BookPDF Available

The institutional economics of corruption and reform: Theory, evidence, and policy


Abstract and Figures

Corruption has been a feature of public institutions for centuries yet only relatively recently has it been made the subject of sustained scientific analysis. Lambsdorff shows how insights from institutional economics can be used to develop a better understanding of why corruption occurs and the best policies to combat it. He argues that rather than being deterred by penalties, corrupt actors are more influenced by other factors such as the opportunism of their criminal counterparts and the danger of acquiring an unreliable reputation. This suggests a novel strategy for fighting corruption similar to the invisible hand that governs competitive markets. This strategy - the ‘invisible foot’ - shows that the unreliability of corrupt counterparts induces honesty and good governance even in the absence of good intentions. Combining theoretical research with state-of-the-art empirical investigations, this book will be an invaluable resource for researchers and policy-makers concerned with anti-corruption reform.
Content may be subject to copyright.
The Institutional Economics of
Corruption and Reform
Theory, Evidence, and Policy
Corruption has been a feature of public institutions for centuries, yet
only relatively recently has it been made the subject of sustained sci-
entific analysis. In an important contribution to this ongoing project,
Johann Graf Lambsdorff shows how insights from institutional eco-
nomics can be used to develop a better understanding of the institutions
necessary to carry out corrupt transactions and those that are helpful in
inhibiting them. He argues that rather than being deterred by penalties,
corrupt actors are more influenced by other factors such as the
opportunism of their criminal counterparts and the danger of acquiring
a reputation of unreliability. This suggests a novel strategy for ghting
corruption the ‘‘invisible foot’ whereby the unreliability of corrupt
counterparts induces honesty and good governance even in the absence
of good intentions. Combining interdisciplinary theoretical research
with state-of-the-art empirical investigations, this book will be an
invaluable resource for researchers and policy-makers concerned with
anticorruption reform.
johann graf lambsdorff is Chair in Economic Theory at the
University of Passau, Germany, and Senior Research Consultant to
Transparency International. He is the founder of the Corruption
Perceptions Index, Which he has orchestrated since 1995.
The Institutional
Economics of
Corruption and
Theory, Evidence, and
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sa
o Paulo
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
Information on this title:
ª Johann Graf Lambsdorff 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 Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloging in Publication Data
Lambsdorff, Johann, Graf, 1965–
The institutional economics of corruption and reform: theory, evidence, and policy/
Johann Graf Lambsdorff.
p. cm.
Includes bibliographical references and index.
ISBN-13: 978-0-521-87275-1 (hardback: alk. paper)
ISBN-10: 0-521-87275-8 (hardback: alk. paper)
1. Administrative agencies Corrupt practices Prevention. 2. Political
corruption Prevention. 3. International finance Corrupt practices Prevention.
4. Institutional economics Sociological aspects. I. Title.
JF1525.C66L36 2007
6 dc22
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication, and
does not guarantee that any content on such websites is, or will remain, accurate or
To Maria, Philipp, Frederik, and Felix
List of boxes page viii
Acknowledegements ix
A roadmap to this book xi
1 Introduction 1
2 Enemies of corruption 27
3 What is bad about bureaucratic corruption?
An institutional economic approach 58
4 The dilemma of the kleptocrat: What is bad
about political corruption? 81
5 Corruption and transactions costs: The rent-seeking
perspective 109
6 Making corrupt deals: contracting in the shadow
of the law 136
7 Exporters’ ethics and the art of bribery 164
8 How confidence facilitates illegal transactions:
An empirical approach 190
9 Corrupt relational contracting 209
10 Concluding thoughts 225
Appendix: Technical details to the Transparency
International Corruption Perceptions Index 236
References 256
Subject index 282
List of boxes
Box 1 Corruption and the size of the public sector page 4
Box 2 Corruption and decentralization 6
Box 3 Corruption and regulatory quality 10
Box 4 Corruption and competition among private firms 12
Box 5 Sources of the 2005 Corr uption Perceptions Index 21
Box 6 Corruption and trust 29
Box 7 Corruption, values, and colonialism 32
Box 8 Corruption and gender 34
Box 9 Corruption and geographic preconditions 35
Box 10 Corruption and official wages 37
Box 11 Corruption and democracy 39
Box 12 Corruption and parliamentarism 41
Box 13 Corruption and voting systems 43
Box 14 Corruption and press freedom 46
Box 15 Corruption and the judiciary 47
Box 16 Corruption and income per head 72
Box 17 Corruption and growth of GDP 73
Box 18 Corruption, productivity, and quality 76
Box 19 Corruption and distorted budget allocation 88
Box 20 Corruption and inequality 91
Box 21 Corruption and the underground economy 95
Box 22 Corruption and investments 100
Box 23 Corruption and the composition of investments 104
Box 24 Different types of corruption and investment 105
Box 25 How to fix a football match, by Declan Hill 169
Box 26 Corruption distorting the private sector 174
Box 27 Hindering corrupt intermediaries 184
Box 28 Fighting corruption with asymmetric penalties,
jointly written with Mathias Nell 229
This book is a twelve-year project, combining research efforts since
1994. At an early stage, some parts of it became highly prominent:
The Corruption Perceptions Index, which was designed in 1995 and is
compiled under my leadership at the University of Passau since then,
on behalf of Transparency International. This index was not pre-
dominantly designed to serve as a tool for alerting politics and the
public at large. Its initial aim was to serve as a starting point for
academic research. Considerable progress has been made on this front
of research since then the empirical investigations reviewed here
provide evidence on this.
Many friends and colleagues have been helpful in providing critical
comments. I would like to thank J. Ahrens, I. Amundsen, L. Bajec,
H. Davoodi, D. della Porta, W. Easterly, B. Efron, G. Engel,
O.-H. Fjeldstad, B. Frank, P. Heywood, A. K. Jain, P. Manow, P. Mauro,
K. Meyer, H. Mo
ller-deBeer, B. Mukherjee, M. Nell, G. Pfeffermann, J.
Pope, S.Rose-Ackerman, M. Schinke,C.Schinke,A. Schmidt, R. Seubert,
V. Tanzi, U. Teksoz, and an armada of anonymous referees. I am grateful
to two anonymous referees for helping me integrate the diverse concepts
presented here into a consistent book.
Some special thanks go to W. Zucchini, who was invaluable in
refining the statistical methodology for the Corruption Perceptions
Index and devoted much time and effort to designing appropriate
algorithms. I owe many thanks to H.-J. Jarchow, who supervised my
habilitation at the University of Go
ttingen. He provided me with the
freedom to devote my efforts to a new area of research and supported
me with his critical accuracy. Further special thanks are due to Peter
Eigen. Without his organizational support through Transparency
International much of my work would not have had the opportunity
to reach out to a broader audience already at its early stages. Similar
thanks go to Frank Vogl, Jeremy Pope, George Moody-Stuart
Lawrence Cockroft, and many others within Transparency Inter-
national. They established a community that injects spirit into what is
otherwise bloodless research. Although this book is dedicated to
academic rigor I hope that some of this spirit remains.
University of Passau, October 06
A road map to this book
Those who are willing to carry out corrupt acts lose the capacity to
commit to honesty. This is the core argument developed and exploite d
in this book. Corrupt actors can neither commit to honestly serving
the public nor credibly promise reciprocity to their corrupt counter-
parts. This thought is at the center of understanding the disastrous
economic and social consequences of corruption. At the same time,
this concept deserves to be placed at the center of reform. Bribe -takers
and bribe-givers have a schizophrenic relationship to honesty. They
betray their superiors and the public but attempt to signal recip rocity
to each other and often fail in doing so. A strategy for reform must
exploit this failure. The Achilles’ heel of corrupt transactions is that a
briber often does not know what he will get in return. This is a crucial
weakness of those who are willing to engage in corrupt transactions.
