscal rule aimed at reversing
the sharp growth of the public payroll in the 1990s, formalizes a similar
requirement under the debt-restructuring agreements. Under this rule, pay-
roll spending (defined as wages, salaries, and pensions) is capped at 50 per-
cent of net revenue for the federal government and at 60 percent for all
subnational governments. Several restrictions for personal management
apply as long as expenditures are above the limits.
Procedural norms have been introduced or reinforced in the LRF for all
levels of government to support the policy rules and to eliminate the defi-
ciencies leading to deficit bias in the past. Major norms include the match-
ing of expenditure commitments with adequate funding for the year in
which they become effective and two consecutive years, prohibition of lend-
ing by public financial institutions to their main shareholders, prohibition
of issuance of commitments in the last year of the term of office for expen-
ditures to take place beyond that year, and inclusion of tax benefits in the
annual budget only if accompanied by measures to offset their budgetary
impact in the following two years.
The new institutional framework contained in the LRF improves the trans-
parency of fiscal activities, especially through comprehensive, timely, fre-
quent, and detailed reporting at all levels of government. Public sector
accounts and statistics have become far more transparent and accurate in
recent years. Budget documentation includes estimates of tax expenditures.
In addition, the authorities are required to present a four-year MBF, along
124 Ilan Goldfajn and Eduardo Refinetti Guardia

with a clear statement of underlying macroeconomic assumptions, which

then serves as the basis for the annual budget proposal.22 A rolling MBF is
an essential ingredient of effective fiscal policy rules, since it helps anticipate
future fiscal trends and possible sustainability problems, and gauge the tim-
ing and size of required adjustments.
Perhaps the most important innovation in this rules-based context has
been the introduction of the primary budget balance as an operational legal
target. The target was first used in the Fiscal Stabilization Program of 1998.
Since then, the coverage of the legal target in the LDO has been expanded
to include the social security system, the central bank, and federally owned
enterprises. The LDO, submitted by the executive for approval by the leg-
islative branch as the basis for the annual budget proposal, specifies the pri-
mary balance target for the coming year, along with indicative targets for the
next two years. Given the macroeconomic assumptions, also contained in
the LDO, the primary balance target must be consistent with the debt lim-
its ratified by the Senate.
An important practical innovation in the budget process after 1998 has
been the adoption of more realistic assumptions for budget preparation in
order to facilitate compliance with the primary balance target. This has
helped correct the previous deficiency of the budget process in Brazil, illus-
trated by the low level of execution of the approved investment budget and
by the shortfalls in achieving the primary surplus target prior to 1998.23 The
results indicate that the budget was relevant neither for defining spending
constraints nor for allocating public expenditure.
The situation was exactly the opposite after the introduction of the rules-
based system, and the definition of a legal target for the primary surplus:
more realistic assumptions for revenue and nondiscretionary expenditure
increased the consistency between the primary surplus considered in the
budget proposal and the result observed. Since September 1998 all the quar-
terly fiscal targets set for the primary result of the consolidated public sector
have been met.
Noncompliance with the rules is subject to corrective action and possible
sanctions. Any excess over the debt limit prescribed for a given level of gov-
ernment has to be eliminated within one year. While the excess persists, new
financing and discretionary transfers from the federal government are pro-
hibited; in addition, noncompliance may result in the banning of new debt
and denial of credit guarantees. A list of the governments that have exceeded
the limit has to be published by the finance ministry on a monthly basis.
Public officials in noncomplying governments are liable to criminal prose-
cution. The Fiscal Crime Law details penalties for mismanagement, ranging
from fines to loss of job and ineligibility for public office for a maximum of
five years, to imprisonment.
In sum, besides all the improvements in fiscal transparency, the new rules-
based system is founded on two premises: the design of a new federal fiscal
 Fiscal Rules in Brazil 125

system, aimed at maintaining fiscal discipline at the subnational level and

prohibiting future bailouts, and the attainment of public sector solvency
through a legal target for the primary surplus – consistent ex ante with a grad-
ual but significant convergence to debt sustainability.

Public debt sustainability

The government is considered to be solvent if the present discounted value

of its current and future primary expenditure is no greater than the present
discounted value of its current and future path of revenue, net of any initial
indebtedness. A government’s debt position is considered to be sustainable
if it satisfies the present value budget constraint (i.e. it is solvent) without
any major correction in the future that would not be feasible or desirable for
economic or political reasons.24 The solvency condition, derived under con-
stant values for growth and interest rates and primary surplus–GDP ratio,
is also a condition for sustainability since, by construction, it does not
require a major change in future variables (except for an improvement in the
debt structure) to satisfy the intertemporal public sector budget constraint.
Arguably, the institutional framework implemented in recent years has con-
tributed to the solvency of the public sector in Brazil.
Nevertheless, whether in fact the fiscal policy and procedural rules
enshrined in the LRF encourage fiscal discipline entails a quantitative assess-
ment of debt sustainability. Such an assessment requires the calculation of a
quantitative baseline scenario, under a set of plausible macroeconomic and
institutional assumptions.25 The assumptions in our baseline scenario are
conservative: annual growth rate of 3.5 percent, below the potential output
growth for Brazil, currently estimated by some analysts26 at about 4.5 per-
cent with recent data on productivity and labor force growth; real interest
rate of 9 percent; constant real exchange rate; and recognition of the remain-
ing hidden liabilities (including housing subsidy liabilities under the fund to
compensate for wage fluctuations) estimated at approximately 0.65 percent
of GDP between 2004 and 2007 – given no adverse court rulings. In this
baseline scenario, the authorities are assumed to target a primary budget sur-
plus of 4.25 percent of GDP – as specified for 2004–06 in the present LDO.
Nominal and real interest rates are defined by the implicit internal public
debt interest rate. This implies a 9 percent real interest rate, consistent with
an even higher real SELIC rate (about 10 percent). This is a conservative
assumption in the light of prevailing fundamentals – a healthy banking sys-
tem, floating exchange regime, and sound fiscal policy framework – which
warrant a lower interest rate. The assumption regarding the recognition of
hidden liabilities accentuates the decline in the debt ratio after 2006.
Under these hypothetical assumptions, the net public debt ratio declines
substantially from close to 60 percent of GDP at present to 45 percent of
GDP in 2011 (Table 8.4). The results show that with a primary surplus
126 Ilan Goldfajn and Eduardo Refinetti Guardia

Table 8.4 Brazil: baseline scenario for public sector debt, 2002–11

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Net debt
(% of GDP) 55.5 58.7 57.5 55.9 54.6 53.4 51.4 49.4 47.3 45.2
Primary surplus
(% of GDP) 4.0 4.4 4.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2
Past hidden debts1 0.4 0.3 0.7 0.6 0.6 0.6 0.0 0.0 0.0 0.0
Inflation rate2 10.2 12.8 7.3 4.7 4.0 4.0 4.0 4.0 4.0 4.0
Real GDP
growth rate 1.9 –0.2 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5
Nominal
interest rate3 19.2 23.4 15.0 13.4 13.4 13.4 13.4 13.4 13.4 13.4
Real interest rate 8.2 9.4 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
Nominal
depreciation 52.3 18.2 3.9 2.5 2.5 2.5 2.5 2.5 2.5 2.5

Notes
1
Recognition of hidden liabilities net of privatization proceeds.
2
Implicit interest rate on internal net debt.
3
Change in GDP deflator.
Sources: Authors’ calculations.

slightly above 4 percent of GDP, the debt ratio should start declining over
the medium term in line with the debt ratio targeted by the Senate resolu-
tion. While the direction of the scenario is plausible, if the primary surplus
is in fact sustained over a prolonged period, the actual fall in the debt ratio
is probably understated due to the endogeneity of the growth rate with
regard to the interest rate and the debt ratio – absent in the analysis. Indeed,
a marked decline in the debt ratio can be expected to ease real interest rates,
which in turn will affect favorably growth performance, compounding the
downward effect on the debt ratio and so forth.
In any event, alternative scenarios for GDP growth and real interest rates
may produce different outcomes. A sensitivity test performed with a varying
mix of interest and growth rates indicates the primary surplus target required
to stabilize the debt ratio over the next decade (Table 8.5). Under this crite-
rion – less ambitious than of reducing the debt ratio – the baseline scenario
assumptions for interest and growth rates would imply a primary surplus tar-
get of merely 2.4 percent of GDP.
Besides the interest and growth rates, the exchange rate is probably the
most important input for debt sustainability calculations. The path of the
nominal exchange rate depreciation only affects the debt ratio insofar as it
exceeds inflation (measured by the GDP deflator), in terms of its relative
impact on the debt stock and pass-through to GDP. If the nominal rate
path generates a consistent depreciation, it will eventually lead to a higher
 Fiscal Rules in Brazil 127

Table 8.5 Brazil: primary surplus required to stabilize

the debt–GDP ratio (in percent of GDP), 2002–12

Real growth rate Real interest rate

6 8 9 10 12

1.5 2.5 3.2 3.6 4.0 4.8

2.5 2.0 2.7 3.0 3.4 4.2
3.5 1.4 2.1 2.4 2.8 3.6
4.5 0.9 1.6 1.8 2.3 3.1
5.5 0.4 1.1 1.3 1.8 2.5

Sources: Authors’ calculations.

inflation rate. Yet the probability of a future real exchange rate depreciation
from the present levels is low, as the real exchange rate is undervalued well
below the average level recorded over the last 15 years. In addition, under
the present inflation-targeting regime, monetary policy is geared toward
avoiding inflationary pressures, increasing the likelihood of a nominal
exchange rate appreciation, especially as the current real effective exchange
rate seems to be below the longer-term trend of the real rate.
On balance, for the foregoing reasons, the risk of an outcome worse than
the baseline scenario is relatively small over the next decade. However,
admittedly, if unfavorable contingencies were to occur, it would become
necessary to implement a stepped-up adjustment effort and to again raise
the primary surplus target, above the increases from 1999 until 2003, which
would require much-needed reforms in the public sector. Beyond the
horizon of our scenario calculations, the public pension system, in combi-
nation with future demographic pressures, is a source of increasing fiscal
stress that over the long term jeopardizes public debt sustainability. Also, the
distortions in the tax system constitute a constraint on growth performance,
thus compounding the public debt burden. For these reasons, the authori-
ties – with the support of a widening political consensus – have launched
both pension and tax reforms.

Conclusions

The institutional reforms introduced since 1998, particularly the LRF, have
brought about a major improvement in fiscal performance and have
strengthened the outlook for fiscal sustainability in Brazil. The main argu-
ments and findings of the chapter lead to the following conclusions.
Important steps taken in the late 1990s, notably, the debt-restructuring
agreements with state and local governments – the agreement with the state
of São Paulo being the best example – have facilitated the introduction of a
128 Ilan Goldfajn and Eduardo Refinetti Guardia

rules-based fiscal policy. In fact, the agreements have brought about a regime
change, replacing the past practice of periodic bailouts, which have provided
powerful incentives to avoid the political cost of the fiscal adjustment with
a coordinated approach to subnational fiscal discipline. With these and
other innovations, including recognition of hidden liabilities, Brazil has
established the basis for fiscal discipline.
Fiscal rules alone cannot compensate for bad fiscal management or elim-
inate structural imbalances. However, accompanied by enhanced trans-
parency and improved management practices, as enacted in the LRF and
related statutes, Brazil’s policy rules do contribute to sound fiscal behavior
and to overall policy credibility. In particular, the implementation of a pri-
mary balance target has played an important role in reinforcing the effec-
tiveness of the inflation-targeting framework.
In the context of the LRF and the other innovations, Brazil is in a position
to ensure public debt sustainability over the medium term. Under reason-
able and even conservative assumptions regarding the real interest rate, GDP
growth and real exchange rate, the debt ratio should start declining over the
next few years. The key condition for this result is to maintain a sizable pri-
mary surplus, that is, of 4 percent of GDP or more.
Although it is possible to construct unfavorable scenarios in which the
debt ratio explodes, sensitivity tests on alternative growth and interest rates
suggest that the debt ratio can be stabilized even in the case of a transitory
adverse shock in the relevant determinants. The risk of a worse-case scenario
is small: further permanent real exchange depreciation is unlikely; real inter-
est rates are on a declining trend, though still very high compared to other
emerging markets; an eventual recovery in the world economy would sup-
port growth rates closer to the potential rate. On all these counts, further
appreciation, interest rate decline, and a rebound of activity can be expected
over the medium run. Nonetheless, if an unlikely negative scenario were to
materialize, further correction in the public sector balance would require
accelerated implementation of reform in public pensions and taxation envis-
aged by the present government.

Notes

1. We would like to thank Amaury Bier, Armínio Fraga, Katherine Hennings, George
Kopits, Joaquim Levy, Helio Mori, Pedro Malan, Roberto Pires Messenberg, Daniel
Sonder, and the Economic Department at the Central Bank of Brazil for valuable
comments and suggestions. All remaining errors are the authors’ responsibility.
2. See, for example, the calculations in Goldstein (2003).
3. For a discussion of numerous distortions in the tax system, see Rezende da Silva
(2001).
4. The constitution establishes that “the law shall not injure the vested right, the
perfect juridical act and the res judicata.” See Article 5, sec. XXXVI, Constitution of
the Federative Republic of Brazil. Additionally, Article 1 of Constitutional
 Fiscal Rules in Brazil 129

Amendment 32, of September 11, 2001, which modifies Article 62 of the

Constitution, prohibits the issuance of provisional measures by the president of the
republic requiring the seizure of goods, bank deposits or other financial assets.
5. See the medium- to long-term scenario calculations of the primary balance of the
public pension system for private and public sector employees (incorporating
demographic trends) prepared by the World Bank (2000a).
6. As a matter of fact, the evidence suggesting that institutions matter is stronger
than the evidence on the mechanisms by which these institutions matter; see
Poterba and von Hagen (1999).
7. For a definition of fiscal policy rules, as compared to procedural rules, and a review
of experience mainly in advanced economies, see Kopits and Symansky (1998).
8. For an alternative view, arguing against the deduction of these assets, see Favero
and Giavazzi (2002).
9. Although the gross debt stock is more frequently used for exercises of debt dynam-
ics because the data on subnational governments are difficult to collect and the
quality of government assets is difficult to measure, the net debt concept reported
here is appropriate in the case of Brazil. Intergovernmental debt has been consoli-
dated out and the assets that are used for calculating net debt are liquid and can be
readily used to redeem gross debt. These assets can be liquidated to finance deficits
without affecting the gross debt stock. In this respect, the net public debt concept
is close to the true measure of public sector net worth. Incidentally, this treatment
is consistent with the revised Government Finance Statistics Manual; see IMF (2001a).
Liquid assets are particularly suitable for redeeming debt at short notice. However,
from a medium-term perspective, less liquid assets clearly ought to be taken into
consideration (in symmetry with the accounting of less liquid liabilities, that is, gov-
ernment debt that does not mature in the short term). In the case of Brazil, the assets
owed to the government, included in the net government debt calculations, are
effectively available for payment of fiscal expenses. In particular, the deposits of the
social security system, the tax collected by all government levels but not yet trans-
ferred to the treasuries, and the demand deposits of all levels of government are very
liquid. Other assets (investments of several funds, various government credits, and
credit to public enterprises) are less liquid, but not necessarily of lower quality.
10. All debt ratios are calculated as percent of GDP valued with the deflator at year’s end.
11. Each of these estimates is to be interpreted as the partial effect (on the debt ratio)
of the depreciation and the recognition of skeletons since 1994, while everything
else is held constant.
12. Under Central Bank Resolution No. 2827, outstanding loans of any publicly
owned and private bank to the public sector may not exceed 45 percent of the
bank’s net worth.
13. However, the Treasury is authorized to roll over existing liabilities held by the
Central Bank of Brazil.
14. The debt growth observed in 2001 was mainly associated with the impact of the
exchange rate depreciation.
15. Between 1987 and 1997, at least four major rescue programs refinanced subna-
tional governments and their enterprises.
16. See Bacha (1994) and Cardoso (1998).
17. The State of São Paulo, the most important state of the federation, signed the first
agreement in December 1997; the last agreement was signed in May 2000. Only
two small states, Tocantins and Amapá, with virtually no debt outstanding, did
not sign agreements.
130 Ilan Goldfajn and Eduardo Refinetti Guardia

18. This is in contrast to the autonomous approach, whereby the initiative comes from
individual subnational governments; see Kopits (2001a).
19. As a complementary law, any modification of the LRF requires approval by a qual-
ified majority of Congress.
20. In case of significant economic instability or drastic changes in monetary or
exchange rate policy, the federal government can submit a proposal for changing
these limits to the Senate. Thus changes in monetary or exchange rate policy
affecting fiscal performance will trigger an extension of the time allowed for
adjustment to the ceiling.
21. The limits are set at 350 percent of revenue for the federal government, 200 per-
cent for the state governments, and 150 percent for the local governments – to be
met by the end of 2015.
22. For a comprehensive assessment of progress in fiscal transparency in Brazil, see
International Monetary Fund (2001b).
23. For the federal government, the execution rate of approved investment outlays
rose from 46 to 74 percent in 1995–97 to 90 percent in 1998.
24. See International Monetary Fund (2002d).
25. For the underlying methodological discussion, see Goldfajn (2002).
26. See Silva Filho (2002).
9
Fiscal Rules in Mexico:
Evolution and Prospects
Andrés Conesa, Moisés J. Schwartz, Alejandro Somuano,
and J. Alfredo Tijerina1

Introduction

Poor fiscal management, over time, usually results in unsustainable public

debt, high inflation and interest rates, and weak economic performance.
Rules restricting fiscal management have thus been regarded as a means to
promote discipline, stability, and growth. The economic literature suggests
that rules are more likely to increase welfare than discretionary policy
because they reduce uncertainty (Kydland and Prescott 1977; Barro and
Gordon 1983a). Of course, in countries that face pronounced business
cycles, fiscal rules need to be sufficiently flexible to buffer unanticipated
shocks to the economy.
Rules have been applied to both monetary and fiscal policies. Currently
the most widely adopted monetary rule consists of various forms of inflation
targeting, supported by central bank autonomy. Under this rule, the central
bank is committed to the pursuit of price stability, independently from the
political cycle. Fiscal rules are more heterogeneous and complex than mon-
etary rules, as they impose restrictions on a more diverse set of variables and
practices, encompassing a wide spectrum of public finance management.
Fiscal policy rules are applied to key performance variables (fiscal balance,
public sector debt, and so on), while procedural rules are usually applied to
the budget process, including transparency requirements.2
Fiscal rules are neither recent nor exclusive to the central government.
Balanced-budget rules have been in place at the subnational level since the
mid-nineteenth century in some federal countries (the United States and
Switzerland), and adopted in others after the Second World War (Germany,
Italy, Netherlands, Japan, and Indonesia) as an important component of sta-
bilization programs. The modern history of fiscal policy rules began in 1994
with the Fiscal Responsibility Act of New Zealand (Kopits 2001a). This legis-
lation includes a medium-term balanced-budget target and a public debt
limit; however, its most noteworthy feature consists in transparent account-
ing and reporting obligations that accompany the policy rules. In general,

131
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
132 Andrés Conesa et al.

while some countries place greater emphasis on transparency (Australia, New

Zealand, and the United Kingdom), others stress quantitative fiscal targets
(European Union, Brazil, and Peru). In decentralized systems with
autonomous subnational governments (Argentina and India), the federal
government can set only its own rules, leaving the subnational governments
to their own devices.3
In Mexico, against the backdrop of large fiscal imbalances and significant
public indebtedness, fiscal consolidation and active debt management have
been at the center of the authorities’ economic strategy, especially since the
1995 crisis. Recent trends highlight the effort to strengthen fiscal discipline,
despite the difficulties in containing expenditures and raising recurrent pub-
lic sector revenues (Gil-Díaz and Thirsk 2000). Major difficulties in increasing
tax revenue in Mexico stem from excessive tax preferences and weak admin-
istrative capacity (OECD 1999). The tax reforms of the late 1980s and early
1990s achieved greater tax neutrality but without increased revenue yield
(Gil-Díaz 1995).4 Although nonrecurrent revenue from privatization of gov-
ernment enterprises during the first half of the 1990s helped reduce the pub-
lic debt stock and alleviate the debt-service burden, public finances remain
vulnerable to oil shocks and the business cycle, which have undermined sta-
bility and growth. These weaknesses have encouraged the evolution of fiscal
rules in Mexico. In this chapter we analyze Mexico’s experience with existing
rules and estimate the fiscal effort required to comply with alternative rules.

Recent evolution of fiscal rules in Mexico

Mexico’s public finances are characterized by structural deficiencies,

reflected in sizeable fiscal deficits and high public indebtedness. Arguably, a
common factor in recent economic crises has been the presence of signifi-
cant fiscal imbalances. During 1995–2001, the public sector borrowing
requirement (PSBR) fluctuated around 5 percent of GDP and public sector
debt stood in excess of 40 percent of GDP, excluding contingent liabilities.5
To reverse this trend, the authorities consistently made attempts at cutting
expenditures and at raising revenue, in the context of an increasing com-
mitment to rules-based fiscal discipline, with the support of various social
and political factions.
A distinctive feature of public finances in Mexico is the reliance on highly
volatile revenue. While tax revenues are driven by the business cycle – with
a salutary automatic stabilizing macroeconomic impact – nontax revenues
are highly unstable given that they largely depend on the international
price of oil and include some nonrecurrent income. Likewise, nonprogram-
mable spending, which includes the interest bill, is also highly volatile,
since it depends on interest rates determined in the world financial markets.
In view of the structural deficiencies and the volatility of public sector
revenues, the authorities have launched reform steps to eliminate some
 Fiscal Rules in Mexico 133

distortions and to reduce the potential impact of revenue volatility on the

rest of the economy.
The tax reform of 2002 sought to enhance tax neutrality and distributive fair-
ness, and increase tax collection. While the reform passed by Congress suc-
ceeded in improving neutrality and fairness, lack of legislative support for
repealing preferential treatment under the VAT resulted in less-than-anticipated
revenue.6 Furthermore, failure to enact a comprehensive VAT reform implies
forgone government spending in areas with the highest social return.7

Existing rules and norms

Traditionally, fiscal policy management in Mexico has been guided by con-

gressional authorization of yearly limits on net borrowing by the federal gov-
ernment and by Mexico City, based on the projected fiscal balance of the
respective government; similarly, each state congress must approve net bor-
rowing by the state government. Under the Constitution, the federal and
subnational governments are subject to the so-called golden rule: borrowing
is permitted only to finance public investment. In addition, subnational
governments are constitutionally barred from incurring liabilities with for-
eign entities and from contracting liabilities in foreign currency. Mexico City
may contract foreign debt, but only through the federal government.
At a procedural level, borrowing by subnational governments (including
Mexico City) against participaciones8 (federal transfers) as collateral must be
registered with the purpose of monitoring the evolution of subnational pub-
lic debt and guaranteeing the solvency of subnational governments. Thus,
when a subnational government defaults, its participaciones are used for debt
repayment. Mexico City must additionally obtain authorization from the
federal government to affect its participaciones.
More important, starting in 1998, contingent procedural rules have been
introduced in the annual PEF (Federal Expenditure Budget) to absorb unex-
pected shocks, and to achieve fiscal targets.9 These rules have changed over
time as they are subject to yearly revision by Congress. The PEF for 2003
included the following norms:

● Unanticipated tax revenue shortfalls (below the projected level) during

the fiscal year are subject to compensatory action. Shortfalls related to an
oil price decline must be compensated for with reserves from the Oil
Stabilization Fund; lacking sufficient reserves, further compensation
needs must be met with expenditure cuts. Shortfalls due to other causes
must also be compensated for with expenditure cuts.
● Unanticipated oil revenue shortfalls (below the projected level) are sub-
ject to compensatory action depending on the underlying cause of the
shortfall. Shortfalls due to unanticipated changes in oil price must be
offset with nonprogrammed sale of financial assets by PEMEX (state oil
134 Andrés Conesa et al.

company), and if necessary, with spending cuts and a reduction in its

primary balance. Shortfalls due to lower production volume must be com-
pensated for with spending cuts by PEMEX. Shortfalls due to exchange
rate losses will be reflected in a reduction in the primary balance of
PEMEX.
● Unanticipated revenue gains (above the projected level), regardless of the
nature of the underlying cause, must be utilized as follows: 25 percent to
improve the budget balance; 25 percent to be accumulated in the Oil
Stabilization Fund; and 50 percent to fund public infrastructure in the
states.

As a further incentive to fiscal discipline at all levels of government, partici-

paciones to states and municipalities are determined by collected federal rev-
enue. This allows for a distribution of the risk associated with unexpected
changes in revenue between the national and subnational governments, and
accordingly, for sharing the burden of compensatory action at each govern-
ment level. Thus, the norms prescribed in the PEF for offsetting unantici-
pated deviations from projected revenue changes provide an important tool
for enforcing fiscal discipline.
Compliance with these rules and norms and attainment of the fiscal bal-
ance target approved by Congress have been accompanied by recent steps to
institutionalize the budgetary process and enhance overall fiscal trans-
parency. These steps discourage government entities from inflating spend-
ing needs and thus help reduce the deficit bias – as discussed in Chapters 2
and 5 by Drazen and Schick respectively. For instance, the programming of
the federal budget is now more centralized and, as indicated, expenditures
are limited by the realization of projected revenues for the year. Also, as of
1998, budgetary control and efficiency in public spending are strengthened
through the improved programmatic structure of expenditures for monitor-
ing programs, according to final destination. Moreover, a performance-based
evaluation system, taking into account the cost of public goods and services,
as well as their quality and social impact, has been implemented (Mexico,
SHCP 2002c).
In recent years, Mexico has made remarkable progress toward fiscal trans-
parency.10 Starting in 1999, all information in the Federal Public Account
(Cuenta de la Hacienda Pública Federal)11 and the PEF has been accessible elec-
tronically. Whereas off-budget disbursements had already been eliminated
from the PEF, in 2001, the government submitted a constitutional amend-
ment to formally abolish all such disbursements. In all, since 2001 more
detailed and timely information has been released in monthly reports on fis-
cal developments, with considerable detail on revenue, expenditure, and
indebtedness. At the same time, a new measure of the overall financing
needs of the government, the PSBR, was introduced. Although not a
legally binding measure, the PSBR is an analytically more meaningful and
 Fiscal Rules in Mexico 135

comprehensive indicator of fiscal balance than the official, so-called

traditional, public sector balance.12

Results

These norms and rules have proven rather useful. For instance, in 1998 when
the Mexican economy was shaken by the fall in international oil prices, pub-
lic spending was cut to reach the fiscal deficit target, contributing to an envi-
ronment of certainty and stability. Similarly, during the uncertain evolution
of the international economy in 1999, fiscal authorities implemented a pru-
dent program that later brought about a lower deficit than initially targeted.
Admittedly, fiscal discipline was to an extent achieved at the cost of damp-
ening the effect of automatic stabilizers on the economy. In other words, the
contingent rules may have exacerbated macroeconomic volatility.
In contrast to previous years, 2000 was a year of relatively high oil prices
and economic growth, reflected in the favorable evolution of public
finances. Government spending was consistent with revenue performance
and the fiscal target, allowing for the allocation of larger resources to social
programs as well as the establishment of the Oil Stabilization Fund. Given
the economic slowdown and declining oil prices in 2001, public revenues
were lower than expected; thus, the rules imbedded in the LIF (Federal
Revenue Law) and PEF required adjustment of public spending to meet the
limit on public indebtedness. As a result of the observance of these restric-
tions, in combination with increased expenditure control and improve-
ments in tax collection, the PSBR was reduced from 6.3 percent of GDP in
1998 to 2.7 percent of GDP in 2002.
In sum, the contingent rules helped promote sound public finances
even at times of economic slowdown and uncertainty, including during
oil shocks. In turn, this has allowed the economy to withstand in an
orderly fashion the effects of slowdown and uncertainty in world mar-
kets. Nevertheless, realization of the pending reform agenda, including
the adoption of well-designed fiscal policy rules, should foster economic
growth by conferring increased credibility and efficiency to fiscal policy-
making.

Pending reform agenda

Beyond the existing rules, there is still considerable scope for bolstering fis-
cal discipline in Mexico. A number of proposals included in the fiscal reform
package, known as the “New Distributive Public Finance” submitted to
Congress in April 2001, were intended to address this need.13 The specific
purpose of these proposals was to enhance the efficiency of the budget
process, on the one hand, and to further institutionalize the fiscal policy
framework, on the other.
136 Andrés Conesa et al.

Enhancing the efficiency of the budget process

The following proposals contained in the reform package focus on time
limits, sequencing, and scope of the budget process:

● Advancement of the government submission of the LIF and PEF

to Congress by one month (from the present deadline of November 15 to
October 15); assignment of a specific time period for the government to
consider congressional amendments to the original proposal; and congres-
sional approval of the LIF during its first working session (September 1–
December 15), or during an extraordinary session.
● Congressional approval of the LIF required prior to approval of the PEF.
Proposed congressional modifications must include a cost–benefit analy-
sis; also, any congressional proposal involving increased expenditure
must identify offsetting revenue increases.
● Automatic mechanism to continue the previous year’s budget statutes in
the event of congressional failure to enact the present LIF or PEF by the
end of the current year, to ensure normal government operations until
their approval.
● Authorization of investment programs and projects over a multiyear
period, to guarantee financing for programs or projects whose execution
is longer than one year.

Institutionalizing the fiscal policy framework

A constitutional amendment (Article 126) was also proposed as part of the
reform package. This would require a balance between federal government
expenditures and revenues over time, in order to eliminate the deficit bias.
The amendment would be supported by the following additional measures:

● Mandatory annual convergence to the balanced-budget target over a four-

year period, tantamount to a balanced-budget rule calculated over a four-
year moving average.
● Elimination of off-budget operations, by requiring legislative authoriza-
tion for all expenditures (including primary expenditure proposals by
Congress, as long as they are financed with offsetting revenue increase)
under the PEF.
● Preparation of a rolling medium-term budgetary framework for the LIF
and PEF. As in many OECD and Latin American countries, the annual
budget proposals would be cast in the context of indicative targets and
projections for the main macroeconomic and fiscal variables over a three-
to five-year horizon.
● Earmarking of a predetermined share of the fiscal surplus for reduction of
public debt or accumulation of reserves in anticipation of unexpected
 Fiscal Rules in Mexico 137

adverse shocks, plus constitutional amendment to make permanent

expenditure cuts to offset revenue shortfalls – currently subject to yearly
approval in the PEF.

Simulation of alternative fiscal policy rules

There is no single set of fiscal policy rules valid for all countries. For instance,
in Latin America, whereas Chile follows a structural balance rule, in Brazil,
each government targets a primary surplus predicated on a desired reduction
of the public debt ratio, or alternatively, is bound by the current balance rule
(golden rule). Inspired by some of these examples, we present numerical sim-
ulations to illustrate the effect of adopting various fiscal policy rules.
First, we simulate ex post a structural balance rule and estimate the annual
fiscal adjustment that it would have required. Then, we simulate ex ante two
alternative fiscal rules geared mainly to restoring public debt sustainability:
one involves reduction in the share of public sector absorption of savings to
30 percent by 2030, and the other, convergence of the PSBR to zero by 2030.
Finally, we simulate a balanced-budget rule imposed in terms of the tradi-
tional definition of the fiscal balance, according to the April 2001 proposal.
In order to explore the fiscal implications of these rules, we calculate the
adjustment required under each rule and trace the evolution of the broadly
defined public debt.

Structural balance rule

Since Mexico, like Chile, is exposed to volatility in the prices of nonrenew-
able natural resources, we find it useful to examine the fiscal implications if
Mexico had adopted Chile’s structural balance rule for Mexico over the
period 1991–2002.14 A major advantage of the structural balance is that, as
an indicator of the fiscal stance, regardless of the economic cycle,15 it can be
useful for designing a workable fiscal policy rule. Given a reliable measure of
the structural balance target, calculated in accordance with the potential
growth of the economy and supported by a credible commitment by the
authorities, it can serve as a mechanism to guide fiscal policy – while allow-
ing for the automatic stabilizers to operate. If the economy grows faster than
expected, and public revenues turn out to be larger than those required to
comply with the rule, a surplus position is required. If, on the other hand,
the economy unexpectedly slows down and revenues decline, the authori-
ties are allowed to incur a deficit and still meet the fiscal target.
Although in principle any summary fiscal aggregate, such as the tradi-
tional budget balance, primary balance, or general government balance, can
be calculated as a structural measure,16 the PSBR, the most general
and comprehensive indicator available, is selected for our simulations.17
138 Andrés Conesa et al.

The structural PSBR has been calculated through the following modifica-
tions: subtraction of all nonrecurrent revenues (i.e. withdrawals from the Oil
Stabilization Fund, central bank profits, privatization revenue, and collateral
recovered from Brady Bonds), and removal of the cyclical components from
both revenues and expenditures. The latter adjustment involves removal of
the revenue effect of cyclical fluctuations and oil shocks, as well as of the
expenditure effect transmitted to states and municipalities through partici-
paciones from federal revenue changes.18
As expected, the actual PSBR incorporating nonrecurrent revenue under-
states the fiscal stance, especially in the early 1990s, by a wide margin
(Figure 9.1). The differences between the structural PSBR and the actual PSBR
without nonrecurrent revenue largely reflect cyclical and oil price effects.
For instance, in 2001, due to the economic contraction, the structural PSBR
was larger than the actual PSBR with nonrecurrent revenue, but similar to
the PSBR without nonrecurrent revenue, indicating that the fall in tax rev-
enue due to the economic slowdown was offset by higher-than-anticipated
oil revenue.
The simulation results indicate the annual fiscal adjustment necessary to
achieve a structural PSBR of up to 3 percent of GDP – given by the actual
PSBR averaged over the period over the period 1991–2002 (Figure 9.2). The
simulated annual adjustment should be interpreted as the reduction (expan-
sion) in public expenditure (revenue) required to accomplish the target. The
exercise suggests that, while fiscal policy during 1991–93 could have been
eased, given the relatively modest public sector imbalance, during the
remainder of the decade a significant adjustment, especially during 1996–99,
was necessary to meet the large fiscal costs of the banking sector bailout, in
the aftermath of the 1995 financial crisis.

7
6
5
4
3
2
1
0 Structural PSBR
–1 Actual PSBR with nonrecurrent revenues
–2 Actual PSBR without nonrecurrent revenues
–3
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a

Figure 9.1 Mexico: estimates of PSBR (percent of GDP), 1991–2002

Note: a Estimates.
Sources: Mexico Secretariat of Finance and authors’ estimates.
 Fiscal Rules in Mexico 139

0
1994 1995 1996 1997 1998 1999 2000 2001 2002a
–1 1993
1991
–2
1992

Figure 9.2 Mexico: fiscal adjustment required under structural balance rule (percent
of GDP), 1991–2002
Notes: Positive values indicate fiscal adjustment, while negative values indicate allowance of fiscal
relaxation, to meet the PSBR limit.
a
Estimates.
Sources: Mexico Secretariat of Finance and authors’ estimates.

Rules for debt sustainability

The structural balance rule can be useful for maintaining a stable and cred-
ible fiscal policy over the short to medium run, while allowing for
the operation of automatic stabilizers in the face of cyclical variations
and oil price shocks. However, it may be limited as a tool for ensuring
public debt sustainability over the long run.19 Equally, such a rule may not
help reduce the absorption of financial resources by the public sector.
In fact, in Mexico, the public sector absorbs more than 80 percent of total
financial resources,20 which obviously constrains private investment and
consumption.21
With this in mind, two fiscal rules aimed to reduce public sector’s absorp-
tion of financial savings and to ensure debt sustainability are pertinent: an
absorption rule, which requires a gradual reduction in the share of public sec-
tor absorption of financial savings until it reaches a limit of 30 percent in
2030; and a zero PSBR rule, which pursues a monotonic convergence of the
PSBR until its elimination in 2030.22 Numerical simulations track the expected
evolution of public debt and financial savings absorption under each scenario
from 2002 through 2030. Also, we present the average annual fiscal adjust-
ment required to comply with each rule.23 The simulations are based on a set
of assumptions regarding future macroeconomic, financial, and fiscal trends.
The following macroeconomic and financial assumptions underpin the
simulations. For the period 2002–06, all assumptions are consistent with
the current PRONAFIDE (National Program to Finance Development).
Thereafter, real GDP growth and real interest rates are set at 4 and 5 percent,
140 Andrés Conesa et al.

respectively. The inflation rate is held at 3 percent, and the US price of oil is
constant in real terms. Financial saving – proxied by M4, Mexico’s broadest
monetary aggregate – grows by one percentage point more than real GDP.24
On the fiscal front, we adopt the following assumptions. Again, consistent
with PRONAFIDE, the PSBR is reduced from 3 percent of GDP in 2002 to
1.5 percent in 2006. Non-oil and oil revenues remain constant as a share of
GDP from 2006 onwards, given efficiency gains in tax administration,
and broadly in line with proven oil reserves. Assuming a linear relation-
ship between economic growth and the demand for public services, current
spending apart from social security is projected as a constant share of
GDP. For the social security system (including public pension programs
for private and public employees), assuming an initial aggregate actuarial
deficit roughly equivalent to GDP, subject to reform from the current pay-
as-you-go to a fully funded regime, we calculate that the primary deficit
rises from 0.7 percent of GDP in 2003 to 2.7 percent in 2030. Infrastructure
energy projects financed by the private sector, as well as credit to the private
sector provided by public banks, are assumed to remain constant as a
proportion of GDP. Disbursements stemming from the 1995 banking-
sector bailout are assumed to continue on a monotonic downward trend
until 2025.
Simulation results show the evolution of public debt and the public sec-
tor absorption of financial resources under four different scenarios: (1) base-
line scenario, without PRONAFIDE and without a rule; (2) scenario with
PRONAFIDE but without a rule; (3) absorption rule; and (4) zero PSBR rule
(Figure 9.3). All scenarios, except (1), assume that PRONAFIDE macrofiscal
targets, including the targets for PSBR, are achieved in the 2002–06 period
(Mexico, SHCP 2002a).
The simulations indicate that the public debt is sustainable in scenarios
(2) through (4), declining from 43 percent of GDP in 2003 to 21 percent in
2030 under the public absorption rule, to 12 percent of GDP in 2030 under
the zero PSBR rule, and to 26 percent of GDP without a fiscal rule but with
PRONAFIDE. This last scenario, compared to the baseline, highlights that
the implementation of PRONAFIDE leads to a sustainable debt–GDP ratio
even in the absence of a fiscal rule (Figure 9.3). In essence, this reflects the
frontloaded adjustment through 2006, which more than compensates for
the increasing cost of pensions.
Public debt sustainability, however, does not necessarily imply a signifi-
cant reduction in public sector financial absorption. In the absence of fiscal
rules, but with PRONAFIDE, public sector absorption diminishes from
87 percent in 2002 to only 40 percent in 2030. In contrast, the government
would demand only 30 percent of total financial savings in 2030 under the
public absorption rule and less than 20 percent under the zero PSBR rule
(Figure 9.3). By contrast, in the baseline scenario, the public sector financial
absorption is merely reduced to 70 percent by 2030.
 Fiscal Rules in Mexico 141

Public debt
50

40

30

20

10
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
30
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
20

No fiscal rule (with PRONAFIDE) No fiscal rule (no PRONAFIDE)

Absorption rule PSBR = 0 in 2030

Public sector absorption of financial resources

100

80

60

40

20

0
20 2
03

20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
30
0

0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
20

20

No fiscal rule (with PRONAFIDE) No fiscal rule (no PRONAFIDE)

Absorption rule PSBR = 0 in 2030

Figure 9.3 Mexico: public debt (percent of GDP) and financial absorption (percent of
total financial resources), 2002–30
Sources: Mexico Secretariat of Finance and authors’ estimates.

In order to comply with either fiscal rule, it would be necessary to imple-

ment a lasting fiscal adjustment, either through a permanent revenue
increase or expenditure cut, averaging 0.2 percent of GDP yearly under the
absorption rule, and 0.5 percent of GDP under the zero PSBR rule. However,
adherence to the rules without meeting the PRONAFIDE targets would
require more than doubling the annual average fiscal adjustment, as com-
pared with achieving these initial targets, under each rule.

Balanced-budget rule
As a further exercise, we simulate a balanced-budget rule, in line with the
constitutional amendment proposal of April 2001.25 Although the proposed
142 Andrés Conesa et al.

45
40
35
30
25
20
15
10
5
0
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
30
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
20

Figure 9.4 Mexico: public debt under the balanced-budget rule (percent of GDP),
2002–30
Sources: Mexico Secretariat of Finance and authors’ estimates.

rule is limited to the traditional definition of federal budget operations, the

simulation is performed as above, in terms of the PSBR and the public sector
debt. For this purpose, it was necessary to calibrate the rule to calculate the
corresponding path for the PSBR.26
The main assumptions for the baseline scenario are taken from the previous
exercise. The rule scenario is simulated with and without compliance with
PRONAFIDE through 2006. All fiscal variables are projected as previously,
with the following exceptions. Primary spending other than for social secu-
rity is assumed to grow at 4 percent, that is, the average annual rate of the
last two decades. Social security spending reflects the impact of the 1997
reform of the private employees’ pension regime, but, in contrast with pre-
vious exercises, this simulation excludes the fiscal cost of a future reform of
the public sector programs.
The resulting PSBR estimates are used to simulate the evolution of public
debt ratio for 2002–30 under the balanced-budget rule (Figure 9.4). Clearly,
this rule guarantees debt sustainability, as the debt ratio is reduced to almost
20 percent of GDP. Again, to comply with the rule, the government must
implement a significant fiscal adjustment, either through revenue hikes or
expenditure cuts. The average annual fiscal effort required to achieve the
proposed rule would be 0.3 percent of GDP with PRONAFIDE, and 0.8 percent
of GDP in its absence.

Comparative assessment
A comparison of the ex ante simulation results illustrates the wide dispersion
in the magnitude of the fiscal effort required under various possible fiscal pol-
icy rules (Table 9.1), depending on the stringency of the rule and on the ini-
tial adjustment. The more ambitious the rule is, and the looser the compliance
 Fiscal Rules in Mexico 143

Table 9.1 Mexico: fiscal adjustment required under selected fiscal rules (percent of
GDP), 2002–30

Annual adjustment with Annual adjustment without

PRONAFIDE PRONAFIDE

Financial absorption rule 0.20 0.46

Zero PSBR (by 2030) rule 0.46 1.04
Balanced-budget rule 0.34 0.79

Sources: Authors’ estimates.

with the targets under the current PRONAFIDE program, the higher will be
the annual average effort required to meet the targets under the rule. The
annual fiscal effort needed varies from 0.2 percent to 1 percent of GDP. While
all the rules lead to a marked improvement in the debt position, the zero PSBR
rule is the most effective in this respect, followed by the balanced-budget rule.
Some may argue that the rules discussed may be too stringent – as evi-
denced by the reduction in debt ratio and the concomitant fiscal effort – and
that these could be eased without affecting structural soundness of public
finances. In spite of the convenience of designing rules that allow deviations
when the economy is hit by transitory adverse shocks, public finances in
Mexico are characterized by a substantial gap between permanent revenue
and expenditure. Moreover, Mexico is highly dependent on oil-related rev-
enue, while facing increasing pressures for spending on infrastructure, edu-
cation, health, and poverty abatement. Without questioning the advantages
of a structural or cyclically adjusted balance approach, these conditions sug-
gest the need to be extremely careful when introducing escape clauses in the
design of macrofiscal rules for Mexico.

Concluding remarks

There is general recognition among academics and policymakers not only

of the complexities in the design and implementation of fiscal rules but also
of their potential usefulness in promoting sound public finances. Similarly,
in Mexico there is broad consensus on the need for clear and transparent fis-
cal rules aimed to strengthen public finances and to increase their indepen-
dence from the political cycle, yet keep sufficient flexibility to accommodate
the economic cycle.
In this context, Mexico has adopted a set of mostly contingent procedural
rules helpful in containing an inherent deficit bias. These rules focus mainly
on the planning, debate, and execution of the federal budget, as well as on
specifying contingency action in the event of deviation from revenue pro-
jections. In addition, a current-balance rule and debt limits are in place at
subnational levels of government. As a complement to these rules, further
144 Andrés Conesa et al.

steps in this area are envisaged to strengthen their effectiveness. To this end,
in April 2001, the authorities submitted a number of proposals that seek
further to institutionalize the fiscal policy framework, including through a
federal balanced-budget amendment to the Constitution.
In an attempt to examine the implications of the proposed constitutional
amendment and of other possible fiscal policy rules for Mexico, we per-
formed several quantitative simulations. An ex post simulation of a structural
balance rule illustrates the case for allowing the operation of automatic
stabilizers in the face of periodic cycles and random shocks. Ex ante simula-
tions of rules intended mainly for restoring debt sustainability highlight the
need for a significant effort beyond fulfillment of the current PRONAFIDE
program. In this regard, the most powerful rules are those that target a grad-
ual elimination of the PSBR by 2030 or a continuous balanced budget, as
required in the proposed constitutional amendment. Simulation results
show that, assuming fulfillment of the PRONAFIDE, these two rules imply
an annual adjustment of 0.3–0.5 percent of GDP. In view of both the need
and feasibility of such an adjustment, congressional approval of the reform
proposal of April 2001 would help ensure fiscal sustainability and discipline
on a permanent basis.

