The Architecture of Government

Since the days of Montesquieu and Jefferson, political decentralization has been seen as a force for better government and economic performance. It is thought to bring government “closer to the people,” nurture civic virtue, protect liberty, exploit local information, stimulate policy innovation, and alleviate ethnic tensions. Inspired by such arguments, and generously funded by the major development agencies, countries across the globe have been racing to devolve power to local governments.

   This book reexamines the arguments that underlie the modern faith in decentralization. Using logical analysis and formal modeling and appealing to numerous examples, it shows that most such arguments are based on vague intuitions or partial views that do not withstand scrutiny. A review of empirical studies of decentralization finds these as inconclusive and mutually contradictory as the theories they set out to test. The book’s conclusion – that one can say just about nothing in general about when decentralizing will be beneficial and when harmful – promises to prompt a rethinking of both the theory of political decentralization and current rationales for development aid.

Daniel Treisman is a professor of political science at the University of California, Los Angeles. He is the author of After the Deluge: Regional Crises and Political Consolidation in Russia (1999) and (with Andrei Shleifer) Without a Map: Political Tactics and Economic Reform in Russia (2000). A recipient of fellowships from the John Simon Guggenheim Foundation, the German Marshall Fund of the United States, the Hoover Institution, and the Smith Richardson Foundation, he has published broadly in academic journals, including the American Political Science Review, the American Economic Review, the British Journal of Political Science, and World Politics, as well as policy journals such as Foreign Affairs and Foreign Policy.

---

Cambridge Studies in Comparative Politics

General Editor

Margaret Levi   University of Washington, Seattle

Assistant General Editor

Stephen Hanson   University of Washington, Seattle

Associate Editors

Robert H. Bates   Harvard University
Peter Lange   Duke University
Helen Milner   Princeton University
Frances Rosenbluth   Yale University
Susan Stokes   Yale University
Sidney Tarrow   Cornell University
Kathleen Thelen   Northwestern University
Erik Wibbels   University of Washington, Seattle

Other Books in the Series

Lisa Baldez, Why Women Protest: Women’s Movements in Chile

Stefano Bartolini, The Political Mobilization of the European Left, 1860–1980: The Class Cleavage

Mark R. Beissinger, Nationalist Mobilization and the Collapse of the Soviet State

Nancy Bermeo, ed., Unemployment in the New Europe

Carles Boix, Democracy and Redistribution

Carles Boix, Political Parties, Growth, and Equality: Conservative and Social Democratic Economic Strategies in the World Economy

Catherine Boone, Political Topographies of the African State: Territorial Authority and Institutional Change

Michael Bratton and Nicolas van de Walle, Democratic Experiments in Africa: Regime Transitions in Comparative Perspective

Michael Bratton, Robert Mattes, and E. Gyimah-Boadi, Public Opinion, Democracy, and Market Reform in Africa

Continued after the Index

---

The Architecture of Government

    RETHINKING POLITICAL DECENTRALIZATION

  DANIEL TREISMAN

  University of California, Los Angeles

---

CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi

Cambridge University Press
32 Avenue of the Americas, New York, NY 10013–2473, USA

www.cambridge.org
Information on this title: www.cambridge.org/9780521872294

© Daniel Treisman 2007

This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 2007

Printed in the United States of America

A catalog record for this publication is available from the British Library.

Library of Congress Cataloging in Publication Data
Treisman, Daniel.
The architecture of government : rethinking political decentralization / Daniel Treisman.
   p. cm. – (Cambridge studies in comparative politics)
Includes bibliographical references and index.
ISBN-13: 978-0-521-87229-4 (hardback)
ISBN-13: 978-0-521-69382-0 (pbk.)
1. Decentralization in government. 2. Central-local government relations.
3. Federal government. 4. Comparative government. I. Title. II. Series.
JS113.T74  2007
320.8 – dc22      2006032899

ISBN 978-0-521-87229-4 hardback
ISBN 978-0-521-69382-0 paperback

Cambridge University Press has no responsibility for
the persistence or accuracy of URLs for external or
third-party Internet Web sites referred to in this publication
and does not guarantee that any content on such
Web sites is, or will remain, accurate or appropriate.