Anticorruption can therefore take the tactic of a judo-fighter
someone who exploits his opponents’ weaknesses.
The power of economic thinking started with the concept of the
invisible hand. Competition substituted for benevolence by guiding
self-seeking actors to serve the public. With respect to fighting cor-
ruption we do not have such a powerful mechanism. If something
comes close to it, it is the corrupt actor’s capacity to betray each
other. This betrayal is a good thing. I call this the principle of the
invisible foot.
The willingness to take bribes works agains t the cor-
rupt actors. Anticipating this, even self-seeking individuals may have
reason to commit to honesty rathe r than seek opportunities for bribes.
This book does not provide readers with recipes on how to fight
corruption. Instead of designing a toolbox it rather presents a meth-
odological approach that, I hope, will inspire anticorrupt ion in the
This term was originally invented by Brock and Magee (1984). They used the
term to indicate welfare losses arising in rent-seeking competition. My usage is
different here.
future. This is repeatedly supported by cases and examples. Chapter 1
presents methodological and quantitative details. How is corruption
defined, how is it measured? Why did some older guiding principles
for anticorruption fail? I argue that corruption is not eliminated by
fighting on a different battleground, such as reducing government,
decentralizing public decision-maki ng, privatizing publi c property, or
enhancing competition. Corruption is a genuinely new challenge that
requires novel answers. The invisible hand of competition brings
about good markets but not good governance. This is shown in
Chapter 1.
Chapter 2 surveys some empirical evidence on anticorruption,
including its sometimes poor performance, and starts to present the
core concept of this book. Having dismissed many alternative
approaches to fighting corruption, the principle of the invisible foot is
Chapters 3 and 4 explore that the corrupt actors’ lack of com-
mitment to honesty is at the core of the welfare losses of corruption.
In Chapter 3, this idea is developed for bribe-taking bureaucrats.
There should be no doubt that bribe-taking is often a utility-
maximizing strategy of a self-seeking bureaucrat. But the downside of
one’s willingness to take bribes is that such actors disqualify for
professions where their commitment would be vital.
A similar thought arises for heads of government, as explored in
Chapter 4. They might transfer public funds into their private pock-
ets. But they are not trusted by inves tors if they disrespect law. The
advantage from bribery turns against its actors. This is at the core of
understanding the social costs of grand corruption, the type of cor-
ruption that takes place higher up in hierarchy.
Chapters 3 and 4 at the same time provide reader s with an up-to-
date assessment of research, both empirical and theoretical. Such an
assessment is also provided in the various boxes of this book, which
allow readers to obtain a quick grasp of empirical research. I hope
such details do not distract readers from the core message of the book.
There is a multitude of social costs invoked by corruption. This
ascertains that we cannot avoid the negative consequences of cor-
ruption but must fight corruption itself. We are given no alternative
but to devote our efforts to reducing corruption.
Chapter 5 asks whether we should facilitate or impede corruption.
The answer appears straightforw ard in favor of the latter option.
A road map to this bookxii
But we must note that the traditional rent-seeking theory argued
differently. It opted in favor of facilitated corruption because other-
wise competition for preferential treatment would waste resour ces for
lobbyism, engaging lawyers for lawsuits, or harassing politicians with
public campaigns. I show that this conclusion is misguided our
effort must be directed toward increasing the transaction costs of
Chapter 6 shows how in practice corrupt actors attempt to secure
reciprocity. This chapter might be misunderstood as a how-to-bribe
guide for criminal actors. But learning how to arrange bribes is a
fruitful starting point for reform. We must understand our enemies if
we want to defeat them. We must understand corrupt actors’ temp-
tation to betray each other in order to encourage precisely this
Chapter 7 brings our thought to the international arena. The
invisible hand of competition brings about good markets. But does it
destroy good governance? This proposition, fortunately, would take
things too far. The reason for ethics to survive marke t pressures
relates to the skills required in corrupt transactions. I address this
topic by raising an empirical question: do differences in skills affect
trade? Are some exporters advantaged in entering corrupt markets?
The answer is a clear yes. The skill of bribery is at the core of
understanding some remarkable differences in trade patterns of large
exporting nations. Ethical behavior can survive market pressure.
Whereas some actors may refrain from corruption owing to moral
concerns, others are simply untalented. One application of this finding
relates to corrupt intermediaries. These offer expertise on corrupt
transactions to the untalented. Certification should be offered to those
intermediaries who are willing to commit to anticorruption.
Chapter 8 picks up the international perspective of Chapter 7 and
confronts it with a challenging position. While I claim that transaction
costs of corruption should be increased we hear investors complain
about the unpredictability of corruption. Should we prefer corruption
to be predictable? Is reliability and reciprocity always a good thing?
This is not an academic debate. Politics is often involved in guaran-
teeing international reciprocity even when corruption was involved.
I argue that this practice must be stopped. The unpredictability of
corruption is precisely what may put an end to it. We must make sure
that corruption remains a risky and capricious activity.
A road map to this book xiii
Chapter 9 brings us back to the question of how corrupt actors
reciprocate. It shows that corrupt transactions are often embedded in
regular, legal business transactions. These transactions can provide
the breeding ground by establishing the trust and the reciprocity
necessary to engage in illegal deals. This chapter is crucial for
understanding that there is no ‘‘corrupt marketplace.’’ There is hardly
ever a given demand for corrupt services and supply of such, with a
going rate for bribery being determined in equilibrium. Corruption is
restricted to insi ders with established links. Corruption is open only to
those who exploit long-standing relationships for a criminal career.
Chapter 10 summ arizes. Other guiding principles for anticorrup-
tion such as repression, prevention, or transparency may run out of
steam. The principle of the invisible foot should be at the core of
anticorruption and provide future inspiration. A plethora of building
blocks can emanate from this principle; this book does not try to be
exhaustive in this regard. One focus that deserves recognition is the
design of the legal system. Penalties may mark the starting point of a
corrupt career. An asymmetric design of penalties may avoid this
problem and inhibit corrupt reciprocity. This is only an example for
the overarching principle of the invisible foot. To state it again: cor-
rupt actors can neither commit to honestly serving the public nor
credibly promise reciprocity to their corrupt counterparts. Reform is
about exploiting this handicap.
A road map to this bookxiv
1 Introduction
There are several good protections against temptations,
but the surest is cowardice.
Mark Twain,
Following the Equator, 1897
1.1. Why this book?
Corruption, the misuse of public power for private benefit, turns out
to be a relatively new challenge for social sci ences. It has been an issue
for politics and society for many centuries, but its systematic scientific
treatment is rather novel. However, most researchers consider
corruption to be just another application of preexisting theories
without sufficiently considering their adequacy. This, I believe, is like
putting new wine into old wineskins. Just as wine causes the skins to
burst corruption ruptures preexisting theories. Just as we lose wine in
old skins we may fail to understand corruption without considering
its intrinsic dynamics and logic. Applying old theories then falls short
of an adequate understanding of the phenomenon.
A lecture that I run on the economics of corruption starts with a
game: students are supposed to derive a strategy of how to win a
public tender when they have insufficient funding to take the official
I find myself time and again appalled by the variety of unusual,
innovative, and totally criminal proposals. This is what corruption is
about: someone violates the rules of the game in a way that was not
anticipated by others. To apply models of perfect foresight, rational
expectations, competition with a level playing field, and similar
models are, hence, no longer enlightening. In this spirit, a variety of
I owe this idea to Krassen Stanchev, Institute for Market Economics, Sofia, Bulgaria.
orthodox approaches to corruption appear less useful. Some examples
are provided here.
