ThesisPDF Available

Reducing or Fueling: The Effect of EU International Aid Allocation on the Level of Corruption in Recipient Countries

Authors:

Abstract and Figures

The existing literature on the effect of foreign aid on corruption in recipient countries has been controversial and contains methodological issues regarding the use of instrumental variables. In this thesis, based on the work of Carnegie and Marinov (2017), I utilize the EU Council presidency as the instrument to analyze the effects of EU aid on corruption in recipient former colonies until 2006. The results from the instrument variable approach show that while EU aid allocation reduced most types of corruption, the magnitudes of these effects were quite small. The different effects depending on the measurements of corruption were also revealed in the analysis. These results were confirmed by the following quantitative analyses of the contemporary history of EU sanctions against violations of human rights, democracy, the rule of law, and good governance, as well as the case of DRC in which the 2001 Belgian presidency had a concrete influence on aid allocation. The findings suggest EU aid conditionality does not have substantial effects to reduce corruption in recipient countries in comparison to the effects on other development issues such as democracy promotion and human rights improvement when donor countries have little interest to impose sanctions against corruption.
Content may be subject to copyright.
Reducing or Fueling: The Effect of EU International Aid
Allocation on the Level of Corruption in Recipient Countries
By
Yusuke Ishikawa
Submitted to
Central European University
Department of Political Science
In partial fulfillment of the requirements for the degree of Master of Arts
Supervisor: Professor Gabor Simonovits
Budapest, Hungary
2020
CEU eTD Collection
i
Declaration of Authorship
I, the undersigned …………Yusuke Ishikawa……………………………………. hereby
declare that I am the sole author of this thesis. To the best of my knowledge this thesis contains
no material previously published by any other person except where due acknowledgement has
been made. This thesis contains no material which has been accepted as part of the requirements
of any other academic degree or non-degree program, in English or in any other language.
This is a true copy of the thesis, including final revisions.
Date: ………………………………………………………
Name (printed letters): ………………………………………
Signature: ………………………………………………………
©Yusuke Ishikawa, 2020
All Rights Reserved
CEU eTD Collection
ii
Abstract
The existing literature on the effect of foreign aid on corruption in recipient countries has been
controversial and contains methodological issues regarding the use of instrumental variables.
In this thesis, based on the work of Carnegie and Marinov (2017), I utilize the EU Council
presidency as the instrument to analyze the effects of EU aid on corruption in recipient former
colonies until 2006. The results from the instrument variable approach show that while EU aid
allocation reduced most types of corruption, the magnitudes of these effects were quite small.
The different effects depending on the measurements of corruption were also revealed in the
analysis. These results were confirmed by the following quantitative analyses of the
contemporary history of EU sanctions against violations of human rights, democracy, the rule
of law, and good governance, as well as the case of DRC in which the 2001 Belgian presidency
had a concrete influence on aid allocation. The findings suggest EU aid conditionality does not
have substantial effects to reduce corruption in recipient countries in comparison to the effects
on other development issues such as democracy promotion and human rights improvement
when donor countries have little interest to impose sanctions against corruption.
CEU eTD Collection
iii
Acknowledgements
First of all, I would like to thank Professor Gabor Simonovits for giving valuable advice through
his courses, as well as his supervision, which helped me to write my first quantitative thesis. I
would also like to express my sincere gratitude to thesis buddy, Austėja Trinkūnaitė, who gave
me useful opinions and information from a practitioner’s perspective and, most importantly,
kept me motivated for thesis writing in an extraordinary situation under the COVID-19
pandemic. In addition, I am grateful to Academic Writing instructors Robin Bellers and
Zsuzsanna Toth for their continuous and kind support throughout 2019/20 academic year. I am
sure that I improved my academic writing skills during my study at Central European University
(CEU).
More generally, I would like to express my gratitude to Professor Matthijs Bogaards and
Professor Agnes Batory, who helped me to develop various research plans on anti-corruption
policies throughout the academic year. Furthermore, I wish to thank all my other professors and
friends at CEU for supporting me in numerous regards and discussing a range of intellectually
stimulating topics. I fully enjoyed my intellectual journey at CEU as my second master’s degree,
especially through methodological and philosophical discussion in and out of the classroom.
I dedicate this thesis to my parents, grandparents, and relatives with love and gratitude. Noting
their continuous support in this thesis might not be the best way to express love and gratitude,
but I express them here as this thesis and the degree program at CEU would not be completed
without their unconditional support.