Notes

1. The opinions expressed in this paper are the sole responsibility of the authors and
do not necessarily represent those of the Secretariat of Finance, Mexico. Technical
assistance by A. Téllez, R. Altamirano, and C. Lelong is greatly appreciated.
2. See Kopits and Symansky (1998), Kopits (1999), and Stein (1999) for a detailed
description of the types of fiscal rules and their characteristics.
3. See Craig and Manoel (2002).
4. Efficiency and fairness were improved as top marginal tax rates on corporate
income were reduced from 40.6 percent in 1987 to 35 percent in 2000, and on indi-
vidual income from 55 percent to 40 percent during the same period. Currently,
both rates are set at 34 percent.
5. The present value of various off-budget contingent liabilities is estimated to have
totaled about 90 percent of GDP in 1999; see Santaella (2001).
6. See Mexico, Secretaría de Hacienda y Crédito Público (SHCP) (2002b) for a review
of the major changes under the fiscal reform of 2002.
7. For a comparison of the costs and benefits of VAT reform, see the PRONAFIDE
(National Program to Finance Development) of June 2002 (Mexico, SHCP 2002a).
8. Participaciones are federal transfers directly related to Federal Revenues Collection
(Recaudación Federal Participable), and other formulas usually tied to foreign com-
merce for local governments located at international ports, and to collection of
local taxes. These transfers are unconditional; that is, they do not depend on the
spending destination at the local government.
9. Procedural rules may allow for a reduction in government spending if revenues are
lower than expected. We emphasize the term ‘contingent’ since the adopted rules in
Mexico involve spending adjustments on different sectors depending on the source
of the revenue fall (i.e. tax revenue and oil revenue).
 Fiscal Rules in Mexico 145

10. For a comprehensive assessment of progress in this area, see IMF (2002a).
11. Following the Management Information System (Sistema Integral de Contabilidad
Gubernamental) the Public Account shows the final results on the execution of the
LIF and the PEF.
12. For an initial presentation of the two measures, see the presentation by Gil-Díaz
(2001).
13. See Mexico, Presidencia de la República (2001) for a revised version of the reform
package.
14. Marcel et al. (2001) estimate that Chile achieved a structural surplus for almost
the entire 1992–2000 period.
15. Of course, the fiscal stance is not to be confused with sustainability, which is com-
monly determined taking into account the evolution of the public debt–GDP ratio.
16. There is no consensus on what the structural balance coverage should be. Chile,
for example, calculates it on the overall balance of the central government, rather
than in terms of the public sector borrowing requirement (Marcel et al. 2001).
17. While the PSBR is more comprehensive than the traditional balance, it is still not
completely exhaustive (International Monetary Fund 2002a). In particular, the
PSBR does not include liabilities associated with social security institutions and
subnational governments.
18. A detailed description of the methodology used to estimate the structural PSBR
and the algorithm used to estimate the evolution of the ratio of public debt to
GDP is available from the authors on request.
19. For our purpose, public debt sustainability is defined as a path of the public
debt–GDP ratio that is constant or declining through time.
20. The public sector’s absorption of financial resources is calculated as the ratio of
the cumulative PSBR to the broadest monetary aggregate (M4) in Mexico.
21. From a measurement standpoint, the structural PSBR can be regarded as identical
to actual PSBR without nonrecurrent revenue, as the cyclical component can be
ignored over the long run.
22. This rule does not imply that public sector absorption of savings will be zero.
23. Our simulations do not include explicit random shocks. However, as shown later,
the simulations are undertaken considering alternatively whether the goals in
PRONAFIDE are achieved or not.
24. M4 includes currency held by the public, deposits in domestic banks, domestic
financial assets held by residents and nonresidents, plus deposits held by branches
and agencies of Mexican banks abroad. In a stable macroeconomic environment,
M4 tends to grow at a slightly higher rate than GDP; furthermore, with structural
reforms that contribute to financial deepening, more resources are available to
finance public debt.
25. The main difference, for computational convenience, is that the proposal allows
for a temporal deviation from the rule and this exercise does not.
26. For participaciones, based on estimated tax revenue, we calculate federal transfers
to states and municipalities and add them to the interest cost of public debt to
obtain nonprogrammable spending. We then estimate the trajectory of program-
mable spending consistent with a balanced budget. Next, we calculate the fiscal
balance, subtracting total spending from total revenue, and add the other com-
ponents to obtain the PSBR.
10
Fiscal Rules on the Road to an
Enlarged European Union
Fabrizio Coricelli and Valerio Ercolani1

Introduction

The Stability and Growth Pact (SGP) has come under increasing pressure as
the fiscal position of several European Union (EU) member countries has dete-
riorated with the slowdown of their economies (European Commission
2002a). Proposals for revising the fiscal rules contained in the SGP have
advanced in two main aspects: first, the budget deficit limit should apply to
the cyclically adjusted balance, and second, public investment should be
excluded from the computation of the balance. Indeed, in a recent proposal,
the European Commission (2002c) suggests that the deficit be adjusted for the
economic cycle. The proposal also gives more weight to public investment
and the cost of structural reforms in assessing the fiscal position of countries
with a public debt stock below 60 percent of GDP. These revisions would also
provide more flexibility to most countries which recently completed, or are
currently candidates for, EU accession and which have moderate debt ratios
and high public investment.
Interestingly, the issue of EU enlargement, particularly concerning the
fiscal requirements for accession countries (ACs), has been totally neglected
in the debate, with some notable exceptions (Buiter and Grafe 2002). This is
surprising if one considers that accession countries average high structural
deficits and more than three times the ratio of public investment to GDP as
member countries. One possible reason for such neglect is the widespread,
but mistaken, view that as long as they stay outside the euro area, new
members will not be subject to penalties and thus should not regard the
3 percent of GDP limit on budget deficits – an obligation for all member
countries – to be binding. Although financial penalties for exceeding the
limit are not applicable to ACs, they risk losing access to the Cohesion Fund.
Another reason for ignoring the implications of a revised SGP for enlarge-
ment may be the view that persistent budget deficits can support growth in

146
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Fiscal Rules on the Road to an Enlarged EU 147

countries with relatively low levels of income per capita, contrary to

evidence of an adverse effect of budget deficits on growth (among others,
Easterly and Rebelo 1993).
This chapter uses the same methodology as applied to member countries in
the past to analyze the fiscal position of selected ACs, consisting of new mem-
bers (Hungary, Poland, Slovenia) plus a candidate (Romania), characterized by
different stages of reform and different levels of income per capita. Even with
its limitations, this is one of the first attempts to compute cyclical and struc-
tural deficits for a set of ACs with a view to designing appropriate fiscal rules.
The analysis points to three main features in these countries: structural
deficits well above the limit of 3 percent of GDP, with the exception of
Slovenia; procyclical fiscal stance; and significant output volatility coupled
with a high level of public investment, including EU accession-related expen-
diture. Hence the greater likelihood of surpassing the deficit limit.
Without a revision of the existing fiscal rules, ACs will probably undertake
belated and inefficient adjustments, proving very costly to their economies.
The main limitations of the current rules, namely the procyclical bias of the
ex post budget deficit limit and the inclusion of all public investment in the
deficit, appear in full light in the case of ACs. Nevertheless, these limitations
affect all member countries and enlargement could provide an opportunity
to modify the rules. Although the deficit limit is not procyclical per se – as
in principle there is scope for freely using automatic stabilizers – the actual
implementation of fiscal policy within the EU implies that budgetary plans
are agreed upon with the European Commission on the basis of forecasts on
the state of the economy. Accordingly, if a country starts from a deficit posi-
tion, it will agree to a plan of gradual deficit reduction. A significant forecast
error (consisting of an unanticipated slowdown), however, can push the
actual deficit well above 3 percent of GDP.2 In such a situation, the country
is required to make an adjustment, which indeed will imply a procyclical fis-
cal policy. Moreover, it is likely that under pressure a member country will
adjust by resorting to one-off measures and creative accounting, rather than
by correcting its structural position.
In this chapter we suggest a new rule that focuses on ex ante limits to expen-
diture.3 Abstracting from the important issue of the role of debt ceilings dis-
cussed at length by Buiter and Grafe (2002), the suggested rule would be
countercyclical by design, fully transparent, and free from ambiguous inter-
pretation. Furthermore, the rule would require governments to generate sur-
pluses in periods of growth above trend, and would allow for deficits in
periods of economic slack. As long as a country were to follow the rule based
on the estimated potential output growth and the targeted inflation rate, no
ex post adjustment would be required. The suggested rule, which is similar to
that adopted by the United Kingdom, implies a discipline that is even tighter
than prescribed by the SGP. Furthermore, the rule could accommodate spe-
cific needs for public investment, as well as accession-related expenditures.
148 Fabrizio Coricelli and Valerio Ercolani

The rule is fully consistent with the theory of optimal tax smoothing.
Whereas changing the existing rules may entail risks, a refusal to revise them
can pose an even more serious threat, as the entire architecture of fiscal
discipline can be put into question by persistent deficits well above the limit.

Computing the structural balance

A major obstacle to computing the structural balance, that is, the actual
balance adjusted for the impact of the economic cycle on the budget, is the
lack of long time series for estimating trend GDP and for capturing the
sensitivity of expenditure and revenue to the cycle. In spite of this, even con-
sidering only the 1990s, it is possible to obtain useful results. With all their
specific problems during the 1990s, ACs display a cyclical behavior similar
to EU member countries. As shown in Coricelli (2002), for some ACs, the
degree of comovement of their output with that of the EU is very high,
although the amplitude of the movements is higher. In other words, one can
identify a common cycle in candidate and EU countries, although ACs
display much higher output volatility.

The “gap” plus elasticity approach

The European Commission (1995) calculates the structural or cyclically
adjusted balance (CAB) from the difference between the actual budget
balance (ABB) and the cyclical component of the balance (CBB) for each EU
member country. The methodology underlying the CBB is based on the
computation of the output gap4 and of elasticities linking revenue and
expenditure to the cycle.5
The cyclical component of the budget (CBB) is simply the difference
between the cyclical components of revenue (CR) and expenditure (CE). In
turn, cyclical revenue is given by

CR 
TY  *

i

t
Ti; Y * gapt 
TY  *

c

t
Tc; Y * gapt 
TY  *

ind

t
Tind; Y * gapt


SSC
Y 
*

t
ssc; Y * gapt (10.1)

where Ti stands for individual taxes; Tc for corporate taxes; Tind for indirect
taxes; SSC for social security contributions; Y for GDP; and
for the elastic-
ity of each revenue component to GDP. Further, cyclical expenditure is

CE 
ub; ur *
ur; Y * gapt (10.2)

where
ub; ur is the elasticity of unemployment benefits to the rate of

unemployment and the
ur; Y is the elasticity of unemployment to output
 Fiscal Rules on the Road to an Enlarged EU 149

(a sort of inverse Okun-coefficient). All revenue and expenditure items are

multiplied by the percent gap between actual and potential GDP. For most
member countries the cyclical component of the budget balance oscillated
by 1.5 percentage points of GDP up and down during the period 1960–97,
and was seldom above 3 percent of GDP. Only in Finland and Sweden did
the cyclical component reach more than 4 percent of GDP during the deep
recession in the early 1990s (Buti and Sapir 1998).
In addition, the impact of the cycle on the budget balance can be usefully
summarized by the so-called marginal sensitivity of the budget (SB),
comprised of the marginal sensitivities of revenue (SR) less expenditure (SE).
With Rc denoting the total revenues that are potentially affected by the
cycle, and with
Rc;Y the weighted average of the individual elasticity of such
items with respect to GDP changes, we obtain

SR 
Rc; Y
RY 
c

For member countries, SR was on average 0.5 in 1997,6 implying that one
percentage point of negative output gap causes a revenue reduction of
0.5 percent of GDP.
Unemployment benefits represent the only component of public expen-
diture that automatically responds to movements in output, whereby SE is
much smaller than SR. With
ub;Y (
ub; Y 
ub; ur *
ur; Y) denoting the elasticity
of unemployment benefits with respect to GDP, we can define SE as

SE 
ub; Y

For EU countries, SE has been on average around 0.1 percent of GDP. In other
words, one percentage point of negative output gap causes a 0.1 percent
of GDP increase in expenditure. Thus SB is dominated by the revenue
component.
Combining SB with output volatility provides two important indicators
for monitoring fiscal policy used by the European Commission, namely the
cyclical safety margin (CSM) and the minimal benchmark (MB). Output
volatility is used to define a worst-case scenario, by taking for each member
country the average of the two worst indicators among the following three
(over the period 1970–2001): the largest recorded negative output gap, the
arithmetic average of the negative output gaps above 4 percent; and the
average volatility of output gaps, measured as twice their standard deviation.
The resulting measure of output gap variability is then multiplied by the SB
to obtain the CSM, which is an indicator of the highest level of cyclical
deficit that can be achieved by a country. This index is clearly an increasing
function of SR, SE, and output volatility.
150 Fabrizio Coricelli and Valerio Ercolani

The difference between the deficit limit of 3 percent of GDP and the CSM
defines the MB. This indicator measures the maximum structural deficit that
a country can sustain without inhibiting the functioning of automatic
stabilizers during recessions. Clearly, countries with high CSM should aim at
structural surpluses in order to avoid being forced to implement a procyclical
adjustment during a downturn.

Fiscal indicators for accession countries

To calculate fiscal indicators used in monitoring fiscal policy in the EU, we
focus on Hungary, Poland, Slovenia, and Romania.7 Estimates of potential
output growth are very similar to those obtained in other studies, using dif-
ferent methodologies. Except for Slovenia, we used observations over a
period that includes the prereform regime. Although the pre and postreform
output are not strictly comparable, we believe that information of past
trends is valuable, especially because countries entered their reform period
with different initial conditions. Considering only the postreform period
carries the risk of interpreting growth from the recovery of lost output as
potential growth. Nevertheless, we tried several specifications, focusing on
the 1990s. Estimates of the CBB generally decreased, while those of the CAB
increased. In all, the main results of the analysis of fiscal stance and the
structural position remain unchanged.8
Apparent revenue elasticities were computed separately for each main
revenue item – as the absence of long time series precluded regression
estimates. On the expenditure side, the sensitivity of unemployment bene-
fits to changes in output, through the effect of output on unemployment,
were calculated. By and large, ACs were found to have characteristics simi-
lar to member countries. The similarity increases with the stage of reform
and development of the countries examined, which has an impact on the
magnitude and structure of fiscal variables. For instance, the main differ-
ences were found for Romania, which is lagging in the reform process and
is much poorer than the new members.
One of the reasons for the similarity in outcomes is that the government
size of ACs, measured as the ratio of revenue or expenditure to GDP, is
comparable to that of EU countries (Table 10.1). It is well known that the
main cyclical impact on the budget takes place through revenue with a GDP
elasticity of revenue generally close to one. As a result, a 1 percentage point
deviation of GDP from its trend (or potential) level has an impact on the
budget of a magnitude close to the ratio of revenue to GDP. The impact
through unemployment benefits is as small in ACs as in EU countries, with
the exception of some Scandinavian economies. Thus, the elasticity esti-
mates suggest that the structure of fiscal systems of ACs is broadly similar to
that of EU countries.
However, the sensitivity to the cycle of the budget (SB) for the ACs exam-
ined is on average 0.35, roughly 0.2 percentage points below the EU average
 Fiscal Rules on the Road to an Enlarged EU 151

Table 10.1 Accession countries: fiscal elasticities, minimal benchmark, and cyclically
adjusted balance

Rc;Y SR
ub; Y, SE SB CSM MB CAB (in 2000)

Hungary 1.00 0.32 0.08 0.40 3.02 0.02 4.10

Poland 0.99 0.29 0.06 0.36 3.16 0.16 4.25
Romania 0.98 0.25 0.06 0.31 4.15 1.15 2.00
Slovenia 0.99 0.34 0.08 0.42 0.75 2.24 0.76
EU average 1.00 0.50 0.10 0.60 1.60 1.40 0.60

Sources: Authors’ calculations for ACs, and European Commission (2002a) for EU average.

(Table 10.1).9 Romania displays a much smaller sensitivity, mainly due to a

lower ratio of revenues to GDP, consistent with the view that convergence
in fiscal structure is still underway in that country.
Given that output volatility is much higher in ACs (Table 10.2), while their
budget sensitivity is not so far from that of member countries, we can con-
clude that for ACs it would be much harder to keep the budget deficit within
the 3 percent of GDP limit during an economic downturn as illustrated by
a relatively high CSM (Table 10.1). In addition, comparing the CAB with the
actual balance, it is apparent that the budget deficit in ACs is almost entirely
structural in nature. This fact held true for most of the 1990s (Appendix).10
Given that the public investment–GDP ratio has been much higher in ACs
than in EU countries (Table 10.2), a large component of the structural deficit
is linked to public investment.11 This raises the issue of the relevance of the
“golden rule” for ACs that need large investment in infrastructure.

Procyclical fiscal stance

Arguably, the forecast of higher average growth for ACs during the catching-
up process would permit them to satisfy the deficit limit. However, their
fiscal performance in recent years seems to suggest that such a statement is
not warranted, as fiscal policy in ACs tends to be procyclical.
During the 1990s, an increasing structural imbalance more than compen-
sated for the positive effects of economic growth in the ACs as illustrated by
the procyclical stance (Figure 10.1). Figure 10.1(a) displays an inverse rela-
tionship between the structural balance as a ratio to GDP and the output
gap, indicating that in periods of growth below trend, fiscal policy was tight-
ened, while the opposite holds for periods of growth above the trend.
Figure 10.1(b) confirms this result, by linking the structural balance to the
GDP growth. The structural deficit has remained high during periods of
growth, while during the sharp economic downturn of the early 1990s, lack
of access to borrowing induced a significant tightening of fiscal policy.
For countries like Slovenia, in which there was little cyclical fluctuation of
 152
Table 10.2 Accession countries: output and fiscal indicators (percent of GDP), 2000

GDP Trend Output Output Budget Total Total Interest Public

Growth GDP gap volatility1 balance revenue expenditure expenditure investment
growth

Hunagary 5.2 3.9 1.5 3.7 3.5 40.6 45.1 6.1 7.1
Poland 4.1 4.2 3.6 4.5 3.2 39.4 42.6 2.7 3.1
Romania 1.6 1.0 5.3 6.2 3.7 31.5 35.1 4.9 3.1
Slovenia 4.6 4.8 1.2 0.4 1.3 42.8 44.1 1.5 4.1
EU average 3.3 2.2 1.1 1.5 0.2* 46.8* 47 3.7** 1.0**

Notes: 1 Standard deviation is calculated for the 1990s in ACs, and for 1970–97 in EU countries.
* Excluding revenues from UMTS.
** Data refer to euro area.
Sources: Authors’ calculations, EBRD and IMF for ACs; European Commission (2002a) for the EU average.
 Fiscal Rules on the Road to an Enlarged EU 153

(a) 3

1
Changes in CAB

0
–15 –10 –5 0 5 10
–1

–2

–3

–4
Output gaps

(b) 4

0
–15 –10 –5 0 5 10
CAB

–2

–4

–6

–8
GDP growth

Figure 10.1 Accession countries: fiscal stance (percent of GDP), 1990–2000*

Notes: * See Appendix for details on data series. The relationship summarized by the regression
lines is statistically significant at the 3 percent level in both cases.
Sources: IMF and authors’ estimates.

GDP, the latter figure seems more relevant. Budget deficits were on average
well above 3 percent of GDP, except for Slovenia.
More broadly, a procyclical stance has been found in emerging market
economies, especially in Latin America (IMF 2002c) in particular during bad
times, when access to financing deficits may disappear. Although to a much
154 Fabrizio Coricelli and Valerio Ercolani

2.5

1.5

0.5

0
Accession Latin Emerging Advanced
countries America Asia countries

Figure 10.2 Volatility of government revenue (standard deviation of revenue–GDP

ratio), 1991–2000*
Note: * 1993–2001 for accession countries.
Sources: IMF and authors’ estimates.

less extent, a procyclical stance also characterized EU countries during the

period 1970–95 (Buti et al. 1997). This effect seems due to an expansionary
fiscal policy during periods of positive output gap, more than compensating
for the positive effects of automatic stabilizers (Buti and Sapir 1998). Less
clear-cut are the views on the stance during recessions. Procyclical fiscal
stance implies the absence of tax smoothing. Indeed, for ACs and Latin
American countries during the 1990s, the volatility of the revenue – GDP
ratio has been much higher than in advanced economies (Figure 10.2).
Will the adoption of the existing rules overcome the procyclical stance of
fiscal policy displayed by ACs? As discussed in the next section, the deficit
limit tends to induce procyclical adjustments by requiring a country to
reduce the deficit during a marked, and often unanticipated, slowdown of
the economy. The problem faced by Germany and France in 2003 is bound
to be more serious for ACs as the likelihood of surpassing the limit is higher
than for member countries, due to a much higher output volatility in the
former.

Rules for an enlarged European Union

Previous sections lead to three main results. First, ACs display very large
structural deficits; when combined with high output volatility, these deficits
cause major difficulties in satisfying the deficit limit. Second, AC fiscal
policy has been highly procyclical. Third, public investment is much higher
in ACs than in member countries.
 Fiscal Rules on the Road to an Enlarged EU 155

Assessing the existing rules

What happens when candidate countries become EU members? They are
immediately subject to the SGP, although financial penalties for nonobser-
vance apply only when they join the euro area. Meanwhile, failure to satisfy
the deficit limit may trigger an interruption of payments from the Cohesion
Fund.
What would be the rationale for applying the deficit limit to new mem-
bers? Among other factors, the justification of the ex post deficit limit of
3 percent of GDP is associated with two main elements. First, this limit was
roughly equivalent to the public investment–GDP ratio of EU members
during 1960–90 (Buiter and Grafe 2002). Second, and more relevant to the
current debate, given the output volatility in EU countries, breaking the
limit would be rather exceptional for countries running deficits of about
1 percent of GDP. Indeed, with a budget sensitivity of 0.5 to the cycle, in
order to exceed the limit a country would have to experience a negative
output gap of 6 percent if it maintained a balanced structural budget.
How can one reconcile this view with the fact that in 2002 Germany’s
deficit approached the 3 percent of GDP limit as a result of a real growth rate
roughly 2 percentage points below expectations? The point is rather trivial.
With a unitary GDP elasticity of revenue, in the short run the decline in rev-
enue will equal the shortfall in growth. By contrast, expenditure is planned
ex ante on the basis of expected growth, irrespective of any value of output
gap. An unanticipated contraction in output may lead to a deficit above the
limit whatever the value of the output gap is. One cannot blame a country
for misbehavior when a deficit emerges because of a forecast error. Asking
the country to make the adjustment after observing ex post the higher deficit
makes fiscal policy procyclical. The argument is very simple, but reveals a
basic drawback in the ex post rule, namely the contradiction between the
ex post evaluation and the expenditure plan that is based, by necessity, on
the ex ante forecast. This further underscores a flaw in the view that the
deficit limit represents a wide margin of flexibility during economic cycles.
In other words, given the unitary GDP elasticity of revenue, and largely
exogenous expenditure, what matters in the short run is the actual change
in GDP rather than the output gap, which is affected by past output perfor-
mance. This line of reasoning is even more relevant in the case of ACs. In
1998–99, during the period of the Russian crisis, Estonia displayed a swing
of 10 percentage points in GDP growth, with dramatic budgetary conse-
quences, irrespective of the measured output gap.
One solution could be to consider the condition of new members as
exceptional, and thus tolerate deficits above the limit for a transition period.
However, exceptions to a rule may induce other countries to claim excep-
tional circumstances as a justification for excessive deficits. Furthermore, the
idea of considering new members “special” could create a two-tier EU.
156 Fabrizio Coricelli and Valerio Ercolani

Excusing new members for fiscal indiscipline may in fact harm their
economies, with adverse consequences on the external current account
balance, macroeconomic volatility, and economic growth.
We suggest that the existing fiscal rules in the EU can be improved, with
benefits applying not only to ACs, but also to all member countries. In sum,
the existing rules are inefficient in two respects. First, the ex post deficit limit
induces procyclical adjustments of fiscal policy during a marked slowdown,
but does not exert sufficient pressure for adjustment during good times.
Second, public investment is considered the same as any other expenditure,
subject to the overall limit on the budget deficit. The proposed rule tackles
the procyclical bias and can also accommodate the “golden rule.”
The suggested rule is a binding commitment that by design is counter-
cyclical. Furthermore, focusing on nominal expenditure, the rule is trans-
parent and free of variant interpretations. Finally, the rule makes fiscal policy
consistent with the underlying potential growth and inflation targets of dif-
ferent countries. Specifically, for euro area members, the inflation target is
set by the European Central Bank (ECB).
The suggested rule is a steady state solution. Starting from large structural
deficits there is initially an issue of convergence toward a structural balanced
budget, amended eventually with some form of golden rule. This realistically
implies a framework for gradual adjustment, as it is already the case for mem-
ber countries. The recent proposal of the European Commission (2002c) sug-
gests an adjustment in the structural deficit of 0.5 percent of GDP per annum
for member countries that do not comply at present with the SGP.

An ex ante “structural close-to-balance” rule

We suggest an ex ante balanced-budget rule of the following form:

Rs(1  *t  1)  Rsc{1 
Rc;y [(1  g*t  1)(1  *t  1)  1]}  Rft  1  Eft  1, (10.3)

where Rcs is the structural component of revenue linked to the cycle at time
t; Rs is revenue not linked to the cycle at time t (capital revenue); gt* 1 is the
growth in potential output; t*  1 is the inflation target (set by the ECB for the
euro area); Rtf  1 is the nominal value of the total structural revenue for
t  1, forecast at time t; and Etf  1 is the nominal value of expenditure (exclud-
ing unemployment benefits) for the year t  1, decided at time t.
Therefore, for the year t  1, the rule requires expenditure to match the
aggregate revenue, consisting of revenue not linked to the cycle in t updated
on the basis of targeted inflation, plus the structural component of revenue
linked to the cycle in t, increased by the growth rate of potential output and
the target rate of inflation.12
Once the level of nominal revenue has been established, the rule defines
the expenditure level, excluding unemployment benefits; in this way, the
 Fiscal Rules on the Road to an Enlarged EU 157

cycle cannot influence public expenditure. This amount should be consis-

tent with the medium-term fiscal framework for EU member countries.
Estimates of potential output growth and the inflation target have to be
agreed upon with the EC. The planned expenditure and revenue forecast of
each national government should also be approved by the Commission.
Denoting with Rt  1 the nominal level of revenue actually collected dur-
ing the year t  1, and with Et  1 the actual nominal level of expenditure
(excluding unemployment benefits) at year t  1, the balanced-budget rule
is satisfied if (a) Etf  1  Et  1  Rtf  1  Rt  1, when actual growth coincides
with the estimate of potential growth and the inflation target is met, or if
(b) Etf  1  Et 1, but Rtf  1  Rt  1, when either actual growth deviates from
potential growth, or inflation is different from target, or both. The rule is not
met if (c) Etf  1  Et  1 and/or Rtf  1  Rt  1, and none of the conditions in
(b) obtains. In this case financial penalties would automatically be applied.
In condition (b) there is a budget deficit when the actual output growth is
lower than the potential growth rate, and a surplus when actual is above
potential growth.13 However, as long as the country has followed the
announced rule, no corrective action is required and no penalty applies.
Thus fiscal policy is fully countercyclical, the budget is on average balanced
by design, and government debt is stabilized in the medium run. Therefore,
the rule is an authentic structural close-to-balance rule.
The rule satisfies two main objectives. First, it is fully consistent with the
philosophy of the SGP, which establishes that a common monetary policy has
to be underpinned by a common fiscal target, though not necessarily the same
level of budget imbalance. This is achieved by assuming a target on the struc-
tural budget balance and a fiscal behavior over the cycle consistent with the
underlying potential growth rate and the inflation target of the ECB. Second,
the rule is countercyclical by construction. In this way, the rule avoids the
undesirable outcome of forcing countries to make an adjustment during an
economic downturn, possibly worsening the economic outcome for the EU as
a whole. At the same time, the rule induces the necessary adjustment during
good times, forcing the countries to run surpluses in those periods.
Another feature of the rule is that countries that have a higher potential
growth can run higher deficits during recessions. To illustrate the statement,
consider two countries with different potential growth but suffering a
recession of the same intensity. Assuming unitary GDP elasticity of revenue
and setting planned expenditure growth in line with the rate of potential
output growth, the country with higher potential growth will display a
larger deficit as its expenditure will increase faster, while its revenue will
decline in the same proportion, as in the other country. In summary, the rule
accommodates different realities of the countries. This may be particularly
relevant for ACs that are characterized by higher rates of potential growth.
Moreover, the suggested rule is not subject to the traditional criticism that
spending norms do not relate to fiscal variables, namely, the budget balance,
158 Fabrizio Coricelli and Valerio Ercolani

which can lead to negative externalities (Buti et al. 2003). Under a structural
close-to-balance rule, the budget is to be balanced over the cycle, and thus
negative externalities are minimized.
One could object that the suggested rule is more difficult to implement
than the current rule. In fact, this can hardly be true. On the revenue side,
the calculation of the cyclical revenue necessary to establish whether a coun-
try deviated from tax neutrality is exactly the same as that currently used
by the Commission for the computation of the CAB, replacing, however, the
output gap with the actual GDP growth rate. Regarding the expenditure side,
there is no implementation problem. The fiscal authority has only to verify
whether the nominal level announced at time t is actually spent at t  1. No
interpretation problem can arise. This kind of rule is in fact the one normally
used in the budget process, for example in the United Kingdom (United
Kingdom H.M. Treasury 1998). Upon application within the EU, the rule
would have important implications from a political economy point of view.
Indeed, fiscal rules would be shifted to the national authorities, as it should
be in an economic and monetary union but not a political union. The
budget process would be controlled by auditors approved by member coun-
tries. The role of the Commission and other European institutions would be
limited to coordinating the process and verifying the auditors’ reports.
Deviations from the rules should automatically trigger a punishment
procedure. A by-product of the expenditure audit is that the budget process
and all expenditure centers will become more transparent, and thus reduce
the scope for creative accounting. National authorities will be responsible
and accountable for each country’s fiscal policy. They cannot blame the
Commission for their own inefficiencies.

Possible limitations
The suggested rule identifies unemployment benefits as the only component
of expenditure directly linked to cyclical fluctuations. However, besides
cyclical fluctuations, nominal expenditure is affected by other “nonstruc-
tural” variables as well. For instance, unexpected inflation would require
higher nominal expenditure to implement investment plans decided at
time t. Similarly, changes in the nominal interest rate, domestic or foreign,
would affect interest expenditure on debt planned at time t.
The rule would apply only when these variables do not change signifi-
cantly; this is clearly a limitation, but current fiscal practices in the EU could
be equally criticized because they focus on nominal deficits. Even the recent
proposal put forward by the Commission focuses only on the effects of the
economic cycle on the budget balance.
There are two alternatives to cope with the effects of changes in inflation
and interest rates. One would be to complement the rule with an extraordinary
 Fiscal Rules on the Road to an Enlarged EU 159

fund, created in agreement with the Commission, from which the

government can draw resources to keep the original expenditure plan; by
contrast, in the event that actual expenditure was below the planned level,
the windfall should be deposited in the fund. The other alternative would
consist of adjusting nominal expenditure for the effects of inflation and
interest rate changes.
It is clear that setting nominal expenditure has drawbacks, as unexpected
changes in inflation and interest rates would modify the expenditure plans.
This limitation is shared with most other rules, including those on budget
balance.

Application to accession countries

The structural close-to-balance rule that we suggest is applicable in the short
run to those ACs that do not have large structural deficits. In most countries
examined, structural deficits over the past four years averaged at least 3 per-
cent of GDP. Thus a balanced-budget rule would require a large fiscal adjust-
ment, hardly achievable in the short run, especially if one takes into account
pressures on expenditure stemming from structural reforms and EU acces-
sion. In addition, the ratio of public investment to GDP in these countries
is much higher than in current EU members.
The rule could thus be modified, in accordance to a sort of “golden rule”:

Eft  1  Rft  1  x, (10.4)

whereby a permanent structural deficit x could be accepted until ACs

completed their economic transition. Such a deficit should be related to an
agreed-upon level of public investment. Nevertheless, the need for fiscal
discipline is even stronger in these economies, which are still in the category
of emerging markets. Good growth prospects should permit the achieve-
ment of the close-to-balance structural position within the medium-term
framework.

Conclusions

This analysis raises several doubts on the efficiency of existing fiscal rules in
the EU in light of the enlargement. Notwithstanding data limitations, the
conclusions about the prevalence of large structural deficits and high public
investment ratios in ACs seem rather robust. Similarly, evidence on
procyclical fiscal policy is quite convincing. Existing rules have the unfortu-
nate property of reinforcing the procyclical stance during bad times and
160 Fabrizio Coricelli and Valerio Ercolani

providing little incentive for surpluses during good times. Furthermore,

these rules are likely to affect ACs adversely, by applying a common limit to
budget deficits irrespective of their stage of development, potential growth
rate, or public investment needs.
This chapter suggests some elements for a new type of rule, focusing
primarily on the cyclical aspect. The proposed rule is consistent with the
views recently expressed by the IMF (2001d) and is similar to that adopted
by the United Kingdom. Buiter and Grafe (2002) discuss more generally a
rule that takes into account debt sustainability problems. Over the longer
term the suggested rule and its implications for fiscal performance should be
linked to the dynamics of public debt.

Appendix

Data sources
International Monetary Fund, International Financial Statistics, for nominal
GDP, real GDP, and unemployment rate; Government Finance Statistics
Yearbook, for government revenue. European Bank for Reconstruction and
Development (EBRD 2001)14 for government balance. National sources for
unemployment benefits.

Methodological issues
Data on tax revenue cover the period 1990–2000 for Hungary, 1995–2000 for
Poland, 1992–99 for Romania, and 1992–2000 for Slovenia. In calculating
revenue elasticities with respect to GDP we identified several outliers: for
Hungary at the beginning of the 1990s, and for Poland and Romania at the
end of the decade, in particular 1999. No outliers were detected for Slovenia.
The values of these elasticities appear stable over the sample, except for these
outliers.
The series for unemployment benefits cover the period 1993–2000 for
Hungary, 1993–99 for Poland and Slovenia, and 1995–99 for Romania. On
the expenditure side, elasticity estimates appear less stable than for revenue.
Nevertheless, the impact of expenditures on the final calculation of the CAB
is much smaller, and thus the potential error much less serious.
Because of the lack of disaggregated data on revenue for the period
1990–94, the CBB for Poland in that period has been calculated by multi-
plying SB with the corresponding output gaps. The same approach has been
followed for Romania for the year 2000. Annual estimates of ABB, CBB, and
CAB for each AC are shown in Figure 10.3.
 161

(a) Hungary
4

0
% of GPD

–2

–4

–6

–8

–10
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
ABB CBB CAB

(b) Poland
4
3
2
1

0 –
% of GPD

–1

–2

–3
–4
–5
–6
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
ABB CBB CAB
162 Fabrizio Coricelli and Valerio Ercolani

(c) Romania
4

0
% of GDP

–2

–4

–6

–8
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
ABB CBB CAB

(d) Slovenia
1

0.5

0
% of GDP

–0.5

–1

–1.5

–2
1992 1993 1994 1995 1996 1997 1998 1999 2000
ABB CBB CAB

Figure 10.3 Estimates of budget balance, 1990–2000

Sources: IMF and authors’ estimates.

Notes

1. We thank George Kopits and Lucjan Orlowski for very helpful comments and
suggestions. This chapter was first presented at the annual meetings of the
Association for Comparative Economic Studies, Washington, January 3–5, 2003.
 Fiscal Rules on the Road to an Enlarged EU 163

2. A significant slowdown of the economy relative to the forecast is what matters,

and not the occurrence of a recession.
3. In the recent literature, the main reference on expenditure rules is Kopits
and Symanky (1998). Other contributions are Mills and Quinet (2002), IMF
(2001d), Brunila (2002), von Hagen (2002), and Fitoussi and Creel (2002). In some
cases, the overall expenditure rule is combined with the “golden rule” for public
investments.
4. The HP filter (Hodrick and Prescott 1980) was used by the Commission for esti-
mating trend GDP until 2002, when it was replaced with the production function
method. For EU countries the discrepancy between the results of different
methodologies is not very large.
5. The approach consists of regression estimates of the response of major revenue and
expenditure items to changes in GDP. The methodology recently adopted by the
OECD is based on a two-step procedure, in which the relationship between GDP
and each revenue or expenditure item is computed by regressing first each item
with respect to the macroeconomic variable directly linked to it, and then by
regressing the macroeconomic variable with respect to GDP (van den Noord 2000).
6. We computed SR as an average of values for the 1990s, rather than for a single
year.
7. We applied the HP filter to calculate trend output for ACs. For 2002–04, EU fore-
casts of GDP were used to avoid the end-point bias typical of the filter. In line
with European Commission (1995), we used a value of 100 for the Lagrange mul-
tiplier. For Slovenia, data for only ten years were available. Nevertheless, the sta-
bility of the growth rate of Slovenia in recent years indicates the presence of very
little cyclical fluctuation (Appendix). For Hungary, see Darvas and Simon (2000).
8. It should be also noted that the HP filter gives high weight to the 1990s. See King
and Rebelo (1993) for a technical discussion.
9. Such a lower sensitivity is found as well in other studies. For Hungary, see Csajbók
and Csermely (2002).
10. One could object that this result depends on the methodology used, and possibly
on large measurement errors, in both the output gap and the cyclical components
of the budget. However, given the elasticity of revenue, the highest possible cycli-
cal impact on SR would be obtained by taking total revenue, implying only minor
changes in the calculation of the structural balance.
11. A disaggregated analysis (European Commission 2002a, chapter V) shows that the
distribution of public expenditure is similar for ACs and EU countries, except for
public investment, which is three to four times higher in ACs.
12. The assumption of constant structural level of revenue is consistent with the the-
ory of “optimal tax smoothing” (Barro 1979). Of course, adjustments to tax rates
could be easily introduced, although they should be approved by the competent
authority.
13. The rule refers to the growth rate of potential output, not to the output gap. This
rationale underlies the DFI (discretional fiscal indicator) built to detect discre-
tional fiscal behavior (Buti and van den Noord 2003).
14. The EBRD database is used because it contains a longer series than those available
from the European Commission (2002b). Nevertheless, even with the latter, our
main results do not change significantly.
11
Fiscal Rules for Economies with
Nonrenewable Resources:
Norway and Venezuela
Olav Bjerkholt and Irene Niculescu1

Introduction

Recent studies have singled out resource-abundant economies as having

weak economic performance over the last three decades; the greater the
resource endowment, the lower the per capita growth rate.2 Analysts offer a
variety of reasons to explain the low growth rate in these economies.
Resource abundance has led to a shift away from competitive manufacturing,
with a consequent loss of growth-inducing externalities. Critics have also
argued that resource abundance may blunt the incentives to save and invest,
partly due to the slow development of financial institutions.3 Other expla-
nations center on phenomena derived from pervasive rent seeking, as a large
amount of resources is channeled through the public sector. For oil-
producing countries a notable feature has been the extreme procyclicality of
government expenditure with respect to oil price fluctuations. A more
sophisticated growth-theoretic argument is that the resource abundance has
led these economies to overshoot in expanding demand beyond the steady
state path, requiring subsequent adjustment from above.4
A central question is whether fiscal rules could have helped these
economies make better use of resource gains and achieve higher growth.
Although it is doubtful that rules alone could have prevented mistaken
policy choices especially during resource booms, appropriate fiscal rules may
have helped to shift the policy focus from a shortsighted use of resource rev-
enues toward the long-term pursuit of fiscal sustainability and growth. This
chapter begins with a brief survey of salient characteristics of economies
with nonrenewable resources, followed by a discussion of the risks they face.
After considering the potential role of fiscal rules in these economies, we
examine the cases of two large oil producers, Norway and Venezuela. In
particular, we look at the evolution of the rule recently adopted in Norway

164
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Economies with Nonrenewable Resources 165

limiting the cyclically adjusted nonresource budget deficit to the return on

accumulated proceeds from resource extraction. Lastly, we present recent
innovations in Venezuela, aimed at temporarily separating stabilization and
saving functions and combining an oil stabilization fund with other
macrofiscal rules. The chapter concludes with a comparative assessment of
fiscal rules in the two countries.

Economies with nonrenewable resources

Economies well endowed with a nonrenewable resource, especially the oil-

producing countries, extract a substantial resource rent from a resource that
is sufficiently scarce in an oligopolistic market, dominated by the
Organization of Petroleum Exporting Countries (OPEC). A large proportion
of the resource rent normally accrues to the government through ownership,
royalties, and taxes; therefore, fiscal policy assumes a crucial role in the
economy-wide and intertemporal distribution of the rent.
As the export price is volatile, the resource rent fluctuates in ways that are
difficult or impossible to predict. Occasional periods of windfall gains may
lead to overestimation of future income, as happened in oil-producing coun-
tries when the oil prices rose in the 1970s and early 1980s.5 Given the dearth
of experience of many of these countries, government spending decisions
were often made in response to short-term fluctuations in petroleum earnings.
Thus, commodity price volatility exercised a destabilizing influence on the
economy, and fiscal policy assumed an additional role, namely, managing
risks associated with the exploitation of the natural resource.6
Economies endowed with natural resources differ with regard to resource
dependency. At one end of the spectrum are those with a large resource
endowment, upon which export earnings and government revenues are totally
dependent, as in some oil-rich countries in the Middle East. The challenge
for these economies is their transformation from total resource depen-
dence toward a long run, end-of-resource sustainability. At the other end
are economies with limited resource dependency that does not warrant
particular fiscal consideration. In the middle range are some Latin American
countries that are substantially but not entirely resource dependent.
Typically, an economy endowed with a valuable natural resource evolves
through three phases. An initial phase of exploration and early extraction is
followed by a mature phase with a stable production level, and eventually,
succeeded by the terminal phase when production and revenue decline. The
early boom phase may bring large fiscal deficits as substantial costs are
incurred. The mature phase harvests the gains by using liquidated natural
wealth – mainly through the fiscal channel – for either consumption or sav-
ing. The adjustment needed to adapt to declining levels of resource earnings
in the terminal phase is facilitated if earlier proceeds have been saved.
166 Olav Bjerkholt and Irene Niculescu

Since nonrenewable resources are part of the national wealth, when such
resources are extracted and sold on the world market, the resource rent is
counted as income in the national accounts. Yet in a real sense it is not
income, but liquidated or transformed wealth. Hence, resource-abundant
economies may not be as rich as they seem from resource exploitation, nor
do they save on a net basis as much as it appears. Furthermore, the resource
rent differs from other incomes because the part of it that accrues to the gov-
ernment does not have the contractionary effect on the domestic private
sector as other revenues. This implies that a (cyclically adjusted) balanced
budget carries a different meaning for economies with a large net fiscal
income derived from natural resource rent; a balanced budget will be expan-
sionary or contractionary depending upon the fluctuation of resource prices
or other factors determining the net fiscal resource income. Price volatility
renders a strict balanced-budget policy highly procyclical and destabilizing.
Resource revenue tends to bring about shifts in the relative size of the
tradable and nontradable sectors, as discussed in the booming sector or
Dutch disease literature.7 This results in real appreciation and loss in com-
petitiveness. The use of resource revenue thus entails macroeconomic adjust-
ment costs, which are necessary for the benefits of the natural resource
wealth to be enjoyed, but which also carry the risk of overshooting and pol-
icy failure. However, the adjustment costs can be eased over the medium
term through gradual phasing in and proper signaling, depending mainly on
the industrial structure and on the flexibility of the labor market.
Conversely, the adjustment can result in macroeconomic havoc if fiscal
impulses are driven by the fluctuations in resource revenue. The smoothing
of expansionary and contractionary impulses, however, may not suffice. The
deindustrialization resulting from using a liquidated nonrenewable resource
eventually has to be reversed by a costly reindustrialization process. The
growth-theoretic literature suggests that the loss of industrial clusters may
imply long-term reduction in endogenous technological progress. The larger
the resource sector, the larger the potential gains, but the larger also the
exposure to risks inherent in the extraction and use of the resource.

Cash-flow versus net-worth risks

The focus on risk management in oil-exporting countries emerged as a

consequence of the drop in oil prices in the mid-1980s. Whereas the ensu-
ing liquidity problem drew attention to the uncertainty in earnings over a
short span of time, for most economies with substantial resource earnings
the resource base is likely to last for decades, requiring a long-term policy
perspective. At that time, the gap between expected and actual earnings in
the short term reflected uncertainty in cash flow rather than uncertainty
in net worth.8
 Economies with Nonrenewable Resources 167

Uncertainty affects resource earnings in two ways. First, uncertainty in

short- to medium-term fluctuations in earnings cannot be estimated with
precision. Second, uncertainty also affects the total (discounted) value of the
nonrenewable-resource extraction. While cash-flow risk is associated with
short-term price volatility, net-worth risk arises because the long-term trend
in prices is uncertain, information on cost of extraction from new fields is
insufficient, and the stock of reserves and thus overall production potential
is unknown. However, short-term fluctuations in revenue need not affect
economic stability, including the levels of domestic consumption and
investment, to the extent the government can resort to short-term borrow-
ing. Thus the policy regime should aim at coping with net-worth uncer-
tainty, but remain robust enough to accommodate cash-flow fluctuations.
Of the three major sources of uncertainty, prices and costs are more sig-
nificant than reserves. Uncertainty about the stock of reserves may well be
large, but it comes into play at the far end of the resource horizon and has
a relatively minor affect on the net worth as long as the rate of price increase
is lower than the discount rate. By contrast, changes in the price trend and
the cost level will have a direct effect on earnings in the near future. In fact,
the future price of the resource is determined by market prices and produc-
tion costs of alternative products, which is largely a technological issue.
Governments can manage net-worth uncertainty through the combina-
tion of a depletion policy to maximize the net worth, an appropriate royalty-
cum-tax regime, sales of property rights, and investment of proceeds in
activities unrelated (or possibly negatively related) to the resource market. In
any case, it is prudent to adopt a risk-averse attitude with regard to spend-
ing oil revenues.9 A sanguine spending policy implies reallocation of
resources from manufacturing to service industries, a process that can be
costly to reverse. If the consumption level is raised in anticipation of future
resource revenues and it turns out that the net worth of the resource sector
is less than assumed, then a painful readjustment will have to be endured.
To lower consumption may be difficult enough, but reindustrialization to
strengthen export capacity is even more difficult. By implication, it is nec-
essary to avoid resource dependence, while setting aside a major part of the
earnings for investment abroad, until the net-worth risk diminishes.
The option of raising the royalties and lowering the tax rate helps dampen
the influence of price fluctuations on revenues. Selling the resource on long-
term contracts reduces the risk, but is feasible only for some commodities,
say for natural gas, but not for oil. The net-worth risk can be reduced by
speeding up depletion and by reducing the reserves left to compete with sub-
stitutable products in the future. Even if rapid depletion (usually subject to
limitations) increases costs, the present value of reserves might still increase.
The option of diversifying through transformation of the natural resource
wealth into alternative assets, thus reducing resource dependence, calls for
large-scale sales of property rights and purchase of financial and real assets.
168 Olav Bjerkholt and Irene Niculescu

Although a country with a large resource endowment can hardly achieve the
sale of a major part of its endowment (lacking sufficient market breadth),10
even a smaller disposal would be a step toward risk spreading, leaving the
country with the task of investing substantial funds over a limited period of
time in financial assets abroad and/or real assets at home. For proper risk
spreading, the value of these assets should be negatively correlated with the
resource market. However, domestic investment opportunities may be lim-
ited by absorption capacity and foreign investments may be exposed to
political risk.

Is there a role for fiscal rules?