There were some economists who started with the indisputable
notion that corruption in the form of bribery represents a mutually
beneficial exchange. Microeconomists consider such an exchange to
be desirable and inevitable; functionalists assume that its mere
existence indicates its useful function. Given that briber and bribee
are better off after striking a corrupt deal, on what grounds can we
claim that the deal is detrimental to economic well-being? But this
notion disregards how corrupt ion constrains decision-making. When
officials cannot credibly promi se to reject side-payments from clients,
they are not trustworthy at the outset and may not be employed in the
first place. Corruption turns out to be harmful even to those who have
the chance of striking illegal deals.
For example, it may well be worthwhile to construct good-quality
roads. But the government may choose to cancel the project if bad
quality is expected to result from bribes being paid to inspectors. Or
imagine that a fair and efficient tax system should be established, but
tax collectors cannot be kept from taking bribes in exchange for
turning a blind eye to underreporting. A country may have to
continue living with the old system. If a state auditor cannot
guarantee that she will not fake reports in exchange for a bribe, her
contribution loses value. She may not be hired in the first place even
though an honest exchange would have been favorable to all.
Other researchers argued that instead of fighting corruption itself
one should combat its causes, of which they claimed excessive
government intervention, market restrictions, and a burdening bureau-
cracy to be most prominent. These arguments have been pointed out by
early writers (Bayley 1966; Nye 1967; Huntington 1968; Leff 1964;
Morgan 1964) and still make their way into modern economic
textbooks such as Mankiw (2000: 123). Corruption is then nothing
else but a symptom of inadequate state intervention (Ades and Di
Tella 1999). This transforms the problem into something which is
more akin to economic theories. State intervention is widely dealt
with in economics. The standard recipe for containing corruption
would be to get rid of government intervention. Take the case of
Philadelphia’s Department of Licenses and Inspections where
officials accepted money from plumbing contractors in exchang e
for a quick approval of job-site work.
2 Institutional economics of corruption and reform
A standard ‘tip’ was $20, a source said, and it could grow if a plumber was
in a bind of some kind. ‘‘A lot of it would occur when a plumber would
need to close an excavation hole where they’d buried pipe, and it couldn’t
be closed until an inspector approved it, ‘‘the source said.’’ So you could
stand around with your crew waiting, or you could page an inspector and get
him out there real quick, and thank him for it.’’ . . . the payments to inspectors
have been suspected for years but that they were hard to crack since those
paying the bribes were happy for the speedy service. (Philadelphia Daily
News, March 14, 2001: ‘‘Plumbers Allegedly Bribed Inspectors’’)
The case reveals how regulation to obt ain an inspector’s approval
induced corruption. But the case shows at the same time that simple
recipes for cracking down on government regulation are not feasible.
Inspections are necessary so as to guarantee the delivery of proper
quality, and their abandonment is likely to do more rather than less
harm, maybe even increase corruption further.
One of the biggest cases of systematic corruption also related to
market distortions: in the Iraqi Oil-for-Food program between 1995
and 2003, oil was allowed to be sold only in exchange for
humanitarian goods. The extreme public desire for much-needed
goods not only provided ample opportunities to mark up prices but it
also led to high-ranking UN officials turning a blind eye to massive
According to an estimate, Saddam Hussein’s regime was
able to collect as much as US$1.8 billion. Of the 4,500 private firms
involved in the program, close to half were involved in the payment of
bribes. One paradigmatic case relates to a truck being sold by Daimler
Chrysler. While the regular price would have been US$130,000, the
company charged US$143,000 and passed on US$13,000 to a Swiss
bank account of an Iraqi official. Likewise, oil left the country too
cheaply and kickbacks were paid in exchange. This case well fits
standard economic modeling on the distortionary effects imposed
by market restrictions. Such restrictions create opportunities for
systematic corruption. But at the same time, the common economic
advice to abolish market restrictions is far from obvious. The standard
economic recipe would be to prevent the UN Security Council from
imposing trade restrictions as a way of sanctioning countries; this
is not at all a suggestion that will gain undisputed approval.
The full report by the Volcker Commission is available at Accessed
November 2006.
Introduction 3
The experience from the Iraqi Oil-for-Food program will rather lead to
considerations of how better to monitor the purchases and control
These two cases are representative of many other incidences of
corruption. Regulation is often an integral and much-needed part of
government. Suggestions to avoid regulation are more revealing of a
writer’s negative attitude toward government, in general, rather than
a useful contribution to reform.
For the last decade, most economists have been much less lenient
on corruption than their predecessors and have clearly emphasized its
adverse welfare consequences. But the remedies suggested have been
embedded into econom ic orthodoxy. The thrust of some approaches
has been to be critical of government in toto. If corruption involves a
self-seeking government whose members attempt to enrich them-
selves, one needs to crack down on the government itself; see Becker
(1994), and for a critical review see Orchard and Stretton (1997).
Boyko et al. (1996) suggest that privatization is a means of
reducing corruption and increasing efficiency at the same time. A
downsized ‘‘grabbing hand regime’’ would have less opportunities for
milking the citizenry (Shleifer and Vishny 1998). This argument is
well embedded into economists’ belief in the marke t and distrust
toward politicians, suggesting that corruption can be contained by
minimizing the public sector. However, the findings reported in Box 1
are not supportive of this approach.
Box 1 Corruption and the size of the public sector
It has been suggested that the overall size of the government
budget relative to GDP may be positively correlated with levels of
corruption. This is shown by LaPalombara (1994: 338), who uses
a sample of countries in which Scandinavian countries are
disregarded by assuming them to be an exception. The reverse
finding is reported by others. Elliott (1997: 182–3) reports for a
sample of eighty-three countries that the size of the government
budget relative to GDP decreases with levels of corruption. This is
supported by Adsera et al. (2000). Gerring and Thacker (2005:
245–6) report insignificant results. Graeff and Mehlkop (2003)
observe that corruption significant ly decreases with government
size in the high-income countries.
4 Institutional economics of corruption and reform
These considerations suggest that a more promising focus
would be on particular types of government expenditures in their
potential to cause corruption. In this respect it is suggested that
redistributive activities as opposed to other government activities
are more likely to cause corruption. La Porta et al. (1999: 242)
show a positive correlation of the total government transfers
and subsidies relative to GDP with corruption. However, the
variable correlates too closely with total government expenses,
bringing about the aforementioned problems. In sum, there is no
convincing evidence on the size of government expenses as a cause
of corruption.
Elliott (1997) concludes that types of government activities may
be more important than the size of their budgets. Regressing
corruption on the government’s budget (relative to GDP) might
also be affected by reverse causality: corrupt governments have
difficulties in obtaining funding, be it through taxation or loans.
See Box 21 for respective evidence. This lack of resources then
forces them to operate on a rather small budget. Another criticism
of the hypothesis put forth by LaPalombara is provided by Husted
(1999: 342, 350, 354). He argues that governments are larger in
societies characterized by a greater acceptance of authority. Such
acceptance would be a cultural deter minant of both corrupt ion
and the size of the government budget.