CEU eTD Collection
iv
Table of Contents
Declaration of Authorship ........................................................................................................................ i
Abstract ................................................................................................................................................... ii
Acknowledgements ................................................................................................................................ iii
Table of Contents ................................................................................................................................... iv
List of Figures and Tables ....................................................................................................................... v
List of Abbreviations .............................................................................................................................. vi
Chapter 1: Introduction ........................................................................................................................... 1
Chapter 2: Foreign Aid and Corruption................................................................................................... 5
2.1 Theories of Anti-Corruption and Applications to the Effects of Foreign Aid ............................... 5
2.2 Quantitative Evidence of the Effect of Foreign Aid on Corruption ............................................ 10
2.2.1 Positive Effect of Foreign Aid .............................................................................................. 10
2.2.2 Negative Effect of Foreign Aid ............................................................................................ 11
2.3 Summary...................................................................................................................................... 16
Chapter3: Methodologies of Corruption Studies: Instrumental Variables ............................................ 17
3.1 Existing Instrumental Variable Approaches ................................................................................ 17
3.2 EU Council Presidency as the Instrument Variable .................................................................... 19
3.3 Summary...................................................................................................................................... 21
Chapter 4: Analysis ............................................................................................................................... 22
4.1 Variables and Model .................................................................................................................... 22
4.1.1 Variables ............................................................................................................................... 22
4.1.2 Model .................................................................................................................................... 24
4.2 Results ......................................................................................................................................... 25
4.3 Discussion.................................................................................................................................... 28
4.3.1 Little Interest to Impose Sanctions against Corruption ......................................................... 28
4.3.2 Case Study 2001 Belgium Presidency and EU Aid to DRC .............................................. 31
4.4 Summary...................................................................................................................................... 35
Chapter 5: Conclusion ........................................................................................................................... 37
Appendix ............................................................................................................................................... 40
Appendix A. List of States ................................................................................................................ 40
Appendix B: List of Variables ........................................................................................................... 42
Appendix C: Distributions of ICRG and V-dem (Political Corruption) score .................................. 43
Appendix D: Replication Data .......................................................................................................... 43
Bibliography .......................................................................................................................................... 44
CEU eTD Collection
v
List of Figures and Tables
Figure 1: Public Sector Corruption, Scale: 0 (low) - 1 (high)....................................................8
Fugure 2: Year-By-Year Effects of EU Aid Allocation on Corruption (t through t +5) .........27
Figure 3: Corruption in DRC (1960-2006) ..............................................................................32
Figure 4: Distributions of ICRG and V-dem (Political Corruption) score...............................43
Table 1: Summary of the Foreign Aid Literature................................................................13-15
Table 2: Descriptive Statistics...................................................................................................24
Table 3: IV Estimates of the Effects of EU Foreign Aid on Corruption
in Recipient Countries.................................................................................................26
Table 4: EU Sanctions to Africa, the Caribbean and the Pacific Countries..............................30
CEU eTD Collection
vi
List of Abbreviations
ACP: Africa, the Caribbean and the Pacific
CC: Control of Corruption (World Governance Indicators)
CPI: Corruption Perception Index
DID: Difference-in-differences method
DRC: Democratic Republic of Congo
EU: European Union
GMM: Generalized Method of Moments
ICRG: International Country Risk Guide
IMF: International Monetary Fund
IV: Instrumental variable
MCC: Millennium Challenge Corporation
NIP: National Indicative Program
ODA: Official Development Assistance
V-Dem Institute: Variety of democracy Institute
WGI: World Governance Indicators
CEU eTD Collection
1
Chapter 1: Introduction
Scholars of political economy have debated the relationship between foreign aid and corruption
in recipient countries particularly since the late 1990s when international organizations started
to pay attention to fighting against corruption in recipient counties. While some scholastic
literature supports that foreign aid can reduce corruption in recipient countries, other literature
denies the positive effects by pointing out null significance or effect of foreign aid on increasing
the level of corruption. For example, Okada and Samreth (2012) find that foreign aid can reduce
corruption, but the extent of the effects depends on donor countries. Asongu (2012), in turn,
argues aid can increase the level of corruption when the cases are limited to African countries.
However, the existing studies do not assign foreign aid randomly, and consequently,
these studies fail to consider bias caused by a variety of donor motivations (Carnegie and
Marinov, 2017). Although many researchers attempt to overcome such endogenous problems,
they only provide limited or no justification in their studies. Many of the instrumental variables
in the existing studies are selected just based on previous applications without justifying
requirements for their use. Reflecting such limitations, this thesis will extend “a novel approach”
developed by Carnegie and Marinov (2017, p.671) European Union (EU) Council Presidency
as the instrumental variable who examine the effects of foreign aid allocation on the level of
democracy and human rights in recipient countries.
Not only does this thesis develop an approach to analyze the relationship between aid
and corruption in recipient countries based on Carnegie and Marinov (2017), it uses up-to-date
innovative indicators from the V-Dem (Varieties of Democracy) Institute. V-Dem successfully
develops indicators which can measure specific aspects of political institutions such as
executive, public sector, judicial, and legislative corruption (V-Dem Institute, 2019), yet only
a limited amount of extant scholarly quantitative literature in corruption studies is based on
CEU eTD Collection
2
these indicators. With theoretical justification, this thesis will utilize these newly-developed
indicators 2OLS analyses of the relationship between foreign aid and corruption.