Discussion of the use of resource revenue revolves mainly around the level
of permanent income, defined as the expected return on the remaining
resource reserves (plus the return on accumulated assets from resource
revenue). This is the income level that obtains without reducing the overall
resource wealth over time. In the early phase, with low extraction, the
permanent-income rule thus calls for borrowing against future earnings,
which in turn are used for amortization in the mature to terminal phases.
Permanent income is, however, an elusive concept. As the valuation of the
natural resource wealth in practice fluctuates over time, the permanent
income must be expected to vary over time.11
Simple intertemporal optimization models support using the permanent-
income rule, thus leaving all generations with the same benefit from the
nonrenewable resource.12 Equalizing incomes equalizes marginal utilities
and appears fair from an intergenerational perspective. If the international
rate of interest is higher than the rate of technological progress plus the pop-
ulation growth rate, the preferences for equity among generations within a
Ramsey-type growth model require transfers from later to earlier genera-
tions. Abstention from consumption, obviously, increases the stock of
reserves available to later generations. If the preference for intergenerational
equity is high enough and the interest rate is not too high, the optimal solu-
tion is for consumption to grow over time, but less than production.
However, the permanent-income rule is limited in several respects. First, the
rule does not pay sufficient attention to the uncertainty of future revenue
flows: it counts resources in the ground as equally certain as holdings of risk-
less financial assets. Second, the rule ignores future expenditure commit-
ments, including contingent liabilities associated with the social security
system, which tend to expand over time. Social transfers are broadened to
become universal benefits. Demographic pressures further boost benefits. And
third, in spite of its intended fairness, the permanent-income rule does not
take into account the costs of industrial restructuring mentioned earlier.
Arguably, uncertainty in these areas calls for a cautious use of resource revenue
whereby spending is allowed to increase only as the uncertainty vanishes.
 Economies with Nonrenewable Resources 169

Thus, fiscal policy in economies with nonrenewable resources must contend

with the additional tasks of phasing in the resource revenue, reallocating
over time the resource earnings relative to the depletion and earnings profile,
and protecting against the destabilization and possible default due to unful-
filled expectations. Questions about the rationale and design of rules-based
fiscal policy to cope with these tasks follow naturally from the earlier men-
tioned considerations. Specifically, can fiscal rules be designed in line with an
optimality consideration? The short answer is that a rule can be derived for
setting government expenditure, taking into consideration the expendi-
tures in the previous year, the macroeconomic adjustment costs of chang-
ing the expenditure level, and adverse price movements reflected in default
risk.13
If there is no claim on optimality grounds, however, what then is the role
of a fiscal rule? Can fiscal rules help in providing a better policy framework
and focus attention on the longer-term issues that so easily fade into the
background?14 As suggested earlier, a balanced-budget rule might not be a
desirable option for two reasons. First, an economy drawing down natural-
resource wealth, taking the depletion rate as given, may have good reasons
for intertemporal redistribution of the liquidated wealth, including to gen-
erations living beyond the terminal phase of exploitation. Second, even
absent intertemporal concerns, a balanced-budget rule would need to be
modified (by specifying it in terms of, say, a structural balance) to avert the
procyclical impact of high volatility in resource earnings; in other words,
revenue use would have to be decoupled from the current resource earnings.
An alternative approach pursued in a number of resource-abundant
economies (most recently adopted, for example, in Ecuador) consists in
accumulating part of the resource revenues in a stabilization fund, for
smoothing the impact of short-term volatility, or in an endowment fund, for
promoting long-term sustainability.15 A fund can make the treatment of
resource earnings more visible, if subject to strict transparency requirements
and if transfers between the fund and the government budget are embedded
in a coherent macroeconomic framework. Because a basic function of the
fund is to cover the budget deficit (while reflecting changes in government
net financial worth), the fund cannot be separated from budgetary opera-
tions. Thus, potentially, the fund becomes the centerpiece of any effort to
establish fiscal rules to ensure stability in the short run and sustainability in
the long run.

Resource endowment and socioeconomic setting

For a better understanding of the evolution of rules-based policies in Norway

and Venezuela, it is useful briefly to review the experience and setting of
each of these major oil-endowed economies. A comparison of key social,
economic, and institutional characteristics is particularly relevant.
170 Olav Bjerkholt and Irene Niculescu

Although a latecomer as a petroleum producer (commencing in the early

1970s), Norway has become the second largest oil exporter in the world,
with large petroleum earnings relative to a small population. Oil production
in Norway started as part of a diversified economy with public finances
characterized by high levels of taxation and expenditure. A significant
manufacturing sector was, unlike in most OECD economies, complemented
with abundant fish, hydroelectric power, and forest resources. The emergence
of the oil and gas sector crowded out manufacturing activity, though with-
out creating much unemployment.
Prior industrialization of the economy helped prevent greater oil depen-
dency during the period of most turbulent price fluctuations. The increase
in oil revenue over time, particularly throughout the 1990s, supported a
sharp increase in government commitments to generous welfare services
and in labor costs required to produce such services. The establishment of
the State Petroleum Fund (SPF) in the early 1990s was an attempt to come
to terms with these commitments through a long-term policy for using the
remaining petroleum wealth.
Social homogeneity, relative income equality, and general support of the
welfare state, which was part of the post–Second World War political legacy,
may have helped cope with the early challenges of the petroleum era. By the
same token, the accelerated aging of the population provides an added
dimension to the judicious medium- to long-term management of the
resource wealth.
By contrast, in Venezuela, large-scale oil drilling began in the 1920s amidst
a poor rural setting as the country was in the process of political unification.
In 1958, Venezuela became one of the founding members of OPEC estab-
lished to coordinate production in the international market. Through the
1960s, oil-export revenue supported urbanization, industrialization, and
overall economic development.
The oil boom of the 1970s strengthened the dependence of public
finances and of the economy on oil exports. The unfavorable consequences
of oil dependence became evident when oil revenues fell in the mid-1980s.
Since then, a short-term view has prevailed in policy decisions, and reforms
to reduce oil dependence have generally been postponed or only partially
implemented. Over the last two decades, the Venezuelan economy has
exhibited to an extreme extent the stylized features discussed earlier for
emerging market economies with nonrenewable resources, resulting in a
very poor overall performance.
Non-oil activity was incapable of replacing the dynamic effect of oil
exports and fiscal profligacy. Public spending has deteriorated in quality, and
remained above a sustainable level because non-oil revenue could not offset
the fall in oil revenue; these trends have resulted in persistent financial
deficits since the mid-1980s. Erratic fluctuations in the oil market, in public
finances, and in the real exchange rate provoked a fall in public and private
 Economies with Nonrenewable Resources 171

Table 11.1 Norway and Venezuela: selected indicators, 2000

Norway Venezuela

General indicators
Population (millions) 4.5 24.1
Old-age dependency ratio (percent of population) 15.4 4.4
GDP (billions of US$) 169.0 120.0
GDP per capita (thousands of US$) 37.3 4.9
External current account balance (percent of GDP) 14.0 10.9
Unemployment rate (percent) 3.3 13.2
Inflation (percent) 3.0 16.2
Oil sector indicators
Production (millions of barrels per day) 3.3 3.2
Exports (millions of barrels per day) 3.1 2.8
Reserves (billions of barrels) 23.0 76.8
Reserves (years of production) 19.1 65.8
Oil revenue/fiscal revenue total (percent) 25.2 49.8
Oil exports/total exports (percent) 47.0 73.1
Oil GDP/total GDP (percent) 24.4 22.7
Oil investment/total investment (percent) 18.7 39.3

Sources: IMF and World Bank.

investment, particularly in the manufacturing sector, with adverse conse-

quences on growth, productivity, formal employment, and poverty; thus
overall economic activity has been insufficient to absorb a relatively young
labor force.
A comparison of the Norwegian and Venezuelan economies, summarized
by basic indicators, reveals important differences, particularly in terms of
GDP per capita, the relation between oil production and reserves, and the
share of oil revenue in overall government revenue (Table 11.1). However,
equally important are the underlying demographic and social conditions.
Arguably, contrasting socioeconomic environments are critical determinants
of the design of the present institutional arrangements for the conduct of
fiscal policy, as discussed in the following sections.

Norway’s “bird-in-the-hand” approach

Successive oil shocks called attention to the macroeconomic exposure of

resource abundance and led to the creation of the State Petroleum Fund
(SPF) as the basic tool for making the use of oil revenues more transparent,
facilitating the decoupling of revenue use from revenue inflow, and ensur-
ing an appropriate long-term allocation of revenue. Formally, the SPF can be
viewed as a government account held with the central bank. The inflow
accruing to the SPF is the government’s oil-related net cash flow, in the event
172 Olav Bjerkholt and Irene Niculescu

of a budget surplus, while the outflow consists of transfers to the government

budget to cover the oil-corrected deficit. The central bank invests the SPF funds
abroad, increasingly in major stock markets.
Initially, long-term considerations for transfers from the SPF were used to
set target levels in the macroeconomic surveys presented every fourth year
to the parliament.16 The final proposal for the use of oil revenues, incorpo-
rated in the annual budget bill, was expressed by the target level with
changes due to cyclical considerations and “extraordinary transfers” decided
in the budget process. The aim of the procedure was thus to ensure that the
long-term policy considerations would provide a baseline for current budget
decisions, with some latitude for modification for short-run reasons. The
baseline was meant to prevent irresponsible fiscal decisions, but without
encroaching on the parliamentary debate and approval. Any parliamentary
decision with expansionary fiscal implications made outside the budget
session had to be accompanied with an explicit decision to withdraw trans-
fers from the SPF. However, this procedure never came into play as intended.
The SPF (without any accumulated capital) existed only on paper until 1996,
the first year with a fiscal surplus since 1990. Thereafter, the SPF increased
substantially, and by the end of 2001 it had a value close to 60 percent of
non-oil GDP (about US$80 billion).
In 2001, a new regime was enacted, supplementing the SPF with an
explicit fiscal rule to replace the political-institutional procedure, intended
to strengthen the decoupling from current revenue. This rule is based on a
“bird-in-the-hand” approach, whereby the use of oil revenue must be
determined by the liquidated resource wealth accumulated in the SPF.17 The
underlying arithmetic splits total revenue into oil-related revenue (R1) and
other revenue (R2), and total expenditures likewise into oil-related expendi-
ture (C1) and other expenditure (C2). The design of the fund requires identi-
fication of the oil-related components in the budget, which in principle
comprise the government share of the petroleum rent, though in practice
the more directly observable oil-related net cash flow (R1
C1) is used.18 Oil-
related revenue consists of taxes and royalties, return on accumulated assets,
plus other oil-related revenue. The oil-related expenditure is comprised of
items related to oil development and production, including government
capital injections. The overall surplus is thus

S  R1  R2
C1
C2. (11.1)

The oil-corrected deficit (D2) is

D2  C2
R2  R1
C1
S. (11.2)

For the SPF, established at t  0, the stock of accumulated capital at the

 Economies with Nonrenewable Resources 173

beginning of period t is given by the equation

Ft  (1 rt
1) Ft
1  (R1, t
1
C1, t
1)
D2, t
1. (11.3)

To fulfill the requirements of decoupling and gradual phase-in, the fiscal rule
sets the target for the oil-corrected deficit equal to the real return on the cap-
ital of the petroleum fund:

D2, t*  rt (Ft) (11.4)

where rt is the actual rate of return on the fund portfolio.

Condition (11.4) is a sustainable rule for determining the use of oil
revenue, as it takes into account future uncertainties to a large extent. In
essence, the stochastic properties of the oil price fluctuations, that is,
whether the oil price fluctuates as a random walk, is mean reverting, or fol-
lows a more intricate stochastic process, become negligible. The rule works
to ensure a gradual phasing in of oil revenue, thus limiting the macroeco-
nomic impact. Although the rule may seem overly conservative, especially
in comparison with the permanent-income rule, it can accommodate the
explicit and implicit future fiscal commitments built up over time.19
However, the rule as set out in (11.4) is crude in the sense that it pays no
attention to the cyclical situation and ignores the risk of unanticipated
changes in the return on the SPF portfolio.20 Therefore, a modified rule,
accommodating cyclical fluctuations and uncertainty in the return, deter-
mines instead the target in terms of the cyclically adjusted oil-corrected deficit21

D2, t# rt*(Ft) (11.5)

where rt* is a set rate of return on the fund portfolio.22 The modified rule
includes in addition an escape clause stating that abrupt changes in the target
deficit, say by a fall in stock values, are to be smoothed out over a number of
years. Hence, utmost attention is paid to smooth phasing in of the oil revenue.
A remaining question is whether the rule will contribute to macroeco-
nomic stabilization or will accommodate a continued fiscal expansion and
thus an overheating of the economy. The theoretical prediction is that the
expansion will lead to increasing costs through wage increases and currency
appreciation. The ensuing shift away from the production of tradables due
to reduced cost competitiveness will eventually take place in step with the
fiscal expansion and when the boom peters out. Overshooting in the wage
level is to be expected, followed by increased unemployment at a later stage.
The gradualism inherent in the rule – notwithstanding its slight procyclical
bias during prolonged expansions or recessions – is, however, key to letting
the adjustment run with the least possible overheating and overshooting.
Moreover, the primarily long-term role of the fiscal rule is well complemented
174 Olav Bjerkholt and Irene Niculescu

by the inflation-targeting regime that has been adopted as the context for
monetary policy.
The combination of the fiscal rule with an oil fund has become the cor-
nerstone of Norway’s fiscal policy, developed on the basis of a wide political
consensus following three decades of oil production. Although well designed
for Norway, the remaining problem with the fiscal rule may well turn out to
be political. In particular, it may be increasingly difficult to prevent increases
in public expenditures for laudable purposes above and beyond what the
rule prescribes, especially during a prolonged period of high oil prices and/or
high returns on the assets accumulated in the SPF.23

Venezuela’s “birds-in-the-bush” approach

In Venezuela, the level of proven oil reserves and the associated risk
imply that fiscal policy should be broadly in line with long-term fiscal sus-
tainability. Although the Norwegian approach has been taken into account
in Venezuela’s recent fiscal reform, initial socioeconomic conditions impose
certain limitations on the design of the fiscal policy framework. Indeed, a
different approach to oil-risk management is called for, particularly because
the extreme oil dependence exacerbates the requirements for stabilization;
lacking stabilization, oil dependence is likely to affect economic growth
adversely. More important, unlike in Norway, Venezuela’s non-oil fiscal rev-
enue alone is insufficient to support the country’s development needs in
education, health, social programs, and infrastructure.
Achieving stability has been a major challenge for Venezuelan policy-
makers. Attempts to deal with the volatility of oil revenue by creating or
reforming various oil stabilization funds (1990, 1998, 1999, and 2001) have
been disappointing. The first stabilization fund was never really imple-
mented; the second never accumulated funds since it came into effect dur-
ing a period of very low oil prices. The third fund accumulated deposits
(US$7 billion in savings in 2001), but at a low benchmark price (US$9 per
barrel) with half of the oil revenue above that price to be saved in the fund,
which meant that an excess of savings had been accumulated when prices
began to decline in 2001. The fund was reformed at the end of that year and
no contributions were required in 2002.
The most recent fiscal reform envisaged a two-pronged approach at the
outset by initially assigning the stabilization and saving functions to two
separate funds, thus deferring the full implementation of an integrated sav-
ings fund to the future. The reform is based on the 1999 Constitution and
the 2000 Organic Law of Financial Administration of the Public Sector
(LOAF). The constitution states that the revenue from exploitation of the
subsoil must be used first and foremost to finance fixed investment, and
education and healthcare expenditure. Accordingly, the LOAF establishes
the Macroeconomic Stabilization Fund (FEM) and the Intergenerational
 Economies with Nonrenewable Resources 175

Savings Fund (FAI). The FEM is intended to stabilize public expenditure and
thus to contribute to macroeconomic stability, and the FAI to maintain fiscal
long-term sustainability and intergenerational equity – both funds to be
supported by pending legislation.
Agreement has been reached on the design of simple and symmetric norms
for contributions and withdrawals: a “moving average” benchmark for the
FEM and a dynamic “permanent income” method for the FAI. The stabiliza-
tion norm establishes a benchmark based on a moving average of past oil rev-
enue: when actual revenue is above the benchmark, the excess is deposited
in the FEM; when actual revenue is below, a withdrawal is permitted to meet
the fiscal adjustment requirement, with a ceiling of up to 50 percent of the
assets accumulated in the fund. Contributions and withdrawals from the FEM
are integrated within the budget. Any additional transfer, due to unforeseen
increase or fall in oil revenue, has to be approved by the national assembly.
The savings norm, based on the principle of intergenerational equity, is
intended to bring about an intertemporal redistribution of oil revenue by
saving part of the revenue in the FAI during the maturity phase in order to
guarantee future generations a similar per capita revenue from the fund.
Estimates of permanent revenue, based on projections of production, prices,
costs, and taxation, vary over time, as pointed out earlier, according to new
information available on the evolution of these variables, policy decisions,
and new estimates. Thus the norm can be considered to be dynamic in the
sense that the estimates need to be reviewed regularly. Transparency in the
application of the norm is to be ensured by an independent technical board,
created by law, responsible for estimating the permanent level of oil revenue.
Political support for the FAI is rather timid. Heeding critics’ arguments
that this is not the appropriate time for full and immediate implementation
of a savings mechanism, the LOAF requires that, during the first ten years of
operation, a declining portion of the contribution to the FAI be deducted for
investments with intergenerational characteristics, particularly in high-yield
non-oil-related projects (in infrastructure, education, and healthcare), so as
not to concentrate the burden of the reform on present taxpayers. As a mat-
ter of fact, the size of the present non-oil deficit (about 10 percent of GDP)
implies that in the foreseeable future, Venezuela will need to follow such a
“birds-in-the-bush” approach.24
Like Norway’s SPF, the FAI is expected to spread the risk of the oil wealth
by investing the accumulated funds in a diversified portfolio of long-term
financial assets abroad, not correlated with the value of oil reserves, placed
in escrow for a 20-year period to insure an adequate buildup. In the future,
the FAI is envisaged to absorb the stabilizing function and, in this respect,
the FAI will be similar to the actual SPF. During the transition, a connection
between the FEM and the FAI has been established by setting a ceiling on
the accumulation in the FEM, above which the excess savings will be trans-
ferred to the FAI or used for public debt restructuring.
176 Olav Bjerkholt and Irene Niculescu

Differences in initial conditions, as compared to the Norwegian case,

argue also for a different set of macrofiscal rules geared to the objectives of
stability and sustainability. In Venezuela, it is necessary to cope with the
prevailing deficit bias, asymmetric liquidity constraint, and procyclical
borrowing, as well as to safeguard the effectiveness of the stabilization fund.
The 1999 Constitution provides the basis for the necessary legal and proce-
dural elements to promote these goals by establishing through the LOAF a
multiyear budgetary framework (MBF) and fiscal rules that require the
observance of ordinary balance and limits on current expenditure and
borrowing. In addition, the LOAF sets the principles that should govern
fiscal management: efficiency, solvency, transparency, and responsibility.
Specifically, the LOAF regulation defines the constitutional rule of ordinary
balance at the central government level over the medium term. For this pur-
pose, ordinary revenue is given by current revenue adjusted by contributions
and withdrawals from the FEM, which accumulates surpluses during booms
and finances deficits in recessions. Ordinary spending is equivalent to cur-
rent revenue, thus preventing cuts in infrastructure investment in reces-
sions. This is, in essence, a medium-term golden rule, adjusted for net
transfers from the oil stabilization fund.
The LOAF provides transitory provisions for a gradual implementation of
the MBF and the macrofiscal rules. The proposed phase-in includes a
transition period 2002–04 for the reform of public finances, including tax
reform, to reduce the non-oil deficit in convergence to the ordinary-balance
rule by the end of the period. For the period 2005–07, the first MBF, cover-
ing three years and based on oil-related and macroeconomic projections,
will come into full force.
In spite of the critical assessment of stabilization funds in solving the fis-
cal problems of economies with nonrenewable resources,25 and the earlier-
mentioned preference for expenditure limits by some experts, the prevailing
trends in Venezuela’s public finances suggest that a stabilization fund is a
promising tool in the recently enacted institutional framework. In
Venezuela, non-oil public sector deficit has been a primary source of money
creation and of macroeconomic disequilibria.26 The stabilization fund, in
combination with the ordinary balance rule, is crucial for preventing fiscal
procyclicality, which an expenditure rule alone cannot achieve.
Furthermore, an adequate implementation of the FEM is necessary along
with effective macroeconomic policy coordination. As regards the latter, the
Constitution requires the executive branch and the central bank to formu-
late an annual policy agreement, presented along with the annual budget
law to the national assembly. The Annual Policy Agreement Framework Law
should be fully phased in by 2005, when the first MBF comes into effect.
In sum, full commitment to the fiscal rules and the oil funds should sig-
nificantly strengthen Venezuela’s public finances. The FEM and the annual
policy agreement constitute the instruments for stabilization in the short
 Economies with Nonrenewable Resources 177

term; as a complement, the FAI and the fiscal rules contribute to sustain-
ability in the long term. Although the Constitution and the LOAF establish
an adequate institutional framework in which Venezuela can move toward
macroeconomic stability and fiscal sustainability, a broad social consensus
and the political will are yet to emerge for the implementation of the
reforms and the commitment to comply with that framework.

Concluding remarks

The experience of economies with abundant nonrenewable resources often

points to an overall lack of fiscal discipline, as well as weak and inefficient
resource management. The central challenge has been to move from one-
sided concern with the use of resource revenue for enhancing expenditure
toward the pursuit of macroeconomic stability and sustained growth. A well-
calibrated combination of resource management, possibly through a
resource account or fund, and of macrofiscal rules, can promote the realiza-
tion of these goals. However, there is no unique recipe for coping with
resource-related uncertainties. Management of resource-related revenue and
outlays, as well as of the resource fund, should take into account socioeco-
nomic and institutional conditions. Although in principle various options
for fiscal rules may be considered as long as they are conducive to sustain-
ability and growth, rules have to be designed and made operational for the
circumstances of each country.
The example of Norway illustrates the advantages of a “bird-in-the-hand”
approach consisting of an oil savings account plus a macrofiscal rule that
limits the (cyclically adjusted) non-oil fiscal deficit to the return on assets
accumulated from liquidated resource wealth. This approach decouples the
use of revenues from fluctuations in the net revenue derived from the
resource depletion. Norway focuses primarily on net-worth risk through sav-
ing oil revenue in a unique fund and setting a simple and transparent rule
to accommodate fluctuations. Such an approach is applicable in economies
where the fiscal situation allows the effective operation of a savings fund,
built up largely from sales of a significant part of the resource wealth.
The recent Venezuelan fiscal reform can be characterized as a “birds-in-
the-bush” approach. Against the backdrop of poor macroeconomic perfor-
mance, lack of fiscal discipline, and a decline in GDP per capita during the
past two decades, Venezuela has had to assign priority to macroeconomic
stability through the management of cash-flow risk. In addition, faced with
the need to finance acute requirements for spending on education, health-
care, social programs, and infrastructure, Venezuela has been obliged to
forego at this time the creation of an oil savings fund. Instead, it has opted
to temporarily separate stabilization and saving functions, while combining
them with a rules-based fiscal framework, leaving for the future the gradual
implementation of a savings fund.
178 Olav Bjerkholt and Irene Niculescu

Fiscal reforms along these lines may serve to depoliticize the use of non-
renewable resource wealth. Only a broad political support or consensus at
the outset can help achieve these reforms. Such support is facilitated by the
transparency, simplicity, and common-sense appeal of fiscal rules and
resource funds. But this support must be sustained over a prolonged period.
Although temporary abandonment of fiscal rules or violation of rules may
seem reasonable in exceptional circumstances – most commonly, during an
economic downturn – either would weaken the credibility of the rules-based
framework. In neither Norway nor Venezuela can unlimited support and
credibility be taken for granted; they must be earned and maintained
through steadfast implementation.

Notes

1. We are grateful to George Kopits for suggesting a discussion of Norwegian and

Venezuelan experiences in an integrated context. We are also grateful for comments
from Sheetal Chand, Amalia Lucena, Oscar Salcedo, and Fernando Villasmil.
2. See, for example, Sachs and Warner (1995); Gylfason et al. (1999); and Mayer et al.
(1999).
3. As argued by Gylfason et al. (1999).
4. Rodríguez and Sachs (1999).
5. Gelb and associates (1988) discuss the fate of selected countries through the
oil booms and after. The Venezuelan case is also discussed in Rodríguez and Sachs
(1999) and in Hausmann et al. (1993). The overestimation of future oil prices
was shared by international forecasters; the mainstream of international oil
forecasters predicted in 1980–81 that the oil price in 2000 would be about US$200
a barrel.
6. Lack of attention given to the fundamental uncertainty in long-range projections
in oil-producing countries before the lessons learned in the early 1980s was wide-
spread, for example, in the projections for Mexico in Brailovsky (1981).
7. Influential early contributions are Corden (1984), and Neary and van Wijnbergen
(1986).
8. The argument follows Norman (1982).
9. Although this reasoning implies that the rate of depletion cannot be completely
separated from the use of resource revenue, we consider the depletion rate as
given.
10. See Hausmann et al. (1993: p. 127–8).
11. Cappelen and Gjelsvik (1990) studied counterfactually the consequences of a
permanent-income rule applied to Norway in comparison with the “bird-in-
the-hand” rule discussed later.
12. As in Engel and Valdés (2000).
13. Although apparently addressing cash-flow risk rather than net-worth risk,
Hausmann et al. (1993) derive a spending rule for Venezuela in the form of a con-
cise and transparent mathematical formula – a rare example of a spending rule
derived from optimality considerations and yet simple enough to be adoptable by
an enlightened political regime. Alternative approaches developed, for example,
in Aslaksen et al. (1990), or the certainty equivalence approach in Aslaksen and
Bjerkholt (1986) are less amenable to operational policy conclusions.
14. For a persuasive argument in this direction, see Kopits (2001a).
 Economies with Nonrenewable Resources 179

15. A stabilization fund scheme was proposed for Indonesia in Kopits et al. (1993).
16. Such macroeconomic surveys have been prepared since the 1950s, in recent
decades also including macroeconomic projections 40–50 years ahead as
background for discussing long-term policy issues.
17. Norwegian Ministry of Finance (2001a,b).
18. In the short run the cash flow may differ a great deal from the resource rent, but
in accumulated terms they will approach each other.
19. Nevertheless, the IMF staff questioned whether the rule will be able to cope with
future fiscal needs: “Absent an early pension reform, a desire to maintain the
current share of non-pension public spending in GDP in the future – let alone
to allow it to rise in line with the demands of demography – would force Norway
to violate the fiscal rule and could thus result in the exhaustion of the fund before
mid-century.” (IMF 2001c).
20. Evidence of this risk is the loss of $10 billion incurred by the SPF in 2001.
21. The difference between D2, t* and D2, t# also smoothes out income and outlay
items, which can vary substantially over time, without cyclical importance.
22. The rate has been set in the Norwegian Ministry of Finance (2001b) at 4 percent
under the assumption that it is to remain more or less constant over time.
23. Rødseth (2001).
24. Adapting the metaphor, this approach is based on the expectation of future
returns, as compared to the more conservative Norwegian approach which is
predicated on returns on already accumulated capital.
25. See Davis et al. (2001).
26. See Niculescu (1999).
Part III
Design Issues at the
Subnational Level
12
Subnational Fiscal Rules:
A Game Theoretic Approach
Miguel Braun and Mariano Tommasi1

Introduction

This chapter presents a skeptical view of fiscal rules for subnational govern-
ments from the perspective of a broad game theoretic approach. The gist of
our argument is that the root of fiscal problems lies in political institutional
factors, such as the incentive for fiscal profligacy at the subnational level
caused by inadequate federal tax-sharing schemes, or the incentive for pub-
lic spending caused by principal–agent-type problems – for instance, incum-
bents’ tendency to increase public spending during election years.2 Fiscal
rules that do not address these underlying issues have a limited capacity to
solve fiscal problems, and might even be counterproductive.
Following the definition in Kopits and Symansky (1998), a fiscal policy
rule is a long-lasting constraint on fiscal policy, expressed in terms of a sum-
mary indicator of fiscal performance, such as annual limits on the govern-
ment budget deficit, borrowing, spending, or debt. Fiscal rules can be viewed
as a special case of fiscal institutions. A broader definition of fiscal rules
would include both policy rules in the narrow sense and budget procedures.
In the rest of the chapter we use fiscal institutions to refer to the narrower
definition involving numerical limits.
Although fiscal rules are increasingly championed as a key policy instru-
ment in achieving fiscal discipline (Kopits 2001a), it is not immediately clear
why enacting or signing a law, pact, constitutional amendment, or interna-
tional treaty that sets certain numerical targets will constrain economic and
political actors. In fact, a review of the empirical literature on the effective-
ness of fiscal rules in promoting fiscal discipline in the next section indicates
that the evidence is inconclusive.
Rules clearly contribute to fiscal discipline if there is a benevolent, all-
powerful external enforcer to sanction deviations, but this is rarely the case.
Absent strong, independent enforcement, it is not obvious that rules can
alter behavior. For instance, in the case of subnational fiscal rules being
imposed by the central government, the latter may have electoral incentives

183
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
184 Miguel Braun and Mariano Tommasi

not to punish a region that has violated the rule. Furthermore, a weak central
government might be unable to muster the political power necessary to
punish a deviating province. In many federal systems, the central government
is not necessarily benevolent, nor all-powerful; therefore, its capacity to act
as the enforcer of subnational rules is questionable.
The chapter explores the theoretical grounds for the usefulness and rele-
vance of rules, especially at the subnational level of government. Instead of
considering the central government as the perfect enforcer of subnational
rules, we focus on the (more interesting) case in which reliance on some
form of self-enforcement is necessary. This requires discussion of the game
theoretic foundations of rules. The policy reform strategy that emerges from
our approach is illustrated with a brief analysis of the case of Argentina.

Evidence on the effectiveness of rules

Finding convincing empirical evidence on the effect of fiscal rules on fiscal

performance is a challenging task. Changes in fiscal rules are not common
at the national level, and when they occur, they are often accompanied by
other reforms (financial liberalization, privatization, etc.) that affect fiscal
outcomes. This limits the explanatory potential of within-country time-
series regressions and cross-country regressions, because a study that found
a correlation between fiscal rules and favorable outcomes without control-
ling for other reforms would be biased. A related limitation of studies on the
effectiveness of fiscal rules arises from the question on the rules’ origin. If
the rules reflect voters’ preferences, and at the same time voters desire fiscal
austerity, then any study finding a correlation between rules and outcomes
would be contaminated by omitted-variable bias.
Poterba (1994), Alt and Lowry (1994), and Bohn and Inman (1996)
compare subnational fiscal rules within the United States. According to
Poterba and Reuben (1999), the argument in favor of this approach is that
US states operate in a homogeneous legal environment and face similar fiscal
shocks, while they exhibit substantial differences in their budget rules, fiscal
institutions, and fiscal outcomes. In fact, these studies document a negative
correlation between the stringency of state balanced-budget laws and the
average size of state budget deficits.
However, these studies do not address the issue of causality. Budget rules can
be endogenous in the sense that legislators and voters can change them.
Hence, it is possible that states in which voters have a preference for fiscal pru-
dence tend to have low deficits and balanced-budget laws.3 Again, in this case,
the correlation between fiscal rules and fiscal outcomes would simply reflect
the fact that both variables are jointly explained by an omitted variable.
In fact, there is evidence that this issue is critical. Poterba and Reuben
(1999) use instrumental variable estimation to attempt to solve the endo-
geneity issue in regressions of debt yields on fiscal institutions, but find that
 A Game Theoretic Approach 185

their point estimates become insignificant when instrumenting with proxies

for voter preferences. Thus the jury is still out on whether “good” state fiscal
rules cause better fiscal outcomes in the United States.
Similarly, in a recent survey of cross-country empirical studies, Kennedy and
Robbins (2001) report that the evidence on high-income countries is not con-
clusive. Fiscal consolidation has occurred in countries with or without fiscal
rules, implying that rules are not a necessary condition for fiscal adjustment.
In addition, since fiscal rules at the national level in industrialized countries
have not been in place for long, they have not yet been seriously tested,
because there has not been a major recession in these countries since the rules
have been in place. Preliminary evidence for Argentina, Brazil, Colombia, and
Mexico (Braun and Tommasi 2002) shows that the recent trend toward imple-
menting fiscal rules has had limited success in promoting fiscal discipline.

Where do rules come from?

The predominant assumption among economists seems to be that fiscal

rules are decided by a benevolent social planner (or by some exogenous
enforcement agent) so as to limit the tendency toward fiscal imprudence.
However, an important body of economics and political science literature
increasingly considers rules (and institutions in general), as the outcome of
complex intertemporal noncooperative interactions or games. Calvert
(1995) shows that in order to analyze the effectiveness of institutional
constraints on individual behavior, the key is to explain how institutions
can drive individuals to choose cooperative actions regardless of their indi-
vidual interest. Furthermore, it is essential to understand how institutions
can become self-enforcing, in the sense that mechanisms within the political
system encourage rules to be respected.
For an analysis of the logic and implementation of fiscal rules, it is
necessary to identify the “origin” or nature of the rules, since the underly-
ing problems being addressed via the rules, as well as the complementary (or
alternative) mechanisms, might differ case by case. Furthermore, the
enforcement technologies might also differ. We can best illustrate with some
examples of possible origins of fiscal rules: (1) In a strongly unitary system,
the central government imposes a fiscal rule (say, no borrowing) on subna-
tional governments; (2) Citizens impose rules on policymakers (e.g.
Proposition 13 in California); (3) The present generation of political actors
imposes rules on future generations of actors (for instance, in constitutions);
(4) A government (or unified actor) imposes rules upon itself, primarily to
signal commitment to external actors, for example, to financial markets; and
(5) A fiscal pact among subnational units sets certain fiscal rules, such as in
the EU and Brazil.
While some of these rules are imposed unilaterally, others are agreed upon
multilaterally by contracting parties. In the first set of rules an actor in a
186 Miguel Braun and Mariano Tommasi

position of authority tries to regulate other actors’ future behavior. These

vertical rules can be analyzed on the basis of conventional principal–agent
literature on optimal contracts. However, as we argue later, this scenario is
rarely the case. In federations, the central government is composed of many
actors, which often are representatives of subnational and sectoral interests,
and therefore the decision-making process of the national government rep-
resents also a noncooperative interaction, embedded in the larger game
played by different levels of government and groups.
On the other hand, horizontal rules, such as in (5), are compacts among
actors that assume mutual compromises: “I will keep my fiscal house in
order, in exchange for your doing the same.” It is a situation comparable to
private contracts enforced by the courts, though in the fiscal realm enforce-
ment can be problematic. It is in these sets of rules that the noncooperative,
game theoretic approach to institutions is most relevant. Why do subna-
tional governments comply with pacts? Who enforces sanctions when a
deviation occurs? How do economic shocks affect the incentive to cooper-
ate or to subvert institutions? These and other related questions should be
addressed to evaluate the effectiveness of a proposed set of fiscal rules. In the
rest of the chapter we focus mainly on horizontal rules because they capture
perhaps the most crucial aspects of subnational fiscal rules.

Can rules matter?

Rules are supposed to alter behavior, or at least to reinforce and insure cer-
tain behavioral patterns. To show their effectiveness, it is necessary to prove
that, in their absence, fiscal outcomes would have been worse. Thus we
explore under what conditions rules might have an impact on behavior,
considering that actions of fiscal authorities are the outcome of a noncoop-
erative game.4 These conditions involve either exogenous enforcement or
multiple equilibria.
If there is no outside enforcement and if the underlying game has a unique
equilibrium, rules will either be

● irrelevant (not binding): the actions taken in equilibrium without rules

lead to equilibrium values of the relevant fiscal variables which are within
the bounds permitted by the rules; or
● ignored: the underlying equilibrium leads to values outside the set
permitted by the rules, and the rules have no power to alter behavior.

In the case of games with multiple equilibria, rules might help players
coordinate equilibria which, hopefully, are Pareto dominant. Rules in this
case can be seen as a coordination device. An example taken from Przeworski
(1997) can help illustrate this point. Take three different game theoretic
setups representing different structures of conflict, in which A and B
 A Game Theoretic Approach 187

Case 1 Case 2 Case 3

A B A B A B
A 1,1 2,2 A 2,2 4,1 A 2,2 1,1
B 2,2 3,3 B 1,4 3,3 B 1,1 3,3

Figure 12.1 Outcomes under alternative policy choices

Source: Przeworski (1997).

represent actions by political groups, and 4
3
2
1 represent prefer-

ences over policy outcomes. The payoff pairs that constitute equilibrium
outcomes are shown in bold letters (Figure 12.1).
Cases 1 and 2 represent the situation of games that have only one equi-
librium. In Case 1, rules are not necessary, because the Pareto optimal out-
come is a unique Nash equilibrium, and is arrived at by individual
rationality. In this case, policymakers may still sign an agreement, or imple-
ment a rule, but it is simply a reflection of political reality, and is redundant.
In Case 2, exogenously enforced rules would be necessary to implement
B,B, which is preferred by both but is not an equilibrium in the one-shot
game. Rules would have to act so that individual incentives are curtailed.
That is, rules must somehow alter the game (specifically, the payoffs or fea-
sible strategies) so that what before was not an equilibrium now becomes
one. This case could be appropriate for analyzing many of the problems
faced by fiscal policy rules. The typical common pool problem faced by sub-
national jurisdictions due to credit market externalities is a case in which
cooperation is not a Nash equilibrium. If everybody else is running a low
level of debt, then an individual province has the incentive to defect and
run a high level of debt at a low cost. The key lesson from this example is
that for fiscal rules to matter, they must be enforceable, meaning that they
somehow affect the fiscal game so as to curtail individual incentives to defect
from the desired policy.
Hence, a key issue in determining whether fiscal rules can be an effective
mechanism to achieve fiscal discipline and macroeconomic stability is the
extent to which rules can be enforced. In the case where rules are intended
to change individual behavior toward cooperation, it is unlikely they will
work unless they include effective mechanisms encouraging the relevant
actors to comply with the rule. For instance, if a federal fiscal agreement stip-
ulates that all subnational governments must limit the growth of primary
spending to a specified annual rate and leaves enforcement to the courts, the
result may be ambiguous. In a country in which the courts lack political
independence, their decisions can be subverted, and thus the rule is unlikely
to encourage adjustment.
Absent a perfect external enforcer, institutions – and fiscal rules in
particular – must become self-enforceable (Calvert 1995). This means that
188 Miguel Braun and Mariano Tommasi

enforcement must come from within the game, that is, from changes in
payoffs, feasible strategies, or other fundamental conditions (including
automatic financial sanctions for noncompliance) that affect individual
behavior. Thus, rules can have an impact when there are multiple equilibria,
as in Case 3. In other words, rules can act as a coordinating device, so that
the groups can select the B,B equilibrium that is preferred by both.
The arguments presented so far show that fiscal rules can be redundant in
some cases,5 and insufficient to alter irresponsible fiscal behavior in others,
if not accompanied by deeper institutional changes substantially affecting
what we may call the fiscal game. However, unlike this static game, fiscal
games are not stationary, since many variables do change exogenously or
endogenously. The international prices of commodities exported and taxed
in certain regions of the country might change, or the dynamics of inflation
or of debt accumulation might put the country so close to the brink that the
payoffs of cooperation increase substantially. In these transitions, fiscal rules
can also play the role of coordinating device toward a new cooperative equi-
librium, coordinating the level at which fiscal variables will stabilize. This
highlights the relevance of endogeneity problems in assessing the impact of
fiscal rules, and implies that more than just rules are necessary for fiscal
consolidation.
In addition, even if we abstract from nonstationarities, games among
subnational and national governments are repeated games, in which fiscal
decisions are made time after time. Repeated games are much more likely to
fall under Case 3 as the normal form of a much larger game that might
consist of the repetition of the game in Case 2. In that sense, we concur with
Drazen in Chapter 2 that rules and reputation should be seen as complements:
rules (or institutions more generally) might help to coordinate actions
toward cooperative play, given the multiplicity of equilibria in repeated
games.
Therefore, we need to look into all the factors affecting the likelihood of
achieving a cooperative equilibrium, and consider the role that fiscal policy
rules can play in that broader context. Indeed, in many countries (including
Argentina, Brazil, Colombia, and Mexico), federal fiscal interactions are an
important part of the broader political game in which economic and social
policy decisions are made. At the same time, fiscal decisions are the outcome
of a subgame affected by the set of institutions (including rules) influencing
the overall political game.
In order to improve fiscal and macroeconomic performance, as well as to
attain goals such as adequate provision of social insurance and development
of human capital, it is important to have an understanding of the determi-
nants of poor performance, and to attempt solutions that tackle core reasons
behind noncooperative behavior by different levels of government. This
requires deeper (political!) country-specific analysis than standard
recommendations.
 A Game Theoretic Approach 189

In the rest of the chapter we provide a generic sketch of underlying prob-

lems of decentralized countries. We illustrate the approach with the case of
Argentina to expose the difficulties of achieving an adequate diagnosis, as
well as the type of solution required. In this case, fiscal rules are only one
component of a broader strategy of institutional reform.

Deficient outcomes and underlying problems

The problems underlying poor macroeconomic and fiscal performance can

be classified into two categories: principal–agent problems and cooperation
problems. Principal–agent problems encompass the relation between citizens
and their elected representatives. Imprudent fiscal behavior is often the
result of the actions of public officials who are not maximizing the welfare
of their constituents. Cooperation problems arise from the game played by
multiple subnational and national political actors, all maximizing objectives
that, to some extent, include the welfare of their own electors. For brevity,
we concentrate here on the second type of intergovernmental problems.
Table 12.1 summarizes fiscal policy problems identified in the political
economy literature over the last couple of decades.6 The problems are clas-
sified according to those faced by any country (or smaller political unit) even
if it is “unitary,” and the additional problems faced by a federation of states
or countries.
In general, a country that considers the introduction of subnational fiscal
rules faces major challenges. These involve suboptimal macrofiscal out-
comes, such as excessive subnational deficits, excessive indebtedness, insuf-
ficient subnational tax revenue, distortionary national and subnational
taxation, procyclical fiscal behavior, rigidities in the tax structure, and inad-
equate risk sharing. Many of these outcomes can be understood as the result
of noncooperative actions by national and subnational authorities. In this
respect a typical, but by no means unique, example of moral hazard at the
subnational level consists in the adoption of a lax fiscal stance (especially
when hit by permanent shocks), in expectation of a federal bailout in a fiscal
crisis.7 This opportunistic behavior of subnational governments is often met
by opportunism on the part of the federal government, whose willingness
to engage in bailouts may depend on the political alignment of the
subnational government in question, or on the exchange of a bailout for
favorable votes on a major national issue.8
This behavior reflects transactional cooperation problems. If there were no
transaction costs in intergovernmental relations, more efficient policy adjust-
ments could be made, via adequate intergovernmental compensatory
arrangements. This reasoning, adopted from transaction cost economics,9 is
useful in understanding many of the rigidities built into federal fiscal arrange-
ments as protection from the type of opportunistic actions described. As a
case in point, the federal revenue guarantees given to the provinces in
190

Table 12.1 Fiscal policy problems and their consequences

Underlying Size of public Deficit/debt Allocative

problem spending accumulation inefficiency

Unitary system
Lack of Too high Debt Underinvestment
representation of because voters accumulation in infrastructure
future prefer current excessive and long-term growth
generations spending
Political rotation Spending increases Debt increased Maybe unnecessary
(Political cycle, before elections before elections volatility of politically
strategic use to limit capacity sensitive components of
of debt) of new party to spending
change policy
Common pool Too high because Excessive debt Procyclicality
(political cleavages, groups do not accumulation,
interest groups, etc.) internalize full delayed stabilization
costs of additional (dynamic common
spending pool)
Principal–agent High spending Corruption,
(fiscal illusion, due to rent inefficiency
delegated authority, extraction by
information rents) policymakers
Federation
Common pool High spending Excessive debt
across jurisdictions as local accumulation, delayed
(dispute over governments do stabilization (dynamic
federal transfers, not internalize common pool)
local spending, etc.) costs of added
spending
Credit market If subnational Procyclicality of spending
externalities borrowing if more credit is available
constraints are lax, during good times
aggregate debt
may increase more
Financial bailouts Incentives to incur Reduced incentives for
in subnational local governments to
debt are spend efficiently
strengthened
increasing
aggregate debt
Structure of federal Flypaper effect Incentives for inefficient
transfers (low leads to higher national taxation at the
incentives to tax, spending national level. Increased
etc.) procyclicality of spending
if transfers to subnational
governments are from
earmarked taxes

Sources: Authors’ classification.

 A Game Theoretic Approach 191

Argentina in the 1999 and 2000 fiscal pacts led to costly negotiations during
the 2001 crisis, putting at risk the macroeconomic policy stance, which even-
tually collapsed in December 2001, as described by Webb et al. in Chapter 15.
These rigidities share some similarities with rules narrowly defined. They
can be understood in a framework developed by Tommasi and Spiller (2000),
which indicates that when the determinants of intertemporal cooperation
(summarized by a high discount factor) are favorable, optimal policies
obtain. These policies will not be subject to the opportunistic behavior of
the different players; instead, they will be flexible enough to accommodate
changing economic circumstances. If, on the other hand, intertemporal
cooperation is not feasible (given a low discount factor), then two things
might happen. Either the outcome will be subject to the ability of players to
shift at any particular opportunity (in a volatile economic environment), or
the players will impose rigid rules to prevent such opportunistic behavior
(given the prevalence of acute conflicts of interest among players). Either of
these regimes delivers lower welfare than obtained with a high discount
factor – as envisaged by our approach.

Determinants of cooperation

Suboptimal actions reflect noncooperative behavior by subnational and

national authorities. The deeper question, then, is what determines (beyond
the discount factor) the degree of cooperation in intergovernmental rela-
tions. More generally, cooperation depends on factors that affect the payoffs
in each stage of the game, and on those that affect the intertemporal prop-
erties of the game.10 Determinants of cooperation include both variables per-
taining to the fiscal system, and variables pertaining to the political system,
broadly defined. (The partition is arbitrary, since there are institutional fea-
tures that are both fiscal and political).
Before discussing specific fiscal and political variables, let us consider a
more abstract listing of variables affecting the likelihood of cooperation in
these types of games (Spiller and Tommasi 2001).

● Payoffs in each stage of the game: mostly related to characteristics of the

fiscal system.
● Number of political actors with power over a given decision: the larger the
number of players, the smaller the set of other parameters for which coop-
eration obtains.
● Length of the horizon / patience of key political actors: the likelihood of coop-
eration increases with patience.
● Intertemporal linkages among key political actors: the intertemporal pattern
of interactions among specific individuals in formal political positions
(legislators, governors, government officials) matters for generating coop-
erative outcomes.
192 Miguel Braun and Mariano Tommasi

● Characteristics of the arenas where key political actors undertake their

exchanges: the complex intertemporal exchanges required to implement
effective public policies can be facilitated by the existence of an institu-
tional framework.11
● Timing and observability of moves: cooperation is harder to sustain if there
is opportunity for unilateral moves difficult to observe or verify. This
relates to transparency as emphasized in the literature on fiscal rules,12
and more specifically to available bailout channels. In our application to
the Argentine case we refer to the national executive’s discretion in the
geographical allocation of spending.
● Availability of enforcement technologies: technologies that facilitate cooper-
ation include independent courts, professional administration, and/or
supranational organizations.13

Key variables of the fiscal system (as distinct from abstract variables), that affect
the likelihood of cooperation, mostly through their effect on stage-payoffs,
include the allocation of revenue sources and spending responsibilities
among subnational governments, differences in tax capacity and spending
needs of subnational units, degree of vertical fiscal imbalance, and features
of the intergovernmental transfer system. The vertical structure of the fiscal
system is a crucial determinant affecting the payoffs from cooperation.14 As
Eichengreen and von Hagen (1996, 2000) emphasize, the larger the vertical
fiscal imbalance (the smaller the capacity of the subnational government to
react by its own means in case of fiscal stress), the more likely is noncoop-
erative subnational fiscal behavior.
Key variables of the political system are those that impinge upon the qual-
ity of intergovernmental decision-making in fostering cooperation. Roughly
speaking, there are two interrelated arenas: one is intergovernmental rela-
tions per se (such as the Premier’s Conference in Australia) and the other is
the very structure of the federal government.