Overall, there is little correlation between the overall size of the
public sector and corruption, as shown in Box 1. Privatization may
have its clear economic advantages, but its effect on containing
corruption appears ambiguous. This might be owing to privatized
firms experiencing a ‘‘privatized’’ form of corruption. The bribes
formerly taken from public servants would then be requested from the
private firms’ staff. Privatization also does not provide a guarantee
that the newly founded units are no longer serving politically
motivated interests. Similarly, whether a downsize d government is
less capable of milking the citizenry is equally questionable: privatized
firms can be equally exposed to public interference and demands for
bribes. What was formerly taken from state-owned enterprises is then
extorted from private firms. More often than not, private firms pay
more in bribes than their well-connected state-owned counterparts
(Lambsdorff and Cornelius 2000: 76–7). Finally, many transition
Introduction 5
economies experienced massive corruption in the course of privatization
programs. This may be another reason why downsizing the public
sector does not help in reducing corruption, at least not in the
transition period. Long-term positive effects from privatization may
certainly be possible, where competitive press ures are superior in
avoiding inefficiencies and corruption, as opposed to bureaucratic
control. But such advantages are likely to require best practice in the
process of privatization.
On a similar note, some authors assume that decentralization could
be a means for reducing corruption by ripping the state off its
extortionate capacities and bringing government closer to the people.
But the alternative to a large centralized public sector is sometimes a
weak local go vernment that is captured by strong local players. It
requires little imagination that such a regime may be equally
unattractive to investors, and similar ad verse effects on welfare are
quite likely to arise. As shown in Box 2 a simple economic ‘‘recipe’’
like decentralization does not unequivocally ameliorate the problems
of corruption. The pros and cons of decentralization are an important
issue. But they are the wrong battleground if one aims at containing
One issue highlighted by Box 2 is that arguments pertaining to
decentralization seem to be dependent on how decentralization is
precisely quantified. Apart from this, one cannot exclude that certain
cultural determinants drive both decentralization and the absence of
corruption. Countries characterized by civic cooperation and trust
amongpeople as well as those with well-developed subnational unitsmay
be in a position to decentralize and lower corruption at the same time.
Box 2 Corruption and decentralization
Some authors observe a positive correlation between corruption
and a country’s size, measured by total population (Fisman and
Gatti 2002; Root 1999; Treisman 1999). These correlations are
robust to the inclusion of further variables. This might be taken as
an indicator in favor of decentralization. Smaller countries might
be in a better position to establish a decent administration and to
monitor their politicians. Using the results from a cross section
of countries might be taken as an indicator that decentralizing
government power could be a means to curb corruption.
6 Institutional economics of corruption and reform
But Knack and Azfar (2003) provide a clear warning against
these findings. They show that the correlation between corruption
and population size results from sample selection problems.
Ratings on corruption are only provided for those countries in
which multinational investors have sufficient inte rest. These tend to
be large nations and, among the small nations, only those that are
well governed. Knack and Azfar (2003) conduct regressions for
larger samples of countries and observe that the relation between
corruption and population disappears. Damania et al. (2004) show
that population density decreases corruption in a sample of sixty-
nine countries; it remains to be seen whether this finding survives
the test for sample selection, as proposed by Knack and Azfar.
Another variable for measuring the extent of decentralization
is presented by Huther and Shah (1998) and Fisman and Gatti
(2002). The authors interpret the share of subnational expenditures
in total public spending as a measure of decentralization. In a
sample of eighty countries, this index correlates positively with
various measures of good governance. Huther and Shah report a
correlation with lack of corruption larger than 0.5. However, the
authors do not include further explanatory variables. One cannot
exclude that more developed countries are less corrupt and more
decentralized at the same time. Biased coefficients are therefore
possible. The approach by Fisman and Gatti (2002) makes use of
the same vari able on decentralization yet tests whether the outcome
is robust to the inclusion of further variables. For a wide range of
specifications, they find that fiscal decentralization in government
spending is significantly associated with lower corruption. The
authors also suggest that corruption may be larger when spending
is decentralized, while revenue collection remains in control of the
central government. They base their empirical findings on levels
of corruption in local states of the United States. Arikan (2004)
employs various measures on decentralization and observes mostly
an insignificant relationship to corruption. A high ratio of non-
central government employment to total government employment,
however, seems to go along with lower levels of corruption.
Treisman (1999) takes a more direct approach to investig ating
the effect of decentralization. Rather then regressing corruption on
total population, he distinguishes between federal and centralized
Introduction 7
Box 2 (Cont.)
states. He reports significant evidence that federal states are more
corrupt than centralized ones. But Treisman (1999) argues that
this relationship falls to insignificance when other variables are
included. Adsera et al. (2000) and Panizza (2001) also fail to
obtain a significant impact. Damania et al. (2004) even report a
significant impact of federalism in reducing corruption. On the
contrary, Goldsmith (1999: 878), Kunicova (2002), and Kunicova
and Rose-Ackerman (2005) claim federalism to increase corrup-
tion, even when controlling for GDP per head. In a more recent
publication, Gerri ng and Thacker (2004) are also supportive of a
significant advers e impact of federalism on corruption. They dis-
tinguish between nonfeder al, semifederal, and federal states and
mix these charac teristics with the extent of bicameralism where no
or only a weak upper house exists, where the upper house is not
dominated by a lower house, and where nondominance goes along
with a different partisan distribution between the houses. The
authors find evidence against federal states and in favor of unitary
governments throughout a variety of regressions.
Testa (2003) investigates differences between unicameral systems
and bicameral systems. She shows for a cross section of forty-three
democracies that bicameralism lowers corruption in rather eth-
nolinguistically homogenous states. But bicameralism increases
corruption in countries with a high level of ethnolinguistic frac-
tionalization. The suggested reason for this finding relates to
bicameralism hindering lobbyism (and corruption) by doubling the
legislators that a lobby must buy. But where two chambers differ
in politics, which is likely to arise in countries with high levels of
fractionalization, legislators are used to seeking compromises and
lobbyism may require few resources. The extent of fractionaliza-
tion is also investigated by Alesina et al. (2003). They show that
countries characterized by ethnic, linguistic, or religious fractio-
nalization are rated worse by PRS/ICRG with respect to the
political instability related to corruption.
Many economists point to one major cause of corruption: bad
regulation. Ill-designed institutions are considered to be at the
frontline of assigning adverse incentives to policy-makers, bureaucrats,
8 Institutional economics of corruption and reform
and the public in general. Box 3 reviews studies that are supportive of
a close association between bad regulation and corruption. Such a
viewpoint would accept that government serves useful functions and
that, thus, downsizing government is not the vision for reform.
Reform should rather avoid complicated rules or those that are
difficult to administer and align with individual decision making.
From this perspective, some ‘‘good’’ regulation may even be helpful in
containing corruption. For example, privatization in Eastern Europe
involved bribery because there was too little ‘‘good regulation,’’ that
is, too few legal requirements that restricted corrupt deals.
As a result, detecting bad regulation and misdirected state
intervention can be helpful in becoming aware of areas where
corruption is likely to occur. However, bad regulation and corruption
are quite often two sides of the same coin. When local firms are given
preferential treatment in public tenders, this may induce corruption,
but it may also be the outright result of strong private interests that
capture public funds. In other cases, corruption causes bad regula-
tions, and not the other way round.
Quite striking is an example from Pakistan. The gold trade was formerly
unregulated and smuggling was common. Shortly after Benazir Bhutto
returned as Prime Minister in 1993, a Pakistani bullion trader in Dubai
proposed a deal: in return for the exclusive right to import gold, he would
help the government regularize trade and make some further private
payments. In 1994, the payment of US$10 million on behalf of Ms. Bhutto’s
husband was arranged. In November 1994, Pakistan’s Commerce Ministry
wrote to the bullion trader, informing him that he had been granted a
license to be the country’s sole authorized gold importer a profitable
monopoly position (The Straits Times, Singapore, February 1, 1998, ‘‘Paper
Trails Points to Illicit Bhutto Hoard,’’ and June 2, 1998, ‘‘The Scandals’’).