---

To Alex and Lara

---

Contents

List of Figures and Tables			page xi
Preface			xiii
Glossary of Main Notation Used			xv
1		INTRODUCTION	1
		1.1  A Quick Look Back	6
		1.2  The Arguments	11
		1.3  A Note on Methods: Formal Modeling	15
		1.4  Overview	19
2		THE POLITICAL PROCESS	21
		2.1  Defining Decentralization	21
		2.2  Modeling Politics	27
3		ADMINISTRATIVE EFFICIENCY	53
		3.1  Optimal Scale	55
		3.2  Heterogeneous Tastes and Policy Differentiation	59
		3.3  Costs of Organization	62
		3.4  Conclusion	72
4		COMPETITION AMONG GOVERNMENTS	74
		4.1  Competing for Mobile Residents	75
		4.2  Competing for Mobile Capital	87
5		FISCAL POLICY AND REDISTRIBUTION	104
		5.1  The “Common Pool”	106
		5.2  The “Soft Budget Constraint”	108
		5.3  Decentralizing Redistribution	131
6		FISCAL COORDINATION AND INCENTIVES	137
		6.1  Vertical “Overgrazing”	139
		6.2  Fiscal Decentralization and Incentives	146
7		CITIZENS AND GOVERNMENT	156
		7.1  Civic Virtue	156
		7.2  Accountability	164
		Appendix: Modeling Retrospective Voting with Distributive Politics	184
8		CHECKS, BALANCES, AND FREEDOM	193
		8.1  Freedom	194
		8.2  Policy Stability	201
9		ACQUIRING AND USING KNOWLEDGE	209
		9.1  Information	209
		9.2  Policy Experimentation	222
10		ETHNIC CONFLICT AND SECESSION	236
		10.1  Satisfying Limited Demands for Autonomy	238
		10.2  Splitting the Prizes of Politics	242
		10.3  Restraining the Central Government	243
		10.4  Socializing Politicians	244
		10.5  Stimulating Growth of Small Ethnic Parties	244
		10.6  Conclusion	245
11		DATA TO THE RESCUE?	247
		11.1  The Quality of Government	251
		11.2  Economic Performance	258
		11.3  Ethnic Conflict	262
		11.4  Democracy	264
		11.5  Stable Policies	267
		11.6  Conclusion	268
12		CONCLUSION: RETHINKING DECENTRALIZATION	270
		12.1  Possible Objections	275
		12.2  Explaining Decentralization’s Appeal	283
		12.3  A New Agenda?	289
References			295
Index			321

---

Figures and Tables

Figures

2.1.		The political process	29
2.2.		The revenue (Laffer) curve	48
3.1.		Cost per resident of providing g units of a local public good to μ residents	57
3.2.		Cost per resident of providing g units of local public good ω = {1, 2} to μ residents	57
3.3.		Communication costs and the number of tiers	66
6.1.		Effects of fiscal decentralization on local and central bribe rates	152
11.1.		Decentralization and corruption, 1980s–1990s	253
11.2.		Decentralization and inflation, 1980s–1990s	261
11.3.		Decentralization and democracy, 1980s–1990s	265

Tables

2.1.		Types of decentralization	28
9.1.		Expected payoffs under decentralization in locality n	220
9.2.		Expected payoffs under centralization, pA = pB ≡ pn	221
9.3.		Expected payoffs under centralization, pA < pB < 1/2	221

---

Preface

This book has been a long time in the making. It began as an empirical project. I wanted to see what difference decentralized political institutions make for economic performance and the quality of government. The common presumption in Western democracies seemed to be that devolving power to autonomous local governments produced a number of important benefits. In the developing world, international aid agencies were backing reforms to decentralize responsibilities and resources in an ever-lengthening list of countries. Studying the politics of postcommunist Russia and reading about Latin America, I had grown skeptical that powerful local governments were quite as unmixed a blessing as was generally believed. I thought I would examine the empirical record.