The case of Argentina

In Argentina, provinces have strong constitutional rights. Expenditure is

fairly decentralized and tax collection centralized. Although almost 65 per-
cent of provincial spending is financed out of a common pool of taxes, there
is wide variation, with 10 out of 23 provinces financing less than 20 percent
of their spending with their own revenues. The main channel for adjusting
to this vertical fiscal imbalance is the Federal Tax-Sharing Agreement, a
convoluted process involving endless bargaining and leading to consider-
able problems.
Tommasi et al. (2001) discuss deficiencies of fiscal federalism in Argentina
which have been identified in the literature: high subnational deficits,
increasing indebtedness of subnational governments, procyclical finances of
 A Game Theoretic Approach 193

subnational governments, bailouts, inefficient local taxation, poor tax collec-

tion, inefficiencies in the expenditure structure, and inadequate provision of
local public goods. These features, in turn, are the outcome of a noncooper-
ative federal fiscal game. Noncooperative fiscal behavior in Argentina can be
explained by mapping the abstract elements identified earlier into the
historical and present characteristics of federal and political institutions.
The environment for intergovernmental transactions has been less than
appropriate in Argentina.15 The allocation of policy jurisdictions has evolved
in a peculiar way throughout the twentieth century, while intergovernmental
agreements have lacked the necessary institutional support and enforcement.
Intergovernmental competition and conflict have prevailed over coopera-
tion in areas of concurrent jurisdiction. Given political and economic
instability since the 1930s (including numerous military “interruptions” and
high, volatile inflation), political actors have adopted a particularly myopic
perspective. National and subnational governments have not invested in
building intergovernmental institutions, and they have attempted to protect
their own interests without collaborating with one another.
This legacy has deeply affected the revenue-sharing arrangements.
Changes in the bargaining strength of political actors have been reflected in
shifts in the system of transfers. For instance, when power was concentrated
in the national government (mostly during military regimes) the distribu-
tion of taxes shifted toward retention at the national level. Conversely, with
the return to democracy, elected governors managed to secure a larger share
of transfers for the provinces. These modifications in favor of the subna-
tional governments were, in general, accompanied by an element of
interprovincial redistribution, subject to coalition politics at the time (Saiegh
and Tommasi 1999).
There have been mutually reinforcing interactions that arose from the
pattern of intergovernmental relations and from the overall capacity to
implement efficient intertemporal exchanges (Spiller and Tommasi 2001).
The combination of a lack of legislative incentives, disproportionate power
held by provincial leaders and scope for the national executive to act oppor-
tunistically, have deeply affected intergovernmental relations.
In Argentina, “executive federalism” (Watts 1999) has been prevalent: that
is, the president and the provincial governors play a dominant role in
intergovernmental relations. However, deals struck between national and
subnational executives can easily be altered at either the legislative or the
implementation stage. This is compounded by the fact that the governors’
influence in national policymaking enables them to secure occasional
legislative benefits at the expense of the national or other subnational units.
On the other hand, during the implementation stage, the president can eas-
ily unravel agreements reached or ratified in the national legislature.16 Thus,
these agreements are often reached outside the national legislature, and they
tend to incorporate very rigid rules in order to prevent their modification.
194 Miguel Braun and Mariano Tommasi

The broader approach that we suggest for institutional reforms to improve

fiscal performance of subnational (and national!) governments can be illus-
trated with the Argentine case. Despite the overall usefulness of the
approach, its detailed implementation is a complex matter, and many of the
details can be improved upon. It must be stressed that any recipe for address-
ing federal fiscal problems is likely to be country specific and even context
specific within a given country.
The evolution of underlying political exchanges requires consideration,
since it provides the foundation that reforms must build upon. In Argentina,
there have been successive rounds of fiscal pacts in 1992, 1993, 1994
(enshrined in constitutional amendment), 1999, and 2000, as well as con-
tinuous negotiations and modifications among pacts. The pacts have
attempted to address the underlying problems and main concerns of the
times, as well as the concerns of the different parties against the oppor-
tunism of others. The difficulty has been enforcement. A specific example of
failure has been the repeated attempt to replace the inefficient provincial
turnover sales tax. Replacing it would require actions leaving the complying
province vulnerable to future opportunism by the national government and
noncomplying provinces.
There is near unanimity among observers and political actors that
Argentina should reform its federal fiscal system. Major lending organiza-
tions and other specialists have produced reform proposals to correct
deficiencies we have described. Some of these proposals even include fiscal
rules narrowly defined. However, these “recipes for change” are not fol-
lowed. As we have argued, even if there are efficiency gains, transactional
problems (especially the enforceability of those transactions) make the
reforms difficult to accomplish.
Therefore, a feasible and sustainable reform strategy must focus on a
higher level of institutional reform, including the structure and process of
intergovernmental decision-making. In the spirit of the above intertempo-
ral cooperation framework, we should find a way of moving to the first best
policies (obtained with a high discount factor), rather than to the subopti-
mal opportunistic or rules-based outcomes. Conventional wisdom among
economists is in favor of rules – that is, moving from political opportunism
to rigid rules, taking as given the underlying political incentives. We argue
instead in favor of changing the political governance structure.
For Argentina, we suggest a broad strategy that should include the
following elements:

(1) Reform of the electoral mechanisms to lower the dependence of national

legislators on local party elites, which is a key aspect of the poor trans-
actions environment.
(2) Reform of the instruments of legislative interaction between the presi-
dent and congress to improve enforcement and lead to more efficient
agreements.
 A Game Theoretic Approach 195

(3) Reform of the budget process to curtail some executive discretion (as it
seems to have been instrumented in Mexico), limiting the ability to per-
form bailouts.
(4) Reform of intergovernmental relations.
(5) Reform of the tax-sharing agreement to improve the Wicksellian con-
nection between the taxes raised and the public goods consumed within
each jurisdiction.
(6) Macrofiscal rules to guide through the transition toward a more cooper-
ative (and more sustainable) fiscal stance.

Looking into the political feasibility of reforms, we can observe that (1) is
very difficult, so we can discard it for now. Equally, there was a window of
opportunity for (2) at the beginning of 1999, but it was lost. Reforms (3) to
(6) present a compact package, all of which could be undertaken under the
roof of the still unfulfilled 1994 constitutional mandate, which calls for a
new Coparticipation Law. It explicitly includes several conditions that act as
constraints to prevent opportunism and to enforce efficient reforms.
Specifically, the Constitution requires the creation of a Federal Fiscal
Institution (FFI; Órgano Fiscal Federal) to implement the law.
In spite of its imperfections, this mandate provides a good opportunity to
redefine the governance structure of federal fiscal relations. The Federal
Fiscal Institution could function as a formal arena in which to decide issues
of fiscal federalism, and thus provide intergovernmental relations with the
necessary flexibility to adjust automatically to changing circumstances. A
federal fiscal law is an incomplete contract reflecting political agreement at
the time of its creation. If the law were to specify a set of mechanisms to dis-
tribute taxes (i.e. a “rule”) it would still always require adjustments to chang-
ing circumstances (Saiegh and Tommasi 2000). If no explicit change were
made to the federal governance structure, the contract would be “com-
pleted” under the “default” extant governance structure. But, as we have
argued, the features of the extant governance structure earlier described are
precisely the underlying determinants of the many economic inefficiencies
observed today. That is why this higher-level institutional innovation would
help “complete” the contract with better ex post decision procedures.
Thus the FFI should provide the proper channel for intergovernmental
dealings, outside the national legislature, away from informal and
unenforceable executive–executive deals, especially removed from excessive
ex post discretion in the hands of the national executive. This would help
minimize the overlapping and contradictory patterns between national and
regional policies and facilitate decision-making in a foreseeable setting, and
allow for more effective intergovernmental relations and for policy adjust-
ments to changing circumstances.17
Reforms in the federal decision-making procedures, coupled with changes
in the transfer system that substantially increase the Wicksellian connection,
would constitute a profound redefinition of fiscal federalism in Argentina.
196 Miguel Braun and Mariano Tommasi

Provinces would assume greater individual responsibility and acquire more

collective power in federal decisions. Only in the context of this broader
reform scheme does it make sense to discuss the very important details of
the fiscal rules that are necessary.
Argentina’s recent experience with fiscal rules provides support for this
approach. Faced with a deteriorating budget balance and growing debt
payments, Congress approved the Fiscal Solvency Law in September 1999.
The law set a budget balance target by 2003 and a ceiling for the growth of
expenditures for the central government. However, contrary to the optimism
expressed by some observers, the rule has been broken every year.
At the subnational level, several governments followed the national exam-
ple and also passed fiscal solvency legislation. These statutes contain char-
acteristics favored by the recent literature, such as limits on government
debt and requirements regarding the timely and accurate publication of
information. However, compliance with these laws in terms of debt and
deficit performance has been uneven. Only five out of eleven provinces that
imposed a hard budget constraint actually fulfilled their commitment in
2000. Out of the five that complied with the law, two of these, Córdoba and
Tucumán, had been achieving the objective stated in the law for several
years, so the law appears more like a reflection of preexisting underlying
political agreements. With respect to limits on expenditure, three out of
eight provinces that imposed limits did not abide by them in 2000.
The bottom line is that Argentine provincial and central government’s
enacted fiscal rules that did not affect the underlying political equilibrium
(Case 2), and therefore were not effective to achieve fiscal discipline.

Conclusions

Work on this chapter began with the idea of identifying the circumstances
in which subnational fiscal rules might be called for, and analyzing the
details of the design of rules. However, a review of the theoretical founda-
tions of fiscal rules, as well as an examination of past experience, including
in Latin America, led us to a more skeptical view, which warrants a broader
investigation before getting into the practical details of design issues.
Our analysis suggests that policy actions are the outcome of a political
game in which relevant actors negotiate decisions over time. The quality of
the resulting policies for social welfare will depend on the extent to which
the political game facilitates efficient cooperative exchanges.
In order to improve fiscal performance in a sustainable and efficient man-
ner, one needs to focus on the determinants of fiscal behavior, including
those that lead to noncooperative outcomes. As illustrated by the case of
Argentina, the main deficiencies need to be identified, together with a com-
plete diagnostic of its fiscal and political determinants. On the basis of such
 A Game Theoretic Approach 197

diagnostics, one should attempt to correct institutional arrangements as

deeply as possible. In that context, fiscal rules can be a useful instrument to
accompany the transition path. However, an excessive focus merely on nar-
rowly defined rules might lead to a risky sense of security.

Notes

1. We received valuable comments from the discussant, Paul Boothe, and from Allan
Drazen and George Kopits, as well as helpful research assistance from Emmanuel
Abuelafia.
2. See Alesina and Perotti (1995) for a survey of political determinants of fiscal
problems.
3. Peltzman (1992) shows that voters in US states tend to reward fiscal prudence.
4. For brevity in this section we refer to a game across subnational governments; the
national actor and vertical considerations are included later in the chapter.
5. Kopits (2001a) describes US fiscal and monetary discipline in the 1990s as an
example of successful discretionary policymaking. This could be viewed as
corresponding with the game in Case 1.
6. For a classic survey, see Alesina and Perotti (1995), as well as Drazen in Chapter 2.
7. This is, of course, very common throughout Latin America, as discussed in Kopits
et al. (2000) for Argentina and Brazil; Nicolini and others (2000) for Argentina;
Dillinger (1997), and Bevilaqua (2002) for Brazil; Hernández-Trillo et al. (2000) for
Mexico; Echavarría et al. (2000) for Colombia; and Gamboa (1997) for a compar-
ison of Brazil, Mexico, and the United States.
8. The list of examples of inefficiencies induced by noncooperation in federal fiscal
games is quite long, including procyclical subnational fiscal behavior, excessive
public employment, inability to reform an inefficient tax system, and underin-
vestment in the capacity to raise taxes efficiently; see Sanguinetti (2001) and
Saiegh and Tommasi (1999, 2000).
9. This way of reasoning has been dubbed “transaction cost politics” by North (1990)
and Dixit (1996).
10. More properly, if we look at the overall intertemporal game expressed in normal
form, the possibility of cooperation depends on the elements of the description
of the game (payoffs of the stage of the game, timing of moves, information struc-
ture, etc.).
11. The landmark in the literature is the depiction of the US Congress by Weingast
and Marshall (1988).
12. See, for instance, Kopits and Craig (1998).
13. The potential role of supranational organizations (such as the European
Commission) as enforcers of agreements in Latin American countries requires
more detailed analysis.
14. For brevity in the text we refer to “cooperation” and “lack thereof” as alternatives,
even though actual fiscal games tend to be more continuous.
15. For a more detailed analysis, see Braun and Tommasi (2002).
16. This is achieved mainly via executive decrees and ample discretion in budget
implementation.
17. The details of the structure proposed for this FFI can be found in Iaryczower et al.
(2000).
13
Rules-Based Adjustment in
a Highly Decentralized Context:
The Case of India
Kalpana Kochhar and Catriona Purfield1

Introduction

India is on the brink of emerging as a major force in the global economy.

Chances are that any analysis of India today will inevitably revolve around
the rapid growth in per capita income, the significant reduction in poverty,
and the remarkable performance in exporting information technology and
services during the last decade. Although some restrictions remain on trade
and on both capital inflows and outflows, India has become considerably
more globally integrated than it was only a decade ago.
At the same time, India has one of the largest and most intractable fiscal
imbalances in the world. Although India has largely avoided the most visible
adverse macroeconomic effects to date, the fiscal imbalances are taking their
toll in terms of foregone growth by narrowing the room for macroeconomic
policy maneuver and the scope for further structural reforms and external
liberalization.
Although the negative effects of large deficits and accumulated debt of the
public sector are clearly recognized,2 fiscal consolidation presents an
especially thorny challenge in India’s highly decentralized federal system.
Because the large fiscal imbalances are evenly attributable to the central and
the state governments, adjustment has to take place at both levels of
government. However, several factors greatly complicate this task. These
include the structure of assigned spending and taxing powers, wide regional
disparities in income and social and physical infrastructure and regional bar-
riers to trade. Together with coalition governments that need the support of
regional and state-level political parties, these factors make fiscal adjustment
in India complicated.
Following several failed attempts at reducing imbalances in the past, the
Indian authorities are moving towards a rules-based approach to fiscal

198
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
 Adjustment in a Highly Decentralized Context 199

adjustment in the hope of overcoming political obstacles. The Fiscal

Responsibility and Budget Management Law (FRBM), enacted in August
2003 at the central level, along with parallel legislation in a few states, is a
milestone that should facilitate external liberalization and high growth in
India’s path to fiscal sustainability.
This chapter reviews India’s fiscal development, its macroeconomic con-
sequences, the challenges of fiscal adjustment in a highly decentralized fed-
eral system, and the strategy needed to achieve fiscal sustainability in this
context. It further examines the adequacy of the new fiscal responsibility
framework to meet this challenge.

Fiscal developments and macroeconomic consequences

Evolution of fiscal imbalances

Government finances in India have deteriorated progressively since the mid-
1990s and its fiscal imbalance now ranks among the worst in the world.
Since the end of the 1990s, the general government deficit has exceeded 10
percent of GDP and the primary deficit has averaged 4 percent of GDP. While
India’s primary deficit has traditionally exceeded those incurred in Turkey
and Argentina, recently this margin has increased substantially as both
countries have begun sizeable fiscal adjustment programs in the aftermath
of major macroeconomic crises. Persistent primary deficits and the narrow-
ing gap between real growth and interest rates have driven the accumulation
of general government debt (Figure 13.1).3 Consequently, the public debt

12 12

10 10

8 8

6 6

4 4

2 2

0 0

–2 –2
2

3
/9

/9

/9

/9

/9

/9

/9

/9

/0

/0

/0

/0
91

92

93

94

95

96

97

98

99

00

01

02
19

19

19

19

19

19

19

19

19

20

20

20

Figure 13.1 India: growth-interest rate differential (in percent), 1992–2003

Source: IMF staff estimates.
200 Kalpana Kochhar and Catriona Purfield

Table 13.1 India: central and state government finances (in percent of GDP),
1998–2002

1998 1999 2000 2001 2002

Est.

Overall balance
General government
8.4
9.7
9.8
10.4
10.2
Central government
5.3
5.5
5.7
6.2
6.1
State governments
3.1
4.1
4.2
4.2
4.1
Primary balance
General government
3.2
4.0
4.0
4.2
4.0
Central government
0.6
0.7
0.9
1.3
1.2
State governments
2.5
3.3
3.1
2.9
2.7
Debt
General government 66.9 69.1 71.4 75.8 80.6
Central government1 59.3 59.0 59.2 61.0 57.0
State governments, net2 7.6 10.2 12.1 14.8 23.6
State governments, gross3 19.2 21.5 22.9 25.3 34.1
Contingent liabilities
Guarantees 9.8 10.8 11.9 11.6 NA
Central government 4.4 4.3 4.2 4.2 NA
State governments 5.4 6.5 7.7 7.4 NA
State electricity board arrears 0.5 0.7 1.0 1.1 NA
Unfunded pension liabilities (NPV) NA NA NA NA 6.7

Notes
1
From 2000, state borrowing from the National Small Savings Funds is included as state govern-
ment liabilities.
2
Excludes state obligations to the central government.
3
Includes loans from the central government to states.
Sources: Ministry of Finance, Reserve Bank of India, and IMF staff estimates.

burden now exceeds 80 percent of GDP – with a rise of more than 10 per-
centage points recorded over the last five years – among the highest in the
developing world.4
Both the central government and state governments have contributed to
the deterioration. Although most of the deterioration has been attributable
to the central government, state imbalances have played an increasing role.
As in most federal systems, state governments in India have an important
role in implementing government policies.5 Even though there has been lit-
tle change in the responsibilities of states in the latter part of the 1990s, the
growth in the state deficit has outpaced that of the central government
(Table 13.1).
The deterioration in central and state government finances reflects a
combination of an eroding revenue base and mounting spending pressures.
General government revenue collections have fallen to slightly below
18 percent of GDP in 1998–2000. The trend decline in revenue mainly
reflects poor tax performance, in particular the low buoyancy of the tax
 Adjustment in a Highly Decentralized Context 201

system, which is narrowly based on indirect taxes and manufacturing activ-

ity, as well as the impact of trade liberalization on customs collections and
tax concessions. In addition, states have failed to offset declining central
government transfers by raising their own revenue.
Meanwhile, expenditures have risen sharply in large part because of the
inability of states to withdraw subsidies to farmers and households, to limit
water and food consumption, and to reduce the size of government employ-
ment (Lahiri 2000). The combination of mounting wage and pension costs
(Shome 2000), subsidy outlays, and interest payments has resulted in a
marked rise in current spending (IMF 2003). Both tiers of government have
sought to offset these pressures by compressing capital outlays, which now
stand barely above 3 percent of GDP, or one half of the level reached a
decade earlier.

Macroeconomic consequences
Judged by the sheer size and persistence of fiscal imbalances and indebted-
ness, arguably the fiscal situation is the single biggest threat to stability in
India. Yet, interestingly, the adverse macroeconomic developments expected
from such sizable imbalances have not materialized. There is little evidence
of inflation pressures and hardly any indication of conventional (price-
based) “crowding out”, and the external balance is strengthening. These
observations have prompted some to argue that the concern about fiscal
deficits in India is misplaced.
A critical reason why India has managed to avoid a disruptive crisis and
serious macroeconomic difficulties relates to the profile of public debt. First,
from the perspective of the recent large emerging-market crises (Argentina,
Brazil, Turkey), the proportion of government debt financed externally is
small and is largely on concessional terms, thus avoiding the dangers asso-
ciated with “original sin.”6 Second, not only is public debt mostly in domes-
tic currency, but also its maturity structure has been significantly lengthened
(Table 13.2). The authorities thus have managed to reduce the rollover risk
without the usual side effects of a rise in long-term rates. This has been pos-
sible due to the contribution of prolonged weakness in private investment
and thus credit demand since 1997 and the availability of ample liquidity as
the RBI attempted to encourage investment.
Nevertheless, the disquiet with India’s fiscal situation has some basis. First,
there is clear evidence that interest rates have come down markedly since
1997 (Figure 13.2) partly as a result of cyclical and structural factors. The
global disinflationary trends, India’s trade liberalization, and large food grain
stocks have helped dampen inflationary expectations. In addition, the sig-
nificant progress in deepening the government securities market has also
helped bring about a secular decline in interest rates. However, the fall in
interest rates on newly issued government debt has not been translated into
an equivalent decline in the nominal effective interest rate on government
202 Kalpana Kochhar and Catriona Purfield

Table 13.2 India: maturity structure of central government securities, 1997–2002

Year Weighted Range of Weighted average Weighted average

average yield maturity of maturity maturity of
(in percent) new issues (years) outstanding stock
(years) (years)

1997 12.0 3–10 6.6 6.5

1998 11.9 2–20 7.7 6.3
1999 11.8 5–19 12.6 7.1
2000 11.0 2–20 10.6 7.5
2001 9.4 7–30 14.3 8.2
2002 7.3 8–29 13.8 8.9

Source: Reserve Bank of India.

16 16

14 Ten-year govt. 14
bonds
12 12

10 91-day Treasury 10
bills
8 8

6 6

4 WPI inflation 4

2 2

0 0
1

3
/9

/9

/9

/9

/9

/9

/9

/9

/9

/0

/0

/0

/0
90

91

92

93

94

95

96

97

98

99

00

01

02
19

19

19

19

19

19

19

19

19

19

20

20

20

Figure 13.2 India: nominal interest and inflation rates (percent per annum),
1991–2003
Source: IMF staff estimates.

debt.7 Moreover, the real effective interest rates on government debt actually
doubled to nearly 7 percent in 2001–02. Consequently, India now is dan-
gerously close to an internal debt trap where new debt has to be incurred
just to service the previously contracted debt.
Another manifestation of “crowding out” is the slump in public capital
expenditure (Figure 13.3). As pointed out earlier, public infrastructure spend-
ing bore the brunt of efforts to contain the extent of fiscal deterioration.
Although public investment cuts do help narrow (or contain) the deficit in
 Adjustment in a Highly Decentralized Context 203

110 110

100 100

90 90

80 80

70 70

60 60

50 50

40 40
1990/91 1992/93 1994/95 1996/97 1998/99 2000/01

Figure 13.3 India: ratio of capital to current spending (1990/91
100), 1991–2000

Source: IMF staff estimates.

the near term, they can have serious deleterious effects on private invest-
ment, as suggested in several empirical studies (Ahluwalia 2002a,b; IMF
2002d; RBI 2002).
The relatively large pool of private sector saving, which at 26 percent of
GDP compares very favorably with rates in most high-growth Asian coun-
tries, has helped finance India’s fiscal imbalances. However, much of these
savings are accessible to the government through mandatory statutory liq-
uidity ratios and tax-preferred small savings schemes whose proceeds must
be invested in government securities. Indeed, more than 90 percent of
household financial saving is now being used to finance the gap between
public sector investment and saving, up from 65 to 70 percent in the mid-
1990s, leaving very few resources for the private corporate sector.
The large and compliant public sector banking system plays a key role in
preventing the spillover of large deficits into macroeconomic difficulties.
The government requires banks to hold 25 percent of their deposits in the
form of government securities. However, partly owing to low credit demand
from the private sector, banks have currently invested nearly 42 percent of
their deposits in government securities. More importantly, public sector
banks have increased their investments with duration of over three years –
such investments currently account for nearly 75 percent of these banks’
total investments in securities – while the bulk of their liabilities (nearly
80 percent) are of one- to three-year maturity. Thus, these banks face signif-
icant interest rate risk – an increase in interest rates would result in a signif-
icant erosion of profitability, and could weaken their capital position and
their ability to effectively intermediate credit.8
The fiscal situation is a constraint on India’s progress toward global
integration. Notwithstanding the significant reduction in tariffs and
204 Kalpana Kochhar and Catriona Purfield

Table 13.3 Selected regions: volatility of tax revenue, GDP, and terms of trade
(in percent), 1990–99

Tax revenue GDP growth Change in terms

growth of trade

India1 4.5 1.6 11.5

Latin America 11.8 4.6 10.7
Other developing countries 13.0 4.9 12.0
Industrial countries 3.6 2.0 4.0

1
Note: Estimate for 1990–2002.
Source: IMF staff estimates.

quantitative restrictions of the past decade, India’s trade regime is one of the
most restrictive in the world. However, further reduction and simplification
of tariffs has stalled because of the adverse implications of the revenue loss
from further cuts in customs duties. In addition, some critics believe that
one of the main impediments to investment (including foreign direct invest-
ment) is the acute infrastructure bottleneck, a direct consequence of the
large fiscal imbalances. Lahiri (2000) notes that the precarious fiscal position
is holding up the progress of banking reforms and is slowing down the
introduction of capital account convertibility.
The large public indebtedness allows little room for maneuver in the face
of shocks. With a relatively closed capital account and sizable pool of captive
domestic savings, India has so far been able to manage a relatively large debt
burden thanks to the low volatility of macroeconomic variables (Table 13.3).
However, as India becomes progressively more open, the experience of other
emerging market countries (explored in Chapter 3 by Hausmann) points to
a jump in macroeconomic volatility and increasing difficulty in managing
fiscal imbalances. A relatively mild example of such a situation is the chal-
lenge facing the RBI in coping with the surge in capital inflows of the past
12–18 months by allowing a modest appreciation of the rupee combined
with significant sterilized intervention. Over time, in the face of growing
and more volatile capital inflows, the scope for continued sterilized inter-
vention becomes narrower as the stock of government securities dwindles.
However, the RBI has had to rule out the issuance of its own bills, as has been
done in several emerging market economies to facilitate sterilization, given
the large deficits and borrowing needs of the government.
For all these reasons, in spite of the success in preserving macroeconomic
stability, a lax fiscal policy stance has exacted a cost in terms of growth and
poverty alleviation. India’s impressive economic growth of the mid-1990s –
fruits of the reforms of the early 1990s – has slowed significantly in recent
years. Following very weak performance in the immediate aftermath of the
earlier balance of payments crisis, growth recovered to in excess of 7 percent
 Adjustment in a Highly Decentralized Context 205

35 12.0

30 10.0
25
8.0
20
6.0
15
4.0
10

5 2.0

0 0.0
1990/91 1992/93 1994/95 1996/97 1998/99 2000/01
Public sector investment (left scale)
Private sector investment (left scale)
Real GDP growth (right scale)
General government deficit (right scale)

Figure 13.4 India: government deficit, investment, and growth (in per cent of GDP),
1991–2002
Source: IMF staff calculations.

in the subsequent four years. Real fixed investment rose by nearly 40 per-
cent between 1993 and 1995, led by more than 50 percent growth in
industrial investment.9
Beginning in 1997, the economy entered a prolonged growth slowdown.
To be sure, the slowdown reflected cyclical and exogenous factors, but the
persistent weakness suggests that the large fiscal imbalances (Figure 13.4)
and the loss of momentum on structural reforms held back economic activ-
ity. The deterioration of state finances is of particular concern in this regard
(Srinivasan 2001). High interest rates in the mid-1990s and the progressive
compression of capital spending also resulted in lackluster private invest-
ment activity. As a result, India has lost ground relative to other emerging
market countries, in terms of per capita growth and poverty alleviation.

A new institutional approach to adjustment

Cognizant of the need to restore fiscal sustainability, given the factors earlier
described, India has enacted the FRBM, following three years of debate and
consensus building. The new fiscal responsibility legislation (FRL) estab-
lishes a broad framework for the conduct of fiscal policy by setting a
medium-term target to guide fiscal policy formulation. The framework
places increased emphasis on transparency in budget formulation, imple-
mentation, and assessment. As such, it merges aspects of the various
206 Kalpana Kochhar and Catriona Purfield

rules-based approaches of advanced countries such as Australia, New Zealand,

the United Kingdom, and the European Union.
India’s new fiscal framework possesses some features that set it apart from
similar frameworks in other emerging market economies. Most important,
although the aggregate state deficit is almost half of the general government
deficit, the new framework is valid for only the central government. The
FRBM requires the central government to eliminate the “revenue deficit”
(broadly equivalent to the current deficit), of 4.25 percent of GDP, by March
2008.10 Fiscal policy rules tend to be most successful when they fix targets
that are appropriate to each level of government. The exclusion of
subnational governments from India’s framework could complicate the
achievement of fiscal sustainability (Kopits 2001b). Other country experi-
ences, cast in a game theoretic context – outlined in Chapter 12 by Braun
and Tommasi – suggest that subnational governments may not voluntarily
follow the example of the national government, as illustrated by the case of
Argentina. To date, only 4 of the 28 states in India have enacted FRLs
(Box 13.1). A bottom-up approach to fiscal adjustment may prove difficult
to coordinate, especially in a federal system as decentralized as India’s, if

Box 13.1 India: State-level fiscal responsibility legislation

Recently, various states have introduced FRL to help strengthen their finances.
As of end-2003, Karnataka, Kerala, Punjab, and Tamil Nadu enacted FRL and a
similar bill is before the state parliament of Maharashtra. All of these states have
adopted a comprehensive framework, which combines explicit fiscal targets
with requirements for greater transparency in budget formulation implemen-
tation.
The fiscal targets set under these laws vary by state and are often more
stringent than those established under the FRBM. For example, in Karnataka,
the legislation requires both the elimination of the current deficit and a reduc-
tion in the overall deficit to 3 percent of Gross State Domestic Product (GSDP)
by March 2006. It sets annual interim targets to guide the convergence to these
end-goals and stipulates that the state’s debt burden should not exceed 25 per-
cent of GSDP by March 2015. By contrast, in Punjab, the FRL introduces ceil-
ings on the growth of the overall and current deficits and on official debt and
guarantees. Some states also limit borrowing to investment financing.
To strengthen the institutions, state FRLs provide for the establishment of a
medium-term budget framework. Multiple-year budgeting was first introduced
in Punjab in 2002. A medium-term fiscal reform program sets out the macro-
economic assumptions underpinning the budget forecasts as well as an evalu-
ation of the fiscal outturn. In Karnataka, the documents accompanying the
budget also require an assessment of fiscal sustainability. To help enforce
compliance, the state FRLs provide for mid-year reviews of budget implemen-
tation and for the introduction of remedial measures in the event of slippage.
 Adjustment in a Highly Decentralized Context 207

institutional changes that encourage states to adopt a cooperative behavior

do not accompany the FRBM.
Indeed, recent efforts by the central government to induce fiscal consoli-
dation at the state level have so far generated little improvement. The
Medium-Term Fiscal Reforms Facility (MTFRF), established in 2000 to offer
financial incentives for states to reform, has so far failed to meet its targets.
Under the program, 12 states signed memoranda of understanding with the
central government and 21 states formulated medium-term fiscal reform
plans to eliminate the combined state current deficit by 2006. Three years
into the program, the states had secured a 9 percent reduction in the current
deficit-revenue ratio compared to a 15 percent target. While structural prob-
lems have undoubtedly hampered consolidation efforts, probably the MTFRF,
which only provided 0.5 percent of GDP to share among all the states over
the five-year period, was insufficient financial incentive for reform.
By setting the current balance as the medium-term target, the FRBM effec-
tively subjects fiscal policy to a golden rule. The central government can run
a deficit to finance investment. Thus the new rule has the advantage of safe-
guarding capital expenditure from bearing the brunt of the adjustment effort
as it occurred in the past.
The procedural arrangements governing convergence to the FRBM target
and measuring and assessing compliance are not yet in place. To ensure the
quality and durability of the adjustment, FRLs (or their supporting regula-
tions) generally contain explicit accounting procedures. In India, account-
ing and definitional procedures underpinning the FRBM, as well as the
annual path for convergence to the end-March 2008 target, have been dele-
gated to supporting rules, to be announced in the context of the 2004/05
budget. Instead of establishing an independent agency to collect data and
assess compliance with the law, the controller general of accounts will con-
tinue to compile data on budget implementation using existing accounting
definitions, while a special unit with the ministry of finance will analyze and
prepare reports on fiscal performance and compliance. To ensure higher
transparency in budget formulation and implementation, the FRBM imposes
increasingly stringent reporting standards. The executive must present to
parliament three annual reports covering the medium-term fiscal and debt
outlook, the macroeconomic assumptions underpinning the budget, and
the key fiscal policy measures to be taken over the coming year, as well as
quarterly reports on budget implementation.
Finally, enforcement of the FRBM will rely on the loss of reputation that
the government experiences from not implementing the framework. There
are no financial or judicial penalties for breaching the current balance target –
the law requires only that the government report to parliament the reasons
for the overrun. Breaches of the ultimate medium-term target, or of the
annual targets set under the supporting rules, are permitted for reasons of
natural disaster, national security, or other exceptional circumstances
208 Kalpana Kochhar and Catriona Purfield

specified by parliament. The minister of finance is only required to report to

parliament on the extenuating circumstances after missing the targets.
However, the new law requires parliament to support corrective action
through tax or expenditure measures during the budget implementation
process to avert or correct deviations from the fiscal target.

The challenge of achieving fiscal sustainability

The key objective of the FRBM is to achieve a prudent level of debt that is
consistent with fiscal sustainability. We undertake a quantitative assessment
to determine whether the FRBM framework is adequate to achieve this goal.
However, the framework does not explicitly define what level of debt is sus-
tainable. We adopt a simple definition whereby fiscal policy is considered
sustainable if it is consistent with halting the rise in the general government
debt burden and placing India’s debt dynamics on a downward path.

Simulations
Whether the FRBM secures a reduction in the debt depends on the imple-
mentation of the new framework in a multi-tiered federal system. As a first
approximation, the path of general government finances can be simulated
using specific macroeconomic assumptions that mirror India’s performance
over the past decade (Table 13.4). The simulations assume that the central
government adjusts its policies to meet the FRBM target (Table 13.5).11
However, this requires greater adjustment than that suggested by the present
level of the current deficit because the proportion of central government rev-
enue shared with states is netted out of the central government coverage.
The exact magnitude of adjustment depends on the share of revenue-raising
measures in the overall adjustment effort. Revenue measures are assumed to
generate about 60 percent of the adjustment.12 This assumption implies that
a total adjustment of 4.8 percentage points – comprising an increase in gross
revenue of 2.7 percent of GDP and a reduction in expenditure of 2.1 percent
of GDP – will be required to meet the current deficit target. The assumption
is that the government will channel half of the revenue savings into capital
expenditure so that the overall deficit adjusts by 2.1 percent of GDP.13

Table 13.4 India: macroeconomic assumptions (in percent), 2004–08

1993–2003 1999–2003 2004–08

Real GDP annual growth rate 6.0 5.4 6.1

Nominal effective interest rate on
general government debt 8.7 9.0 8.7
Real effective interest rate on
general government debt 2.7 5.2 4.5

Source: IMF staff estimates.

 Adjustment in a Highly Decentralized Context 209

Table 13.5 India: central government finances under the FRBM1 (in percent of GDP),
2003–08

2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 Cumulative

change
2003/04–
2007/08

1 Net Revenue (2–3) 9.0 8.8 9.0 9.4 10.1 11.0 2.2
2 Gross revenue 11.2 11.1 11.4 11.9 12.7 13.8 2.7
3 Tax share to states 2.3 2.3 2.3 2.4 2.6 2.8 0.5
4 Expenditure (56) 15.0 14.8 14.4 14.6 14.6 14.9 0.1
5 Current expenditure 13.3 13.0 12.5 11.7 11.1 11.0
2.1
6 Capital expenditure 1.7 1.8 1.9 2.9 3.5 4.0 2.2
7 Overall balance (1–4)
6.0
6.0
5.4
5.2
4.6
4.0 2.1
8 Current balance (1–5)
4.3
4.3
3.5
2.3
1.0 0.0 4.3
9 Gross change (2–5)
to meet FRBM target — 0.1 0.8 1.3 1.5 1.2 4.8

1
Note: Excludes commercial departments: net of transfers to states.
Source: IMF staff calculations.

The success of the adjustment in stabilizing and reducing the general

government debt burden depends on the crucial willingness of state gov-
ernments to contribute to the adjustment effort. To illustrate this point, we
construct two medium-term scenarios (Table 13.6 and Figure 13.5). In the
first, states would not change their fiscal behavior but continue to sustain a
1
primary deficit in the order of 2 2 percent of GDP, which given the already
sizeable state debt burden implies a rising interest burden and state deficit.14
The adjustment in central government finances would prove insufficient to
offset the widening state level deficit, and the general government debt
1
burden would rise to 882 percent of GDP by end-March 2008. In the second
scenario, all state governments follow the example of the central govern-
ment and adjust policies to eliminate their current deficit by March 2008,
channeling half of the savings into capital spending. By end-March 2008,
the combined state deficit would be near 3 percent of GDP; the rising trend
in the general government debt burden would reverse by 2005/06 and fall
to about 82 percent of GDP by 2008.
The success of the FRBM will also depend on the evolution of growth and
interest rates over the coming four years. The simulations conducted so far
assume a positive differential between growth and interest rates. As a result,
the debt-stabilizing primary deficit is about 1 percent of GDP.15 Were the dif-
ferential between nominal growth and interest rates to disappear, as hap-
pened between 1999 and 2002, the primary deficit would fully devolve to
additional debt. If the central and state governments eliminated their cur-
rent deficit as earlier discussed, the increase in interest costs (ceteris paribus)
 210
Table 13.6 India: general government finances under the FRBM (in percent of GDP), 2003–08

2003/04 2007/08 Cumulative adjustment

Base year No state State No state State Difference

adjustment adjustment adjustment adjustment

Central government overall balance
6.0
4.0
4.0 2.1 2.1 0.0
State government overall balance
4.0
5.2
2.7
1.2 1.3
2.5
General government overall balance
10.0
9.1
6.6 0.9 3.4
2.5
General government debt 83.0 88.4 82.3 5.4
0.7
6.1

Source: IMF staff calculations.

 211

7 7 6 6
Central government deficit State government deficit
6 6
5 5
5 5
4 4
4 4

3 3 3 3

2 2
2 2
1 1
1 1
0 0

–1 –1 0 0
3

8
/0

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0
02

03

04

05

06

07

02

03

04

05

06

07
20

20

20

20

20

20

20

20

20

20

20

20
Current deficit With adjustment
Overall deficit Without adjustment

12 12 90 90
General government deficit General government debt

10 10 88 88

86 86
8 8

84 84
6 6
82 82

4 4
80 80

2 2
78 78

0 0 76 76
3

8
/0

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0
02

03

04

05

06

07

02

03

04

05

06

07
20

20

20

20

20

20

20

20

20

20

20

20

Overall deficit with state adjustment

Overall deficit without state adjustment With state adjustment
Primary deficit with state adjustment Without state adjustment
Primary deficit without state adjustment

Figure 13.5 India: government debt dynamics under the FRBM (in percent of GDP),
2003–08
Source: IMF staff estimates.
212 Kalpana Kochhar and Catriona Purfield

Table 13.7 India: effect of elimination of growth-interest rate differential on govern-

ment debt under the FRBM (in percent of GDP), 2003–08

2003/04 2007/08

Base No state State

year adjustment adjustment

Central government current balance
4.3
2.0
2.0

Central government overall balance
6.0
6.0
6.0
State government overall balance
4.0
5.8
3.0
General government overall balance
10.0
11.8
8.9
General government debt 83.0 96.1 89.5

Source: IMF staff calculations.

would cause the central government to overshoot the FRBM target and the
overall debt burden would reach almost 90 percent of GDP by end-March
2008 (Table 13.7).
These simulations highlight three important challenges facing FRBM
implementation in this highly decentralized setting. First, in a federal sys-
tem such as India’s, the ultimate goal of fiscal sustainability hinges critically
on the behavior of states. The simulations highlight how the absence of a
commensurate adjustment by states can jeopardize fiscal sustainability even
when the central government achieves the FRBM target. Given the substan-
tial role states play in implementing government expenditure, coordinated
adjustment with the states is needed to safeguard against moral hazard on
the part of the states, which spills over onto the overall risk premium. The
challenge in implementing the FRBM will be to find a way to bring the states
into the adjustment effort.
Second, not only is the envisaged adjustment large by international
standards, but various pressures may also complicate debt stabilization. For
1
example, guarantees have the potential of adding 112; percent of GDP to
16
actual debt ratio. The net present value of unfunded civil service pension
liabilities is estimated at nearly 612; percent of GDP (India, Ministry of
Finance 2001), and the transition to a funded contributory civil servant pen-
sion scheme will initially entail additional outlays. Public investment also
needs to rise by 4 percent of GDP yearly to meet the targeted reduction in
poverty and increase in literacy ratios. Meanwhile, any additional steps to
liberalize trade tariffs will compound the adjustment effort, given the loss of
customs revenue, while financial and capital account liberalization will
make tax bases more mobile and elusive.
Third, the simulations highlight how susceptible debt sustainability is to
the evolution of growth and interest rates. The reduction in government
deficits, financing needs, and debt burden should help create the conditions
for expansion of private sector credit growth, private investment, and
 Adjustment in a Highly Decentralized Context 213

growth. However, progress in implementing various structural reforms,

including in the areas of taxation, trade, labor markets, and agriculture will
also be necessary to sustain high growth rates. Successful implementation of
the FRBM could influence market expectations about the future course of fis-
cal policy and therefore have a dampening effect on interest rates.
Experience elsewhere shows that interest rates are initially slow to fall as
markets first wait for firm evidence regarding the durability of the switch in
fiscal policy. However, once markets are convinced, the fiscal contraction
has been associated with declining interest premia and expanding economic
activity in many cases.17 The challenge will be to convince markets that the
FRBM represents a decisive shift in fiscal regime so as to realize gains in terms
of lower real interest rate premia and to help set off a virtuous cycle of higher
growth and improved debt dynamics.

Implications
Although the passage of the FRBM is an important step in the right direc-
tion, it may not be sufficient to address these challenges. The challenge of
orchestrating an adjustment of the magnitude required for fiscal sustain-
ability and, by corollary, convincing markets of the commitment to the
reforms, will move forward once the government announces the procedural
rules underpinning the framework. However, the crucial issue of coordinat-
ing the adjustment effort with states remains outside the FRBM. Indeed, the
central government will have to induce states to undertake a complemen-
tary adjustment. This will likely require a two-pronged approach: on one
hand, the central government should offer incentives to states to adopt
complementary policies, and on the other, it should harden state budget
constraints by tightening their access to deficit financing.
To help induce compliance with the FRBM, the central government could
explore ways to make state assistance conditional on fiscal adjustment. Here
the lack of success in securing a significant adjustment under the existing
MTFRF suggests that a voluntary approach may not work. The central gov-
ernment may need to expand the size and scope of the existing facility by
offering enhanced financial incentives to assist states in their retrenchment
efforts. Alternatively, in view of the sizeable and often unsustainable debt
burden in many states,18 the government could follow the example of Brazil
in the late 1990s and offer a debt-restructuring package that is conditional
on the realization of specific state-level fiscal retrenchment goals.19 In addi-
tion, the transparency requirements of the FRBM require replication at the
subnational level to increase accountability by disseminating sufficient
information to allow markets and the central government to evaluate state
performance effectively.
The central government will also have to take measures to coerce more
sustainable fiscal behavior on the part of states. In particular, it will need to
reform the system of fiscal relations so that states have stronger incentives
214 Kalpana Kochhar and Catriona Purfield

to control expenditure and raise revenue while facing hard budget

constraints. Creating a closer link between expenditure and revenue-raising
decisions can improve incentives. Steps to reduce state dependence on
shared revenue and central government transfers will encourage states to
adopt measures to expand their revenue base to meet expenditure needs.
The introduction of a uniform subnational VAT can play a critical role in
expanding the states’ revenue base by enabling a smoother process of bring-
ing services into the tax net and by removing barriers to interstate trade.
Likewise, the integration of plan and nonplan budgets and increased state
autonomy over the allocation of these resources will allow greater efficiency
in expenditure. However, these measures will only be successful in securing
prudent state behavior if the central government takes steps to control the
states’ access to budgetary financing. Tighter controls on state borrowing
from national small savings schemes, plan and nonplan loans, and market
borrowing need to supplement recent measures to restrict the use of off-
budget guarantees. State financing should be determined for the state sector
as a whole in a manner that is consistent with overall fiscal sustainability.
The challenge of achieving the magnitude of adjustment necessary to sta-
bilize and reduce India’s debt burden requires a plan for fiscal and structural
reform. Supporting policies to guide the convergence to the current balance
target and to ensure the quality, durability, and credibility of the framework
must accompany the FRBM. Since India’s tax revenue – GDP ratio is low by
international standards (IMF 2002d), measures to broaden the tax base and
increase tax collections – such as those outlined in Report of the Task Force
on Direct and Indirect Taxes, also known as the Kelkar Task Force (India,
Ministry of Finance and Company Affairs 2002a, b) – will play an integral
role in any convergence plan. The formulation and legislative approval of a
convergence plan should facilitate greater political awareness and support
for the much-needed structural reform. The medium-term rolling budget
framework covering both the macroeconomic outlook and ministry budgets
will also help reorient the budget process toward medium-term fiscal objec-
tives, inject realism into the budget estimates, evaluate the impact of pro-
posed measures, and identify potential expenditure pressures.
The challenge of convincing markets and the private sector that the FRBM
represents a permanent shift in the conduct of fiscal policy will lie in
strengthening budget institutions and transparency. For a reputation-based
approach to work, greater transparency – including in the formulation of pol-
icy and in budget implementation – will be key to ensuring that the new rules
become self-enforcing. While the new quarterly implementation reports pre-
pared by the ministry of finance improve the dissemination of fiscal data,
markets would welcome confirmation of compliance with the numerical
targets and accounting standards by an independent audit authority.
The adoption of explicit definitions for the core target in the new rules
will also assure markets of the quality of the adjustment effort. In particular,
 Adjustment in a Highly Decentralized Context 215

an exact accounting definition of capital expenditure, including how the

government plans to classify state debt swaps and public–private partner-
ships, would help assure markets that there will be limited opportunities for
creative accounting by redefining current spending as public investment.
Further clarification of the criteria that determine whether the government
can use the escape clauses for noncompliance would also increase trans-
parency. Finally, actual implementation of corrective action to avert poten-
tial deviations will underscore the credibility of the government’s
commitment to fiscal retrenchment.