When monopoly rights are given in exchange for bribes, it is rather
corruption that drives market distortions. Claims that the monopoly
right sho uld be abandoned so as to get rid of corruption appear
misplaced, because at the core of the problem would be criminally
innovative politicians and businesspeople, and their capacity of
inventing bad regulations. A final concern: the difference between
‘‘bad’’ and ‘‘good ’’ regulation is far from obvious. One criterion could
be whether regulation creates opportunities for corruption. But in this
case the argument becomes circular and we are not provided with a
Introduction 9
causal theory of corruption. Overall, looking for ‘‘bad’’ regulation
provides some hints for detecting corruption but it falls short of an
overarching approach to reform.
Box 3 Corruption and regulatory quality
Broadman and Recanatini (1999) show for a sample of transition
economies in Europe and Central Asia that higher barriers to
market entry lead to higher corruption. Djankov et al. (2002) are
equally concerned with the nature of entry regulation. They
determine the number of procedures required for starting a new
business for a cross section of seventy-one countries, along with the
necessary time and official costs. The authors find a strong corre-
lation of these variables with a country’s level of corruption for a
variety of specifications and control variables. Svensson (2005: 29)
finds a positive correlation between corruption and the number of
business days needed to obtain legal status. These findings support
the argument that entry regulation often does not serve to correct
for market failure but brings about problems of its own.
Treisman (2000) finds that ‘‘state intervention’’ tends to
increase corruption. The former variable is measured by a sub-
jective index compiled by IMD . But as other explanatory variables
enter into the regression, the relationship breaks down. Another
correlation between corruption and a meas ure of policy distortion
for thirty-nine countries is presented by the World Bank (1997:
104, 168). Unfortunately, the study lacks a precise definition of
policy distortions. Also, the robustness of the results is not tested
by including further explanatory variables.
Gerring and Thacker (2005) report a positive correlation
between regulatory quality and absence of corruption. Ades and Di
Tella (1997; 1999) provide a more detailed analysis of policy dis-
tortions. The authors use an index that measures ‘‘the extent to
which public procurement is open to foreign bidders’’ and another
index that measures ‘‘the extent to which there is equal fiscal
treatment to all enterprises.’’ Both variables, and also a corruption
variable, are taken from the survey by IMD. Both variables sig-
nificantly explain the level of corruption, even controlling for
other explanatory variables. This leads the authors to conclude that
policy intervention causes corruption. Goel and Nelson (2005)
10 Institutional economics of corruption and reform
observe a positive association between corruption and government
regulation of and involvement in the financial sector, the latter
index being from the Heritage Foundation. The finding for a
sample of sixty-three countries is robust to various tests.
Many authors acknowledge that corruption may cause policy
distortions and not vice versa, bringing about problems of simul-
taneity bias. Ades and Di Tella (1997) claim that their instruments
for policy distortions ascertain the direction of causality. Cer-
tainly, policy distortions and corruption are quite often just two
sides of the same coin. In this case, instruments have to carry a
heavy burden.
A simple correlation for a sample of twenty-six African coun-
tries is provided in Lambsdorff and Cornelius (2000). They show
that corruption is positively associated with the degree to which
‘‘government regulations are vague and lax.’’ These results are
interesting in shifting the focus away from the total burden of
regulation to their application. Vagueness and lax application of
regulation supplies public servants with the bureaucratic discretion
necessary for requesting bribes. Clear rules might present a burden
to business but would not necessarily trigger corruption. However,
the regressions are not yet controlled by further variables, neither
are they extended to a broader sample of countries.
In a similar vein, Gatti (1999) argues that a highly diversified
trade tariff menu fuels bribe-ta king behavior, whereas uniform
trade tariff rates limit public officials’ ability to extract bribes from
importers. She reports a positive associ ation between the standard
deviation of trade tariffs and the level of corruption for a small
sample of thirty-four countries. Causality may be difficult to
ascertain, because corrupt public servants may impose diversified
tariffs so as to be in a better position to ask for bribes.
Some researchers claim that corruption simply mirrors the absence
of economic competition; see Box 4 for a review of eviden ce.
Competition among suppliers drives down prices. In public procure-
ment, for exampl e, the resulting rents for private firms decrease.
Consequently, public servants and politicians have less to ‘‘sell’’ in
exchange for bribes, reducing their motivation to start with a corrupt
career. On the contrary, where competition is restricted, profits
increase and politicians can grasp the opportunity to assign these
Introduction 11
profits, in exchange for a share. Again, we would be favorable to this
general associ ation. However, the argument ad ds little to the overall
economic dispute as to whether restrictions on competition can in rare
instances be beneficial and sheds no light on the process of dealing
with natural monopolies. Above that, the argument can easily suffer
from reverse causality: the prospects of corrupt income may motivate
private firms to pay bribes and politicians to offer market restrictions.
The case on p. 9 is illustrative of this. We would end up in a vicious
circle. Instead of being provided with ideas for refor m, we would
observe rather that encouraging competition and reducing corruption
can be two sides of the same coin.
On the other hand, competition may sometimes increase rather
than decrease corruption. Where companies compete with quality
rather than with prices, competition may force firms to myopic
behavior. Instead of cultivating a high-quali ty reputation they would
rather bribe inspectors, inducing them to turn a blind eye to the
delivery of substandard quality.
Box 4 Corruption and competition among private firms
Government restrictions on economic freedom are likely to reduce
competition and thus encourage corrupt ion. Henderson (1999)
argues that corruption is negatively correlated with different
indicators of economic freedom. This result is largely supported by
Goldsmith (1999: 878) for a sample of sixty-six countries, where
the regression is controlled for GDP per head, and by Paldam
(2002) who includes further explanatory variables in a sample of
seventy-seven countries. Goel and Nelson (2005) report similar
findings. Such arguments, however, might be tautological. The
Heritage Foundation’s Economic Freedom measure, for example,
includes an assessment of corruption. This implies that a measure
of the dependent variable is placed on the independent variable
side of the equation (Sandholtz and Gray 2003).
Graeff and Mehlkop (2003) relate corruption to the sub-
components of the index of Economic Freedom by Gwartney and
Lawson (2000) for a sample of up to sixty-four countries. Con-
trolling for a variety of further influential variables they find that a
variety of these subcomponen ts is insignificant. An assessment of
the legal security of private ownership rights, the viability of
12 Institutional economics of corruption and reform
contracts, and the rule of law is found to lower corruption, parti-
cularly in rich countries. Another aspect of economic freedom that
is associated with less corruption is the freedom of exchange in
capital and financial markets. The argument on this latter sub-
component, however, may be disputed. This variable includes the
percentage of deposits held in private banks and the percentage of
credit given to the private sector. Instead of being a cause of cor-
ruption such indicators may rather result from a low-corruption
environment. Interestingly, the freedom of citizens to own foreign
currency bank accounts domestically and abroad is found to
increase corruption, at least in the poorer countries of the inves-
tigated sample. The authors conclude that not all aspects of eco-
nomic freedom deter corruption because some regulation may
increase the transaction costs of corrupt deals. In a related inves-
tigation, Neeman et al. (2003) argue that financial openness is
detrimental to development because the income from corruption is
allowed to be allocated abroad rather than being re-invested in a
country; see also Box 16 for details.