The ensuing chapters will discuss to what extent EU international aid allocation has an
effect on the level of corruption in recipient countries which were former European colonies
(Appendix 1 for the list of countries), which contributes to the corruption studies in terms of
specific findings about effects of EU foreign aid allocation. Due to the use of the instrumental
variable, EU council presidency, this study limits the scope of analysis to former European (UK,
France, Spain, Italy, Netherland, Belgium, Portugal) colonies. To the best of my knowledge
this is the first study to analyze the effects of EU foreign aid on corruption in former colonies.
Studies of the effects of EU foreign aid on corruption in recipient countries are limited to
corruption in EU member states or candidate countries (e.g., Vachudova, 2009; Fazekas and
King, 2019). Also, while there are some attempts to study the effect of multilateral aid on
corruption in recipient countries, few scholars attempt to examine the effect of each multilateral
institution (e.g., World Bank, International Monetary Fund, European Union). The thesis will
contribute to the anti-corruption literature by identifying the specific effects of EU foreign aid
on corruption in recipient countries of former European colonies.
In terms of defining corruption, although there is no universally agreed definition, this
thesis will use the following: The abuse of public power for private gain. This definition is used
not only because it is close to other common academic definitions of corruption (e.g., “Abuse
of public office for private gain”: Kaufmann, 1997, p.141), but also because it can exclude
broader definitions. The definition in this thesis is narrower than another famous definition of
corruption by Transparency International, “the abuse of entrusted power for private gain”
(Lovitt, 2004, p.1), as this definition also includes corruption in the private sector. In short, this
thesis will exclusively focus on the effects of corruption such as bribery, extortion, and
CEU eTD Collection
3
embezzlement in the public sector (Rose-Ackerman and Palifka, 2016), as other broader
concepts regarding corruption such as money laundering (Bullough, 2018) will not be covered .
In terms of structure, this paper starts with a literature review on major theories of
corruption principal-agent theory and collective action theory and applies the theories to the
relationship between foreign aid allocation and corruption in recipient countries (Chapter 2).
This section provides theoretical foundation of the discussion of the effects of foreign aid on
corruption, which many relevant quantitative articles fail to provide. Then, this paper
deliberates on the results of existing quantitative articles on the relationship between foreign
aid and corruption in terms of methodologies, the scope of their analyses, and their findings
(Chapter 2). Next, the thesis discusses the use of instrumental variables in the analyses of
foreign aid including the instrumental variable in the thesis (EU Council presidency; Chapter
3). I will show that EU Council presidency rotation from 1965 to 2006 will work as the
instrumental variable better than other instrumental variables in the study of EU foreign aid and
corruption. EU Council Presidency until 2006 had been randomly assigned based on
alphabetical order and this perfectly satisfies the requirements as an ideal instrumental variable.
Subsequent to 2006, the new rotation principle of the EU Council Presidency was introduced
at the beginning of 2007, which is no longer regarded as an as-if randomization (Angrist and
Pischke, 2014; Carnegie and Marinov, 2017).
As a main theme, this thesis will conduct 2OLS analyses using the instrumental variable
(Chapter 4). It further qualitatively assesses the effects of EU aid allocation on corruption not
only from the analysis of previous EU sanctions over years but also from the case of aid
programs to DRC which is a rare typical case which had a concrete influence over EU aid
policies using the Council presidency position council conclusions of 12330/01, 13802/01,
and 15078/01 (European Commission, 2001a; 2001b; 2001c), as well as resumed EU aid
distribution (starting from 2001 and boosted from 2003: Arnould and Vlassenroot, 2016;
CEU eTD Collection
4
Bobineau and Gieg, 2016) after the 2001 Belgian EU Council presidency (Chapter 4). I will
show while EU aid allocation until 2006 reduced most types of corruption, the magnitudes of
these effects were quite small. The results will provide strong policy evidence for EU citizens
and policy makers about foreign aid allocation in the sense that they should reconsider how to
provide foreign aid in order to ensure corruption control in recipient countries and not allow
public officials in the countries to behave in a corrupt way.
CEU eTD Collection
5
Chapter 2: Foreign Aid and Corruption
This chapter will review the existing literature of both theories and quantitative findings of the
effects of foreign aid on corruption. I will first discuss theories of corruption and these
applications to the relationship between aid and corruption. Then I will review the quantitative
evidence of the effects of aid on corruption, which examines the effects using a number of
different methodologies, such as the instrumental variable approach.