Conclusions

Although India has largely avoided the disruptive macroeconomic conse-

quences of large fiscal deficits, there is no room for complacency. The deficits
and accumulated debt are taking a significant toll on the economy in terms
of foregone growth and poverty reduction. Recently some observers have
been highlighting India’s highly favorable demographic trend. Because the
working age population has a relatively high propensity to save, they note,
India should benefit from highly favorable saving trends that will help fuel
rapid growth. This argument rests on the critical assumption that the econ-
omy will be able to create the conditions necessary for the gainful employ-
ment of a large number of people joining the workforce. Unfortunately, the
track record for employment creation in the past decade has not been stel-
lar. Notwithstanding the favorable demographic trend, India’s medium-term
economic prospects depend critically on progress with the closely inter-
twined tasks of fiscal consolidation and structural reform. Together, such
actions would strengthen the public finances, improve the efficiency of
resource allocation, and generate a virtuous economic cycle to increase
growth.
Fiscal policy formulation in India is entering uncharted territory. To
restore fiscal sustainability, the government will have to orchestrate a daunt-
ing fiscal adjustment in a highly decentralized setting. To do this, it has cho-
sen to introduce a rules-based framework for fiscal policy. Partly reflecting
the constitutional arrangements in India, the coverage of this new frame-
work is limited to the central government. State finances are excluded even
though they have been a major force in driving the deterioration in general
government finances in the latter part of the 1990s.
Other highly decentralized emerging market economies that have
attempted to implement large fiscal adjustments have encountered major
difficulties in coordinating the adjustment effort with subnational tiers of
government. With the exception of Brazil, where the federal government
adopted a top-down approach to setting targets and administrative controls
on subnational governments under a uniform set of rules, these attempts
have met with failure. India in some sense will be a test case as to whether
216 Kalpana Kochhar and Catriona Purfield

a bottom-up approach to consolidation can succeed. This chapter under-

scores the point that the sheer size of the task ahead requires a correspond-
ing effort on the part of all state governments.
In line with Chapter 12 by Braun and Tommasi, we highlight the need for
institutional changes in the federal system if a voluntary cooperative
approach to adjustment is to work. State governments are responsible for a
wide array of expenditure responsibilities, and revenue sharing and central
government transfers comprise a major share of state revenue. Ultimately,
whether or not the FRBM succeeds in securing durable fiscal sustainability
boils down to the need to address the aggregate imbalance between expen-
diture needs and revenue capacity. If the central and state governments are
to share the cost of eliminating this imbalance, a comprehensive overhaul
of the system of federal relations must create incentives for states to behave
more responsibly.
As India becomes more globally integrated, a comprehensive review of the
federal arrangements and interstate relations will become necessary. Recent
research has focused on the question of whether all federal structures are
equally capable of sustaining a market economy (Weingast 1995), arguing
that only “market preserving federalism” can work in a competitive econ-
omy with an open capital market. Market-preserving federalism requires a
federal government that ensures that goods and factors are mobile across
subnational jurisdictions, and a stable policy environment for markets to
function and investment to take place. In this case, there would be limited
revenue sharing among levels of government and all levels of government
would face limited access to capital markets. With the federal government’s
retaining monopoly over money creation and responsibility for price stabil-
ity, the constraints on access to financial markets imply that the ability of
subnational governments to borrow is limited. Taken together, these would
ensure that all levels of government effectively face a hard budget constraint.
In India, key changes in federal arrangements, as the process of global inte-
gration continues, include removal of all remaining restrictions on the free
movement of goods, services, and factors of production across state lines,
and a reconsideration of the intergovernmental transfer mechanism.

Notes

1. The views expressed are those of the authors and do not necessarily reflect those
of the International Monetary Fund. Montek Ahluwalia, Vijay Kelkar, and George
Kopits provided useful comments on an earlier draft.
2. The Prime Minister’s Economic Advisory Council wrote in February 2001: “The
fiscal situation is undoubtedly one of the most serious weaknesses in the economy
at present and corrective action in this area is urgently needed.”
3. For a discussion of the relative impact of growth and interest rates on central
government debt, see Rangarajan and Srivastava (2003).
 Adjustment in a Highly Decentralized Context 217

4. Various observers have argued that India has sizable assets, likely to be in the
range of 25–30 percent of GDP, which reduce India’s net debt burden. However,
the official debt statistics understate the true extent of the government’s debt
obligations, which have been compounded by a proliferation of off-budget
claims, including guarantees, public enterprise arrears, and unfunded pension
liabilities. Thus, even taking into account public assets, a proper accounting of
contingent liabilities would yield an estimated net debt in a similar range of
gross debt.
5. There are three tiers of government: the central government, an intermediate tier
of 28 states and 7 union territories (5 are governed by central government
appointees), and constitutionally mandated local bodies. The assignment of
expenditure responsibilities and of tax bases among the different tiers is broadly
comparable to that of other federal systems. Two broad mechanisms govern the
sharing of resources between the central government and the states. A Finance
Commission (FC) is appointed every five years to recommend how taxes are to be
shared between the center and the states and among the states. A combination of
central government grants and borrowing fill in the vertical imbalances that
remain after revenue sharing. Two agencies split responsibility for grant alloca-
tions. The main “plan” grant is for implementation of state-level development
plans approved by the Planning Commission (PC). The PC also provides specific
earmarked grants for central government–sponsored schemes. The FC recom-
mends grants-in-aid to help fill residual gaps in the nonplan budget. The
Constitution permits domestic borrowing, which is subject to central approval if
a state has outstanding obligations to the central government.
6. Eichengreen et al. (2003b) show that “original sin” (borrowing in foreign currency
to finance the government) is the cause and not the consequence of volatility in
output and capital flows.
7. Defined by dividing actual interest payments in a given year by the outstanding
liabilities at the end of the previous year. See Rangarajan and Srivastava (2003).
8. Patnaik and Shah (2002) simulated a large positive interest rate shock over a year
for a sample of 42 major domestic banks in India. Twenty-five banks had “reverse”
exposures (i.e. they could lose between 25 percent and 105 percent of their equity
capital); only nine banks were adequately hedged.
9. Acharya (2001) characterized this as a “manifestly a boom time for the Indian
economy.”
10. A parliamentary committee replaced the annual targets on the overall deficit and
debt–GDP ratio contained in the original FRBM bill with a single medium-term
target for the current deficit. Some commentators have argued that the removal
of the more specific targets significantly weakened the FRBM. Others, however,
point to the experience of some countries in Latin America that have encountered
difficulties in implementing FRLs with precise yearly numerical targets.
11. Under a scenario where there is no adjustment by either level of government, debt
could exceed 100 percent of GDP by 2008/09.
12. Following current practice, states receive one-fifth of total gross central govern-
ment revenue collections.
13. The simulations do not calculate the dynamic impact of increased investment on
growth. The incremental capital output ratio is close to 4 percent, implying that
if the current effective tax burden remains unchanged, the increase in govern-
ment revenue collections should be sufficient to service the extra debt given the
current effective interest rates.
218 Kalpana Kochhar and Catriona Purfield

14. This implies that the higher level of shared taxes received from the central gov-
ernment as a result of its effort to increase revenue to meet the FRBM target would
be translated into additional expenditure.
15. Calculated as PBt  Dt (i
g


g
)/(1  i 
 g
), where PB is the debt sta-
bilizing primary balance in period t, Dt is the stock of general government debt,
both in percent of GDP, i is the nominal effective interest rate on general gov-
ernment debt, g is the real GDP growth rate and
is inflation measured by the
GDP deflator.
16. About 7.5 percent of GDP represents state government guarantees, which are
mainly for infrastructure projects. Given their social characteristics, they are likely
to devolve to the state governments.
17. See Alesina and Ardagna (1998) for examples of expansionary fiscal contractions,
and Hemming et al. (2002) for a review of the effectiveness of fiscal policy in
stimulating economic activity.
18. See Prasad et al. (2003) and Verma and Singh (1999).
19. See Bevilaqua (2002) for further details.
14
Fiscal Rules for Subnational
Governments:
Lessons from the EMU
Fabrizio Balassone, Daniele Franco, and Stefania Zotteri1

Introduction

The fiscal rules adopted in the context of Europe’s Economic and Monetary
Union (EMU) have extensive implications for European Union (EU) govern-
ments at all levels. This chapter focuses on the impact of EMU fiscal rules on
the relationship between national and subnational governments, with
particular reference to five member countries. Ultimately, the objective is to
derive lessons, especially for emerging market countries.2
To begin with, three critical areas can be identified. First, as EMU rules
apply to the general government, compliance depends on the behavior of
all levels of government, yet the central government alone is held account-
able. This asymmetry weakens the central government position vis-à-vis
subnational governments concerning the responsibility for compliance, and
therefore increases the need for rules that apply explicitly to lower government
tiers within each country.3
Second, EMU rules demand that the overall budget (both current and
capital) balance over the medium term. However, providing an adequate
level of public infrastructure at the subnational level may be difficult with-
out subnational authorities’ recourse to deficit financing.
Third, in order to reconcile fiscal soundness and flexibility in bad times,
the implementation of EMU rules implicitly relies on the cyclically adjusted
balance. At the subnational level, the latter may be difficult to measure;
moreover, other devices designed to allow for flexibility may prove incon-
sistent with the EMU fiscal framework.
This chapter examines some solutions for making fiscal decentralization
compatible with EMU fiscal rules and compares these possible solutions
with the approaches of five EU member countries. These countries are
diverse in terms of institutional tradition, as well as size, population, and
economic development. Three countries, Austria, Belgium, and Germany

219
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
220 Fabrizio Balassone et al.

have a federal institutional structure – in Belgium, this is a relatively recent

development. Italy and Spain have substantially increased decentralization
in recent years.4

EMU rules and fiscal federalism

Fiscal rules summarized

EMU fiscal rules are designed to ensure that national policies maintain a
sound fiscal stance while allowing sufficient margins for flexibility in bad
times.5 On the one hand, fiscal sustainability is a central tenet of the EMU:
it is a precondition for financial and monetary stability. On the other, flex-
ibility is needed for stabilization policy, and has become more important
with the establishment of the EMU, as member states can no longer rely on
either a monetary policy or exchange rate adjustment tailored to national
needs.
The Treaty of Maastricht states that budget deficits cannot be larger than
3 percent of GDP unless they arise under exceptional circumstances, such as
deep recessions; they remain close to the limit; and the excess deficit lasts
only for a limited period of time.6 If the deficit exceeds the limit and these
conditions are not met, the deficit is deemed “excessive” and it triggers a
procedure for the adoption of corrective measures.
The Stability and Growth Pact (SGP) specifies what is meant by “excep-
tional” and “limited period.” A recession is considered exceptional if real
GDP contracts by 2 percent, though a milder recession may also be consid-
ered exceptional if, for example, it happens abruptly. The excess above the
prescribed deficit limit must be reabsorbed as soon as the exceptional
circumstances have expired.
The pact further specifies that each country should aim for a medium-term
objective of a budgetary position “close to balance or in surplus.” According
to the EU Council of Economy and Finance Ministers (ECOFIN), compliance
with the pact should be assessed by taking into account the cyclical position
of the economy. In practice, the SGP requires that each member state target
a cyclically adjusted balance and let automatic stabilizers or discretionary
action operate symmetrically around it. The lower this budget balance with
respect to the deficit threshold, the more leeway for countercyclical policy
without the risk of an excessive deficit. Past experience suggests that in the
majority of EU countries a cyclically adjusted deficit between 0 and 1 percent
of GDP should be acceptable (Buti et al. 1997).7 Compliance with the deficit
threshold, as well as the 60 percent of GDP limit on the debt–GDP ratio,
would prevent the public finances from becoming unsustainable. Countries
with a debt ratio above the limit should also reduce it at a satisfactory pace
towards the threshold. An increase in the debt ratio during recessions should
be avoided.8
 Lessons on Subnational Rules from EMU 221

Each EU member state must submit its multiyear budgetary plan in a

stability or convergence program. These documents are updated annually
and reviewed by the European Commission to assess their consistency with
the rules (Cabral 2001). There is a midyear examination of public finances
and an ex post evaluation of results, as compared to the planned targets.
ECOFIN can make recommendations to governments on the need to adopt
corrective measures.
A country in excessive deficit is required to adopt corrective measures
according to a fixed timetable. Failure to comply may require payment of a
non-interest-bearing deposit. Should the excessive deficit persist, the deposit
is converted into a fine after two years. Sanctions may also damage reputa-
tion, which can translate into a higher risk premium on government securi-
ties. The public nature of the exercise can contribute to the effectiveness of
the control exerted on budgetary policy by the market.
In the 1990s this framework proved effective in constraining deficit and
debt levels; however, its effectiveness in shaping the fiscal policy of euro-area
members has not yet been proved. In particular, EMU rules have not yet
been tested by severe recessions or large-scale asymmetric shocks. The slow-
down of 2002–03 has so far been problematic for countries (France, Germany,
Greece, Italy, and Portugal) which had not already reached a close-to-balance
position (Buti et al. 2003). Moreover, it is not clear that the automatic stabi-
lizers, envisaged to work freely around a predefined cyclically adjusted target,
would provide a sufficient degree of cyclical smoothing, in view of the larger
requirements of EMU fiscal stabilization (Brunila et al. 2001). The funding of
public investment may also prove difficult (Balassone and Franco 2000).
Another problem is whether budgetary procedures and institutions at the
national level are consistent with the constraints imposed by the EMU
framework (Hallerberg et al. 2001).

Subnational government behavior

The problem of monitoring the soundness of public finances of lower gov-
ernment tiers arises in countries which are not centralized (Ter-Minassian
1997). Within each country, monetary and financial stability is a public
good to which the national government and all subnational governments
contribute by maintaining a sustainable fiscal position. There is an incentive
for each subnational government to exploit, as a free rider, the benefits
accruing from the discipline of others without itself complying with the
rules. This may create a double cost for the other entities: the free rider’s
excessive indebtedness can put pressure on interest rates to rise and this can
result in expensive bailouts.
In principle, this problem can be dealt with through both market-induced
discipline and special regulations. However, the effectiveness of the market
in inducing fiscal discipline requires certain conditions: no government
222 Fabrizio Balassone et al.

body should have privileged access to the market; the market should have
access to all the information necessary to evaluate the financial conditions
of each government; bailing out troubled governments should not be
allowed; and public authorities should react to market signals (Lane 1993).
These conditions are difficult to attain and unlikely to apply simultaneously.
Creative accounting and window dressing may hinder a reliable assessment
of subnational government finances. The no-bailout clause may lack
credibility, especially in those countries where the public sector plays an
important role in providing public services and goods. In addition, the
reaction time of decentralized fiscal authorities may be excessively long
(Blondal 1999). Consequently, market rules are widely supplemented by
regulations in most countries.
Excluding pervasive administrative controls, which by their very nature
are incompatible with a federal structure, two solutions may be considered:
the cooperative management of indebtedness and the introduction of rules
and sanctions for noncompliance.9
With cooperative solutions, all levels of government must be involved in
formulating policy objectives and responsible for their attainment. The
incentive problem is addressed through moral suasion and peer pressure.
Cooperation may require protracted negotiations to the detriment of the
effectiveness of economic policy, especially when many bodies are involved.
On the other hand, cooperative solutions can permit greater flexibility in
dealing with unexpected circumstances.
Rules directly modify the incentive faced by governments. They can bring
benefits in terms of transparency and speed, and can increase the
predictability of government behavior, thereby reducing uncertainty in the
economic environment. But they also raise some problems, such as
credibility of rigorous application, in particular in resisting bailouts, and the
possibility of efficient monitoring to avoid creative accounting.
For these reasons, several countries have adopted eclectic approaches that
combine rules with various forms of cooperation based on peer pressure
(Ter-Minassian and Craig 1997; Banca d’Italia 2001). Some administrative
controls are also frequently used. In some countries, fiscal targets are speci-
fied by the law; in others, they are the outcome of budget procedures in
which both cooperation and controls may be present. In countries charac-
terized by a high degree of decentralization, borrowing is generally permit-
ted to any government tier. The rules generally limit the overall size of the
deficit (either directly, or indirectly via thresholds for interest outlays) and
allow indebtedness for certain purposes only (usually public investment).
The constraint on indebtedness generally applies ex ante: possible overshoots
may be compensated for in subsequent financial years. Further budgetary
flexibility is sometimes provided by the so-called rainy-day funds.10
In assessing the compatibility between fiscal decentralization at the
national level and the rules introduced at the supranational level, three
 Lessons on Subnational Rules from EMU 223

issues gain prominence in the EU context: the asymmetric structure of

incentives and constraints provided by EMU rules with respect to different
government tiers; the absence of special provisions for capital outlays; and
the need to avoid inducing procyclical behavior.
First, EMU rules may exacerbate the internal free-rider problem, as they
introduce an asymmetry in the structure of constraints and incentives faced
by national and subnational governments. While compliance with fiscal
rules applies to general government, no specific responsibility is assigned to
subnational governments. Officials of the central government of each member
state sit on the ECOFIN and commit the state to the common policies. The
central government bears the costs of noncompliance, in terms of both loss
of reputation and possible financial sanctions.
Second, reflecting the lack of a federal authority with the power to enforce
fiscal discipline, EMU fiscal rules are tighter than those generally introduced
at the national level, with respect to the funding of capital outlays and the
effects of the economic cycle on the budget.
Third, the adoption of internal rules which are less flexible than EMU rules
may imply that the existing equilibrium between the national and the
subnational governments turns out to be no longer appropriate. More specif-
ically, the flexibility allowed to decentralized governments, in terms of
deficit financing for exceptional circumstances or capital spending, may be
inconsistent with EMU fiscal rules. However, reducing this flexibility may be
problematic.
The call for close-to-balance or surplus under the SGP implies that most cap-
ital outlays have to be funded out of current revenue. Hence it is no longer pos-
sible to spread the cost of an investment project over all the generations of
taxpayers who benefit from it. This discourages the undertaking of large pro-
jects producing deferred benefits and entailing a significant gap between cur-
rent revenue and current expenditure. The disincentive is even stronger during
the transition toward a balanced budget, when, in order to keep the flow of
investment unaltered, the gap tends to grow. This effect may be stronger for
subnational governments, where investment spending can fluctuate consider-
ably over time and the cost of projects may easily exceed available current rev-
enue.11 Available evidence confirms the link between fiscal consolidation and
cuts in capital spending. In 1992, the actual deficit exceeded the deficit limit
in nine EU countries. In 1997, for all these countries but Greece the ratio was
at or below the threshold; all had reduced the investment–GDP ratio; all but
Greece and the Netherlands had lowered the investment–primary outlay ratio.
Over the same period, investment ratios increased in three of the six countries
that had met the deficit criterion in 1992.12
Economic theory has long maintained that the stabilization function
should be a central government responsibility due to externalities and
spillover effects (Musgrave and Musgrave 1984). Also, subnational tax bases
should be chosen to minimize sensitivity to the economic cycle, to avoid
224 Fabrizio Balassone et al.

procyclicality at the subnational level. This is often ignored in practice since

rainy-day funds or the possibility of compensating for deficits in one year
with surpluses in another provides flexibility for the effects of the cycle at
the subnational level. These mechanisms are not consistent with EMU rules,
which reconcile soundness and flexibility by aiming for a cyclically adjusted
balance, subject to a ceiling on the nominal deficit. This practice is valid not
only ex ante but also ex post. Moreover, EMU rules require estimates of
cyclical developments and their effects on the budget, which may be
unavailable at the subnational level (or, if available, may be biased because
of measurement problems due, for example, to factor mobility).

What solutions are there?13

Adapting existing national rules

Fiscal rules in place in most countries set flexible ceilings for the budget
deficit at the subnational level. The ceilings may exclude capital expenditure
(under the golden rule) or may apply only on an ex ante basis (i.e. if the
deficit overshoots the ceiling, the overrun can be compensated for in the fol-
lowing years). In some cases (e.g. some US states) the deficit overshoot must
be financed through recourse to specially constituted rainy-day funds, rather
than to the market.
Adapting these solutions to the scenario created by EMU rules appears eas-
ier in terms of the asymmetric incentive issue and the financing of investment
than with respect to the effects of the cycle. The incentive problem may be
tackled by introducing a rule – possibly established in a constitutional amend-
ment – that gives equal responsibility for compliance with EMU fiscal rules to
all government tiers. This should be supplemented by a peer-review system –
whose feasibility and effectiveness would clearly be sensitive to the levels of
governments involved – and by credible sanctions for noncompliance.
Concerning capital outlays, the adoption of the golden rule for lower
government tiers would have to be accompanied by an overall ceiling on
investment expenditure by subnational governments. The deficit thus
allowed would have to be compensated for by a central government surplus
with a generous enough margin to allow for countercyclical measures, so as
to take into account the need for the cyclically adjusted budget of the
general government to be close to balance or in surplus.
Allocating among decentralized bodies the overall financing allowed for
investment programs is difficult because subnational governments vary
widely in population, infrastructure, overall receipts, and so on. Thus a
cooperative approach could be contemplated. Decentralized governments
cooperating to define overall budget targets would acquire greater responsi-
bility by aiming consistently at the targets set and reaching agreement on
the allocation of resources. Moreover, the peer pressure for compliance
generated in a cooperative framework could be strengthened by allocating
 Lessons on Subnational Rules from EMU 225

any sanction handed down by EMU to the bodies responsible for the
overshoot.14
With regard to the absorption of cyclical effects on the budget, applying
ceilings that are valid only ex ante is clearly in contrast with European legis-
lation, which is based on ex post limits. On the other hand, setting limits that
are valid also ex post would have shortcomings too. The limits would have
to be decided on a case-by-case basis in connection with each government’s
budget sensitivity to the cycle and the size of expected downturns.
Differences may prove politically difficult to justify. In addition, the neces-
sary information may be unavailable or unreliable. Furthermore, without a
target in terms of a cyclically adjusted balance (whose definition would pose
the same difficulties as defining nominal ceilings), subnational governments
may also tend to target the deficit ceiling during favorable cyclical phases,
in order to minimize fiscal effort. This would result in procyclical policy
behavior, and distort the allocation of resources.
Rainy-day funds could lessen the incentive problem by increasing the
visibility of imprudent fiscal behavior. However, under the European System
of National Accounts (ESA95), the rainy-day funds would have to be treated
as a financing item, raising the same difficulties concerning the definition
of deficit ceilings.15
One possible solution is a careful selection of tax bases, with subnational
government revenue properly supplemented by transfers, so as to minimize
the cyclical sensitivity of subnational budgets. This would have to be sup-
ported by the requirement to balance the budget in nominal terms.
However, this solution may not be consistent with a high degree of decen-
tralization, which would imply autonomy with respect to the provision of
public services and therefore with respect to the level of resource use.

Extending the SGP

Application of the pact to subnational governments would clearly eliminate
any asymmetry in incentives faced at different government tiers. However,
the financing of local investment expenditure through local taxation could
pose particular problems, especially when unusually expensive projects
could lead to expenditure peaks.
Moreover, the large number of bodies involved could make monitoring
particularly costly. As already pointed out, the evaluation needed for the
cyclical adjustment of government data could be especially problematic.
Extending EMU rules to only the larger decentralized (i.e., regional) govern-
ments could be a solution, provided smaller governments have limited
autonomy. Otherwise, the cost of adjustment would merely be shifted from
the central government to the larger subnational governments.

Market for deficit permits

The solution of externality problems by creating appropriate ownership rights
and allowing them to be freely traded was first put forward by Coase (1960).
226 Fabrizio Balassone et al.

Casella (1999) suggests this approach to fiscal discipline within the EMU.
Comparing the negative externality produced by governments running exces-
sive deficits to that caused by environmental pollution, Casella suggests using
the machinery developed in environmental economics to limit deficit levels.
Once the overall ceiling on permits and their initial allotment is set, market
incentives would produce, through free trade, the most efficient allocation in
relation to the financial needs of the various governments in any given year.
The total volume of permits issued can depend on the national economic
cycle, so as to allow both a “structural” margin for investment and a variable
margin to absorb the cyclical impact on the budget. Borrowing and bond
issues lacking debt permit coverage would be prohibited.
The scheme would apparently address the problems identified in recon-
ciling fiscal decentralization at the national level and EMU fiscal rules.
However, it is also subject to three major difficulties. First, while its effec-
tiveness requires that the deficits of the various governments generate the
same externality and are thus perfect substitutes, it must be recognized that
the risk of triggering a financial crisis is not uniform across governments.
If this risk were the function of a single variable, for example, the debt stock,
then one would merely have to make the value of the deficit permits of the
governments inversely proportional to their debt. However, the risk depends
on a number of factors,16 complicating the determination of the value of the
permits held by each government.
Second, the efficiency of the market for permits would depend on the level
of competition. This makes the mechanism ill-suited to situations in which
the number of governments is small (within the euro area there are only
twelve players, all very different in size).17
Finally, there is no easy way to determine the initial allotment of permits.
The possible criteria (GDP, population) would produce greatly differing alloca-
tions. If the demand for permits exceeded the supply, then the countries with
an allotment greater than their requirement would enjoy positional rents.
The first two objections appear more cogent for a permit market among
member states at EMU level than for one among subnational governments
within each country. Presumably, the risk associated with each entity’s bud-
get deficit is more uniform within countries than between countries: the size
of the governments is smaller, and in many cases they have only recently
acquired the power to issue their own debt. Moreover, the number of mar-
ket operators would be vastly greater. Of course, an extensive market could
entail high administrative costs. The third difficulty, the initial allotment of
permits, would depend on political influence at the national level as much
as at the EMU level. It would be compounded, at least initially, by subna-
tional governments’ problems in adapting to the new procedure.
Apart from these difficulties, the permit system seems better suited to
financing investments than to buffering the budgetary effects of the
business cycle. Trading in permits could contribute to greater efficiency in
 Lessons on Subnational Rules from EMU 227

the allocation of public investment. The financial needs associated with

investment projects could be planned and projects modulated as a function
of available resources. As to the cyclical effects, however, the initial allot-
ment would necessarily be based on the forecast of national economic devel-
opment. The emergence of a discrepancy in the course of the year could
result in an overdemand for permits, which would penalize governments in
regions where the cyclical impact was worse than expected.

Assessment
Each of the above solutions has drawbacks. In light of this, a combination
of these approaches could be calibrated to differences in the size of subna-
tional governments.
For larger subnational governments, a domestic replica of the SGP may be
a feasible solution to both the asymmetric incentive problem and the need
to buffer cyclical effects. For relatively large regions, the lack of data is solv-
able and the small number of entities involved allows effective peer pressure
to supplement sanctions in the incentive problem. The need to spread invest-
ment costs over a number of years can be addressed by a “compensated”
golden rule, that is, with an overall deficit cap to be compensated for by a
central government surplus. Cooperation may lead to an inefficient alloca-
tion of borrowing for investment projects if conventional criteria (e.g. size of
population), which do not reflect the returns of the different projects, were
used for the distribution of borrowing allowances. In this case, the introduc-
tion of a market for borrowing rights could induce more efficient outcomes.
For smaller governments, a careful selection of tax bases can largely iso-
late their budgets from cyclical effects provided that they enjoy a limited
degree of revenue autonomy. In this case, the requirement to keep the bud-
get balanced in nominal terms – a less sophisticated rule than the one envis-
aged in the SGP – could be used to solve the incentive problem without any
risk of inducing procyclical behavior, thereby avoiding the data problem and
reducing the monitoring difficulties arising in a “large number” context.
Again, a “compensated” golden rule could be used in order to avoid an
undue compression of capital outlays.

Solutions adopted by EU member countries

The structure, responsibilities, and means of financing subnational govern-

ments in EU member countries are diverse (Smith 1996; J. Fischer 2001).
They reflect national history, traditions, and specific political and cultural
features. In general, subnational financial autonomy is relatively high in fed-
eral states, such as Austria, Belgium, and Germany, and in the Nordic
countries. Italy and Spain have experienced a shift from a centralized to a
decentralized structure over the last quarter century.
228 Fabrizio Balassone et al.

In 1995, the ratio of subnational government expenditure to GDP ranged

from 6 percent in Belgium to 32 percent in Denmark. However, a low/high
ratio does not imply a low/high degree of financial autonomy. For example,
in Italy, the Netherlands, the United Kingdom, and Germany the
subnational expenditure ratio was of the same order (ranging from 10 to 14
percent), but subnational governments’own revenue was less than 3 percent
in the first three countries, while it covered almost all the outlays in
Germany.18 On average, subnational governments in EU countries balance
their budgets (J. Fischer 2001), but only after sizeable transfers from the
national government.
National governments usually set constraints on subnational government
finances. In a number of countries, a golden rule is in place. In others,
borrowing at the subnational level has to be authorized by the finance min-
ister. In France, Ireland, and the United Kingdom, the central government
can directly restrict borrowing by lower levels of government (Hallerberg et
al. 2001). In Sweden a strong constraint on subnational government
finances was introduced in 2001: subnational governments are required
to balance their operations every year, and when a deficit is recorded, the
balance has to be restored within two years. In addition to limits on
borrowing, EU countries frequently adopt explicit coordination agreements
among different government tiers, though generally they are not legally
binding. Only Finland does not have either coordination procedures or
restrictions from the central government.
Attention to the problems highlighted in the previous section was
especially high in five EMU countries: Austria, Belgium, Germany, Italy, and
Spain. During the 1990s, the share of subnational government taxes in
general revenue was relatively stable in Austria, Belgium, and Germany,
while it grew substantially in Italy and Spain.19 Interestingly, none of the five
countries has decided either to replicate fully the SGP or to introduce a
market for tradable deficit permits for subnational governments. Austria,
Belgium, Italy, and Spain introduced explicit internal stability pacts,
although each with different characteristics and scope. Germany has long
considered reforms, but so far its structure remains unchanged. Nonetheless,
the agreement reached in March 2002 by the federal and regional govern-
ments on their joint responsibility for the commitments arising under the
SGP represents a step toward a formal pact.
In all five countries rules are generally based on agreements between the
center and periphery (Figure 14.1). The main exception is Italy, where internal
rules have been largely imposed by the central government, probably due to
the relatively short decentralization experience and the lack of institutional
representation for regional governments at the national level – as in the
Austrian and the German parliamentary chambers. Nevertheless, Italy also is
moving toward a framework where negotiation, coordination, and consensus
between center and periphery are more important; for example, formal rules
 Lessons on Subnational Rules from EMU 229

Imposed Agreed

Explicit Italy* Austria

Belgium

Spain

Implicit Germany*

Figure 14.1. EU member countries: internal stability pacts

*
Note: Countries with flexibility for investment expenditures.

are being supplemented by informal procedures to reach agreements in the

health sector. Some margin for flexibility is allowed for investments (this is
not the case in Austria, Belgium, and from 2002, Spain), but without an
explicit mechanism for taking into account the effects of the economic cycle.
The sensitivity of subnational government budgets to the economic cycle
differs among the five countries and, within each country, among the dif-
ferent subnational tiers. The cyclical component of subnational government
resources lies primarily within tax revenue; therefore, differences in the tax
share of total revenue at the subnational level matter. In addition, sensitiv-
ity depends on which taxes are local and on the way grants are designed.20
In Germany, where a significant share of subnational government revenue
comes from income taxes and VAT-sharing schemes, both regions and
municipalities seem quite exposed to the cycle. In Austria, subregions seem
less exposed to the effects of the cycle than regions: even if the ratio of their
own taxes to overall resources is larger for subregions than for regions, sub-
regional governments’ own tax revenue is more sensitive to the cycle. In
Belgium, the way grants are designed can partially counterbalance the cycli-
cal changes in local revenues. In Spain, tax revenue as a share of local
resources is relatively small so that the effects of the cycle can still be con-
sidered limited. Up to the beginning of the 1990s, this was the case for Italy
as well. In these two countries the problem posed by the cyclicality of rev-
enues will become more pressing as decentralization advances.

Lessons and conclusions

Three critical areas of the interaction between the EMU fiscal rules and fis-
cal decentralization have been identified: the incentive problem posed by
the asymmetric intergovernmental distribution of responsibilities with
respect to compliance with EMU rules, the risk of an undue compression of
capital outlays at the subnational level if borrowing is prohibited, and the
possibility of rule-induced procyclical behavior in subnational governments.
230 Fabrizio Balassone et al.

From an examination of the alternative solutions, the following lessons of

possible relevance for decentralized emerging market economies can be
derived.

● For larger subnational governments, a domestic replica of the SGP, includ-

ing explicit financial sanctions, can be introduced to tackle both the
asymmetric incentive problem and the need to buffer cyclical effects.
● For smaller subnational governments, a careful selection of tax bases can
largely isolate the budget from cyclical effects. In this way the require-
ment to keep the budget balanced in nominal terms – a less sophisticated
rule than that envisaged in the SGP – can be used to solve the incentive
problem without any risk of inducing procyclical behavior.
● To avoid compression of investment outlays, the introduction of a “com-
pensated” golden rule (i.e. a golden rule with an overall deficit cap to be
compensated for with a central government surplus) together with a
cooperative mechanism to allocate borrowing according to the needs of
the different governments, can be considered. If cooperation leads to an
inefficient allocation of borrowing for investment projects, a market for
borrowing rights can be considered as an alternative to induce more
efficient outcomes.

In addition, an analysis of the approaches adopted by EU countries leads to

a number of suggestions concerning the fiscal institutions, including rules,
which strongly reflect country-specific developments.

● No EU country has chosen to replicate the SGP at the national level.

Indeed, in most countries there are no predefined sanctions for subna-
tional governments’ contributing to a slippage with respect to the targets
set for general government.
● Municipalities and other local governments are usually assigned resources
in such a way as to limit the sensitivity of their budgets to the cycle. In
some countries this also applies to larger regions. In general, the share of
subnational governments’ own taxes in total revenue is still relatively low.
This further contributes to insulating these governments from the cycle.
● All countries rely, albeit in varying degrees, on cooperative mechanisms
to tackle the three identified critical areas. Parliamentary chambers rep-
resenting regions and advisory councils including representatives of the
different levels of government play a crucial role. The extent to which
cooperation is supported by some element of control seems to be in
inverse relation with the length of the federalist tradition of the country.

The fiscal framework for subnational governments adopted by five selected

EU member states has so far proved successful in terms of deficit control. In
 Lessons on Subnational Rules from EMU 231

Belgium, Italy, and Spain rising decentralization has not hampered fiscal
consolidation. Between 1993 and 2000, the deficit–GDP ratio has declined
by 7 to 10 percentage points in Belgium, Italy, and Spain (excluding UMTS
receipts). Although the bulk of this adjustment can be attributed to a sharp
drop in the interest rate, reflecting the disappearance of currency risk, sub-
national government finances do not seem to have exercised significant
pressure on fiscal targets in any of these countries. Furthermore, the Belgian
experience indicates that even in a country with significant tensions among
different regions a cooperative approach can be successful.
The question then arises of whether subnational frameworks are adequate
and sustainable, or whether there are some factors that, while not yet pro-
ducing effects, may nevertheless require further changes in the future. In any
event, the robustness of the cooperative mechanisms used to eliminate the
incentive problem posed by EMU fiscal rules needs to be assessed with
respect to both economic and institutional changes, particularly in view of
their interaction.
First, one should consider whether the existing arrangements are adequate
to deal with relevant shocks, such as a deep recession. Cooperative approaches
may require protracted negotiations that prevent rapid adjustment of
revenue and expenditure to new circumstances. This may especially apply
when several governments are involved, to the detriment of the effective-
ness of economic policy.
Second, while at present in most countries (except for Germany) subna-
tional budgets are largely insulated from the effects of cyclical develop-
ments, in the future this feature may vanish if more tax bases are assigned
to lower government tiers. The choice not to replicate nationally the SGP
may reflect measurement problems. The operation of rainy-day funds can be
considered as an alternative.
From an institutional perspective, the highly centralized countries have
substantially widened the responsibilities of regional governments in order
to cope with political pressures related to the cultural diversity of regions.
Economic arguments have not played a primary role. The process is still in
evolution, with relevant changes being introduced in some countries.
Current budgetary frameworks are consistent with limited fiscal
differentiation across regions. In most of the countries considered, regional
governments have limited autonomy in setting tax and expenditure levels.
In this context, tax-sharing agreements are necessarily complemented by
expenditure guidelines, since subnational governments do not have the pos-
sibility of compensating for expenditure increases with higher revenue
growth.21
With a future increase in regional autonomy (likely, for example, in Italy),
expenditure control would become unfeasible. It may be advisable then to
strengthen the role of the rules referring to the budget balance. The ensuing
232 Fabrizio Balassone et al.

tight link between revenue and expenditure decisions is important for an

efficient allocation of resources.
From an economic point of view, the sustainability of current arrangements
may be questioned on allocative efficiency grounds. Due both to the charac-
teristics of the financing systems of subnational governments (which partly
insulate them from cyclical shocks) and to the lack of predefined sanctions, the
central government may end up systematically compensating for the slippages
of other government levels. This may induce a misallocation of resources.
EMU rules have increased the attention to fiscal performance at the sub-
national level. At the same time, they constrain the range of solutions
applicable to regulating these outcomes, and call for clear accountability and
rapid adjustment. In this respect, the introduction of explicit domestic sta-
bility pacts mimicking the SGP has marked advantages over less formal
cooperative mechanisms. The introduction of predefined rules and sanc-
tions may redress the incentive structure facing politicians and induce faster
adjustments (Kopits and Symansky 1998). It may also increase transparency
and allow better control of policy implementation on the part of the
electorate and financial markets.
While all countries are taking steps toward supplementing cooperation
with some rules, in countries such as Italy and Spain, where decentralization
is relatively recent and where the number of subnational governments is
large, rules can be especially useful. Indeed, the reform recently enacted in
Spain seems to reflect this. Explicit rules for budgetary targets have been
introduced to replace bilateral negotiations. Flexibility will be reduced with
respect to both investment and the economic cycle.
The feasibility of a rules-based approach depends on the adoption of uni-
form accounting and statistical standards, such as those embodied in ESA95.
This was also a key element of the cooperatively defined rules adopted in
Austria. Unfortunately, this principle seems to meet difficulties in countries
whose decentralization is recent. In Italy, for example, it is sometimes
claimed that the constitution gives regions autonomy also with respect to
accounting practices. In Spain the recent reforms explicitly acknowledge the
need to improve the degree of transparency of subnational government
finances by requiring that a central statistical archive be set up to ensure that
fiscal information supplied by the regions meets the same standard as those
of the central government.

Notes

1. We are grateful to Pablo Hernandez, Anton Matzinger, and Karsten Wendorff, and
to the discussants David Colmenares and Jaime René Jiménez, for useful comments
on previous versions. The views expressed are those of the authors and do not
reflect the position of the Banca d’Italia.
 Lessons on Subnational Rules from EMU 233

2. In this respect, the present chapter seeks to complement Chapter 7 by Buti and
Giudice.
3. See Balassone and Franco (2001).
4. For a more detailed description of the solutions adopted in these countries, see
Balassone et al. (2003).
5. The economic policy framework of the EMU is extensively examined in Buti and
Sapir (1998), Buti et al.(1998) and Brunila et al. (2001).
6. These three conditions can make the limit threshold extremely binding. See Buti
et al. (1997).
7. The issues related to the choice of the medium-term fiscal target are also
examined in Barrell and Dury (2001).
8. Balassone and Monacelli (2000) analyze the implications of this provision for
stabilization policy.
9. Rules may obviously be the outcome of a cooperative decision-making process,
but once defined they avoid the need to search for a consensus about each
budgetary policy issue. Ter-Minassian and Craig (1997) note that there is some
scope for cooperation also in a rules-based approach.
10. For example, this is the case in the United States. For a detailed analysis, see
Knight and Levinson (1999) and McGranahan (1999).
11. According to Einaudi (1948: 318, our translation), “Building a school may be an
extraordinary effort for a small town, an ordinary one for a big city.” See also
Pigou (1928: 717).
12. Balassone and Franco (2000) also examine the possibility of introducing the
golden rule in the EMU framework. Support for the link between fiscal consoli-
dation and investment cuts is also found by de Haan et al. (1996) for a sample of
OECD countries.
13. We do not take into account the possibility of the Commission’s directly moni-
toring the fiscal outcomes of subnational governments. While apparently appro-
priate in a context in which the budgetary results of large regions are more
relevant than those of some member states, this solution would be politically
problematic as it would radically alter the relationship between the EU and the
member states. Also, the monitoring of a large number of bodies could hamper
the effectiveness of procedures that have been set up to deal with 15 governments
at most.
14. The overall ceiling and the sanctions should avoid the risk of a “pork-barrel”
effect. See, for example, Chari and Cole (1993).
15. In order to avoid this problem, the ESA95 would have to be modified. Resources
drawn from rainy-day funds would have to be treated as government revenue
rather than as a financing item. This would amount to a redefinition of EMU fiscal
rules and of the relevant measure of budget balance.
16. For instance, the risk may depend on the degree of exposure of the banking sys-
tem and the degree of international openness; see Hernández-Trillo (1995).
17. The problem could be attenuated by a continuous double auction market; see
Friedman and Rust (1993).
18. Data on the composition of expenditure by level of government can also be
misleading as a measure of the degree of decentralization given areas of concur-
rent responsibility among different tiers of government.
19. On the relationships among different government levels, see Balassone et al.
(2003).
234 Fabrizio Balassone et al.

20. For example, in Germany current transfers from regions to municipalities mainly
depend on the level of the regions’ tax revenues. Moreover, capital transfers from
regions to municipalities often depend on the budgetary situation of the regions.
21. Introduction of expenditure guidelines to replace the SGP would be problematic
in three respects: institutionally, the question would arise of who is to set the
reference growth rate of expenditure for each sovereign country; if the externali-
ties of fiscal policy are determined by deficit and debt levels, rules should refer to
these parameters and not to expenditure; and operationally, as member countries
can change their tax systems, the need for fiscal rules would remain.
15
Rules for Stabilizing
Intergovernmental Transfers
in Latin America
Christian Y. Gonzalez, David Rosenblatt, and Steven B. Webb1

Introduction

Traditional theory of fiscal federalism assigns the role of macroeconomic

stabilization to the federal government (Musgrave 1959; Oates 1972). One
fundamental justification is that monetary control is exercised only at the
federal level. Even if an economic disturbance is symmetric across regions,
then there is the complication of coordinating fiscal responses by subna-
tional jurisdictions. States or provinces represent economic areas with
completely open trade and capital accounts within a monetary union.
Countercyclical fiscal policies at the subnational level can lead to offsetting
capital flows that limit the impact of attempts to reduce local unemploy-
ment (Oates 1972). For similar reasons, subnational governments face
difficulty responding to localized asymmetric shocks with fiscal policy, since
the economic impact of the policy is muted. However, instead of dealing
with asymmetric shocks, this chapter examines how intergovernmental
transfers affect the division of the burden of stabilization across the levels of
government when the nation as a whole faces economic fluctuations.
In addition to the traditional theoretical analysis, there is the empirical
observation that federal governments often have cheaper and more stable
access to financial markets than subnational governments. This is particu-
larly the case in emerging market economies. Thus it may be more efficient
for sovereign countries, rather than their subnational governments, to
borrow for consumption smoothing. Intergovernmental transfers that shield
subnational governments from part or all of the impact of an economic
downturn shift the burden of borrowing to the federal level during the
downturn.
On the other hand, during upswings lasting only two to three years,
subnational governments have rapidly increased spending. Given the high
share of wages allocated to subnational governments mainly to provide social

235
G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets
© International Monetary Fund 2004
236 Christian Y. Gonzalez et al.

services, these expenditure increases are difficult to reverse when the boom
ends. Superior fiscal planning and fiscal responsibility laws at the subna-
tional level can alleviate this problem.2 Many automatic transfer systems are
linked to a fixed percentage of current federal revenue. A stabilizing rule –
withholding revenue increases during boom periods – could assist subna-
tional governments in avoiding the boom–bust trap.
This chapter examines federal fiscal performance during the recent eco-
nomic cycles of three large nations in Latin America. Whereas Argentina and
Colombia have stabilizing federal transfer rules that go beyond maintaining
a fixed share of revenue, Mexico has alternative arrangements (or no
arrangements), which have performed no worse in recent years.3 We analyze
how the rules or lack thereof have affected fiscal outcomes, and what
alternative arrangements can be created for managing the uncertainty of
revenue sharing.

Conceptual basis

When establishing a system of subnational fiscal rules, policymakers face a

variety of questions. Should states or provinces fire teachers or cut their
salaries during economic downturns in order to comply rigidly with balanced-
budget rules? If not, is it better for the states or provinces, rather than the
federal government, to borrow during the downturn? The other alternative
is to adopt a set of subnational stabilization funds (“rainy day” funds).
Another issue is whether it is credible for the central government to commit
to not transferring additional funds during the downturn. And a further
question, on the upside of the cycle: are subnational governments able to
generate surpluses during the boom periods? The focus in this chapter is on
these public finance problems rather than regional employment or develop-
ment issues. The basic overriding question is whether stabilizing intergov-
ernmental transfers can serve a complementary role to national/ subnational
fiscal rules, budget planning, and debt management. In particular, can they
help deal with the boom–bust cycles so prevalent in emerging market
economies, particularly in Latin America?
Ideally, the following conditions would be in place before establishing a
stabilization rule for national-to-subnational fiscal transfers:

● Subnational governments are credit constrained – rationed in some way

out of the market or confronted by a substantially higher cost of borrow-
ing than the national government.
● The national government possesses stable access to credit and quality
debt management.
● There are no severe structural fiscal imbalances, either within or across
levels of government. In other words, neither level of government faces
unsustainable cyclically adjusted fiscal deficits. In addition, subnational
 Stabilizing Intergovernmental Transfers 237

governments do not spend excessively with financing by automatic

federal transfers.4

The federal level should play the predominant role in macroeconomic

stabilization. In these circumstances, it would be efficient for the federal
level to stabilize federal transfers during the downturn, so that it would be
the one to borrow. However, if the federal government is on the brink of
insolvency, or if there are long-term structural imbalances at the subnational
level being financed by the current system of transfers, then stabilizing
transfers may only complicate or even worsen the initial problems.
An additional consideration is whether the scale of a full guarantee
imposes undue risk for the central government. The scale of the guarantee
is a function of the degree of fiscal independence of subnational govern-
ments. If transfers comprise a high share of GDP or central government
revenue, then the cost of the full insurance during the downturn could
become excessive and even unbearable. The center may wish to limit its
liability with some form of escape clause in the event that the economic
recession is deeper or more persistent than originally predicted, as illustrated
by the case of Argentina.
There could be reasons, however, to have a rules-based guarantee even in
less-than-ideal conditions. Analogous to the design of deposit insurance for
banking institutions, governments should avoid implicit guarantees for
transfers. While the three conditions mentioned earlier might not hold fully
in any Latin America country, a federal commitment to no additional trans-
fers during the downturn lacks credibility.5 In addition, during upswings,
federal withholding of a portion of automatic transfers could inhibit desired
increases in subnational expenditure. There could be a case, then, for some
sort of stabilizing rule to complement national and subnational efforts to
establish fiscal policy rules consisting of, for example, limits on budget
deficits or debt at both levels of government.

Argentina

Fiscal performance
During the 1990s, Argentina’s economy grew rapidly, on average, but suf-
fered a sharp reversal in 1995, and since 1999 it has suffered periods of con-
traction and stagnation reflected in fluctuating fiscal performance. The
federal government reached a small surplus in 1993, followed by a deterio-
ration, and after the 1995 crisis, deficits have been close to 2 percent of GDP.
Provincial deficits, on aggregate, averaged 1 percent of GDP during the
1990s, albeit fluctuating over the economic cycle. Preliminary estimates
show that they may have reached nearly 2 percent of GDP in 2002, while
the federal deficit may have been twice that amount on an accrual basis.6
238 Christian Y. Gonzalez et al.