Ades and Di Tella (1995) test the influence of two other indi-
cators of competition. These are taken from the survey by IMD. A
subjective index of ‘‘market dominance’’ measures the extent to
which dominance by a limited number of firms is detrimental to
new business development. Another index of ‘‘anti-trust laws’’
measures the effectiveness of these laws in checking non-
competitive practices. The authors conclude that the less compe-
titive a market environment, the higher will be the extent of
corruption. However, the authors note the problems of causality
and acknowledge that corruption may provide incentives for
politicians to support monopolies.
One measure of competitive pressures is the integration of a
country into the global economy. If competition reduces corrup-
tion, then increased openness to international trade and invest-
ment should go along with less corruption. A report in Foreign
Policy (2001) indeed found that increased globalization is asso-
ciated with less corruption. However, the study neither controls
for further variables nor provide s insights into a possible caus-
ality. Sandholtz and Gray (2003) report that the more international
Introduction 13
Box 4 (Cont.)
organizations a country belongs to, and the longer it has been a
part of the major international institutions such as the United
Nations, GATT/WTO and the IMF, the lower its level of cor-
ruption. Furthermore, they report that corruption decreases with
other factors of openness such as international telephone minutes
per capita and air freight per capita. Responding to the criticism by
Knack and Azfar (200 3), as described on p. 7, Sandholtz and Gray
(2003) show that their results are not affected by sample selection
Ades and Di Tella (1995; 1997; 1999) demonstrate that open-
ness, defined as the ratio of imports to GDP, is negatively asso-
ciated with corruption. They apply corruption data from BI (in a
cross section of fifty-five countries) and IMD (in a cross section of
thirty-two countries). With both approa ches the results are robust
to the inclusion of further explanatory variables. The authors
conclude that international economic competition reduces cor-
ruption. A similar finding is reported by Sung and Chu (2003) and
Gerring and Thacker (2005). However, Treisman (2000) did not
find significant evidence for such an impact using the TI-index.
Another possible measure of the extent of competition in a
country can be derived from the number of years it has been open to
trade, as assessed by Sachs and Warner (1995). Treisman (2000) and
Leite and Weidemann (1999) provide evidence that this variable
negatively and significantly impacts on the level of corruption. In line
with this thought, one may conjecture that liberalization does not
immediately reduce corruption. In fact, Tavares (2005) claims that
the immediate effect of liberalization is to increase rather than
decrease corruption. Once tariffs are reduced below 40 percent,
nontariff barriers relate to less than 40 percent of imports, a socialist
economic system is abandoned, no major black market premium is
paid for the exchange rate, or major exports are no longer a state
monopoly, the level of corruption as measured by Political Risk
Services/International Country Risk Guide (PRS/ICRG) increases.
Given that our knowledge on trends in levels of corruption is still
limited, certainly, the results may require further validation.
14 Institutional economics of corruption and reform
Other researchers were engaged in trying to find ‘‘shortcuts’’ in
the fight against corruption. One of these shortcuts is to assume that
some forms of corruption are more tolerable than others. Krueger
(1974: 302), for example, prefers favouritism in government over
competitive forms of rent-seeking. Tullock (1980b: 109–11) considers
nepotism to be less problematic as compared with competitive
lobbying (I will deal with this issue in more detail in Chapter 5).
Unusual as these arguments may appear, they tend to provide excuses
for some types of corruption instead of providing us with an approach
for reform. Along similar lines, Murphy et al. (1993: 413) argue that
the problems with corruption are mitigated when corrupt rulers can
collect bribes efficiently. Perfecting corruption rather than fighting it is
the avenue suggested for reform.
But such conclusions are misleading when the underlying model is
too limited for an adequate discussion of a mul tifaceted problem. In a
similar spirit it is sometimes assumed that the adverse effects of
corruption relate largely to the accompanying uncertainty. Bribers
have no legal recourse after making the payment and face threats of
demands for further bribes. Some authors assume that it is this type of
uncertainty that deters investors, less so corruption itsel f. But the
more predictable form of corruption is likely to bring about other
disadvantages. A predictable type of corruption less forces investors to
seek legal alternatives, facilitating the further spread of this type of
corruption (this issue will be examined in Chapter 7). Corruption has
a large variety of disastrous effects that make it difficult to prefer one
type to another. Clearly, corruption is a complex phenomenon and its
diverse variants are likely to bring about quite different effects. But
whether this allows us to rationally prefer one type to another is likely
to remain open to dispute.
1.2. Defining corruption
Definitions of corruption can be discussed at length without
necessarily providing an actual value added to the reader. Still some
researchers display their endeavors in this area. They are willing to go
into time-consuming debate and are fierce in preferring one approach
to another. Such debate, however, tends to absorb much of the energy
that is desperately needed elsewhere. Recognizing this, some colleagues
Introduction 15
have started to avoid definitions of corruption, claiming that most
cases of corruption are unambiguously perceived by most observers.
This is somewhat along the lines of Weber’s (1920: 30) definition of
the spirit of capitalism. He rejects a definition and claims that this
term is composed by the various fragments and conceptions provided
in his subsequent writing. This is similar to the problems faced by
United States Supreme Court Justice Potter Stewart in 1964 when he
argued ‘‘I can’t define pornography, but I know it when I see it.’’
In such a perspective, a definition of corruption would be rendered
at the end of the reading, rather than being provided upfront. I tend to
share this notion and propose the reader might simply skip the
following paragraphs. For the sake of completeness, I nonetheless
provide a rather uncontroversial, rough sk etch and description of
what should be understood by the term corruption. Some readers may
profit more from this readi ng after having completed the book.
Corruption is commonly defined as the misuse of public power for
private benefit. The term ‘‘private benefit’’ relates to receiving money
or valuable assets, but it may also encompass increase s in power or
status. Receiving promises for future favors or benefits for relatives
and friends may also be considered a private benefit. With regard to
favors for relatives and friends, the terms nepotism and favoritism are
also common.
‘‘Public power’’ is exercised by bureaucrats, appointed to their
office, and by politicians, who are elected to their position. Such
public power is exercised in a variety of sectors, such as the judiciary,
public procurement, business regulations and granting of permits,
privatization, foreign exchange (including customs, trade permits, and
international financial transactions), taxes (including the granting of
tax exemptions), police, subsidies, public utility (water, electricity,
telephone, garbage collection, health care), and government services
(health, education).
The term ‘‘misuse’’ can either relate to a behavior that deviates
from the formal duties of a public role (elective or appointive), that is
in contrast to informal rules (established by publi c expectations or
somewhat standardized as codes of conduct), or, more generally,
where narrow interests are followed at the expense of the broader
interests of the public at large. In a functioning government system
these definitions fall into one: public interests are supposed to feed
into the public’s expectations vis-a
-vis office holders. These, in turn,
16 Institutional economics of corruption and reform
are supposed to define formal obligations in line with the public’s
interest; see Figure 1.1. However, corruption is about government
failures, which provides some problems to our definitions. Corr uption
as a real-world phenomenon thus destroys the foundation on which
the just given formal definition rests.
When those who dictate formal obligations operate without regard
to public expectations, a definition of corruption can no longer relate
to formal obligations. This problem is sometimes observed for
parliamentarians. They are reluctant to follow a rigid rule for
themselves and rather oppose the related legislation. This renders the
parliamentarians’ own bribe-taking as permissible. Hence, the taking
of bribes would not violate the law, but clearly it is in contradiction
to public expectations. A definition that sticks narrowly to formal
obligations would therefore be of little use. Instead, a definition
would have to refer directly to public expectation s. This, certainly,
comes at the cost of reduced precision of the term. In some rare
instances such as civil war or ethnic or social cleavages, the public
may even no longer develop consistent and generally shared
expectations vis-a
-vis the operation of public office holders. Corrup-
tion must then refer to an even more abstract term: the public
interest. It goes without saying that such a definition tends to become
vague. But under such rare conditions, public interest would be the
only adequate reference for establishing a meaningful understanding
of corruption.