   Having collected data about the structure of government in countries around the world, I set aside what I thought would be a couple of months to work through the logic of the arguments about decentralization I would use the data to test. Five years went by. Along the way, I became convinced that – with one exception – there was no compelling reason to think that decentralized political institutions have any predictable effect at all. The one more persuasive argument – that some kinds of decentralization slow the pace of policy change – had no implications about whether decentralization was good or bad: It could be either, depending on what kind of change was being prevented.

   During this unplanned journey into the recesses of institutional theory, I have benefited tremendously from conversations with and suggestions from a great many colleagues, all of whom are, of course, blameless for any defects in the final result. Some read bits and pieces, some listened and responded, others suggested directions worth exploring. I am grateful to Yoram Barzel, Pablo Beramendi, Tim Besley, Richard Bird, Thierry de Montbrial, J. R. DeShazo, Alberto Diaz-Cayeros, Tim Frye, Stephan Haggard, Stephen Hanson, Michael Hechter, Torben Iversen, Edgar Kiser, Herbert Kitschelt, Anirudh Krishna, Margaret Levi, Kirstie McClure, Scott Morgenstern, Aseem Prakash, Antonio Rangel, Karen Remmer, Jonathan Rodden, Ron Rogowski, Gérard Roland, Tom Romer, Ken Scheve, Matt Singer, David Soskice, Mariano Tommasi, Michel Treisman, Barry Weingast, Susan Whiting, Erik Wibbels, and two anonymous readers, as well as seminar participants at Princeton; the University of Washington; Duke University; University of California, San Diego; and the American Political Science Association meetings. I imposed more than once on the intellectual firepower of Andrei Shleifer, George Tsebelis, and Jean-Laurent Rosenthal, and I am grateful for their generosity. I owe a great debt to my collaborator, Hongbin Cai, with whom I have been working on the logic of decentralization; Hongbin’s contributions are very evident in this book. Yi Zhang, Ani Sarkissian, Linda Choi, Matias Iaryczower, and Rolf Campos provided excellent research assistance. I thank Margaret Levi and Lew Bateman at Cambridge University Press for their patient interest in the manuscript and the John Simon Guggenheim Foundation, the German Marshall Fund of the United States, and the UCLA Academic Senate and Social Sciences Division for financial support.

   I am grateful to my family for continuing to ask how the book is coming along. I thank my wife, Susi, for her encouragement and companionship. I dedicate this book to Alex and Lara, whose lives have overlapped with its gestation. Although they do not yet know how to spell decentralization, both are strong believers that many decisions in the Treisman household – especially those concerning ice cream and bedtime – would be better made if decentralized to those with the most direct interest in the outcome.

---

Glossary of Main Notation Used

M (m = 1, 2, . . . M)	indexes citizens
J (j = 1, 2, . . . J)	indexes tiers of government
N (n = 1, 2, . . . N)	indexes governments within a tier
I (i = 1, 2, . . . I)	indexes government official within a particular government
T (t = 1, 2, . . . T)	indexes time period
W (w = 1, 2, . . . W)	indexes public good or service provided by a particular government
l	labor supply of individual
L ≡ Σmlm	total labor supply
k	capital endowment of individual
K ≡ Σmkm	total capital endowment
I	public infrastructure investment
g	subnational provision of public goods or services
G	central provision of public goods or services
H(.),h(.),q(.),v(.),z(.)	increasing, concave subutility functions
t	subnational lump-sum tax level
T	central lump-sum tax level
τ	subnational income tax rate
Τ	central income tax rate
r	according to context: cash transfer; interest rate; share of shared tax
R	government revenue
y	income, output of individual
Y ≡ Σmym	total income, output
c	government consumption
s	consumption by citizens of privately supplied goods
f(.)	increasing, concave production function
γ(g–n)	utility of residents of jurisdiction n from externalities from spending in other jurisdictions at the same tier
μ	community size
δ	discount rate