Provincial expenditures increased as a share of GDP during 1991–93,

partly reflecting the final phase of decentralization of secondary education
responsibilities to the provinces. In the mid-1990s, expenditures stabilized
as a share of GDP, implying procyclical real expenditure cuts to respond to
the 1995 recession. In 1999, expenditures increased in relation to GDP, heav-
ily influenced by large expenditure increases in the province of Buenos Aires
(accounting for 38 percent of the nation’s population). While there were
some procyclical cuts during the crisis of 2000–01, they were not sufficient
to stem the rising expenditure–GDP ratio.

Transfers and guarantees

Argentina’s system of intergovernmental transfers is largely based on automatic
revenue-sharing and tax-sharing arrangements. During the 1990s, approxi-
mately 90 percent of federal transfers to the provinces were automatic. The
largest of these transfer programs is the general revenue-sharing pool copar-
ticipación, whereby a percentage of federally collected value-added, income
and asset taxes are shared with the provinces. In addition, there are the
smaller tax-sharing arrangements of fuel taxes that provide funds to finance
specific investment programs.
The 1988 law established that the share of coparticipación distributed to the
provinces – the “primary distribution” – be set as 58 percent of the pool of
federal VAT and income taxes.7 A set of predeductions and the specific
tax-sharing arrangements created a labyrinth of arrangements. For most of
the 1990s there were no stabilizing rules. At the end of the labyrinth, federal
automatic transfers have consumed about one-third of federal current
revenues in recent years,8 representing about 6 percent of GDP. From the
provincial point of view, automatic transfers and discretionary transfers
finance about half of provincial expenditures, on aggregate. However, this
figure is biased toward the three largest provinces and the City of Buenos
Aires (comparable in status to a province). The smaller provinces depend
upon transfers to finance 80–90 percent of their expenditures.
The Fiscal Pact of 1992 represented a watershed agreement between federal
and provincial authorities in a variety of areas, including privatization,
structural reforms, and some minor revisions to the system of transfers. In
addition, it provided a minimum guarantee for monthly coparticipación
transfers. According to the law, the intention was not to respond to external
shocks resulting in a typical three- to four-quarter recession. The guarantee
was established for more short-term declines in monthly federal revenue
since any payments were to be “repaid” to the federal government via reten-
tion of any surpluses above the “floor” during subsequent months.
The Fiscal Pact of 1993 represented another key agreement in a variety of
structural reforms, with a particular focus on gradual reforms of provincial
taxation and deregulation.9 It also raised the minimum monthly floor and
eliminated the federal government’s right to retain the surpluses in later
 Stabilizing Intergovernmental Transfers 239

months, with the proviso that provinces comply with the tax and deregula-
tion clauses. This latter clause also stipulated that the federal government
not try to recover a short-term loan converted into a grant.
In late 1999, another Federal Agreement was reached with the provinces,
focused mainly on the division of transfers, though accompanied by some
general commitments to provincial tax harmonization and fiscal trans-
parency. Despite recession, the minimum floors established during the 1992
and 1993 pacts had long become irrelevant due to substantial average
economic growth in 1994–98. A major component of the agreement was
that during 2000, the provinces would receive a fixed amount in automatic
transfers, to provide revenue predictability, while allowing the federal gov-
ernment to keep a larger share of incremental revenue expected both from
an eventual recovery and an increase in the federal tax ratio. The fixed
monthly transfer was based roughly on the average of the previous two
years.
The agreement also established that during 2001 the provinces would
begin to receive an average of the legal transfer of the three most recent
years, thus paving the way to a moving-average approach. In addition, the
provinces were offered a minimum guarantee for 2001 that was set at a 1 per-
cent increase over the level in 2000. This arrangement represented a sizable
expected loss to provinces in terms of total transfers. As an incentive, debt-
restructuring deals were offered to smaller provinces, along with financing
part of provincial employee pension deficits if reforms were made consistent
with the national system.
In 2000, a more comprehensive Federal Agreement was signed by all
(except one) of the provincial governors. This agreement included several
clauses for provincial reforms toward greater fiscal transparency. The
agreement established a timetable for switching permanently to the moving-
average concept, though still providing guaranteed minimum amounts over
a transition period. For 2001 and 2002, the provinces would receive a fixed
monthly transfer equivalent to the guaranteed minimum for 2001 that had
been stated in the 1999 agreement, which now would be both a floor and a
ceiling for both years, implying an increase of about 1 percent over the
amount in 2000. During 2003–05, the provinces would start to receive a
moving average of the shared revenue in the three most recent years. In the
case of recession, a guaranteed minimum amount had been set, represent-
ing approximately 2.7 percent increase yearly in nominal terms.10
Under this agreement, any major recessions over the period would have
implied that the provinces could break even or come out ahead. As it turned
out, the floor did not strongly favor the provinces during the first half of
2001. In addition, the federal government created a financial transactions
tax with the revenue initially dedicated exclusively to the federal treasury.
However, during the second half of the year, the fixed transfers would have
implied significantly more resources than otherwise would have been the
240 Christian Y. Gonzalez et al.

case, contributing to substantial fiscal, political, and social stress during the
latter part of the year. Ultimately, the federal government was not able to
transfer the full guarantee and arrears accumulated.

Subnational borrowing rules

During the 1990s, provinces had access to credit, mainly by using their
federal transfers as guarantee and automatic payment for debt servicing. The
system has functioned smoothly in terms of effecting payment. However, a
number of provinces have fully committed their future transfers over
particular time periods. In addition, some provinces have issued bonds over-
seas with natural resource royalties as the collateral. Two large subnational
governments – the City and the Province of Buenos Aires – issued general
obligation bonds overseas, with no enhancement.

Alternative stabilization arrangements

In addition to the history of floors, fixed sums and moving averages mentioned
earlier, the federal government has a program of discretionary transfers, intended
to be used for emergency purposes, in the event of asymmetric shocks.
Although the emergencies originally contemplated correspond to natural dis-
asters and other such specific events, in practice, these funds have often been
used for political purposes. In brief, these are not really along the lines of the
fiscal stabilization approach considered in this chapter.
General fiscal stabilization funds do not exist at either the federal or
provincial levels; however, a couple of oil-rich provinces have significant
savings accumulated from oil royalties. In addition, one province consis-
tently has built up significant reserves that could be used during economic
downturns, but has no preestablished rules on when or how to use these
funds during the downturn.

Colombia

Fiscal performance
During the 1980s, Colombia was the good outlier in Latin America, achiev-
ing strong growth and fiscal surpluses in most years,11 thanks to responsible
public management, relative political stability, and expanding oil produc-
tion. While there was some political and fiscal decentralization to munici-
palities, control of finances remained at the center. The 1990s saw reversals
in almost all these dimensions as growth slowed, especially after 1995, in
sharp contrast to the macroeconomic improvements in most of the rest of
Latin America.
As the rise in subnational revenues has fallen short of expenditure growth,
some subnational governments have incurred unsustainable deficits although
aggregate subnational debt did not reach the level in Brazil or in Argentina.
Reflected in sizable transfers to the subnational governments, the fiscal
 Stabilizing Intergovernmental Transfers 241

problem has shifted to the national level. This prompted the national gov-
ernment to initiate a new approach in fiscal reform in 2001.

Transfers and guarantees

Decentralization evolved from the deconcentration of national revenue to
subnational administrative units. Starting in 1968, a departmental fund for
education and health was financed from a fixed percentage of national rev-
enue, while municipalities were assigned 10 percent of the VAT. Although
this revenue earmarking was designed to solve the problem of ad hoc trans-
fers and to supplement inadequate sources of local revenue, ad hoc transfers
remained a problem. A major review of the system of intergovernmental
transfers hardened the subnational governments’ budget constraint and
strengthened their own revenue sources (Bird 1984).
The 1991 Constitution and Law 60 of 1993 moderately expanded the
amount of revenue assigned to departments by broadening the base of the
existing revenue-sharing system (situado fiscal ) to include all recurrent
revenues of the government (VAT, customs, income tax, and special funds).
These statutes committed the national government to increasing the share
of taxes transferred to subnational governments, up to at least 47 percent of
all its current revenue by 2002. This mandate not only took resources away
from the national government but also meant that any revenue from tax
changes would be substantially shared with subnational governments,
weakening the incentive for fiscal adjustment effort.
The 1991–94 reforms also reduced the national government’s discretion in
the distribution of transfers. Law 60 changed the system from paid direct
payments to teachers and health workers to one in which the situado was
transferred directly to each departmental government on the basis of a
formula.12 As a transition measure in 1994–98, each municipality was guar-
anteed at least the amount of VAT transferred in 1992, in constant prices.
These guarantees were applicable to only a limited number of municipali-
ties, and thus did not pose a macroeconomic fiscal problem.
The former approach of fixed revenue sharing did not establish a hard
budget constraint because political pressure achieved supplemental trans-
fers. In 2001 the government was to pass a constitutional amendment that
sets limits to the growth of total transfers. After the transition, the transfers
would equal the moving average of what they would have been in recent
years, as calculated according to the regular formula. During the rather long
transition of 2002–08, however, the transfers would grow annually
2–2.5 percent in real terms – creating both a floor and a ceiling.13 As a ceil-
ing, this approach would let the national government reap a fiscal benefit
(at the margin) from high growth and strong revenue performance. As a
floor, the amendment could present Colombia with problems similar to
those in Argentina if the economy stagnates or declines; the transfers would
still have to grow in real terms even if the tax base declined.
242 Christian Y. Gonzalez et al.

Subnational borrowing rules

Subnational governments have less access to credit than the national gov-
ernment; this has actively but not always successfully restrained subnational
borrowing. Through the 1980s, all subnational borrowing was exceptional and
subject to approval from the Ministry of Finance, given that subnational
entities were appointed representations of the central government, with no
political or fiscal autonomy. The ad hoc approval process gradually allowed
more freedom for domestic borrowing in the late 1980s and the 1990s, as
subnational autonomy increased. Domestic debt of the subnational govern-
ments grew rapidly from 2.6 percent of GDP in 1991 to 4.6 percent in 1997,
especially to the banking sector, and reached the crisis point for several enti-
ties during this period (in 1995, 1998, and 2000).
Witnessing the high rates of growth of subnational debt to domestic banks
in 1993 and 1994 and the debt crises of several subnational governments in
1995, the national government attempted to exert some control over indebt-
edness. On the supply side, the Superintendency of Banks tightened bank-
ing regulations in 1995, slowing the real growth of subnational debt for a
while, but regulations were substantially relaxed in 1996 due to political
pressures, and indebtedness grew again in the following years. Legislation
enacted in 1997 limited subnational borrowing according to capacity-to-pay
criteria, aimed to prevent excessive indebtedness through a system of warn-
ing signals that would prompt direct control from the national government
(Perry and Huertas 1997). The law was frequently violated, however; several
departments and municipalities obtained new credit without required per-
mission. When the Ministry of Finance granted special permission for bor-
rowing on the condition of following an adjustment program, the programs
often failed to deliver the expected results. In spite of the Superintendency’s
new regulations on loan classification and capital-risk weighting, the qual-
ity of subnational loans deteriorated drastically in the late 1990s.
The departments’ debt in Colombia has been problematic partly because
they have little discretion over their receipts or spending, most of which is
devoted to salaries. Neither the departments nor the creditors took sufficient
account of this rigidity in their ex ante evaluations of the ability to pay. In
the case of municipalities, the debt crises were related to runaway expendi-
tures financed with the pledge of increasing transfers. Virtually all depart-
ments have received repeated relief through these means, however, indicating
that budget constraints have softened and a significant moral hazard prob-
lem has developed (Echavarria et al. 2000). Thus, the Colombian experience
with top-down, ex ante controls and repeated bailouts has been a disap-
pointment. It is still too early to know if the current (anti-)bailout laws and
regulations will finally succeed in establishing a hard budget constraint for
subnational governments.
Other than guarantees and debt bailouts with adjustment programs, there
have been no systematic alternative arrangements by or for subnational
 Stabilizing Intergovernmental Transfers 243

governments in Colombia to deal with fluctuations in the tax base and

revenue sharing.

Mexico

Fiscal performance
In the early 1990s, Mexico’s economy grew rapidly, but it suffered a sharp
reversal in 1995. This crisis destroyed the financial system, and paralyzed
credit flows. Although the recorded fiscal accounts appeared to be balanced, in
fact a significant deficit was met with short-term borrowing that contributed
to the 1994–95 financial crisis. Afterwards, fiscal adjustment was accompa-
nied by institutional and policy reforms to improve the functioning of
markets that included shifting from the fixed to a flexible exchange rate
regime and restructuring foreign debt from short to longer term. Also public
spending was decentralized to state and municipal governments.
The financial crisis, and the ensuing increase in interest rates, expanded
the states’ debt stock, which was reduced considerably by the 1996–97
federal bailout package. Aggregate state deficits were close to zero after 1995
and subnational debt declined in real terms. By 1997 subnational govern-
ment debt represented 25 percent of the debt owed or guaranteed by the
Secretariat of Finance and only about 2 percent of national GDP. By the late
1990s only the newly autonomous federal district was expanding its debt in
real terms; most other subnational governments were paying down their
debt or letting it grow only through the inflation indexation of the princi-
pal of the restructured debt (Webb and Giugale 2000). The modest growth
of subnational spending for public services and capital investments (in
health, education, and basic infrastructure) and indebtedness does not yet
pose a major threat to macroeconomic stability.

Transfers and guarantees

As in many other countries, transfers account for 80–95 percent of subnational
government revenues. The two main categories of transfers are participaciones
and aportaciones. The most important element in the Mexican federal transfer
system was, through 1997, the revenue sharing (participaciones), originally
consisting of revenues of states and municipalities collected at the federal level
for administrative efficiency via a fiscal pact. Legally, the federal government
only collects taxes and distributes the proceeds to the subnational govern-
ments. In practice, the federal government writes the formula for distribution
of these funds and augments them from federal resources, like oil revenues, so
they have become more like a transfer of the general revenue-sharing type.
Most of these transfers are distributed under budgetary category or Ramo 28.
The transfers to states from revenue sharing within Ramo 28 were almost six
times as large as the states’ own revenues in 1996. This part of subnational
government revenue is automatically procyclical.
244 Christian Y. Gonzalez et al.

The assignable or shared taxes include mainly the federal income tax, VAT,
and ordinary fees from oil. States receive about 22.5 percent as participaciones,
via three funds comprising the revenue-sharing system. Two funds were used
for distribution to states and the remaining third for transfers to municipali-
ties through state governments but according to a federal allocation formula.
Aportaciones account for just over half of federal transfers; these are allo-
cated to states with earmarking to pay for federal commitments in education
and health, and transferred to the states and municipalities together with
those commitments. These transfers are distributed under Ramo 33, through
nine different funds. The size and allocation of Ramo 33 is specified in the
annual federal budget after congressional debate and approval. While the
major funds are backed by strong political commitment, the size is adjusted
to stay within the limits set in advance in the annual revenue law. This
reduces the likelihood that Ramo 33 could lead to unsustainable spending
and deficits at the federal level. The allocation covers the whole year, pro-
tecting much of the subnational revenue from any macroeconomic fluctua-
tion that was not foreseen in the original draft budget. Adjustments are
absorbed by the rest of the federal budget.
Some Ramo 23 funds relate neither to previous revenues of the states nor
to previous responsibilities of the central government. Once they were in
part under presidential discretion, but then most of them were negotiated
between the states and federal ministries. By 2000, these transfers were
phased out.
In brief, there is no stabilizing component to automatic federal transfers
to the states in Mexico. Discretionary transfers for smoothing the economic
cycle have been limited or recently eliminated.

Subnational borrowing rules

Subnational indebtedness has not been a major threat to Mexico’s macro-
economic stability because its share in the portfolio of the financial system
was relatively small, as a result of two factors. First, subnational governments
have limited borrowing capacity and access to capital markets. Second, the
frequent implicit and explicit federal bailouts have softened subnational
budget constraints; that is, the federal government has absorbed the potential
debts of subnational governments. The second factor, in general politically
motivated, indicates that intergovernmental relationships in Mexico still
embody many ad hoc channels that lead to moral hazard.
Subnational governments can borrow primarily from state-owned devel-
opment banks and commercial banks. Other sources are available but have
rarely been used by the states until recently. Subnational governments can
borrow only in local currency from local residents, and only for productive
investments after receiving authorization from the local congress. Most
loans are collateralized with participaciones, although other revenue flows
can be used, especially for loans to revenue-generating public enterprises.
 Stabilizing Intergovernmental Transfers 245

Before 2000, in case of arrears or default, the federal government would

deduct debt service payments from revenue sharing before the funds were
transferred to states. Municipalities could incur debt, but with state guaran-
tee. For participaciones to be used as collateral, states merely needed to
register the new debt contract with the Secretariat of Finance after receiving
authorization from the local legislature.
To induce market discipline in subnational borrowing, the law was
nominally reformed in 1997, imposing new restrictions on state and local
governments; these took effect in 2000. Subnational governments could
borrow to finance their investment projects, and many still use their federal
transfers as collateral. However, banks could not ask the Secretariat to
discount the corresponding amount from defaulting on the state’s federal
transfers. They had to arrange the collateral according to state debt laws; that
is, both parties had to create a repayment mechanism. In addition, states
were obliged to publish debt statistics, and secure credit ratings from two
international rating agencies. Without guarantees from the federal level,
banks had to begin to evaluate the risk of investment projects.

Alternative stabilization arrangements

With no federal guarantees for the transfers, Mexican states made alterna-
tive arrangements, especially since some state governors represented different
parties than the president (and thus could not count on federal assistance)
and since discretionary transfers and debt bailouts ended. One arrangement
is that the aportaciones adjust automatically to reflect changes in the salaries
of the formerly federal teaching force. While this has some advantages in
protecting states from a cost-side shock, the negotiation of teachers’ posi-
tions (and work rules) at the federal level deprives the states of the control
over this large component of their workforce. Also, the federally negotiated
teachers’ salary has spillover effects on the other salaries that the state must
pay without automatic federal compensation.
To compensate for their restricted access to credit, some states requiring
sharp adjustment thereafter used arrears or floating debt for financing
deficits, but this tactic reached its limit in 2000. However, since the 1995
experience, several states have followed a much more prudent route of run-
ning fiscal surpluses to build reserves that can cover fluctuations of revenue
and outlays during the year and sometimes even during national recessions.
These rainy-day funds do not have explicit deposit and withdrawal rules as
in the US states.

Assessment

Argentina represents a particularly important case, different from other

countries such as Brazil, where most transfers are automatic, set as a per-
centage of federal revenues, without any stabilizing rule. The intention of
246 Christian Y. Gonzalez et al.

establishing a moving average of shared revenue for federal transfers to

provinces has a fundamental logic. In fact, this approach would partially
insure provincial revenues during prolonged downturns (fully insure them
for sudden midyear sharp downturns), and would allow provinces to adjust
gradually to persistent external economic shocks. However, setting fixed
floors for transfers led to serious complications for federal fiscal management
in the second half of 2001.
It might be instructive to return to the initial conditions for setting
stabilization rules for intergovernmental transfers discussed in the introduc-
tion.14 The first condition concerned subnational access to, or cost of,
market borrowing. At the start of the 1990s, provinces had a limited credit
history and limited experience in accessing domestic or international credit
markets. Even the better-performing provinces consistently faced higher
interest rates than the federal government. As for the second condition,
although debt management at the federal level has technically been efficient,
access to credit has been sporadic over the last decade. The inability to bal-
ance the budget resulted in an eventual default by the federal government.
The third condition on the absence of structural fiscal imbalances holds nei-
ther within nor across levels of government, as deficits were persistent at
both levels of government throughout the period, and most provinces
remained overly dependent on federal transfers.
Several key lessons emerge from the recent Argentine experience. One is
that full insurance against the downturn is particularly risky if automatic
transfers comprise a significant share of GDP (6 percentage points in the
Argentine case). Another lesson is that long-term federal–provincial fiscal
imbalances need to be addressed prior to moving toward a system of
stabilizing intergovernmental transfers. Finally, guaranteed minimum
amounts, without escape clauses, are particularly dangerous in the face of an
unstable federal fiscal situation and output volatility.
Although without having experienced severe long-term structural imbal-
ances, Colombia was in the process of establishing the structure of fiscal
federalism during the 1990s. The fundamental alignment of expenditure
and revenue across levels of government was in a state of adjustment, perhaps
making guaranteed transfers somewhat risky.
In 2001, the national government imposed some fiscal rules on the depart-
ments and municipalities, requiring that they reduce the share of wages in
their spending, to increase flexibility, and offering debt rescheduling in
return for fiscal adjustment. The effect of the law will become clear only after
a few years, as it might reduce the likelihood of another round of subna-
tional debt bailouts, or alternatively, it might set a precedent for some
entities to overborrow again with the expectation of a bailout.
On its part, to avoid unsustainable deficits in a downturn, the national
government will need to treat the guarantee obligations like contingent debt,
provisioning for them with a build-up of reserves (including accelerated
 Stabilizing Intergovernmental Transfers 247

amortization to give more headroom with creditors), and preparing to make

necessary fiscal adjustment.
In Mexico, given the difficulty of the national government to stabilize its
own revenue and to meet outstanding contingent liabilities for financial
sector restructuring and public pension reform, guaranteed floors for
revenue-sharing or other transfers to states are probably not advisable. This
requires incentives for the states to exercise precaution, which may help
Mexico to avoid the macroeconomic instability suffered in the past. Since
2000, the new regime for limiting subnational borrowing seems to be off to
a good start. The current transfer system seems to be a relatively sound
way to shield subnational governments partly but not totally from exter-
nal shocks. The system guarantees the states access to a stable share of national
revenue, which gives them unconditional funds for almost half of their total
needs. It also guarantees that education and health transfers, which cover
most of the cost in those sectors, will adjust according to the wage agree-
ments made at the national level. While serious problems remain with the
division of management responsibilities in the social sectors, the system of
mixed guarantees seems sound, giving the states incentives to increase their
own revenues by taking advantage of the sales tax recently approved by
Congress and to create budget stabilization funds with explicit rules.

Conclusion

Floors for subnational transfers are tantamount to a kind of contingent debt

of the central government, though rarely evaluated as such. While there are
good arguments for the central government to make such guarantees,
because the subnational governments are more credit constrained at the
margin, there is a dangerous tendency to overuse such floors in circum-
stances when the national government has little else to offer in intergov-
ernmental bargains.
The moving-average approach for determining transfers reduces the con-
tingent liability problem in that usually the federal government should
break even relative to a fixed revenue-sharing system. Still, there remain
problems with the timing of the start of such a system and the financing of
a stabilization fund, since there could be an unexpected recession in the ini-
tial years. In addition, the federal level needs effective fiscal rules and sav-
ings plans in order not to waste the surplus years.
The scale and flexibility of transfers seems to be critical for which systems
are fiscally safe. If transfers to subnational governments are small in terms
of overall public finances, either because subnational spending responsibil-
ities are small or because they are mostly funded with their own revenues,
then the national government may have the fiscal room and credit access to
guarantee a floor for transfers. If subnational governments have major
spending responsibilities that are mostly funded with transfers, however,
248 Christian Y. Gonzalez et al.

then the public sector will find it too costly to guarantee fully that these
transfers will not be affected by adverse economic shocks.
There are at least four types of arrangements for subnational governments
to protect themselves with a cushion against macroeconomic shocks, and
thus share in the risk of fiscal downturns. First, subnational governments
may maintain some margin to increase their revenue. The political as well
as economic feasibility of the option to increase their own revenue is rela-
tively great in Brazil, at least for the large states; perhaps this is partly because
there never was a federal promise of minimum floors to shared revenue.
Second, they may keep spending flexibility, especially with a deferrable
investment program, without allowing close to 99 percent of revenue to
finance wages and debt service. Well-managed subnational governments
everywhere do this, but they are few. The fiscal responsibility legislation of
Brazil addresses this issue explicitly. Third, they may build a state reserve
fund. At least a few states in Mexico do this, as they have little opportunity
to pursue other alternatives. Fourth, they may establish a secure credit line,
available in times of fiscal distress. Actually, no subnational governments in
our sample have achieved this, except to the extent they can run arrears
(some jurisdictions do this in all three countries) or partially default.
Although none of these options provides a large cushion, together they
can add up to a significant amount so that not all the adjustment burden
needs to fall on the central government. The experience in the three coun-
tries shows that subnational governments have the motivation to develop
and utilize these alternatives only if the easier option of a full federal guar-
antee is not available. Some rules-based burden sharing of the risk of fiscal
shocks seems clearly preferable to an open guarantee that is sure to fail in
extreme circumstances, inflicting damage on the public finance framework.

Notes

1. The authors gratefully acknowledge the comments on country sections from Bill
Dillinger, Joachim von Amsberg, and Zeinab Partow, as well as general comments
from the discussant, Fernando Elizondo. Finally, George Kopits provided invalu-
able advice on revisions of this chapter.
2. Stein et al. (1999) document the procyclical nature of fiscal policy in Latin America
and offer institutional explanations for these results at the national level.
3. For the case of Brazil, see Gonzalez et al. (2002).
4. Kopits (2001a) suggests that a well-designed transfer system, closing vertical imbal-
ances, is a necessary condition for the successful implementation of fiscal rules at
the subnational level.
5. See Tommasi et al. (2001) for a general theory of this incentive problem (and other
incentive problems) with an application to the Argentine case. Also see Sanguinetti
and Tommasi (1998).
6. Provincial budget accounts are on an accrual basis, while federal accounts are
recorded on a cash basis in the historical data of these figures. Provincial public
sector spending varies from about 8 to 35 percent of local GDP.
 Stabilizing Intergovernmental Transfers 249

7. See World Bank (1996), IADB (1997), Schwartz and Liuksila (1997) and Tommasi
et al. (2001).
8. In 2002, the federal government adopted an export tax that would not be shared
with the provinces.
9. It also opened the way for the provinces to transfer their employee pensions to
the national system.
10. It is not clear what the federal government would do with the expected savings
from the lower transfers. A fiscal stabilization fund that would lock up the savings
during the upswing was discussed, but no implementation date established.
Depending upon what growth rates one assumes, over the five-year period, the
provinces would lose transfers that they would otherwise have received (World
Bank 2001).
11. Cyclically induced deficits in the early 1980s were quickly corrected.
12. According to Law 60, 15 percent of the situado is uniformly distributed to each
department and district, and the remaining 85 percent is distributed by a formula
taking into account the current number of students enrolled, the number of
school-age children not attending school, and the number of actual and potential
patients seen by health units.
13. However, there is an escape clause on the ceiling. If GDP growth exceeds 4 percent,
the ceiling will no longer apply and transfers will increase in proportion to
national revenue growth.
14. By comparison, in the United States, there is no general automatic transfer system
(most transfers are either discretionary financing of public works or cofinancing
of social programs like welfare and health care). In Canada, equalization transfers
provide automatic insurance, but asymmetrically: provinces with relatively low
per capita tax bases – and hence not part of the equalization standard – receive
additional funds during the downturn (Courchene 1999).
References

Acharya, S. (2001), “Macroeconomic Management in the Nineties,” New Delhi:

Indian Council of Research on International Economic Relations. Available via the
Internet: http://www.icrier.res.in/pdf/Book1.pdf.
Aghion, P., P. Bacchetta, and A. Banerjee (2001), “A Corporate Balance-Sheet Approach
to Currency Crises,” CEPR Discussion Paper No. 3092, London: Centre for
Economic Policy Research.
Ahluwalia, M. (2002a), “Economic Reforms in India Since 1991: Has Gradualism
Worked?,” Journal of Economic Perspectives, Vol. 16 (Summer), pp. 67–88.
Ahluwalia, M. (2002b), “State-Level Performance Under Economic Reforms in India,”
in A.O. Krueger, Economic Policy Reforms and the Indian Economy, Chicago: University
of Chicago Press.
Albuquerque, P.H. (2001), “Os impactos economicos da CPMF: teoria e evidencia,”
Trabalhos para Discussao 21, Brasilia: Banco Central do Brasil.
Alesina, A. and S. Ardagna (1998), “Tales of Fiscal Adjustment,” Economic Policy,
Vol. 27 (October), pp. 487–546.
Alesina, A. and A. Drazen (1991), “Why are Stabilizations Delayed?” American
Economic Review, Vol. 81 (December), pp. 1170–88.
Alesina, A., R. Hausmann, R. Hommes and E. Stein (1999), “Budget Institutions and
Fiscal Performance in Latin America,” Journal of Development Economics, Vol. 59
(August), pp. 253–73.
Alesina, A. and R. Perotti (1995), “The Political Economy of Budget Deficits,” Staff
Papers, International Monetary Fund, Vol. 42 (March), pp. 1–31.
Alesina, A. and R. Perotti (1996), “Fiscal Discipline and the Budget Process,” American
Economic Review, Papers and Proceedings, Vol. 86 (May), pp. 401–7.
Allan, W. and T. Parry (2003), “Fiscal Transparency in EU Accession Countries:
Progress and Future Challenges,” IMF Working Paper 03/163, Washington:
International Monetary Fund.
Allsop, C. and M.J. Artis (eds) (2003), Oxford Review of Economic Policy: EMU, Four Years
On, Vol. 19 (Spring), pp. 1–213.
Alt, J.E. and R.C. Lowry (1994), “Divided Government, Fiscal Institutions, and Budget
Deficits: Evidence from the States,” American Political Science Review, Vol. 88
(December), pp. 811–28.
Asian Development Bank, Inter-American Development Bank, International
Monetary Fund and World Bank (2001), Social Safety Nets in Response to Crisis:
Lessons and Guidelines from Asia and Latin America, prepared for the Asia-Pacific
Economic Cooperation (APEC) Finance Ministers’ Social Safety Net Initiative.
Available via the Internet: http://www.iadb.org/sds/doc/APECreport.pdf.
Aslaksen, I. and O. Bjerkholt (1986), “Certainty Equivalence Methods in the
Macroeconomic Management of Petroleum Resources,” in J.P. Neary and S. van
Wijnbergen (eds), Natural Resources and the Macroeconomy, Oxford: Centre for
Economic Policy Research and Basil Blackwell.
Aslaksen, I., K.A. Brekke, T.A. Johnsen, and A. Aaheim (1990), “Petroleum Resources and
the Management of National Wealth,” in O. Bjerkholt, Ø. Olsen, and J. Vislie (eds),
Recent Modeling Approaches in Applied Energy Economics, London: Chapman and Hall.

250
 References 251

Aziz, J., F. Caramazza, and R. Salgado (2000), “Currency Crises: In Search of Common
Elements,” in World Economic Outlook: Supporting Studies, Washington: International
Monetary Fund.
Bacha, E. (1994), “O fisco e a inflação: uma interpretação do caso brasileiro,” Revista
de Economia Política, Vol. 14 ( January/March), pp. 5–17.
Balassone, F. and D. Franco (2000), “Public Investment, the Stability Pact and the
‘Golden Rule,’” Fiscal Studies, Vol. 21 ( June), pp. 207–29.
Balassone, F. and D. Franco (2001), “Fiscal Federalism and the Stability and Growth
Pact: A Difficult Union,” in Fiscal Rules: Papers Presented at the Bank of Italy Workshop
held in Perugia, 1–3 February 2001, Rome: Banca d’Italia.
Balassone, F. and D. Monacelli (2000), “EMU Fiscal Rules: Is There a Gap?,” Banca
d’Italia Temi di Discussione No. 375, Rome: Banca d’Italia.
Balassone, F., D. Franco, and S. Zotteri (2003), “Fiscal Rules for Subnational
Governments in the EMU Context,” SIEP Working Paper, Pavia, Italy: Società Italiana
di Economia Pubblica. Available via the Internet: http://www.unipv.it/website
Banca d’Italia (ed.) (2001), Fiscal Rules: Papers Presented at the Bank of Italy Workshop
Held in Perugia, 1–3 February 2001, Rome: Banca d’Italia.
Barnhill, T. and G. Kopits (2004), “Assessing Fiscal Sustainability under Uncertainty,”
Journal of Risk, Vol. 6 (Summer), pp. 31–53.
Barrell, R. and K. Dury (2001), “Will the SGP Ever Be Breached? ” in A. Brunila, M. Buti,
and D. Franco (eds), The Stability and Growth Pact: The Architecture of Fiscal Policy in
EMU, New York: Palgrave.
Barro, R.J. (1977), “Unanticipated Money Growth and Unemployment in the United
States,” American Economic Review, Vol. 67 (March), pp. 101–15.
Barro, R.J. (1979), “On the Determination of the Public Debt,” Journal of Political
Economy, Vol. 87 (October), pp. 940–71.
Barro, R.J. and D.B. Gordon (1983a), “Rules, Discretion and Reputation in a Model of
Monetary Policy,” Journal of Monetary Economics, Vol. 12 ( July), pp. 101–21.
Barro, R.J. and D.B. Gordon (1983b), “A Positive Theory of Monetary Policy
in a Natural Rate Model,” Journal of Political Economy, Vol. 91 (August),
pp. 589–610.
Begg, D. (1998), “Pegging Out: Lessons from the Czech Exchange Rate Crisis,” CEPR
Discussion Paper 1956, London: Center for Economic Policy Research.
Bertola, G. and A. Drazen (1993), “Trigger Points and Budget Cuts: Explaining
the Effects of Fiscal Austerity,” American Economic Review, Vol. 83 (March),
pp. 11–26.
Besley, T. and A. Case (2003), “Political Institutions and Policy Choices: Evidence from
the United States,” Journal of Economic Literature, Vol. 41 (March), pp. 7–73.
Bevilaqua, A.S. (2002), “State Government Bailouts in Brazil,” IDB Working Paper
No. R-441, Washington: Inter-American Development Bank.
Bird, R.M. (1984), Intergovernmental Finance in Colombia: Final Report of the Mission on
Intergovernmental Finance, Cambridge, MA: Harvard Law School International Tax
Program.
Blanchard, O., J.-C. Chouraqui, R. Hagemann, and N. Sartor (1990), “The Sustainability
of Fiscal Policy: New Answers to an Old Question: Analytical and Methodological
Questions,” OECD Economic Studies, Vol. 15 (Autumn), pp. 7–36.
Blondal, J. (1999), “Fiscal Federalism in OECD Member Countries,” in K. Fukusaku
and L.R. de Mello, Jr. (eds), Fiscal Decentralization in Emerging Economies: Governance
Issues, Paris: Organization for Economic Cooperation and Development.
252 References

Bohn, H. and R.P. Inman (1996), “Balanced-Budget Rules and Public Deficits: Evidence
from the U.S. States,” Carnegie – Rochester Conference Series on Public Policy, Vol. 45
(December), pp. 13–76.
Brailovsky, V. (1981), “Industrialization and Oil in Mexico: A Long Term Perspective,”
in T. Barker and V. Brailovsky (eds), Oil or Industry? Energy, Industrialization and
Economic Policy in Canada, Mexico, the Netherlands, Norway and the United Kingdom,
London: Academic Press.
Braun, M. and M. Tommasi (2002), “Fiscal Rules for Subnational Governments: Some
Organizing Principles and Latin American Experiences,” CEDI Working Paper No. 78,
Buenos Aires: Centro de Estudios para el Desarrollo Institucional.
Brunila, A. (2002), “Fiscal Policy: Coordination, Discipline and Stabilisation,” Bank of
Finland Discussion Paper No. 7, Helsinki: Bank of Finland.
Brunila, A., M. Buti and D. Franco (eds) (2001), The Stability and Growth Pact: The
Architecture of Fiscal Policy in EMU, New York: Palgrave.
Buchanan, J. and R. Wagner (1977), Democracy in Deficit: The Political Legacy of Lord
Keynes, New York: Academic Press.
Buiter, W. H. (1985), “A Guide to Public Sector Debt and Deficits,” Economic Policy, Vol.
1 (November), pp. 13–79.
Buiter, W.H. and C. Grafe (2002), “Patching up the Pact: Some Suggestions for
Enhancing Fiscal Sustainability in an Enlarged European Union,” CEPR Discussion
Paper No. 3496, London: Centre for Economic Policy Research.
Burnside, C., M. Eichenbaum, and S. Rebelo (1999), “Prospective Deficits and the Asian
Currency Crisis,” Policy Research Working Paper 2174, Washington: World Bank.
Buti, M. (ed.) (2003), Monetary and Fiscal Policies in EMU – Interactions and Coordination,
Cambridge, UK: Cambridge University Press.
Buti, M. and G. Giudice (2002), “Maastricht Fiscal Rules at Ten: An Assessment,”
Journal of Common Market Studies, Vol. 40 (5), pp. 823–48.
Buti, M. and A. Sapir (eds) (1998), Economic Policy in EMU: A Study by the European
Commission Services, Oxford: Clarendon Press.
Buti, M. and A. Sapir (eds) (2002), EMU and Economic Policy in Europe: The Challenge of
the Early Years, Northampton, MA: Edward Elgar.
Buti, M. and P. van den Noord (2003), “Discretional Fiscal Policy and Elections: the
Experience of the Early Years of EMU,” OECD Working Paper No. 351, Paris:
Organisation for Economic Co-operation and Development.
Buti, M., S. Eijffinger, and D. Franco (2003), “Revisiting the Stability and Growth Pact:
Grand Design or Internal Adjustment?,” CEPR Discussion Paper No. 3692, London:
Centre for Economic Policy. Available via the Internet at www.cepr.org/pubs/dps/
DP3692.asp.
Buti, M., D. Franco, and H. Ongena (1997), “Budgetary Policies During Recessions.
Retrospective Application of the Stability and Growth Pact to the Post-War Period,”
Economic Papers, No. 121, Brussels: European Commission.
Buti, M., D. Franco, and H. Ongena (1998), “Fiscal Discipline and Flexibility in EMU:
The Implementation of the Stability and Growth Pact,” Oxford Review of Economic
Policy, Vol. 14 (Autumn), pp. 81–97.
Buti, M., J. von Hagen, and C. Martinez-Mongay (eds) (2002), The Behaviour of Fiscal
Authorities: Stabilization, Growth, and Institutions, New York: Palgrave.
Caballero, R.J. (2000), “Macroeconomic Volatility in Latin America: A View and Three
Case Studies,” NBER Working Paper No. 7782, Cambridge, MA: National Bureau of
Economic Research.
 References 253

Cabral, A.J. (2001), “Main Aspects of the Working of the SGP,” in A. Brunila, M. Buti,
and D. Franco (eds), The Stability and Growth Pact: The Architecture of Fiscal Policy in
EMU, Basingstoke, Hampshire; New York: Palgrave.
Calvert, R.L. (1995), “The Rational Choice Theory of Social Institutions: Cooperation,
Coordination, and Communication,” in J.S. Banks and E.A. Hanushek (eds), Modern
Political Economy: Old Topics, New Directions, Cambridge: Cambridge University
Press.
Calvo, G. (1994), “Varieties of Capital-Market Crises,” IDB Working Paper No. 306,
Washington: Inter-American Development Bank.
Calvo, G. (1997), “The Unforgiving Market and the Tequilazo,” in J.M. Fanelli and
R. Medhora (eds), Financial Reform in Developing Countries, New York: Palgrave
Macmillan.
Calvo, G. and F.S. Mishkin (2003), “The Mirage of Exchange Rate Regimes for
Emerging Market Economies,” NBER Working Paper No. 9808, Cambridge, MA:
National Bureau of Economic Research.
Calvo, G. and C. Reinhart (2002), “Fear of Floating,” Quarterly Journal of Economics,
Vol. 117 (May), pp. 379–408.
Cappelen, Å. and E. Gjelsvik (1990), “Oil and Gas Revenues and the Norwegian
Economy in Retrospect: Alternative Macroeconomic Policies,” in O. Bjerkholt,
Ø. Olsen, and J. Vislie (eds), Recent Modeling Approaches in Applied Energy Economics,
London: Chapman and Hall.
Cardoso, E. (1998), “Virtual Deficits and the Patinkin Effect,” IMF Working Paper 98/41,
Washington: International Monetary Fund.
Casella, A. (1999), “Tradable Deficit Permits: Efficient Implementation of the Stability
Pact in the European Monetary Union,” Economic Policy, Vol. 14 (October),
pp. 321–62.
Cashin, P., H. Liang, and J. McDermott (2000), “How Persistent are Shocks to World
Commodity Prices?,” IMF Staff Papers, Vol. 47 (December). Available via the Internet:
http://www.imf.org/external/Pubs/FT/staffp/2000/00–02/cashin.htm.
Catão, L. and B. Sutton (2002), “Sovereign Defaults: The Role of Volatility,” IMF
Working Paper 02/149, Washington: International Monetary Fund.
Chamon, M. (2001), “Why Can’t Developing Countries Borrow from Abroad in their
Own Currency?” Harvard University, unpublished paper.
Chamon, M. and R. Hausmann (2002), “Why Do Countries Borrow the Way They
Borrow? ” Harvard University, unpublished paper.
Chari, V.V. and H.L. Cole (1993), “A Contribution to the Theory of Pork Barrel
Spending,” Federal Reserve Bank of Minneapolis Staff Report No. 156, Minneapolis:
Federal Reserve Bank.
Coase, R. (1960), “The Problem of Social Costs,” Journal of Law and Economics, Vol. 3,
pp. 1–44.
Coelho, I., L. Ebrill and V. Summers (2001), “Bank Debit Taxes in Latin America: An
Analysis of Recent Trends,” IMF Working Paper 01/67, Washington: International
Monetary Fund.
Corden, W.M. (1984), “Booming Sector and Dutch Disease Economics: Survey and
Consolidation,” Oxford Economic Papers, Vol. 36 (November), pp. 359–80.
Coricelli, F. (2002), “Shocks, Volatility and Growth in Economies in Transition,”
University of Siena, unpublished paper.
Corsetti, G. and B. Mancowizk (2002), “Nominal Debt and Currency Crises,” Yale
University, unpublished paper.
254 References

Corsetti, G. and N. Roubini (1996), “European Versus American Perspectives on

Balanced-Budget Rules,” American Economic Review, Papers and Proceedings, Vol. 86
(May), pp. 408–13.
Corsetti, G., P. Pesenti, and N. Roubini (1999), “Paper Tigers? A Model of the Asian
Crisis,” European Economic Review, Vol. 43 ( June), pp. 1211–36.
Costello, D. (2001), “The SGP: How Did We Get There?,” in A. Brunila, M. Buti, and
D. Franco (eds), The Stability and Growth Pact: The Architecture of Fiscal Policy in EMU,
Basingstoke, Hampshire; New York: Palgrave.
Courchene, T.J. (1999), “Subnational Budgetary and Stabilization Policies in Canada
and Australia,” in J.M. Poterba and J. von Hagen (eds), Fiscal Institutions and Fiscal
Performance, Chicago: University of Chicago Press.
Craig, J. and A. Manoel (2002), “Budget Management Systems in Anglo-Saxon and
Latin American Countries: A Comparative Assessment,” Washington: International
Monetary Fund, unpublished.
Csajbók, A. and Á. Csermely (eds) (2002), “Adopting the Euro in Hungary: Expected
Costs, Benefits and Timing,” Occasional Paper No. 24, Budapest: National Bank of
Hungary.
Cuddington, J. (1992), “Long-Run Trends in 26 Primary Commodity Prices: A
Disaggregated Look at the Prebisch-Singer Hypothesis,” Journal of Development
Economics, Vol. 39 (October), pp. 207–27.
Cukierman, A. and A.H. Meltzer (1986), “A Positive Theory of Discretionary Policy,
the Cost of Democratic Government, and the Benefits of a Constitution,” Economic
Inquiry, Vol. 24 ( July), pp. 367–88.
Daniel, J. and M. Saal (1998), “Macroeconomic Impact and Policy Response,” in
W.E. Alexander, J. Davis, L. Ebrill, and C. Lindgren (eds), Systemic Bank Restructuring
and Macroeconomic Policy, Washington: International Monetary Fund.
Darvas, Z. and A. Simon (2000), “Potential Output and Trade in a Small
Open Economy,” NBH Working Paper No. 2000/9, Budapest: National Bank of
Hungary.
Davis, J., R. Ossowski, J. Daniel, and S. Barnett (2001), “Stabilization and Savings
Funds for Nonrenewable Resources: Experience and Fiscal Policy Implications,” IMF
Occasional Paper No. 205, Washington: International Monetary Fund.
De Ferranti, D., G. Perry, I. Gill, and L. Servén (2000), Securing Our Future in a Global
Economy, Washington: World Bank.
de Haan, J., J.E. Sturm, and B.J. Sikken (1996), “Government Capital Formation:
Explaining the Decline,” Weltwirtschaftliches Archiv, Vol. 132 (1), pp. 55–74.
Detragiache, E. and A. Spilimbergo (2001), “Crises and Liquidity: Evidence and
Interpretation,” IMF Working Paper 01/2, Washington: International Monetary Fund.
Dillinger, W. (1997), “Brazil’s State Debt Crisis: Lessons Learned,” Departmental
Working Paper No. 17430, Washington: World Bank.
Dixit, A. K. (1996), The Making of Economic Policy. A Transaction-Cost Politics Perspective,
Cambridge, MA: MIT Press.
Dooley, M.P. (1998), “A Model of Crises in Emerging Markets,” International Finance
Discussion Paper No. 630, Washington: Board of Governors of the Federal Reserve
System.
Drazen, A. (2000), Political Economy in Macroeconomics, Princeton, NJ: Princeton
University Press.
Drazen, A. (2001), “The Political Business Cycle After 25 Years,” NBER Macroeconomics
Annual 2000, Cambridge MA: MIT Press.
 References 255

Drazen, A. and P.R. Masson (1994), “Credibility of Policies Versus Credibility of