A world free of corruption is associated with public servants who
intend to serve the public, be it through intrinsic motivations,
incentives, threats of penalties, or peer pressure. Concerning the
institutions that help establish such a behavior, at least four aspects
are commonly emphasized: first, the arms-length principle, stating
that pertinent arguments in public decision-making should not be
Definition of Corru
Public interest
Formal obligations
Expectations vis-à-vis
public duties
Figure 1.1. Defining corruption
Introduction 17
overshadowed by personal relationships and that equality of
treatment for all economic agents should be achieved; second, citizen
participation and involvement, giving people a say in public decision-
making; third, transparent procedures with regard to public decision-
making, limiting the discretion among office holders; and fourth,
competition among office holders, giving voters and clients the chance
to exchange nonperforming individuals with others.
Corruption, certainly, is defined differently in different regions of
the world. Across countries the public forms different expectations
about public roles. How officials should serve the public can be up to
purely local taste. The four aspects mentioned above may obtain
different weightage in different countries. Equality of treatment may
be less relevant in soci eties characterized by strong personal relations,
where relatives and friends expect office holders to provide favorable
treatment. Transparency and participation may be given less
significance in societies that are convinced that only pertinent
arguments are relevant for bureaucrats. What seems to be universal,
though, is that the public comm only considers self-seeking beh avior
by politicians and bureaucrats as corrupt when this goes along with a
blunt neglect of their expectations and interests.
Corruption is an exchange of favors between two actors, an agent
and a client; see (Andrig and Fjeldstadt 2000). The agent is entrusted
with power by her superior, the principal.
The principal delegates a
task to the subordinate, his agent, and sets up the rules as to how this
task is to be fulfilled. The agent is supposed to serve the client in
accordance to these rules. Bribery, extortion, embezzlement, and
fraud in the public sector are variants of corrupt behavior, amounting
to the agent ‘‘defecting’’ from her rule-bound behavior. In the case of
bribery the client acts as a briber and makes a payment (also called
kickback, baksheesh, sweetener, payoff, speed- or grease-money) to
the agent, who then is called a bribee. In return the client obtains an
advantage such as a service or license he is not entitled to obtain, for
example, a tax rebate or a public contract. In the case of extortion the
agent uses her power to extract money or other benefits from the
client. The client may have to pay for a service, although he is legally
entitled to obtain it without such payment. The agent uses coercion,
Throughout the book, I use the female pronoun for agents, supervisors, and
middlemen, and the male pronoun for principals and clients.
18 Institutional economics of corruption and reform
violence, or threats in order to obtain this payment. Embezzlement,in
contrast, is simply theft of public resources by the agent. Without an
involvement by the client a disloyal agent steals from the principal.
Bribery, extortion, and embezzlement imply that the principal’s rules
are trespassed and his interests are hurt. The agent is commonly better
informed about details of her daily tasks and her efforts devoted to
their fulfillment. This implies that she can benefit from informational
advantages. The agent can also actively conceal information from the
principal with the help of trickery, swindle, deceit, manipulation or
distortion of information, facts, and expertise. In this case the term
fraud is used. See Figure 1.2 for an overview.
Corruption must be distinguished from certain other forms of
criminal conduct that involve only private parties. Tax evasion,
contraband, black markets, insider dealings at the stock exchange,
production of counterfeit money, and subsidy fraud can be carried out
without misusing public power. Actors involved in such activities
are private businesspeople, for example, taxpayers, who are not
entrusted with public power.
A wider definition of corruption
would also include this type of behavior. When a private firm’s sales
Provides service/license,
awards contracts
Pays taxes/tariffs
Pays a bribe
Figure 1.2. Corruption in a principal–agent–client model
These activities may also go along with corruption when public office holders are paid
to refrain from prosecution, to grant impunity, or to provide inside information on
criminal opportunities.
Introduction 19
manager takes kickbacks in exchange for contracts, he misuses
‘‘entrusted’ power. But clearly, the position of power was not provided
by the public, suggesting differences to the common definition of
One may distinguish between various forms of corruption based on
other criteria, as well. One criterion is whether the briber or the bribee
obtains the larger benefit from a corrupt deal, depending largely on
which side has the stronger bargaining power. ‘‘Clientilist’’ corrup-
tion takes place if the briber obtains the higher benefit, while
‘‘patrimonial’’ corruption occurs where the bribee obtains the bigger
share. One may equally distinguish between petty and grand
corruption, where the former involves frequent, small payments to
public servants lower in hierarchy, while the latter relates to large,
one-shot payments to higher ranks. The terms ‘‘political’’ and
‘‘administrative’’ corrupt ion are defined according to the key actors,
being either politicians or bureaucrats.
Some behavior would be termed corrupt equivocally by all
observers. But ‘‘gray areas’’ exist where viewpoints differ. Lobbying
is one such gray area. While its presence may suggest that decisions in
the public sector are for sale, it is often legal, carried out in a
transparent and competitive manner, and involves not the
narrow interests of individuals but those of larger business sectors.
This distinguishes it from ordinary types of corruption. Gift-giving to
public servants is another such gray area. While it involves the danger
of dependency and reciprocity by the receiver, it may not require
obfuscation, which is a characteristic of corruption. Gifts, in contrast
to bribes, can always be given in an open, transparent manner.
1.3. Measuring corruption
Given the recent interest in corruption, attempts to quantify the extent
of corruption have become vital. Most prominent among these is the
Transparency International Corruption Perce ptions Index (CPI),
which provides assessments of perceived levels of corruption for a
cross section of countries . The CPI is a composite index, using the
assessments by risk agencies and surveys carried out among elite
businesspeople. While perceptions should never be confused with
reality, the given consensus provides some confidence that the
perceptions gathered are informative on actual levels of corruption.
20 Institutional economics of corruption and reform
The CPI is an annual index, compiled since 1995. I started the
operational work behind the index many years ago at the University
of Goettingen; this work is carried out now at the University of Passau
under my supervision. The index has assumed a central place in
research on the causes and consequences of corruption, based on
regressions for a cross section of countries. An appendix to this
chapter on pp. 236–255 provide s technical detail to the methodology
behind the index. All the data be tween 1996 and 2005 as well as
historical data can be obtain ed at
The goal of the CPI is to provide data on extensive perceptions of
corruption within countries. This is a means of enhancing the
understanding of real levels of corruption and how these differ from
one country to another. In an area as complex and controversial as
corruption, no single source or polling method has yet been developed
that combines a perfect sampling frame, a satisfactory country
coverage, and a fully convincing met hodology to produce compara-
tive assessments. This is why the CPI has adopted the approach of a
composite index. Box 5 lists the various sources that enter the index.
Box 5 Sources of the 2005 CPI
State Capacity Survey by Columbia University (CU), 2003;
ninety-five countries are asses sed by a panel of largely academic
experts from the United States.
The Economist Intelligence Uni t (EIU), Country Risk Service
and Country Forecast 2005, (; 156 countries are
assessed by staff.
Freedom House Nations in Transit (FH), 2005 (www.freedom-; twenty-seven transition coun-
tries are assessed by a panel of experts.
The Institute for Management Development, Lausanne (IMD),
World Competitiveness Yearbook. I use data for 2003–2005
(; fifty-one countries are assessed, based on more
than 4000 annual responses by local executives in top and
middle management of domestic and international companies.
Information International (II), 2003, (www.information-; thirty-one countries are assessed, based on
382 assessments from 165 expatriate business executives.
Introduction 21
Box 5 (Cont.)