Policymakers,” Quarterly Journal of Economics, Vol. 109 (August), pp. 735–54.
Dziobek, C. and C. Pazarbasioglu (1998), “Lessons and Elements of Best Practice,” in
W.E. Alexander, J. Davis, L. Ebrill, and C. Lindgren (eds), Systemic Bank Restructuring
and Macroeconomic Policy, Washington: International Monetary Fund.
Easterly, W. and S. Rebelo (1993), “Fiscal Policy and Economic Growth: An Empirical
Investigation,” Journal of Monetary Economics, Vol. 32 (3), pp. 417–58.
Eaton, J. and M. Gersovitz (1981), “Debt and Potential Repudiation: Theoretical and
Empirical Analysis,” Review of Economic Studies, Vol. 48 (April), pp. 289–309.
Echavarría J., C. Renteria, and R. Steiner (2000), “Decentralization and Bailouts
in Colombia,” Washington: Inter-American Development Bank, unpublished
paper.
Eichengreen, B. (1994), “Fiscal Policy and EMU,” in B. Eichengreen and J.A. Frieden
(eds), The Political Economy of European Monetary Unification, Boulder, CO: Westview.
Eichengreen, B. and R. Hausmann (1999), “Exchange Rates and Financial Fragility,”
NBER Working Paper No. 7418, Cambridge, MA: National Bureau of Economic
Research.
Eichengreen, B. and J. von Hagen (1996), “Federalism, Fiscal Restraints, and European
Monetary Union,” American Economic Review, Vol. 86 (May), pp. 134–8.
Eichengreen, B. and J. von Hagen (2000), “Fiscal Policy and Monetary Union: Is There
a Tradeoff Between Federalism and Budgetary Restrictions? ” NBER Working Paper
No. 5517, Cambridge, MA: National Bureau of Economic Research.
Eichengreen, B. and C. Wyplosz (1998), “Stability Pact: More Than a Minor
Nuisance?,” Economic Policy, Vol. 13 (April), pp. 67–104.
Eichengreen, B., R. Hausmann, and U. Panizza (2003a) “Original Sin. The Pain, The
Mystery, and the Road to Redemption,” Harvard University, unpublished paper.
Eichengreen, B., R. Hausmann, and U. Panizza (2003b), “Currency Mismatches, Debt
Intolerance and Original Sin: Why They Are Not the Same and Why It Matters,”
NBER Working Paper No. 10036, Cambridge, MA: National Bureau of Economic
Research. Available via the Internet: http://www.nber.org/papers/w10036.
Eichengreen, B., R. Hausmann, and J. von Hagen (1999), “Reforming Budgetary
Institutions in Latin America: The Case for a National Fiscal Council,” Open
Economies Review, Vol. 10 (October), pp. 415–42.
Einaudi, L. (1948), Princìpi di scienza della finanza, Torino: Boringhieri.
Elster, J. (1989), The Cement of Society: A Study of Social Order, Cambridge: Cambridge
University Press.
Elster, J. (1995), “The Impact of Constitutions on Economic Performance,” in
Proceedings of the World Bank Annual Conference on Development Economics 1994,
Washington: World Bank, pp. 209–39.
Engel, E. and R.O. Valdés (2000), “Optimal Fiscal Strategy for Oil Exporting
Countries,” IMF Working Paper 00/118, Washington: International Monetary Fund.
European Bank for Reconstruction and Development (EBRD) (2001), Transition Report
2001: Energy in Transition, London.
European Commission (1995), “Technical Note, Table 5: Calculation of the Cyclically
Adjusted Government Budget Balance – 1994,” European Economy, No. 60, p. 47.
European Commission (2000), Public Finances in EMU – 2000, European Economy,
Reports and Studies No. 3, Luxembourg: European Communities.
European Commission (2001a), Public Finances in EMU – 2001, European Economy,
Reports and Studies No. 3, Luxembourg: European Communities.
256 References

European Commission (2001b), “Autumn 2001 Forecasts for 2001 – 2003,” European
Economy, Supplement A, Economic Trends, No. 10/11, Luxembourg: European
Communities.
European Commission (2002a), Public Finances in EMU – 2002, European Economy, Reports
and Studies No. 3, Luxembourg: European Communities. Available via the Internet:
http://europa.eu.int/comm/economy_finance/publications/publicfinance_en.htm.
European Commission (2002b), “Making a Success of Enlargement,” Strategy Paper
and Report of the European Commission on the Progress Towards Accession by Each
of the Candidate Countries, Luxembourg: European Communities.
European Commission (2002c), “Strengthening the Co-ordination of Budgetary
Policies,” Communication to the Council and the European Parliament, COM
(2002)668, 27 November. Available via the Internet: http://europa.eu.int/comm/
economy_finance/publications/sgp/com2002668_en.htm.
Fatas, A. and I. Mihov (2003), “The Case for Restricting Fiscal Policy Discretion,”
Quarterly Journal of Economics, Vol. 118 (November), pp. 1419–47.
Favero, C. and F. Giavazzi (2002), “Why are Brazil’s Interest Rates so High? ” Milan:
IGIER Universita Bocconi (unpublished).
Fiess, N. (2002), “Chile’s New Fiscal Rule,” World Bank, unpublished paper.
Fischer, J. (2001), “National and EU Budgetary Rules and Procedures: An Evolving
Interaction,” in Fiscal Rules: Workshop Papers Presented at the Bank of Italy Workshop
Held in Perugia, 1–3 February 2001, Rome: Banca d’Italia.
Fischer, J. and G. Giudice (2001), “The Stability and Convergence Programmes,” in
A. Brunila, M. Buti, and D. Franco (eds), The Stability and Growth Pact: The
Architecture of Fiscal Policy in EMU, Basingstoke, Hampshire; New York: Palgrave.
Fischer, S. (2000), “Prospects for Latin America in the 2000s,” opening address at the
Latin American and Caribbean Economic Association (LACEA) Conference 2000,
Rio de Janeiro, October.
Fischer, S. (2001a), “Exchange Rate Regimes: Is the Bipolar View Correct?”
Washington: International Monetary Fund, unpublished paper.
Fischer, S. (2001b), “Farewell to the IMF Executive Board,” address given on August 30,
2001. Available via the Internet: http://www.imf.org/external/np/speeches/2001/
083001.htm.
Fitoussi, J.P. and J. Creel (2002), How to Reform the European Central Bank, London:
Centre for European Reform.
Friedman, D. and J. Rust (eds) (1993), The Double Auction Market: Institutions, Theories,
and Evidence, SFI Studies in the Sciences of Complexity, New York: Addison-Wesley.
Flood, R.P. and N.P. Marion (1998), “Perspectives on the Recent Currency Crisis
Literature,” NBER Working Paper No. 6380, Cambridge, MA: National Bureau of
Economic Research.
Gamboa, R. (1997), “El rescate financiero de los gobiernos estatales por el gobierno
federal: Comparación de los casos de Estados Unidos, Brasil y Mexico,” Mexico City:
Bank of Mexico, unpublished paper.
Gavin, M. and R. Hausmann (1999), “Preventing Crisis and Contagion: Fiscal and
Financial Dimensions,” IDB Working Paper No. 401, Washington: Inter-American
Development Bank.
Gavin, M. and R. Perotti (1996), “Fiscal Policy in Latin America: Volatility, Cyclicality
and Modes of Adjustment,” OCE Working Paper, Washington: Inter-American
Development Bank.
Gavin, M., R. Hausmann, and L. Leiderman (1995), “The Macroeconomics of Capital
Flows to Latin America: Experience and Policy Issues,” IDB Working Paper No. 310,
Washington: Inter-American Development Bank.
 References 257

Gavin, M., R. Hausmann, R. Perotti, and E. Talvi (1996), “Managing Fiscal Policy in Latin
America and the Caribbean: Volatility, Procyclicality and Limited Creditworthiness,”
IDB Working Paper No. 326, Washington: Inter-American Development Bank.
Gelb, A.H. and associates (1988), Oil Windfalls: Blessing or Curse?, New York: Oxford
University Press for the World Bank.
Ghosh, A., T. Lane, M. Schulze-Ghattas, A. Bulíř, J. Hamann, and A. Mourmouras
(2002), IMF-Supported Programs in Capital Account Crises, Occasional Paper No. 210,
Washington: International Monetary Fund.
Ghymers, C. (2001), “Macroeconomics and Regional Integration: EMU and the
Stability Programmes. Some Possible Lessons for Latin America,” paper presented at
the ECLAC XIII Regional Seminar on Fiscal Policy, Santiago, Chile, January.
Giavazzi, F. and M. Pagano (1990), “Can Severe Fiscal Contractions Be Expansionary?
Tales of Two Small European Countries,” in O.J. Blanchard and S. Fischer (eds),
NBER Macroeconomics Annual 1990, Cambridge, MA: MIT Press.
Gil-Díaz, F. (1995), “Fiscal Policy and Tax Administration: The Experience of Mexico,”
in Reform of Tax Administration in Latin America: Progress and Perspectives,
Washington: Inter-American Development Bank.
Gil-Díaz, F. (2001), “La propuesta de reforma fiscal para Mexico,” in D. Colmenares (ed.)
El Economista Mexicano: Experiencias de Reformas Fiscales, Vol. 2, pp. 11–28.
Gil-Díaz, F. and W. Thirsk (2000), “La Prolongada Reforma Fiscal en México,” Trimestre
Fiscal (Mexico City, Instituto para el Desarrollo Técnico de las Haciendas Públicas,
INDETEC), Vol. 71 (September–November), pp. 299–327.
Giudice, G. and A. Montanino (2002), “Un Pacte pour la Stabilité et la Croissance
Économique en Europe,” Revue du Marché Commun et de l’Union Européenne, No. 463
(December), pp. 657–60.
Giudice, G. and A. Montanino (2003), “Il Patto di Stabilità e Crescita,” Rivista di
Politica Economica ( July–August), pp. 185–273.
Giudice, G., A. Turrini and J. in’t Veld (2003), “Can Fiscal Consolidations Be
Expansionary in the EU? Ex-Post Evidence and Ex-Ante Analysis,” European
Economy Paper No 195, Luxembourg: European Communities.
Global Insight (2001), Latin America Economic Outlook, 2001Q4.
Goldfajn, I. (2002), “Are There Reasons to Doubt Fiscal Sustainability in Brazil?,”
Technical Note No. 25, Sao Paulo: Central Bank of Brazil. Available via the Internet:
http://www.bcb.gov.br/pec/notastecnicas/ingl/2002nt25fiscalsustainabilityi.pdf.
Goldstein, M. (2003) “Debt Sustainability, Brazil and the IMF,” WP No. 03–1,
Washington: Institute for International Economics. Available via the Internet:
http://www.iie.com/publications/wp/2003/03–1.pdf.
Gonzalez, C.Y., D. Rosenblatt, and S.B. Webb (2002), “Stabilizing Intergovernmental
Transfers in Latin America: A Complement to National/Subnational Fiscal Rules?,”
Policy Research Working Paper 2869, Washington: World Bank.
Gourinchas, P.O., R. Valdés, and O. Landerretche (2001), “Lending Booms: Latin
America and the World,” NBER Working Paper No. 8249, Cambridge, MA: National
Bureau of Economic Research.
Grilli, V., D. Masciandaro, and G. Tabellini (1991), “Political and Monetary
Institutions and Public Financial Policies in the Industrial Countries,” Economic
Policy, Vol. 13 (October), pp. 341–92.
Gylfason, T., T.T. Herbertsson, and G. Zoega (1999), “A Mixed Blessing: Natural Resources
and Economic Growth,” Macroeconomic Dynamics, Vol. 3 ( June), pp. 204–25.
Hagemann, R. (1999) “The Structural Budget Balance: The IMF’s Methodology,” in
Indicators of Structural Balances: Essays Presented at the Bank of Italy Workshop held in
Perugia, 26–28 November 1998, Rome: Bank of Italy.
258 References

Hallerberg, M., R. Strauch, and J. von Hagen (2001), The Use and Effectiveness of
Budgetary Rules and Norms in EU Member States, Amsterdam: Dutch Ministry of
Finance.
Hallerberg, M. and J. von Hagen (1999), “Electoral Institutions, Cabinet Negotiations,
and Budget Deficits in the European Union,” in J.M. Poterba and J. von Hagen
(eds), Fiscal Institutions and Fiscal Performance, Chicago: University of Chicago
Press.
Hausmann, R. and U. Panizza (2003), “The Determinants of Original Sin: An Empirical
Investigation,” Journal of International Money and Finance, Vol. 22 (December),
pp. 957–90.
Hausmann, R., U. Panizza, and E. Stein (2001), “Why Do Countries Float the
Way They Float?,” Journal of Development Economics, Vol. 66 (December),
pp. 387–414.
Hausmann, R., A. Powell, and R. Rigobón (1993), “An Optimal Spending Rule
Facing Oil Income Uncertainty (Venezuela),” in E. Engel and P. Meller (eds), External
Shocks and Stabilization Mechanisms, Washington: Inter-American Development
Bank.
Heller, P.S. (1997), “Fiscal Policy Management in an Open Capital Regime,” IMF
Working Paper 97/20, Washington: International Monetary Fund.
Hemming, R., M.S. Kell, and S. Mahfouz (2002), “The Effectiveness of Fiscal Policy in
Stimulating Economic Activity – A Review of the Literature,” IMF Working Paper
No. 02/208, Washington: International Monetary Fund.
Hemming, R., M.S. Kell, and A. Schimmelpfennig (2003), Fiscal Vulnerability and
Financial Crises in Emerging Market Economies, IMF Occasional Paper No. 218,
Washington: International Monetary Fund.
Hercowitz, Z. and M. Strawczynski (2001), “Cyclical Ratcheting in Government
Spending: Evidence from the OECD,” Discussion Paper Series No. 2001.09, Tel Aviv:
Tel-Aviv University and Research Department, Bank of Israel.
Hernández-Trillo, F. (1995), “A Model-Based Estimation of the Probability of Default
in Sovereign Credit Markets,” Journal of Development Economics, Vol. 46 (February),
pp. 163–79.
Hernández-Trillo, F., A. Díaz Cayeros, and R. Gamboa (2000), “Fiscal Decentralization
in Mexico: The Bailout Problem,” Washington: Inter-American Development Bank,
unpublished paper.
Hodrick, R. and E.C. Prescott (1980), “Post-War US Business Cycles: an Empirical
Investigation,” Discussion Paper No. 451, Pittsburgh, PA: Carnegie Mellon University.
Iaryczower, M., J. Sanguinetti, and M. Tommasi (2000), “Una Propuesta de Reforma al
Federalismo Fiscal Argentino,” CEDI Cuaderno de Opinión No. 8, Buenos Aires:
Centro de Estudios para el Desarrollo Institucional.
India, Ministry of Finance (2001), “An Assessment of Government of India’s
Pensionary Liability,” report of the Working Group on the Assessment of Pensionary
Liability of the Government of India, New Delhi. Available via the Internet:
http://cga.nic.in/pdf/Pen.pdf.
India, Ministry of Finance and Company Affairs (2002a), Report of the Task Force on
Direct Taxes, New Delhi.
India, Ministry of Finance and Company Affairs (2002b), Report of the Task Force on
Indirect Taxes, New Delhi.
Inman, R.P. (1996), “Do Balanced Budget Rules Work? U.S. Experience and Possible
Lessons for the EMU,” NBER Working Paper No. 5838, Cambridge, MA: National
Bureau of Economic Research.
 References 259

Inter-American Development Bank (1995a), “Hacia una Economía Menos Volátil

(Apéndice Técnico),” Informe sobre Progreso Económico y Social en América Latina,
Washington.
Inter-American Development Bank (1995b), Overcoming Volatility in Latin America:
Economic and Social Progress in Latin America: 1995 Report, Baltimore: John Hopkins
University Press.
Inter-American Development Bank (1997), Latin America After a Decade of Reforms,
Report on Economic and Social Progress in Latin America, Washington.
Inter-American Development Bank (1998), “Sustainability of Fiscal Policy in Latin
America,” Washington: IADB Research Network. Available via the Internet:
http://www.iadb.org/RES/index.cfm?fuseaction
LaResNetwork.StudyView
&st_id
51.
International Monetary Fund (1998), World Economic Outlook, Washington.
International Monetary Fund (2001a), Government Finance Statistics Manual,
Washington.
International Monetary Fund (2001b), Brazil: Report on Observance of Standards and Codes
(ROSC) – Fiscal Transparency, Country Report No. 01/217, Washington. Available via
the Internet: http://www.imf.org/external/pubs/ft/scr/2001/cr01217.pdf.
International Monetary Fund (2001c), Norway – 2001 Article IV Consultation, Concluding
Statement of the Mission, Washington. Available via the Internet: http://www.imf.org/
external/np/ms/2001/120601.htm.
International Monetary Fund (2001d), Selected Euro-Area Countries: Rules-Based Fiscal
Policy and Job-Rich Growth in France, Germany, Italy and Spain – Report with
Supplementary Information, prepared by J. Decressin and others, Washington.
International Monetary Fund (2001e), World Economic Outlook, December 2001: A
Survey by the Staff of the International Monetary Fund, World Economic and Financial
Surveys, Washington.
International Monetary Fund (2002a), Mexico: Report on the Observance of Standards and
Code – Fiscal Transparency Module, IMF Country Report No. 02/200 (Washington).
Available via the Internet: http://www.imf.org/external/pubs/ft/scr/2002/cr02200.pdf.
International Monetary Fund (2002b), World Economic Outlook, Washington.
International Monetary Fund (2002c), Assessing Sustainability, Washington. Available
via the Internet: http://www.imf.org/external/np/pdr/sus/2002/eng/052802.pdf.
International Monetary Fund (2002d), “Recent Trends in Growth and Investment,” in
India: Selected Issues and Statistical Appendix, IMF Country Report 02/193,
Washington. Available via the Internet: http://www.imf.org/external/pubs/cat/
longres.cfm?sk
16052.0.
International Monetary Fund, (2003), “State Government Finances,” in India: Selected
Issues and Statistical Appendix, IMF Country Report 03/261, Washington. Available
via the Internet: http://www.imf.org/external/pubs/cat/longres.cfm?sk
16826.0.
International Monetary Fund, Government Finance Statistics Yearbook (GFS),
Washington, various issues.
International Monetary Fund, International Financial Statistics (IFS), Washington,
various issues.
Jeanne, O. (2003), “Why Do Emerging Market Economies Borrow in Foreign
Currency?” IMF Working Paper 03/177, Washington: International Monetary Fund.
Kennedy, S. and J. Robbins (2001), “The Role of Fiscal Rules in Determining Fiscal
Performance,” Canadian Department of Finance Working Paper 2001–16, Ottawa.
King, G. and S. Rebelo (1993), “Low Frequency Filtering and Real Business Cycle,”
Journal of Economic Dynamics and Control, Vol. 17 ( January/March), pp. 207–31.
260 References

Kletzer, K. (1997), “Volatility, External Debt, and Fiscal Risk: Simulations of the Impact
of Shocks on Fiscal Adjustment for Thirteen Latin American Countries,” IDB
Working Paper No. 358, Washington: Inter-American Development Bank.
Kneller, R., M. Bleany, and N. Gemmel (1999), “Fiscal Policy and Growth: Evidence
from OECD Countries,” Journal of Public Economics, Vol. 74 (November), pp. 171–90.
Knight, B. and A. Levinson (1999), “Rainy Day Funds and State Government Savings,”
National Tax Journal, Vol. 52 (3), pp. 459–72.
Kopits, G. (1987), Structural Reform, Stabilization, and Growth in Turkey, IMF Occasional
Paper No. 52, Washington: International Monetary Fund.
Kopits, G. (1999), “Experiencia Internacional con Reglas Macrofiscales: Lecciones para
la Argentina,” paper presented at the Ministerio de Economía, Obras y Servicios
Públicos of Argentina seminar on “Reglas Macrofiscales, Instituciones e Instrumentos
Presupuestarios Plurianuales,” Buenos Aires, May. Available via the Internet:
http://capacitacion.mecon.gov.ar/manuales/reglas%20macro%20fiscales/r3–1.html.
Kopits, G. (2000), “How Can Fiscal Policy Help Avert Financial Crises?,” in Financial
Crisis: A Never-Ending Story: The OeNB’s East West Conference 1999, Vienna:
Osterreichische Nationalbank.
Kopits, G. (2001a), “Fiscal Rules: Useful Policy Framework or Unnecessary
Ornament?,” in Fiscal Rules: Papers Presented at the Bank of Italy Workshop held in
Perugia, 1–3 February 2001, Rome: Banca d’Italia.
Kopits, G. (2001b), “Fiscal Policy Rules for India?” Economic and Political Weekly,
Vol. 36 (March), pp. 749–56.
Kopits, G. (2002), “Central European EU Accession and Latin American Integration:
Mutual Lessons in Macroeconomic Policy Design,” North American Journal of
Economics and Finance, Vol. 13, pp. 253–77.
Kopits, G. and J. Craig (1998), Transparency in Government Operations, IMF Occasional
Paper No. 158, Washington: International Monetary Fund.
Kopits, G. and S. Symansky (1998), Fiscal Policy Rules, IMF Occasional Paper No. 162,
Washington: International Monetary Fund.
Kopits, G., R. Hagemann, L. Molho, and O. Bjerkholt (1993), “Fiscal Rules in
Indonesia: Analysis and Reform Proposal,” Washington: International Monetary
Fund, unpublished.
Kopits, G., J. Jiménez, and A. Manoel (2000), “Responsabilidad Fiscal a Nivel
Subnacional: Argentina y Brasil” in XII Seminario Regional de Politica Fiscal:
Compendio de Documentos, Santiago, Chile: Comisión Económica para América
Latina y El Caribe (CEPAL).
Krugman, P. (1979), “A Model of Balance-of-Payments Crises,” Journal of Money, Credit
and Banking, Vol. 11 (August), pp. 311–25.
Krugman, P. (2000), “Analytical Afterthoughts on the Asian Crisis.” Available via the
Internet: http://www.wws.princeton.edu/~pkrugman/MINICRIS.htm.
Kydland, F.E. and E.C. Prescott (1977), “Rules Rather than Discretion: the Inconsistency
of Optimal Plans,” Journal of Political Economy, Vol. 85 ( June), pp. 473–92.
Lahiri, A. (2000), “Budget Deficits and Reforms,” Economic and Political Weekly, Vol. 35
(November), pp. 4048–54.
Lane, T.D. (1993), “Market Discipline,” IMF Staff Papers, Vol. 40 (March), pp. 53–88.
Lane, P.R. and A. Tornell (1998), “Why Aren’t Savings Rates in Latin America
Procyclical?,” NBER Working Paper No. 6502, Cambridge, MA: National Bureau of
Economic Research.
 References 261

McCallum, B. (1995), “Two Fallacies Concerning Central Bank Independence,”

American Economic Review, Papers and Proceedings, Vol. 85 (May), pp. 207–11.
McGranahan, L. (1999), State Budgets and the Business Cycle: Implications for the Federal
Balanced Budget Amendment Debate, Chicago: Federal Reserve Bank, unpublished.
Mackenzie, G.A. (1998), “The Macroeconomic Impact of Privatization,” Staff Papers,
Vol. 45 ( June), pp. 363–73.
Mackenzie, G.A., P. Gerson, A. Cuevas, and P. Heller (1999), “Pension Reform and the
Fiscal Policy Stance,” IMF Working Paper 01/214, Washington: International
Monetary Fund.
McKinnon, R.I. (1997), “EMU as a Device for Collective Fiscal Retrenchment,”
American Economic Review, Papers and Proceedings, Vol. 87 (May), pp. 227–9.
Manasse, P., N. Roubini, and A. Schimmelpfennig (2002), “An Early Warning Model
of Fiscal Crises,” unpublished paper.
Marcel, M., M. Tokman, R. Valdés, and P. Benavides (2001), “Balance Estructrual del
Gobierno Central. Metodología y Estimaciones para Chile: 1987–2000,” Estudios de
Finanzas Públicas, Santiago: Ministerio de Hacienda de Chile. Available via the
Internet: http://www.dipres.cl/publicaciones/Balance%20MH_1.html.
Mayer, J., B. Chambers, and A. Farooq (eds) (1999), Development Policies in Natural
Resource Economics, Cheltenham, UK: Edward Elgar.
Mercosur (2000), “Comunicado conjunto de los Presidentes del 15 Diciembre de
2000.” Available via the Internet: http://www.mercosur.org.uy/espanol/snor/
varios/com0200.htm.
Mexico, Presidencia de la República (2001), “Última Hoja del Decreto que Reforma los
Artículos 74, Fracciones III, IV y VI, 75 y 126, Adiciona un Segundo Párrafo al
Apartado B del Artículo 72, y Deroga la Fracción VIII del Artículo 74 de la
Constitución,” México City, April.
Mexico, Secretaría de Hacienda y Crédito Público (SHCP) (2002a), Programa Nacional
de Financiamiento del Desarrollo. 2002–2006, Mexico City.
Mexico, Secretaría de Hacienda y Crédito Público (SHCP) (2002b), Aprobación del Paquete
Fiscal. Available via the Internet: http://www.hacienda.gob.mx/ori/index_docs.html.
Mexico, Secretaría de Hacienda y Crédito Público (SHCP) (2002c), Indicador de Gestión,
Subsecretaría de Egresos, Unidad de Contabilidad e Informes sobre la Gestión
Pública, Mexico.
Milesi-Ferretti, G.M. (1997), “Fiscal Rules and the Budget Process,” Giornale degli
Economisti e Annali di Economia, Vol. 56 ( June), pp. 5–40.
Mills P. and A. Quinet (2002), “How to Allow the Automatic Stabilisers to Play Fully?
A Policy Maker’s Guide for EMU Countries,” in M. Buti, J. von Hagen and
C. Martinez-Mongay (eds), The Behaviour of Fiscal Authorities: Stabilization, Growth
and Institutions, New York: Palgrave, pp. 115–29.
Musgrave, R.A. (1959), The Theory of Public Finance: A Study in Public Economy, New
York: McGraw-Hill.
Musgrave, R.A. and P.B. Musgrave (1984), Public Finance in Theory and Practice, New
York: McGraw-Hill.
Mussa, M. (2002), Argentina and the Fund: From Triumph to Tragedy, Policy
Analyses in International Economics 67, Washington: Institute for International
Economics.
Neary, J.P. and S. van Wijnbergen (1986), “Natural Resources and the Macroeconomy:
A Theoretical Framework,” in P.J. Neary and S. van Wijnbergen (eds), Natural
262 References

Resources and the Macroeconomy, Oxford: Centre for Economic Policy Research and
Basil Blackwell.
Nicolini, J. and others (2000), “Decentralization, Fiscal Discipline, and the Bailout
Problem: The Argentine Case,” Washington: Inter-American Development Bank,
unpublished paper.
Niculescu, I. (1999), “Macroeconomic Effects of the Fiscal Domestic Deficit: Structural
Change and Efficacy Loss of the Real Multiplier” (in Spanish), Journal of Central Bank
of Venezuela, Vol. 13(1), pp. 45–79.
Niskanen, W. (1971), Bureaucracy and Representative Government, Chicago: Aldine-
Atherton.
Niskanen, W. (1975), “Bureaucrats and Politicians,” Journal of Law and Economics,
Vol. 18 (December), pp. 617–43.
Norman, V.D. (1982), “Risk Management and Norwegian Petroleum Policy,” paper
presented at the Bergen Conference on Oil and Economics, Bergen, May.
North, D. (1990), “A Transaction Cost Theory of Politics,” Journal of Theoretical Politics,
Vol. 2(4), pp. 355–67.
Norwegian Ministry of Finance (2001a), National Budget 2002. Available via the
Internet: http://www.statsbudsjett.no/2002/english.asp.
Norwegian Ministry of Finance (2001b), Guidelines for Economic Policy. Available via
the Internet: http://odin.dep.no/fin/engelsk/p10003090/index-b-n-a.html.
Oates, W.E. (1972), Fiscal Federalism, New York: Harcourt Brace Jovanovich.
Obstfeld, M. (1994), “The Logic of Currency Crises,” NBER Working Paper No. 4640,
Cambridge, MA: National Bureau of Economic Research.
Organization for Economic Co-operation and Development (1988), OECD Economic
Surveys: Netherlands 1987/988, Paris.
Organization for Economic Co-operation and Development (1999), OECD Economic
Surveys: Mexico, Paris.
Patashnik, E.M. (2000), Putting Trust in the U.S. Budget: Federal Trust Funds and the
Politics of Commitment, Cambridge: Cambridge University Press.
Patnaik, I. and A. Shah (2002), “Interest Rate Risk in the Indian Banking System,”
ICRIER Working Paper No. 92, New Delhi: Indian Council for Research on
International Economic Relations.
Perry, G. and M. Huertas (1997), “La Historia de Una Crisis Anunciada: Regulado el
Endeudamiento de las Municipalidades y los Departamentos en Colombia,” in
G.Aghón and H. Edling (eds), Decentralización Fiscal en América Latina: Nuevos
Desafíos y Agenda deTrabajo, Santiago, Chile: Comisión Económica para América
Latina y el Caribe (CEPAL)/Deutsche Gesellschaft für Technische Zusammenarbeit
(GTZ).
Perry, G. and L. Servén (2003), “The Anatomy of a Multiple Crisis: Why Was Argentina
Special and What Can We Learn from It?,” World Bank Policy Research Working
Paper # 3081, Washington.
Peltzman, S. (1992), “Voters as Fiscal Conservatives,” Quarterly Journal of Economics,
Vol. 107 (May), pp. 327–61.
Perotti, R. (1996), “Fiscal Consolidation in Europe: Composition Matters,” American
Economic Review, Papers and Proceedings, Vol. 86 (May), pp. 105–10.
Perotti, R. (1999), “Fiscal Policy in Good Times and Bad,” Quarterly Journal of
Economics, Vol. 114 (November), pp. 1399–436.
Persson, T. and L.E.O. Svensson (1989), “Why a Stubborn Conservative Would Run a
Deficit: Policy with Time-Inconsistent Preferences,” Quarterly Journal of Economics,
Vol. 104(2), pp. 325–45.
 References 263

Persson, T. and G. Tabellini (2000), Political Economics: Explaining Economic Policy,

Cambridge, MA: MIT Press.
Pigou, A.C. (1928), Public Finance, London: Macmillan.
Poterba, J.M. (1994), “State Responses to Fiscal Crises: The Effects of Budgetary
Institutions and Politics,” Journal of Political Economy, Vol. 102 (August), pp. 799–821.
Poterba, J.M. (1996), “Budget Institutions and Fiscal Policy in the U.S. States,”
American Economic Review, Papers and Proceedings, Vol. 86 (May), pp. 395–400.
Poterba, J.M. and K. Rueben (1999), “State Fiscal Institutions and the U.S. Municipal
Bond Market,” in J.M. Poterba and J. von Hagen (eds), Fiscal Institutions and Fiscal
Performance, Chicago: University of Chicago Press.
Poterba, J.M. and J. von Hagen (eds) (1999), Fiscal Institutions and Fiscal Performance,
Chicago: University of Chicago Press.
Prasad, A., R. Goyal, and A. Prakash (2003), “States’ Debt and Debt Relief,” Reserve
Bank of India, unpublished paper.
Przeworski, A. (1997), “Pacts,” New York University, Department of Politics, unpub-
lished paper.
Radelet, S. and J. Sachs (1998), “The East Asian Financial Crisis: Diagnosis, Remedies,
Prospects,” Brookings Papers on Economic Activity, Washington, pp. 1–87.
Rangarajan, C. and D.K. Srivastava (2003), “Dynamics of Debt Accumulation in India:
Impact of Primary Deficit, Growth and Interest Rate,” Economic and Political Weekly,
Vol. 38 (November), pp. 4851–8.
Reinhart, C.M. (2002), “Default, Financial Crises and Sovereign Credit Ratings,” NBER
Working Paper No. 8738, Cambridge, MA: National Bureau of Economic Research.
Also published in World Bank Economic Review, Vol. 16 (2), pp. 151–70.
Reserve Bank of India (2002), Report on Currency and Finance 2000–01, Mumbai.
Reserve Bank of India (2003), State Finances: A Study of Budgets of 2002–03, Mumbai.
Rezende da Silva, F. (2001), “Brasil a contramano: ajuste fiscal y deterioro de la cali-
dad del sistema tributario brasileno,” in Noveno Congreso Tributario: Trabajos de
Investigacion, Vol. 1, Buenos Aires: Consejo Profesional de Ciencias Economicas.
Rødseth, A. (2001), “The New Guideline for Spending Petroleum Revenue – A Problem
for Stabilization Policy,” Oslo, Norway: University of Oslo, Department of
Economics. Available via the Internet: http://folk.uio.no/asro/Guidelines.pdf.
Rodríguez, F. and J.D. Sachs (1999), “Why Do Resource-Abundant Economies Grow
More Slowly?,” Journal of Economic Growth, Vol. 4 (September), pp. 277–303.
Roubini, N. and J.D. Sachs (1989), “Political and Economic Determinants of Budget
Deficits in the Industrial Democracies,” European Economic Review, Vol. 33 (May),
pp. 903–38.
Sachs, J., A. Tornell, and A. Velasco (1996), “Financial Crises in Emerging Markets: The
Lessons from 1995,” Brookings Papers on Economic Activity, Washington, pp. 147–215.
Sachs, J.D. and A.M. Warner (1995), “Natural Resource Abundance and Economic
Growth,” NBER Working Paper No. 5398, Cambridge, MA: National Bureau of
Economic Research.
Saiegh, S. and M. Tommasi (1999), “Why Is Argentina’s Fiscal Federalism So Inefficient?
Entering the Labyrinth,” Journal of Applied Economics, Vol. 2(1), pp. 169–209.
Saiegh, S. and M. Tommasi (2000), “Le labyrinthe fiscal de l’Argentine et la théorie des
coûts de transaction en politique,” Problémes d’Amérique Latine, Vol. 37 (April–June),
pp. 63–90.
Saint-Paul, G. (1994), “Monetary Policy in Economic Transition: Lessons from
the French Post-War Experience,” European Economic Review, Vol. 38 (April),
pp. 891–8.
264 References

Sanguinetti, J. (2001), “Determinantes políticos e institucionales ligados al desempeño

fiscal. El caso de las provincias argentinas,” Working Paper No. 75, Buenos Aires:
Centro de Estudios para el Desarrollo Institucional.
Sanguinetti, P. and M. Tommasi (1998), “Intergovernmental Transfers and Fiscal
Behavior – Part 1: Insurance Versus Aggregate Discipline,” unpublished paper.
Santaella, J.A. (2001), “La viabilidad de la política fiscal, 2000–25,” in A.M. Fernandez
(ed.) Una Agenda para las Finanzas Públicas de México, Mexico City: ITAM, pp. 37–65.
Schick, A. (1993), “Government Versus Budget Deficits,” in R.K. Weaver and
B.A. Rockman (eds), Do Institutions Matter: Government Capabilities in the United States
and Abroad, Washington: The Brookings Institution.
Schwartz, G. and C. Liuksila (1997), “Argentina,” in T. Ter-Minassian (ed.), Fiscal
Federalism: Theory and Practice, Washington: International Monetary Fund.
Servén, L. (1998), “Macroeconomic Uncertainty and Private Investment in LDCs: An
Empirical Investigation,” Policy Research Working Paper No. 2035, Washington:
World Bank.
Shome, P. (2000), “India: Primary Aspects of a Medium Term Fiscal Strategy,”
New Delhi: Indian Council of Research on International Economic Relations.
Silva Filho, T.N.T. (2002), Estimating Brazilian Potential Output: a Production Function
Approach, Central Bank of Brazil Working Paper Series No.17, Brasília: Central Bank
of Brazil.
Smith, S. (1996), “European Integration and Local Government Finance,” in G. Pola,
G. France, and R. Levaggi (eds), Developments in Local Government Finance: Theory and
Policy, Brookfield, Vt.: Edward Elgar.
Spiller, P. and M. Tommasi (2001), “The Institutional Foundations of Public Policy: A
Transactions Theory and an Application to Argentina,” Working Paper No. 58,
Buenos Aires: Centro de Estudios para el Desarrollo Institucional.
Srinivasan, T.N. (2001), “India’s Fiscal Situation: Is a Crisis Ahead?,” SCID Working
Paper No. 92, Stanford, California: Stanford Center for International Development.
Stark, J. (2001), “Genesis of a Pact,” in A. Brunila, M. Buti, and D. Franco (eds), The
Stability and Growth Pact: The Architecture of Fiscal Policy in EMU, Basingstoke,
Hampshire; New York: Palgrave.
Stein, E. (1999), “Reglas Macrofiscales e Instituciones Prespuestarias,” paper presented
at the Ministerio de Economía, Obras y Servicios Públicos of Argentina seminar on
“Reglas Macrofiscales, Instituciones e Instrumentos Presupuestarios Plurianuales,”
Buenos Aires, May. Available via the Internet: http://capacitacion.mecon.gov.ar/
manuales/reglas%20macro%20fiscales/r4–1.html.
Stein, E., E. Talvi, and A. Grisanti (1999), “Institutional Arrangements and Fiscal
Performance: The Latin American Experience,” in J.M. Poterba and J. von Hagen
(eds), Fiscal Institutions and Fiscal Performance, Chicago: University of Chicago Press.
Stourm, R. (1917), The Budget, a translation from the 7th edn (1913) of Le Budget,
Cours de Finances, Paris, (ed.) W.F. McCaleb, (tr.) T. Plazinski, (introd.) C.A. Beard,
New York: D. Appleton for the Institute for Government Research.
Suescún, R. (2003), “Understanding Procyclical Fiscal Policies,” Washington: World
Bank, unpublished paper.
Summers, L. (2000), “International Financial Crises: Causes, Prevention, and Cures,”
American Economic Review: AEA Papers and Proceedings, Vol. 90, (May), pp. 1–16.
Sundelson, J.W. (1935), “Budgetary Principles,” Political Science Quarterly, Vol. 50
( June), pp. 236–63.
Tabellini, G. and A. Alesina (1990), “Voting on the Budget Deficit,” American Economic
Review, Vol. 80 (March), pp. 37–49.
 References 265

Talvi, E. and C.A. Végh (1996), “Can Optimal Fiscal Policy Be Procyclical?”
Washington: Inter-American Development Bank, Office of the Chief Economist,
and UCLA, unpublished paper.
Talvi, E. and C.A. Végh (1998), “Fiscal Policy Sustainability, a Basic Framework,” IADB
Working Paper No. 372, Washington: Inter-American Development Bank. Available
via the Internet: http://www.iadb.org/RES/index.cfm?fuseaction
LaResNetwork.
StudyView&st_id
51.
Tamirisa, N.T. (1999), “Exchange and Capital Controls as Barriers to Trade,” Staff
Papers, Vol. 46 (March), pp. 69–88.
Tanzi, V. (1989), “Fiscal Policy, Growth, and the Design of Stabilization Programs,” in
M.I. Blejer and K. Chu (eds), Fiscal Policy, Stabilization, and Growth in Developing
Countries, Washington: International Monetary Fund.
Tanzi, V. (1995), “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 1995, Washington: World
Bank.
Tanzi, V. and M. Blejer (1984), “Fiscal Deficits and Balance of Payments Disequilibrium
in IMF Adjustment Programs,” in J. Muns (ed.), Adjustment, Conditionality, and
International Financing, Washington: International Monetary Fund.
Teijeiro, M. (2001), “Una vez más la política fiscal … ,” in M. Lascano (ed.), La
Economía Argentina Hoy, Buenos Aires: Centro de Estudios Públicos.
Ter-Minassian, T. (2002), “Institutional Reforms for Fiscal Sustainability with Special
Reference to Latin America,” Regional Seminar on Fiscal Policy, ECLAC, Santiago,
Chile, January. Available via the Internet: http://www.eclac.cl/de/noticias/pagi-
nas/0/9200/InaugTer.pdf
Ter-Minassian, T. and J. Craig (1997), “Control of Subnational Government
Borrowing,” in T. Ter-Minassian (ed.), Fiscal Federalism in Theory and Practice,
Washington: International Monetary Fund.
Ter-Minassian, T. and G. Schwartz (1997), “The Role of Fiscal Policy in Sustainable
Stabilization: Evidence from Latin America,” IMF Working Paper 97/94,
Washington: International Monetary Fund.
Ter-Minassian, T. (ed.) (1997), Fiscal Federalism in Theory and Practice, Washington:
International Monetary Fund.
Tirole, J. (2002), “Inefficient Foreign Borrowing,” paper presented at the Seventh
Annual Meeting of the Latin American and Caribbean Economic Association
(LACEA), Madrid, October. Available via the Internet: http://www.idei.asso.fr/
Commun/Articles/Tirole/Inefficient-Foreign-Borrowing.pdf.
Tommasi, M., S. Saiegh, and P. Sanguinetti (2001), “Fiscal Federalism in Argentina:
Policies, Politics, and Institutional Reform,” Economia: Journal of the Latin American
and Caribbean Economic Association, Vol. 1(2), pp. 147–201.
Tommasi, M. and P. Spiller (2000), Las fuentes institucionales del desarrollo argentino.
Hacia una agenda institucional, Buenos Aires: UNDP.
United Kingdom, H.M. Treasury (1998), “A Code for Fiscal Stability,” London.
Available via the Internet: http://www.hm-treasury.gov.uk/documents/uk_
economy/fiscal_policy/ukecon_fisc_code98.cfm.
US General Accounting Office (1994), Deficit Reduction: Experiences of Other Nations,
Report No. AIMD-95–30, Washington: Government Printing Office.
US General Accounting Office (2000), Accrual Budgeting: Experiences of Other Nations
and Implications for the United States, Report No. AIMD 00–57, Washington:
Government Printing Office.
266 References

van den Noord, P. (2000), “The Size and Role of Automatic Fiscal Stabilizers in the
1990s and Beyond,” OECD Economic Department Working Paper No. 230, Paris:
Organization for Economic Co-operation and Development.
Verma, A.L. and B. Singh (1999), “Debt Dynamics of State Governments in India:
Dimensions and Analytical Issues,” Reserve Bank of India, unpublished paper.
Vial, J. (2001), “Institucionalidad y desempeno fiscal: una mirada a la experiencia de
Chile de los 90,” unpublished paper.
von Hagen, J. (1992), “Budgeting Procedures and Fiscal Performance in the European
Communities,” Economic Papers, No. 96, Luxembourg: European Communities.
von Hagen, J. (2002), “More Growth for Stability – Reflections on Fiscal Policy
in Euroland,” ZEI Policy Paper, Bonn: Center for European Integration
Studies (ZEI).
von Hagen, J. and I.J. Harden (1994), “National Budget Processes and Fiscal
Performance,” in Towards Greater Fiscal Discipline, European Economy, Reports and
Studies, No. 3, Luxembourg: Office for Official Publications of the European
Communities.
von Hagen, J. and I.J. Harden (1995), “Budget Processes and Commitment to Fiscal
Discipline,” European Economic Review, Vol. 39 (April), pp. 771–9.
von Hagen, J., A. Hughes Hallett, and R. Strauch (2001), “Budgetary Consolidation in
EMU,” European Commission, Directorate General for Economic and Financial
Affairs Economic Paper No. 148, Luxembourg: Office for Official Publications of the
European Communities.
von Hagen, J. and R.R. Strauch (2001), German Public Finances: Recent Experiences and
Future Challenges, ZEI Working Paper No. B13, Bonn, Germany: Center for European
Integration Studies.
Wagner, R.E. (1976), “Revenue Structure, Fiscal Illusion, and Budgetary Choice,” Public
Choice 25 (Spring), pp. 45–61.
Watts, R. (1999), Comparing Federal Systems, 2nd edn, Montreal: McGill-Queen’s
University Press.
Webb, S.B. and M.M. Giugale (eds) (2000), Achievements and Challenges of Fiscal
Decentralization: Lessons from Mexico, Washington: World Bank.
Weingast, B.R. (1995), “The Economic Role of Political Institutions: Market-Preserving
Federalism and Economic Development,” The Journal of Law, Economics and
Organization, Vol. 11 (April), pp. 1–31.
Weingast, B.R. and W.J. Marshall (1988), “The Industrial Organization of Congress (Or
Why Legislatures, Like Firms, Are Not Organized as Markets),” Journal of Political
Economy, Vol. 96 (February), pp. 132–63.
Wildavsky, A.B. (1964), The Politics of the Budgetary Process, Boston: Little, Brown.
Wildavsky, A.B. (1986), Budgeting: A Comparative Theory of Budgetary Processes, New
Brunswick, N.J.: Transaction Publishers.
Williamson, O.E. (1983), “Credible Commitments: Using Hostages to Support
Exchange,” American Economic Review, Vol. 73 (September), pp. 519–40.
Williamson, O.E. (1996), The Mechanisms of Governance, New York: Oxford University
Press.
World Bank (1996), Argentina: Provincial Finances Study, Sector Report No. 15487,
Washington.
World Bank (2000a), Brazil: Critical Issues in Social Security, Report No. 19641-BR,
Washington.
 References 267

World Bank (2000b), World Development Report 2000/2001 – Attacking Poverty,

Washington.
World Bank (2001), Argentina: Provincial Finances Update IV, Washington.
Wren-Lewis, S. (2000), “The Limits to Discretionary Fiscal Stabilization Policy,” Oxford
Review of Economic Policy, Vol. 16 (4), pp. 92–105.
Wren-Lewis, S. (2003), “The Compatibility between Monetary and Fiscal Policies in
EMU,” in M. Buti (ed.), Monetary and Fiscal Policies in EMU – Interactions and
Coordination, Cambridge, UK: Cambridge University Press.
Wyplosz, C. (2002), “Fiscal Policy: Institutions Versus Rules,” CEPR Discussion Paper
No. 3238, London: Centre for Economic Policy Research.
Author Index

Acharya, S., 217 Cardoso, E., 129

Aghion, P., 39 Casella, A., 226
Ahluwalia, M., 203, 216 Cashin, P., 65
Albuquerque, P.H., 79 Catão, L., 34
Alesina, A., 19, 22, 28, 78, 94, 107, 112, Chamon, M., 39, 51, 52
113, 197, 218 Chari, V.V., 233
Allan, W., 80 Coase, R., 225
Allsop, C., 10 Coelho, I., 79
Alt, J.E., 184 Cole, H.L., 233
Ardagna, S., 78, 217 Conesa, A., 6, 131–45
Artis, M.J., 10 Corden, W.M., 178
Aslaksen, I., 178 Coricelli, F., 7, 146–62
Aziz, J., 78 Corsetti, G., 23, 39, 78
Costello, D., 112
Bacha, E.O., 129 Courchene, T.J., 249
Balassone, F., 8, 219–34 Craig, J., 10, 21, 80, 144, 197,
Barnhill, T., 52 222, 233
Barrell, R., 233 Creel, J., 163
Barro, R.J., 16, 52, 131, 164 Csajbók, A., 163
Begg, D., 78 Csermely, Á., 163
Bertola, G., 78 Cuddington, J., 65
Besley, T.A., 10 Cukierman, A., 80
Bevilaqua, A.S., 197, 218
Bier, A., 65 Daniel, J., 79, 128
Bird, R.M., 241 Darvas, Z., 163
Bjerkholt, O., 7, 65, 164–78 Davis, J., 179
Blanchard, O., 80 De Ferranti, D., 54–6
Blejer, M., 78 de Haan, J., 233
Blondal, J., 222 Detragiache, E., 34
Bohn, H., 184 Dillinger, W., 197, 248
Brailovsky, V., 178 Dixit, A.K., 197
Braun, M., 7, 61, 183–97, Dooley, M.P., 78
206, 216 Drazen, A., 5, 15–29, 78, 134, 188
Brunila, A., 112, 163, 221, 233 Dury, K., 233
Buchanan, J., 18 Dziobek, C., 79
Buiter, W.H., 80, 146–7, 155, 160
Burnside, C., 78 Easterly, W., 147
Buti, M., 6, 21, 97–113, 149, 154, 158, Eaton, J., 51
163, 220, 221, 233 Echavarría, J., 197, 242
Eichengreen, B., 39, 52, 64, 99, 107,
Caballero, R.J., 106 113, 192, 217
Cabral, A.J., 112, 221 Einaudi, L., 233
Calvert, R.L., 185, 187 Elster, J., 20, 26, 29
Calvo, G., 10, 51, 77, 78 Engel, E., 178
Cappelen, Å., 178 Ercolani, V., 7, 146–63

268
 Author Index 269

Fatas, A., 10 Knight, B., 233

Favero, C., 129 Kochhar, K., 7–8, 198–218
Fiess, N., 65 Kopits, G., 1–10, 25, 28, 52, 66–80, 95,
Fischer, J., 99, 102, 227, 228 103, 113, 129, 130, 144, 163, 179,
Fischer, S., 78, 107, 113 183, 197, 248
Fitoussi, J.P., 163 Krueger, A., 65
Flood, R.P., 78 Krugman, P., 65, 77, 78
Franco, D., 8, 219–34 Kydland, F.E., 80, 131
Friedman, D., 234
Lahiri, A., 201, 204
Gamboa, R., 197 Lane, P.R., 106
Gavin, M., 51, 55, 64, 65, 80, 104–108, Lane, T.D., 222
112, 113 Levinson, B.A., 233
Gelb, A.H., 178 Liuksila, C., 249
Gersovitz, M., 51 Lowry, R.C., 184
Ghosh, A., 78
Ghymers, C., 109, 112 Mackenzie, G.A., 79, 80
Giavazzi, F., 129 Manasse, P., 34
Gil-Díaz, F., 132, 145 Mancowizk, B., 39
Giudice, G., 6, 21, 64, 97–113, 233 Manoel, A., 21, 144
Giugale, M.M., 243 Marcel, M., 65, 145
Gjelsvik, E., 178 Marion, N.P., 78
Guidotti, P., 112 Marshall, W.J., 197
Goldfajn, I., 6, 114–130 Masson, P.R., 78
Goldstein, M., 128 Mayer, J., 178
Gonzalez, C.Y., 8, 235–48 McCallum, B., 29
Gordon, D.B., 16, 131 McGranahan, L., 233
Gourinchas, P.O., 112 McKinnon, R.I., 98
Grafe, C., 146, 147, 155, 160 Meltzer, A.H., 80
Grilli, V., 19 Mihov, I., 10
Guardia, E.R., 6, 114–30 Milesi-Ferretti, G.M., 22, 23
Gylfason, T., 178 Mills, P., 163
Mishkin, F.S., 10
Hagemann, R., 49, 64 Monacelli, D., 233
Hallerberg, M., 81, 87, 221, 228 Montanino, A., 112
Harden, I.J., 22, 98, 99 Musgrave, P.B., 223
Hausmann, R., 5, 21, 28–52, 80, 107, Musgrave, R.A., 223, 235
108, 112–113, 178 Mussa, M., 64
Heller, P., 10, 78
Hemming, R., 78, 218 Neary, J.P., 178
Hercowitz, Z., 19 Nicolini, J., 197
Hernández-Trillo, F., 198, 233 Niculescu, I., 7, 65, 164–79
Hodrick, R., 163 Niskanen, W., 18, 28
Huertas, M., 242 Norman, V.D., 178
North, D., 197
Iaryczower, M., 197
Inman, R.P., 103, 184 Oates, W.E., 235
Obstfeld, M., 78
Kennedy, S., 185
King, G., 163 Pagano, M., 78, 100
Kletzer, K., 112 Panizza, U., 39, 50–52
Kneller, R., 78 Parry, T., 80
270 Author Index

Patashnik, E.M., 90 Srinivasan, T.N., 205

Patnaik, I., 217 Srivastava, D.K., 216, 217
Pazarbasioglu, C., 79 Stark, J., 112
Peltzman, S., 28, 197 Stein, E., 107, 113, 144, 248
Perotti, R., 22, 28, 64, 94, 112, 113, 197 Stourm, R., 94
Perry, G., 5, 53–65, 242 Strauch, R.R., 90
Persson, T., 10, 19 Strawczynski, M., 19
Pigou, A.C., 233 Suescún, R., 55, 64
Poterba, J.M., 10, 86, 87, 94, 98, 113, Summers, L., 78
129, 184 Sundelson, J.W., 94
Prasad, A., 218 Sutton, B., 34
Prescott, E.C., 80, 131, 163 Svensson, L.E.O., 19
Przeworski, A., 186, 187 Symansky, S., 10, 25, 28, 80, 103, 129,
Purfield, C., 7–8, 198–218 144, 183, 232

Quinet, A., 163 Tabellini, G., 10, 19

Talvi, E., 106, 113
Radelet, S., 78 Tamirisa, N.T., 78
Rangarajan, C., 216, 217 Tanzi, V., 78, 79
Rebelo, S., 147, 163 Teijeiro, M., 51, 68, 79, 80
Reinhart, C.M., 34, 51, 77 Ter-Minassian, T., 110–13, 221,
Rezende da Silva, F., 128 222, 233
Robbins, J., 185 Thirsk, W., 132
Rodríguez, F., 178 Tijerina, J.A., 6, 131–45
Rødseth, A., 179 Tirole, J., 39
Rosenblatt, D., 8, 235–49 Tommasi, M., 7, 61, 183–97, 206, 216,
Roubini, N., 23, 85, 94 248, 249
Rust, J., 233 Tornell, A., 106
Treasury, H.M., 158
Saal, M., 79
Sachs, J.D., 78, 85, 94, 178 Valdés, R.O., 178
Saiegh, S., 193, 195, 197 van den Noord, P., 163
Saint-Paul, G., 106 van Wijnbergen, S., 178
Sanguinetti, J., 197 Végh, C.A., 106, 113
Sanguinetti, P., 248 Verma, A.L., 218
Santaella, J.A., 144 Vial, J., 80
Sapir, A., 98, 112, 149, 154, 233 von Hagen, J., 10, 22, 81, 86, 90, 94, 98,
Schick, A., 6, 65, 77, 81–95, 134 99, 100, 129, 163, 192
Schwartz, G., 113, 249
Schwartz, M.J., 6, 131–45 Wagner, R.E., 18, 107
Servén, L., 54, 64 Warner, A.M., 178
Shah, A., 217 Watts, R., 207
Shome, P., 201 Webb, S.B., 8, 191, 235–49
Silva Filho, T.N.T., 130 Weingast, B.R., 197, 216
Simon, A., 163 Wildavsky, A.B., 84, 112
Singh, B., 218 Williamson, O.E., 90
Smith, S., 227 Wren-Lewis, S., 113
Somuano, A., 6, 131–45 Wyplosz, C., 64, 113
Spilimbergo, A., 34
Spiller, P., 191, 193 Zotteri, S., 8, 219–34
Subject Index

absorption rule, 139, 140, taxation, 79n30

141, 143 uniform import surcharge,
accession countries (ACs) see EU 80n32
accession countries Asian financial crisis, 67, 69,
accountability, 9, 83, 213 75, 120
see also transparency assets, 77n5, 129n9, 167, 168
accounting, 98, 99, 110, 131 asymmetric information, 18, 56, 57,
accrual, 75, 92, 93 58, 66, 67
cash, 93 audit mechanisms, 4–5, 9, 21, 91,
India, 207 158, 214
subnational rules, 232 see also monitoring
see also creative fiscal Australia
accounting budget restructuring, 88
acquired rights, 72–3, 74 monitoring mechanisms, 21
adjustment issues transparency, 132
constraints on adjustment, 72–3 Austria
Europe, 99–101 accounting practices, 232
quality of adjustment, 71–2, 101 federalism, 219–20, 227
allocative inefficiency, 190, 232 subnational expenditure,
Andean Community, 109–10 228, 229
Annual Budget Guidelines Law (LDO), automatic stabilizers, 9, 48, 53, 57–8,
Brazil, 123, 124, 125 61, 63
Argentina balanced-budget rule, 76
budget deficit, 80n35, 237, 246 capital account crises, 70
budget-balance rule, 2, 3, 4 EMU stabilization, 221
consumption cycle, 108 EU accession countries, 147
currency board, 2, 69 Latin America, 108
debt ratio, 30, 41 Mexico, 135, 144
decentralization, 4, 10, 132, 192 positive effects, 154
expenditure control, 72 rule design, 59
external debt, 79n22 Stability and Growth Pact, 102,
failure with subnational fiscal 108, 220
rules, 7, 63 structural balance rule, 137
federalism, 192–6 structural deficit indicator, 150
fiscal crisis, 56, 57, 68, 75, 191
fiscal discipline, 185 balanced-budget rules see
fiscal performance, 237–8 budget-balance rules
fiscal responsibility law, 59, 61 banking
import gap, 44, 45 Asian financial crisis, 69
intergovernmental transfers, 63, Brazil, 119, 120
64, 236, 238–40, 245–6 Colombia, 242
optimal currency area, 52n17 fiscal adjustment, 72
pensions, 74 India, 203
procyclical fiscal policy, 56, Mexico, 140
63, 64 structural reform, 73–4

271
272 Subject Index

Belgium see also budget-balance rules; debt;

debt ratio, 30 deficit bias; government
decentralization, 231 expenditure; primary balance
federalism, 219–20, 227 budget deficit, 9, 15
subnational expenditure, 228, 229 Argentina, 80n35, 237, 246
benefits, 74, 84, 85 Brazil, 116
see also social security; capital account crises, 66, 67, 69
unemployment benefits EMU rules, 99, 107, 220, 221, 223–4
booms, 56, 57, 58, 59, 64, 65n4, 236 EU accession countries, 7, 146, 147,
Brazil, 114–30 151, 154, 155–6, 160
Annual Budget Guidelines Law European countries, 97, 230–1
(LDO), 123, 124, 125 fiscal illusion, 18
banking supervision, 74 fiscal stance, 78n13
budget-balance rule, 2, 3, 137 Germany, 24, 88, 89–90;
capital account crisis, 68, 69 and growth, 146–7
credibility, 49–50 India, 199, 205, 206, 208, 209,
debt ratio, 30, 32, 41, 65n14 211, 215
debt rule, 3, 4 Latin America, 104–6, 107,
debt-restructuring package, 122, 213 109, 110
decentralization, 4, 10, 119, 215 Mexico, 135, 243
domestic currency debt, 43 Norway, 165
enforcement penalties, 58 procedural restructuring, 88
fiscal discipline, 185 procyclical fiscal policy, 57, 63
fiscal impulse, 78n13 Stability and Growth Pact, 102–3, 146
fiscal responsibility law, 6, 61, structural balance rules, 62, 64,
93, 110, 115, 121–4, 128 157, 159
inflation-targeting, 2 Venezuela, 170
pensions, 74 see also debt; deficit bias
primary surplus, 49, 65n13, 76, budget-balance rules, 2, 3, 9, 23
79n21, 114, 116, 118–19, 121 Argentina, 196
quantitative targets, 132 Brazil, 122–3
sanctions, 5 constitutional provision, 26, 27
state revenue, 248 cyclical adjustments, 49
subnational rules, 185, 215 debt structure, 50
taxation, 79n30 Mexico, 6–7, 11n10, 136, 137,
budget 141–2, 143, 144
constraints, xii, 15, 30, 87, 99: political constraints, 91
Brazil, 125; Colombia, 241, 242; resource-abundant countries, 169
India, 213, 214; market-preserving signaling commitment, 76
federalism, 216; Mexico, 244 structural close-to-balance rule,
EMU rules, 219, 223–4, 225, 226 156–8, 159, 160
innovations, 91–2 subnational, 131
maximization, 18–19 Bulgaria, uniform import surcharge,
Mexico, 136, 244 80n32
process, 23, 24, 81, 82–5, 86, 98, 124, business cycle, 40, 44, 97, 108,
214 131, 132
resource-abundant countries, 166
solvency, 125 Canada
stabilizing transfer rules, 236 debt ratio, 30
structural balance computation, equalizing transfers, 249n14
148–54 monitoring mechanisms, 21
 Subject Index 273

capital account crises, xii, 5, 9, 66–9, Mexico, 133, 134, 135, 137–8
72, 75, 76–7 Norway, 60, 170, 171–2, 173, 174
see also financial crises Venezuela, 7, 59, 60, 165, 170, 174–7
capital flight, 70, 74, 75, 84 commons problem, 107
capital mobility, xii, 1–2, 5, 66–80 compliance see enforcement
Caribbean comprehensiveness, 82
procyclical fiscal policies, 53, 54–7 constitutional provision, 4, 5, 9,
volatility, 53 20, 26–7
cash flow, 166, 167 Argentina, 195
central bank autonomy, 131 Brazil, 129n4
Central and Eastern Europe, 1, 7, 103 Mexico, 136, 144
Chile Venezuela, 174, 176, 177
budget-balance rule, 2, 3, 76 constitutionalism, 26–7, 29n12
centralization, 4 consumption
commodity stabilization fund, 59, capital account crises, 70
60, 113n19 Latin America, 106, 108
countercyclical fiscal policy, 56 resources, 167, 168
debt ratio, 41 contagion, 109
fiscal conservatism, 65n10 contingency funds, 2–4
fiscal performance, 93 contingent liabilities, 69, 74, 75,
fiscal risk, 51 92, 94n2
inflation-targeting, 2 India, 200, 217n4
institutional infrastructure, 80n37 Mexico, 144n5, 247
policy guideline, 4, 9 resources, 168
structural balance rule, 4, 62, 63, stabilizing transfers, 8, 61, 247
137, 145n16 see also hidden liabilities
structural surplus, 108, 145n14 convergence
coalition regimes, 81, 87, 88, 89, 198 European, 7, 99, 100, 103
Cohesion Fund, 146, 155 Latin America, 108–10, 111
Colombia see also integration
budget-balance rule, 2, 3 cooperation
commodity stabilization fund, 59, determinants of, 191–2
60, 113n19 fiscal game, 188
debt ratio, 41 intertemporal, 191, 194
debt rule, 3 problems, 189
decentralization, 4, 240, 241 subnational rules, 8, 222, 227, 230,
earmarking of revenue, 72 231, 232
expenditure control, 72 corporate borrowing, 52n11
fiscal crisis, 56, 57, 75 corruption, 93
fiscal discipline, 185 Costa Rica
fiscal performance, 240–1 debt ratio, 41
fiscal responsibility law, 61 countercyclical fiscal policy, 5, 9, 50,
inflation-targeting, 2 53, 57
intergovernmental transfers, 63, 64, commodity stabilization funds, 60
236, 240–3, 246–7 credibility, 58, 59
sanctions, 5 deficit bias, 19
taxation, 79n30 discretion, 4, 9, 80n43
commitment, 24–5, 87–8, 89, 90, 91, EU deficit threshold, 220
98, 123 Latin America, 56
commodity stabilization funds, 9, structural balance rules, 62, 157
59–61, 113n19, 169, 178 subnational level, 235
274 Subject Index

creative fiscal accounting, 15–16, crises, 34

22, 23, 58, 98 EU constraints, 24
EU candidate countries, 147 euro area, 105
expenditure audit, 158 Economic and Monetary Union,
subnational government 98, 107
finance, 222 GDP ratio, 4, 9, 30–2, 38, 40, 45–7:
credibility, xii, 1, 8, 10, 20–1 Brazil, 114, 116–19, 123, 125–6,
Brazil, 49–50, 116, 122, 128 128; India, 7–8, 208, 209, 212;
budget commitments, 87, 90, 91 Latin America, 106; Maastricht
capital account crises, 77 rules, 99, 100, 220; Mexico,
commitment to unchanging rules, 140, 142
25–6, 27 golden rule, 20
constitutionalism, 26 growth of, 15, 28
countercyclical fiscal policy, 58 India, 7–8, 198, 199–202, 208–13,
cyclical adjustments, 49 215, 217n4
discretion, 16 Latin America, 7, 57, 105, 106,
fiscal discipline, 15, 25 107, 110
fiscal expansion, 56 Mexico, 132, 137, 139–41, 142
India, 215 national fiscal council, 111
Mexico, 135 net public debt concept, 129n9
monetary policy, 39 political arrangements, 85
penalties, 29n8 public debt management, 75
predictability, 17 servicing: Brazil, 122; currency
procyclical fiscal policy, 63 devaluations, 73; Latin America,
rule flexibility, 22–3, 58–9 30, 32, 34–5, 40, 43–4, 47–9, 50
signaling, 71, 76 stabilizing transfers, 247
structural balance rule, 62, 64 structural balance rule, 64
transparency, 4, 75, 77 subnational: Argentina, 7, 239, 240,
credit ratings, 30, 31–3, 38, 47, 49, 246; Brazil, 119–20, 121–2, 123,
51, 107 128; Colombia, 240, 242–3, 246;
CSM see cyclical safety margin India, 206, 209, 213; Mexico,
currency 133, 134, 143, 243,
crises, 5, 34, 66, 67–70, 73–4, 75, 104 244–50
depreciation, 40 sustainability, 76, 77, 160: Brazil, 6,
domestic currency debt, 43–4, 50, 114–16, 119, 125–8; definition,
52n11, 201 145n19; India, 208–13; Mexico,
foreign currency debt, 38–40, 47–8, 137, 139–41, 142, 144
52n11, 75, 107 see also budget deficit; deficit bias;
see also exchange rate public sector borrowing
currency boards, 2, 69, 79n22 requirement
current-balance rule, 2 debt rules, 3, 4, 115
see also golden rule Brazil, 123
cyclical safety margin (CSM), 149–50, Mexico, 139–41, 143
151 signaling commitment, 76
Czech Republic, capital account crisis, decentralization, 4, 222, 231
67, 68, 69, 78n12 Argentina, 4, 10, 132, 192
Brazil, 4, 10, 119, 215
debt, 1, 2, 5, 30–52, 190, 222 Colombia, 4, 240, 241
Brazil, 114–30 EMU fiscal rules, 219, 220, 224–5,
ceilings, 147 226, 230
 Subject Index 275

decentralization – continued economic cycles

enforcement, 9–10 EU fiscal adjustment, 146
fiscal, 72 impact on budget, 148, 158
India, 7, 132, 198–218 Mexican fiscal rules, 143
Mexico, 4, 243 procyclical fiscal policies, 53
see also federalism; subnational stabilizing transfer rules, 8
rules subnational governments, 229
deficit bias, xii, 16–19, 22, 28, 53, 64n2 Economic and Monetary Union
Brazil, 123 (EMU), 1, 6, 97–113
fiscal responsibility laws, 61 Central and Eastern Europe, 7
Latin America, 57, 62, 108 subnational rules, 8, 219–34
Mexico, 134, 136, 143 see also European Union; Maastricht
political pressures, 60, 62 Treaty; Stability and
procyclical fiscal policy, 56, 57, 62 Growth Pact
rule flexibility, 59 Ecuador
Venezuela, 176 budget-balance rule, 2, 3, 4
see also budget deficits; debt debt ratio, 41
deficit illusion, 18 debt rule, 3
Denmark dollarization, 2
debt-revenue ratio, 31–2 earmarking of revenue, 72
subnational expenditure, 228 fiscal crisis, 56, 67, 68, 69
depletion policy, 167 fiscal responsibility law, 61
deregulation, 238–9 moral hazard, 74
devaluations, 67, 73 oil stabilization fund, 59, 60, 169
discretion, 1, 16, 17, 108 taxation, 79n23, n30, n31
Argentina, 192, 195 value-at-risk methodology,
countercyclical, 4, 9, 80n43 52n16
cyclical adjustments, 49 EDP see Excessive Deficit Procedure
intergovernmental transfers, 241, education, 54, 55–6, 174, 247
242, 244, 245 electorate support, 8–9, 18
procyclical, 76 EMU see Economic and Monetary Union
dollarization, 2, 40, 48, 51 endowment funds, 169
downturns, 56, 57–8, 107, 151, 236–7, enforcement, 9–10, 21, 90, 186
240, 246 Argentina, 194
see also recession budgeting principles, 82–3
due process in budgeting, 82–3, 85 game theoretic approach,
“Dutch disease”, 89, 166 186, 187–8
India, 207, 214
earmarking revenue, 72, 136–7, lack of, 183
241, 244 Maastricht Treaty, 27
ECB see European Central Bank political commitment, 88
ECLAC see Economic Commission self-enforcement, 185,
for Latin America and the 187–8, 214
Caribbean Stability and Growth Pact, 6
ECOFIN see European Union Council subnational rules, 183, 184, 223
of Economy and targets, 92
Finance Ministers technologies of, 192;
Economic Commission for Latin see also monitoring; penalties;
America and the Caribbean sanctions
(ECLAC), 113n22 entitlements, 85
276 Subject Index

environmental pollution, 226 expectations, 16, 67, 71

equity, intergenerational, 168, 175 credible commitments, 90
Estonia due process, 83
budget-balance rule, 3, 4, 76 expenditure rules, 2, 9, 76, 115–16,
currency board, 2 123, 147
fiscal performance, 93 see also government expenditure
policy guideline, 4, 9
ethical issues, 20 Federal Expenditure Budget (PEF),
European Central Bank (ECB), Mexico, 133, 134, 135, 136–7
98, 156 Federal Revenue Law (LIF), Mexico,
European Commission, 102, 104, 135, 136
113n22, 146, 147, 148–9, federalism, 4, 89, 219–20
158–9, 221 Argentina, 192–6
European Union Council of EMU fiscal rules, 220–4
Economy and Finance Ministers fiscal, 220–4, 235
(ECOFIN), 27, 29n13, 99, 102, 220, market-preserving, 216
221, 223 see also decentralization;
European Union (EU) subnational rules
countercyclical fiscal policy, 53 finance ministers, 23, 58, 81,
deficit constraints, 23–4 86, 87, 89
enlargement, 7, 146–63 financial crises, 5–6
numerical targets, 29n9 contagion, 109
procyclical fiscal policy, 154 procyclical fiscal policy, 56, 64
quantitative targets, 132 vulnerability, xii, 1, 104
rule-based fiscal policy, 1 see also capital account crises;
subnational rules, 8, 185, 219–34 shocks
see also Economic and Monetary financial markets
Union; Maastricht Treaty; reputation, 76
Stability and Growth Pact volatility, 55
exceptionality, 102 financial transactions tax, 74, 239
EU accession countries, 146–63 Finland
Excessive Deficit Procedure (EDP), cyclical component of
21, 101, 102–3 balance, 149
exchange rate subnational expenditure, 228
Brazil, 114–15, 117, 125, 126–7, fiscal discipline, 15, 16, 19,
128, 130n20 24–5, 81
budget balance adjustments, 108 Brazil, 115, 125, 128
debt servicing, 32 budget practices, 82, 86
dollarization, 51 EU accession countries, 159
fiscal risk, 50 fiscal rules impact on, 183, 185
fiscal sustainability, 78n9 Germany, 89
floating, 50–1 Latin America, 104
foreign currency debt, 39, 40, Mexico, 132, 134, 135
52n11 Netherlands, 89
import gap, 44, 45 political commitment to, 87, 98
instability, 109 resource-abundant countries, 177
pegs, 66, 67, 69 Stability and Growth Pact, 103
shocks, 45, 46, 47, 106 subnational governments, 221–2
volatility, 39, 40, 41–2, 47, 52n12 targets, 91
see also currency fiscal illusion, 18
 Subject Index 277

fiscal impulse, 78n13, 166 India, 198–218

fiscal policy Mexico, 131–45
capital account crises, 66–9 numerical, 9–10, 22–4, 98, 99, 111
capital mobility, 66–80 origin of, 184, 185–6
constraints, xii, 15 policy rules, 1–5, 8–9, 22–3, 76,
countercyclical, 5, 9, 50, 53, 57: 115–16, 131, 137–43
commodity stabilization political economy perspective,
funds, 60; credibility, 58, 59; 15–29
deficit bias, 19; discretion, 4, 9, political issues, 85, 87, 88, 91
80n43; EU deficit threshold, procedural rules, 5, 15, 22–4, 81, 82–3,
220; Latin America, 56; 98–9, 123, 131
structural balance rules, 62, 157; reduction of macroeconomic
subnational level, 235 volatility, 53–65
execution procedures, 15 resource-abundant countries,
procyclical, 6, 28, 47, 48, 64, 106–7: 164–79
Argentina, 238; booms, 59; stabilizing transfers, 8, 10, 59,
capital flight, 70; causes of, 56–7; 61, 63–4, 235–49
credibility problems, 63; cyclical volatility, 53–65
safety margin, 150; deficit limit, see also budget-balance rules; debt
155, 156; discretion, 76; effects rules; expenditure rules; “golden
of, 54–6; EMU fiscal rules, 223; rule”; structural balance rule;
EU candidate countries, 7, 147, subnational rules; targets
151–4, 159–60; Europe, 97; fiscal fiscal stance, 78n13
responsibility laws, 61; Mexico, flexibility
243; resource-abundant credibility of rules, 22–3, 58–9
countries, 164, 166; subnational EMU fiscal rules, 219, 220, 223, 224
rules, 225; volatility, 5, 53, 54, intergovernmental transfers, 247
55, 62 foreign currency debt, 38–40,
see also fiscal rules; policy signaling 47–8, 75, 107
Fiscal Responsibility and Budget France
Management Law (FRBM), India, debt, 100
7–8, 199, 205–14, 216 deficit limit, 154
Fiscal Responsibility Law (FRL), 4, fiscal discipline, 99
61, 110–11 subnational expenditure, 228
Brazil, 6, 61, 93, 110, 115, FRBM see Fiscal Responsibility and
121–4, 128 Budget Management Law, India
flexibility, 59 free-rider problem, 19, 221, 223
India, 7–8, 199, 205–14, 216 FRL see Fiscal Responsibility Law
Mexico, 6
New Zealand, 4, 80n43, 116, 131 game theory, 7, 183–97
structural balance rules, 63 GDP see gross domestic product
see also legislation Germany
fiscal rules, xii, xiii, 1–11 balanced-budget rule, 131
Brazil, 114–30 budget deficit, 24, 112n7, 154, 155
commitment to, 87, 88 budget restructuring, 88
definition, 2, 15, 115, 131, 183 debt, 100
effectiveness, 19–22, 82, 184–5 federalism, 89, 219–20, 227
EU enlargement, 146–63 fiscal discipline, 99
European Economic and Monetary subnational expenditure, 228, 229,
Union, 97–113, 219–34 234n20
278 Subject Index

golden rule, 2, 20, 228 EU accession countries, 151, 152,

Brazil, 122–3 153, 159
“compensated”, 227, 230 growth, 43, 54, 55, 139–40, 152, 204
EU accession countries, 151, revenue ratio, 150, 151, 154, 214
156, 159 shocks, 45, 46, 47
India, 207 social expenditures, 56
Mexico, 137 structural balance rule, 62
Venezuela, 176 volatility, 37, 38
government expenditure growth, xii, 4, 55
Argentina, 238 Brazil, 116, 125, 126, 128
Brazil, 119, 121, 125 budget deficits, 146–7
budgeting process, 82, 84 India, 198, 199, 204–5, 209, 212–13
controls on, 72 M4, 145n24
cuts in, 100, 101 Mexico, 135, 139–40
deficit bias, 18, 19, 60–1 Netherlands, 89
emerging economies, 93 resource-abundant countries, 164, 177
EMU fiscal rules, 223, 224 structural close-to-balance rule, 157
EU accession countries, 152 volatility, 54
fiscal illusion, 18 guarantees, 237, 245, 248
“gap” plus elasticity approach,
148–9 hidden liabilities, 114, 115, 117,
Germany, 89–90 125, 128
high spending, 190 see also contingent liabilities
India, 201, 202–3, 208–9, 212, Honduras, decline in GDP growth, 43
214–15, 216 human capital, 54, 55, 188
Mexico, 132, 134, 135, 136, 138 Hungary
Norway, 170, 174 fiscal indicators, 150, 151, 152,
stabilizing intergovernmental 160, 161
transfers, 61, 63, 235–49 uniform import surcharge, 80n32
structural close-to-balance rule,
156–7, 158, 159 IFIs see international financial
subnational, 61, 63, 223–4, 228–9, institutions
231–2, 235–49 IMF see International Monetary Fund
Venezuela, 174–5 implementation, 4–5
see also budget; expenditure rules; import gaps, 44, 45
investment import surcharge, 74–5, 80n32
Gramm-Rudman-Hollings Deficit incrementalism, 84, 85
Reduction Act, 21–2, 25, 27 India
Greece budget-balance rule, 2, 3
fiscal adjustment, 100–1 debt ratio, 7–8, 30
investment, 223 decentralization, 4, 7, 10, 132,
gross domestic product (GDP) 198–218
cyclical component of balance, 149 Fiscal Responsibility and Budget
debt ratio, 4, 9, 30–2, 38, 40, 45–7: Management Law, 7–8, 199,
Brazil, 114, 116–19, 123, 125–6, 205–14, 216
128; India, 7–8, 208, 209, 212; Indonesia
Latin America, 106; Maastricht budget-balance rule, 3, 11n10, 131
rules, 99, 100, 220; Mexico, capital account crisis, 68, 75
140, 142 moral hazard, 74
deficit limit, 53, 146, 155, 220 policy guideline, 4
 Subject Index 279

inflation Latin America, 106

bias, 16–17 Mexico, 139–40
Brazil, 120, 127 shocks, 45, 46, 47, 106
debt servicing, 48 structural close-to-balance
European Central Bank, 156 rule, 158–9
India, 201, 202 volatility, 39, 41–2, 44, 47–8,
Latin America, 106 52n12
Mexico, 140 intergenerational equity, 168, 175
structural close-to-balance rule, intergovernmental relations, 192,
158–9 193, 194, 195
inflation-targeting, 2, 10n9 intergovernmental transfers, 8, 10,
Brazil, 116, 127, 128 59, 61, 63–4, 76, 235–49
European Central Bank, 156, 157 international financial institutions
Mexico, 131 (IFIs), 64
Norway, 174 International Monetary Fund (IMF)
informal rules, 9 capital account crises, 70
information cyclical adjustments, 49
asymmetric, 18, 56, 57, 58, fiscal rules, xii
66, 67 Norway, 179n19
budgeting principles, 82 structural balance framework,
incomplete, 24 63, 64
transparency, 213 structural close-to-balance
innovations rule, 160
budget, 6, 91–2 transparency, 75
institutional, 81, 92–3, 119–25 investment
instability see volatility “compensated” golden rule,
institutional infrastructure, xiii, 227, 230
75, 77, 80n37 deficit limit, 156
institutionalism, 86, 90 EMU fiscal rules, 223
institutions, 85–90, 115 EU accession countries, 146,
fiscal outcomes, 81, 82, 90 147, 151, 152, 154, 159,
innovations, 81, 92–3, 119–25 163n11
institution-building, 9 India, 201, 202–3, 205, 212, 215
international financial, 64 Mexico, 245
self-enforcement, 185 permit system, 227
weak, 107 real assets, 168
integration Venezuela, 170–1
European, 27, 29n9, 111 volatility, 54
Mercosur, 109 see also government expenditure
see also convergence investor sentiment, xii, 67, 69, 70,
interest rate 71, 73
Brazil, 114, 120, 125, 126, 128 Ireland, subnational expenditure,
capital account crises, 69 228
debt servicing, 32, 34, 43–4, Israel, Deficit Reduction Law, 25
48, 50, 73 Italy
Economic and Monetary Union, balanced-budget rule, 131
98, 103 debt ratio, 30
fiscal risk, 34, 35, 36 decentralization, 220, 227, 228,
India, 199, 201–2, 205, 209, 229, 231, 232
212, 213 subnational expenditure, 228–9
280 Subject Index

Japan macroeconomic policy, 2, 9

balanced-budget rule, 131 majoritarian regimes, 81, 87
budget restructuring, 88 market access, 28, 47, 49, 50, 216,
debt ratio, 30 222, 235
Jordan, debt ratio, 30 maturities, 51
MB see minimal benchmark
Keynesian effects, 100 medium-term budgetary frameworks
Korea (MBFs), 121, 123–4, 136, 206
banking restructuring, 74 medium-term expenditure frameworks
capital account crisis, 68, 75 (MTEFs), 91–2
Mercosur, 109
Latin America Mexico, 131–45
debt, 7, 40, 45–7, 48–9 banking restructuring, 74
fiscal performance, 104–7 budget restructuring, 88
fiscal rules, 1, 59–62 budget-balance rule, 2, 3, 6–7,
intergovernmental transfers, 11n10
235–49 capital account crisis, 67, 68, 69,
procyclical fiscal policies, 53, 54–7, 62, 75, 132, 138, 243
64, 106 constitutional provision, 4
stabilizing transfer rules, 8 cost of fiscal transition, 79n27
tax revenue shock, 38, 45, 46, 47 debt ratio, 30, 32, 41
volatility, 28, 47, 49, 53, 54–5, decentralization, 4, 243
104–5, 107 decline in GDP growth, 43
LDO see Annual Budget Guidelines fiscal discipline, 185
Law, Brazil fiscal performance, 243
legislation, xii, 4, 5, 15, 20 fiscal risk, 51
Argentina, 195, 196, 238–9 inflation-targeting, 2
Brazilian Annual Budget intergovernmental transfers, 236,
Guidelines Law, 123, 124, 125 243–5, 247
changing, 21, 25–7 Oil Stabilization Fund, 133, 134,
Colombia, 241, 242 135, 137–8
Indian Fiscal Responsibility and pensions, 74
Budget Management Law, 7–8, PRONAFIDE program, 139–41, 142,
199, 205–14, 216 143, 144
institutionalism, 86 state reserve funds, 248
Mexican Federal Revenue Law, minimal benchmark (MB), 149, 150
135, 136 monetary policy, 1, 48
penalties for law-breaking, 20–1 Brazil, 120, 127, 130n20
Venezuela, 174–5, 176, 177 “fear of floating”, 39
see also constitutional provision; rules, 131
Fiscal Responsibility Law Stability and Growth Pact, 157
liability dollarization, 40, 48 see also inflation-targeting
liberalization, 7, 198, 199, 201, 212 monetary union, 27
LIF see Federal Revenue Law, monitoring, 4–5, 21, 22
Mexico Mercosur, 109
Stability and Growth Pact,
Maastricht Treaty, 21, 27, 97, 99–101, 102, 103
103, 111, 220 subnational rules, 222, 227
see also Economic and Monetary see also audit mechanisms;
Union enforcement; surveillance
 Subject Index 281

moral hazard, 47, 74, 189 Organization of Petroleum Exporting

Brazil, 120 Countries (OPEC), 165, 170
India, 212 original sin, 38–9, 40–3, 47, 48–9,
intergovernmental transfers, 52n12, n13, 201
242, 244 Ouro Preto, Treaty of, 109
MTEFs see medium-term expenditure output gaps, 49, 50, 149, 152, 155
frameworks Pakistan, debt ratio, 30
multiple equilibria Panama, debt ratio, 41
models, 67 Paraguay
frameworks, 71 debt-revenue ratio, 31–2
games, 186–8 Pareto optimality, 186, 187
peer pressure, 107, 109, 222,
national fiscal council (NFC), 111 224, 227
national interest, 97, 98 PEF see Federal Expenditure
natural resources, 7, 76, 137, 164–79 Budget, Mexico
“negative-spotlight” effect, 22, 27 penalties, 20–1, 29n8, 58, 109
net worth, 166, 167, 177 EU accession countries, 146
Netherlands structural balance rule, 62
balanced-budget rule, 131 see also enforcement; sanctions
investment, 223 pensions
political will, 87, 88, 89 Brazil, 115, 127, 128
subnational expenditure, 228 India, 212
New Zealand structural reform, 6, 73, 74
Fiscal Responsibility Act, 4, 80n43, see also social security
116, 131 permanent income, 168, 173, 175
monitoring mechanisms, 21 permit system, 226–7
transparency, 132 Peru
NFC see national fiscal council budget-balance rule, 2, 3, 4
nonrenewable resources, 7, 76, 137, centralization, 4
164–79 debt ratio, 41
Norway fiscal responsibility law, 59, 61
nonrenewable resources, 7, 164–5, import gap, 44, 45
170, 171–4, 177 inflation-targeting, 2
State Petroleum Fund, 60, 170, procyclical fiscal policy, 63
171–2, 173, 174 quantitative targets, 132
taxation, 79n30
OECD see Organization for Economic Philippines, capital account crisis, 68
Cooperation and Development Poland
oil see commodity stabilization funds; constitutional provision, 4
nonrenewable resources debt rule, 3, 4
OPEC see Organization of Petroleum decentralization, 4
Exporting Countries fiscal indicators, 150, 151, 152,
optimal currency areas, 52n17 160, 161
optimal tax smoothing, 147, 163n12 inflation-targeting, 2
Organization for Economic uniform import surcharge, 80n32
Cooperation and policy guidelines, 4, 9, 62
Development (OECD) policy signaling, 5–6, 24–5, 71, 76,
debt, 45–7, 85 80n39, 166
GDP growth, 55 political economy perspective,
tax revenue shock, 38, 45, 46, 47 2, 15–29, 103, 111
282 Subject Index

political issues subnational rules, 225

Argentina, 193, 194 volatility, 5, 53, 54, 55, 62
cooperation, 191, 192 PRONAFIDE program, 139–41, 142,
deficit bias, 17–19, 60, 62 143, 144
political will, 6, 8, 25, 77, 81–94 public sector borrowing requirement
procyclical fiscal policy, 57 (PSBR)
reputation, 24 Brazil, 116
subnational rules, 183–4, 189 calculation of, 145n26
Portugal liabilities, 145n17
budget deficit, 112n7 Mexico, 132, 133, 134–5, 137–8,
fiscal adjustment, 100–1 139–42, 144
poverty, 54, 55–6, 171, 205, 212, 215 public sector liabilities, 73, 75, 77,
prices 116–17, 120
stability, 98, 131, 216 public services, 93
volatility of commodity prices, public spending see government
165, 166, 167 expenditure
primary balance, 9, 48, 49–50, 100
Brazil, 123, 124, 128 “quasi-stabilizers”, 108
India, 200
Mexico, 134 rainy-day funds, 222, 224, 225, 231,
structural balance rule, 137 233n15, 236, 245
see also budget recession
primary surplus, 49–50, 55, 69 Argentina, 237, 238, 239
Brazil, 49, 65n13, 76, 79n21, 114, 116, countercyclical expansion, 5
118–19, 121, 124–7 debt service, 48
Latin America, 107–8 EMU deficit limit, 220
principal-agent problems, 189 Latin America, 110
privatization, 75, 80n33 poverty, 56
Argentina, 238 procyclical fiscal policy, 55, 154
Brazil, 122 Stability and Growth Pact, 102
Mexico, 132 stabilizing transfers, 237, 247
procyclical fiscal policy, 6, 28, 47, structural close-to-balance rule, 157
48, 64, 106–7 structural deficit indicator, 150
Argentina, 238 see also downturns
booms, 59 reindustrialization, 166, 167
capital flight, 70 reputation
causes of, 56–7 fiscal discipline, 15, 16, 24–5
credibility problems, 63 game theoretic approach, 188
cyclical safety margin, 150 noncompliance, 5
deficit limit, 155, 156 policy signaling, 76
discretion, 76 resources, 7, 76, 137, 164–79
effects of, 54–6 see also commodity stabilization funds
EMU fiscal rules, 223 revenue see taxation
EU accession countries, 7, 147, risk
151–4, 159–60 cash flow, 166, 167
Europe, 97 debt, 5, 34, 48, 49, 50–1, 107
fiscal responsibility laws, 61 fiscal, 34–8, 47, 50–1
Mexico, 243 interest rate, 48, 203
resource-abundant countries, net worth, 166, 167, 177
164, 166 oil wealth, 173, 175, 177
 Subject Index 283

risk – continued social impact of capital account

original sin, 40 crises, 66, 72, 77
permit market, 226 social security, 72–3, 168
political, 168 Brazil, 115
sharing, 248 Mexico, 140, 142
spreading, 168 see also benefits; pensions
stabilizing transfers, 237 social welfare, 16, 19, 20, 170
Romania, fiscal indicators, 150, 151, Spain
152, 160, 162 debt ratio, 30, 32
rules see fiscal rules decentralization, 220, 227, 229,
Russia 231, 232
capital account crisis, 67, 68, 69, subnational expenditure, 228, 229
75, 120, 155 spillovers, 97–8, 109, 223
expenditure control, 72 stability, xii, xiii, 1
Brazil, 116
sanctions, 5, 21 capital account crises, 71, 72, 77
Brazilian fiscal rules, 122, 124 central government responsibility,
EMU rules, 97, 102–3, 221, 224–5 223
fiscal responsibility laws, 110–11 internal stability pacts, 228–9, 232
subnational rules, 222, 223, 230, 232 Latin America, 104
see also enforcement; penalties Norway, 173
savings, 175, 203 output, 103–4
securities, 201, 203 price, 98, 131, 216
seignorage, 67 resource-abundant countries, 177
self-enforcement, 185, 187–8, 214 stabilizing transfer rules, 8, 10, 59,
SGP see Stability and Growth Pact 61, 63–4, 235–49
shocks, 9, 45–7, 50, 94, 131, 248 see also commodity stabilization
commodity prices, 60 funds; volatility
EMU fiscal rules, 231 Stability and Growth Pact (SGP),
India, 204 6, 80n43, 97, 101–4, 116, 146
intergovernmental transfers, cyclical stabilization, 108
76, 240, 246 deficit limit, 220
Latin American debt EU accession countries, 155
structure, 106 Germany, 90
Mexican fiscal rules, 132, 135, medium-term equilibrium, 53
136, 143 sanctions, 21
oil, 171 structural close-to-balance
revenue volatility, 38, 45, 46, 47 rule, 157
rule flexibility, 59 structural deficit adjustment, 156
volatility, 55, 104 subnational governments, 225,
vulnerability to, 93 227, 228, 230, 231
see also financial crises see also Economic and Monetary
signaling, 5–6, 24–5, 71, 76, 80n39, 166 Union; European Union
Slovakia, uniform import surcharge, stabilizers, 9, 48, 53, 57–8, 61, 63
80n32 balanced-budget rule, 76
Slovenia capital account crises, 70
fiscal indicators, 147, 150, 151, EMU stabilization, 221
152, 153, 160, 162 EU accession countries, 147
growth rate stability, 163n7 Latin America, 108
social expenditures, 56, 62 Mexico, 135, 144
284 Subject Index

stabilizers – continued Venezuela, 176, 177

positive effects, 154 Sweden
rule design, 59 cyclical component of balance,
Stability and Growth Pact, 102, 149
108, 220 subnational expenditure, 228
structural balance rule, 137 Switzerland, balanced-budget
structural deficit indicator, 150 rule, 131
state enterprises, 75, 93, 120 targets, 29n9, 91–2, 98, 99,
structural balance rule, 64 115, 132
Chile, 4, 62, 63, 137, 145n16 Brazil, 124
ex ante close-to-balance rule, changing the law, 27
156–8, 159, 160 creative accounting, 22
Mexico, 137–9, 144 decentralized governments,
subnational government debt 224, 230
Argentina, 7, 239, 240, 246 fiscal responsibility laws, 111
Brazil, 119–20, 121–2, 123, 128 India, 207–8
Colombia, 240, 242–3, 246 Latin America, 110, 111
India, 206, 209, 213 Mexico, 141
Mexico, 133, 134, 143, 243, Stability and Growth Pact, 102
244–5 see also fiscal rules
see also decentralization tax smoothing, 147, 154, 163n12
subnational rules, 4, 5, 131, 216 taxation
Argentina, 7, 192–6 Argentina, 195, 238, 239
Brazil, 215 Brazil, 6, 115, 127, 128
European Union, 8, 219–34 Colombia, 241
game theoretic approach, 7, 183–97 debt-tax ratio, 30–2, 37–8, 47
intergovernmental transfers, 8, 10, distortionary, 19, 70
59, 61, 63, 64, 235–49 Ecuador, 79n23, n31
see also decentralization; federalism financial transactions, 74, 239
supplementals, 86 India, 200–1, 204, 214
surveillance Latin America, 79n30, 106
Europe, 101, 102, 103 Mexico, 132, 133, 138, 243,
Latin America, 107, 109, 110, 111 244, 247
see also monitoring Norway, 170
sustainability, xiii, 1, 53, 62 progressive income tax, 70
debt, 76, 77, 160 quality of fiscal adjustment, 71, 72
Brazil, 6, 114–16, 119, 125–8 reduction in revenue, 101
definition, 145n19 resources, 167
India, 208–13 revenue volatility shock, 38, 45,
Mexico, 137, 139–41, 142, 144 46, 47
definition, 80n38 structural close-to-balance
EMU fiscal rules, 220 rule, 158
endowment funds, 169 subnational rules, 189, 193, 223–4,
exchange rate stability, 78n9 225, 227, 229, 230, 231
India, 199, 208–13, 215, 216 weak administrative capacity, 73
policy signaling, 76 terms of trade, 9, 50, 204
quality of adjustment, 71 budget balance adjustments, 108
resource-abundant countries, 177 Latin America, 28
Stability and Growth Pact, primary balance, 49
102, 103 volatility, 37, 38
 Subject Index 285

Thailand uniform import surcharge, 74–5,

banking restructuring, 74 80n32
capital account crisis, 68, 75 United Kingdom (UK)
fiscal impulse, 78n13 budget restructuring, 88
time inconsistency, 16–17, 20, 57, 76 debt ratio, 30
trade expenditure rule, 147
India, 201, 203–4, 212 monitoring mechanisms, 21
see also terms of trade structural close-to-balance rule,
transaction costs, 189 158, 160
transparency, xiii, 9, 131–2 subnational expenditure, 228
Brazil, 121, 123 transparency, 132
budgeting principles, 82 United States (US)
commodity stabilization funds, budget-balance rule, 131
169, 178 debt ratio, 30, 32, 40
compliance, 23 deficit restrictions, 98
creative accounting, 22 golden rule, 20
credibility, 4, 75, 77 Gramm-Rudman-Hollings Deficit
fiscal responsibility laws, Reduction Act, 21–2,
110–11 25, 27
India, 205, 207, 213, 214, 215 import gap, 44, 45
Mexico, 134 subnational rules, 184–5
procedural rules, 15 transfers, 249n14
signaling, 6 Uruguay
stability pacts, 232 consumption cycle, 108
subnational accounting debt ratio, 42
practices, 232 US see United States
Venezuelan savings norm, 175
see also accountability Venezuela
trust funds, 90 budget-balance rule, 2, 3, 4
Turkey countercyclical fiscal policy, 56
capital account crisis, 67, 68, debt ratio, 42
69, 70, 75 domestic currency debt, 43
debt ratio, 30 nonrenewable resources, 165,
fiscal consolidation, 78n18 170–1, 174–7
fiscal impulse, 78n13 oil price, 108
oil stabilization fund, 7, 59, 60,
UK see United Kingdom 165, 174–7
uncertainty taxation, 79n30
Mexican fiscal rules, 131, 135 volatility, xii, 1, 5, 6, 9
resource earnings, 166, 167, 168, commodity prices, 165,
173, 177 166, 167
see also volatility debt, 32–4, 35–7, 38,
unemployment 39–40
compensation scheme, 70 EU accession countries, 7, 147,
elasticity of, 148–9 148, 151, 152, 155
inflation bias, 17 euro area, 105
local, 235 exchange rate, 39, 40, 41–2,
Norway, 173 47, 52n12
unemployment benefits, 148, 149, interest rate, 39, 41–2, 44, 47–8,
150, 158 52n12
286 Subject Index

volatility – continued revenue, 38, 45, 46, 47

Latin America, 28, 47, 49, 53, see also stability; uncertainty
104–5, 107
Mexico, 132–3, 135, 137 “Wagner’s Law”, 93
output, 7, 147, 148, 149, 151, welfare maximization, 16–17
152, 155
reduction of macroeconomic, 53–65 zero PSBR rule, 139, 140, 141, 143
resource earnings, 169, 174 zero-based budgeting, 26, 84