Merchant International Group (MI G), Grey Area Dynamics
Rating, 2005 (; 155 countries
are rated by expert staff and a network of local correspondents.
The Political and Economic Risk Consultancy, Hong Kong
(PERC), Asian Intelligence Newsletter. I use data for 2003–2005,
( /); fourteen Asian countries are assessed,
based on roughly 1000 annual respons es by expatriate business
United Nations Economic Commission for Africa (UNECA),
African Governance Report 2005 (,
twenty-eight African countries are assessed by a panel of
roughly 100 resident experts per country.
The World Markets Research Centre (WMRC), 2002, (www.; 186 countries are assessed by staff.
The World Economic Forum (WEF), Global Competitiveness
Report. I use data for 2003– 2005 (; 117
countries were assessed in 2005, based on more than 10,900
responses of local senior business leaders of domestic and
international companies.
As the data collected relates to perceptions rather than to real
phenomena, it has to be considered whether such perceptions improve
our understanding of real levels of corruption. Since actual levels of
corruption cannot be determined directly, perceptions may be all we
have to guide us. However, this approach is undermined at least to
some extent, if the perceptions gathered are biased. Such a pot ential
bias might originate from the particular cultural background of
respondents. For example, Bayley (1966: 721) argues:
The western observer is faced with an uncomfortable choice. He can adhere
to the Western definition, in which case he lays himself open to the charge
of being censorious and he finds that he is condemning not aberrant
behavior but normal, acceptable operating procedure. . . . On the other
hand, he may face up to the fact that corruption, if it requires moral cen-
sure, is culturally conditioned . . . [and] it may be necessary then to assert in
the same breath that an official accepts gratuities but is not corrupt.
22 Institutional economics of corruption and reform
Taking up this perspective, it becomes essential to investigate how
the underlying sample of respondents may define and quantify
corruption. While the sources all aim at measuring the extent of
corruption, the sample design differs con siderably. Basically, three
different types of samples are used. A first group of sources, namely,
CU, EIU, FH, MIG, and WMRC, assemble the perceptions of
nonresidents, turning in their expert opinion with regard to foreign
countries. These assessments are largely carried out by respondents
from developed countries of the western hemisphere such as North
America and Western Europe. These would be the respondents that
are faced with Bayley’s ‘‘uncomfortable choice.’’
A second group of sources, namely, represented only by II, also
assembles the perceptions of nonresidents, but these respondents are
largely from less-developed countries. There is an advantage to
perceptions vis-a
-vis foreign countries because they are not vulnerable
to a ‘‘home-country bias.’’ Such a type of bias would be relevant if
respondents assess their home country purely according to local
standards. Such a standard would be problematic because it can differ
from one country to another, impairing the validity of cross-country
A third group of sources, namely, IMD, PERC, and WEF, gather
assessments made by residents with respect to the performance o f
their home country. These respondents are partly nationals but also
expatriates from multinational firms. While such data m ight be
susceptible to the aforementioned ‘‘home-country bias,’’ they are
not susceptible to introducing an undue dominance of ‘‘western
business people’s’’ viewpoint. Such a viewpoint would be inade-
quate if foreigners lack a proper understanding of a country’s
The data correlate well with each other, irrespective of these
different samples. The high correlations mitigate fears that any of
the aforementioned biases are important to the results. Residents
may therefore have a rather universal ethical standard and
adequately position their country as compared with foreign
countries. Likewise, those respondents who assess foreign countries
seem to have a good grasp of a country’s culture and provide
appropriate assessments in the light of this. In sum, the perceptions
are a helpful contribution to t he understanding of real levels of
Introduction 23
Interestingly enough, the data also correlate well with actual
experience. Experience-based data has been produced by the
International Crime Victim Survey (ICVS). They poll the general
public in more than forty countries. In 1996 they asked ‘‘in some
areas there is a problem of corruption among government or public
officials. During 1995, has any government official, for instance a
customs officer, police officer or inspector in your own country, asked
you or expected you to pay a bribe for his service?’’ While less than
1 percent agreed in most develo ped countries, figures went as high as
one-third in less-developed countries. These data clearly relate to
personal experience. Still, for a sample of forty-three countries the
data correlates well with the sources entering the CPI, commonly with
a coefficient higher than 0.8. This supports the validity of the CPI.
More recently, Gallup Inte rnational incorporated questions on
corruption commissioned by Transparency International in its 2004
survey ‘‘Voice of the People,’’ an annual poll of the general public in
fifty-four countries. Question 5 reads: ‘‘In the past 12 months, have
you or anyone living in your household paid a bribe in any form?’’
Figure 1.3 depicts a scatterplot where the results are portrayed against
international corru
tions index 2005
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Percentage reported to having paid bribes;
global corruption barometer 2004
Figure 1.3. CPI versus reported payment of bribes
24 Institutional economics of corruption and reform
United States
Czech Republic
Equatorial Guinea
Papua New Guinea
United Arab.Emirates
Serbia and.Montenegro
Congo Republic
Congo Democratic Republic
Cote d'lvoire
Figure 1.4. 2005 Corruption Perceptions Index and approximate confidence intervals. The coverage probability is
65–75% (gray lines) or 80–90% (black lines)
the CPI 2004. This reveals a 0.70 correlation coefficient for the data,
again supporting the overall validity of perceptions.
These results also imply that the perceptions are not distorted by
media reports. Respondents might rely on media coverage and reports
obtained from others. Certainly this influence cannot be excluded and
necessarily contributes to perceptions. Yet, in its extreme form such
an influence may suggest that respondents rely only on hearsay. The
potential problem with this influence is that the assessment of a
country might then reflect the quality of the press in uncovering
scandals and particularly its freedom to do so. Countries that suppress
a free press may escape a bad reputation for corruption among their
population. Such an influence would certainly undermine the validity
of the CPI and its usefulness as an aid to understanding real levels of
corruption. However, the CPI tends to correlate positively with press
freedom; see p. 46 for respective results. Perceptions of high levels of
corruption are rather found in countries with little press free dom.
Media reports therefore do not seem to bias perceptions. On the
contrary, perceptions appear to be robust indicators.
Figure 1.4 shows the 2005 CPI, along with the confidence ranges.
These depict the precision of the respective scores. The appendix to
this chapter on pp. 235–255 provides technical details to the CPI.
A similar correlation is reported by Mocan (2004) relating to data on experienced
corruption by the United Nations Interregional Crime and Justice Research Institute
( But she claims that this relationship breaks down once
regressions are controlled for the quality of institutions in a country. She concludes
that perceived corruption relates more to such indicators than ‘‘real’’ levels of
corruption. This conclusion, however, is easily overemphasized. First, the data on
experienced corruption are distributed differently than the CPI. This can easily result
in residuals not being normally distributed. In my own regressions, I chose a more
adequate functional form for the relationship. This produced normally distributed
residuals and the CPI was strongly significant, even when controlling for other
variables on institutional quality. Second, data on reported corruption are not
necessarily ‘‘real.’’ Standards of definition may vary from one country to another.
Minor gifts may already be termed a bribe in Norway, while in Nicaragua facilitation
payments may be considered legitimate. In Norway a payment to the public servant’s
distant relatives may be considered illegitimate, while in Nicaragua only favors going
directly to an official may qualify as a bribe. Third, individual confrontation with
corruption is likely to relate more to the street-level, petty type of corruption as
observed by households; Svensson (2005: 23–4). The Corruption Perceptions Index
(CPI) includes also the extent of grand corruption and focuses on the impact of
corruption on the costs of doing business.
26 Institutional economics of corruption and reform
2 Enemies of