2.1 Theories of Anti-Corruption and Applications to the Effects of Foreign Aid
Although corruption has been analyzed by many scholars across time (e.g., McGovern, 1908;
Leff, 1964; Huntington, 1969), it was not until the late 1970s that scholars started to find
theoretical explanations of corrupt behaviors as well as theoretical suggestions to reduce
corruption. The first pioneering work on political economy of corruption was Corruption: A
Study in Political Economy, written by Rose-Ackerman (1978). This book was innovative in
the study of corruption because it spread one of the most important ideas in corruption studies
that corruption is less likely to occur under some particular institutional settings. In other words,
Rose-Ackerman (1978) overcomes the limitations of previous studies in economics and
political science which had focused on only either individual benefit maximization without
consideration of institutions (economics) or politicians’ likelihood of winning elections
(political science). Using the principal-agent theory, she (1978) succeeds in explaining that
institutions are prone to corruption when principals cannot monitor agents due to information
asymmetry between principals and agents (see also, Pratt and Zeckhause, 1985).
In the 1980s, another pioneering work on political economy of corruption using
principal-agent theory, Controlling Corruption, was published by Klitgaard (1988). Based on
the principal-agent theory in corruption studies by Rose-Ackerman (1978), Klitgaard skillfully
CEU eTD Collection
6
formulates an equation of corruption control as Corruption = Monopoly + Discretion
Accountability (p.75). According to him, corruption can be tackled when the amount of agents’
monopoly (e.g., privatization and deregulation) and discretion (e.g., merit-based selection and
rotation of agents) and when improve agents’ accountability (e.g., legal enforcement and civil
society empowerment) (2018). This book served later scholars and practitioners as a primary
reference book, particularly the part on policy measures which posits the illustrative equation.
A significant number of later approaches to corruption were based on the principal-
agent theory, particularly since the late 1990s when the World Bank and IMF changed their
policies and decided to fight against corruption in developing countries (World Bank, n.d.;
World Bank, 1997; IMF, 2005). The later scholarly study of corruption has provided examples
of applications of the principal-agency theory to specific institutions which are related to anti-
corruption. Anti-corruption scholars have studied, for example, the effects of electoral
institutions (plurality or proportional representation; e.g., Kunicova and Rose-Ackerman, 2005),
type of executive branch (presidentialism or parliamentarism; e.g., Gerring and Thacker, 2004),
and decentralization and federalism (e.g., Peters, 2010), on the level of corruption.
Principal-agent theory can also contribute to understanding the relationship between
foreign aid and corruption. Although existing anti-corruption studies do not offer theoretical
reviews of the relationship between the two from a principal-agent theory perspective,
theoretical consideration in relevant areas EU membership conditionality
1
demonstrates
principal-agent theory is applicable to the relationship between foreign aid and corruption in
recipient countries. Studies of the effect of EU membership conditionality on corruption show
1
Throughout this thesis, the term conditionality refers to negative conditionality (as opposed to positive aid
conditionality). According to Mungiu-Pippidi (2019), negative conditionality means conditionality due to
“violations of association or trade agreements”(p.74), while positive conditionality refers to conditionality
which offers “rewards for partner countries for their reform gestures” (p.75). This is because the idea of
negative conditionality dominated more than positive conditionality in EU aid allocation decision-making
processes over the period of interest in the thesis.
CEU eTD Collection
7
imposing conditionality, through increasing accountability, led EU candidate countries to
establish necessary institutions to fight against corruption in the pre-accession stage. For
example, EU member states required Romania during its accession process to set up anti-
corruption agencies and increase transparency in judicial process (Szarek-Mason, 2010, ch.4).
2
Applying this conditionality arguments to foreign aid, principal-agent theory indicates that
foreign aid help recipient countries to reduce the level of corruption through imposing anti-
corruption conditionalities which are similar to conditionalities for joining the EU.
However, against principal-agent theory’s expectations, many developing countries
have struggled to fight against corruption since the 1990s. As Figure 1 shows, corruption in the
public sector from 1960 to 2019, the levels of corruption across the world have not significantly
improved regardless of anti-corruption reforms since the 1990s. These realities of anti-
corruption reforms led a growing body of anti-corruption scholars to reconsider traditional
theory critically. The movement dates to around 2010 and an increasing number of scholars
propose different alternative theoretical frameworks to analyze corruption control.
Within these alternative theoretical frameworks, the most prominent theory proposed as
an alternative to principal-agent theory is collective action theory.
3
Anti-corruption scholars
from the collective action perspective argue mainly two limitations of principal-agent theory in
the study of corruption: principals’ interests and collective action dilemma for principals. First,
2
These EU membership requirements on corruption control originate with the Copenhagen Criteria. It was
formulated in 1993 and was developed through the discussion in 1997 and 2000 by describing more detailed
criteria of corruption control and other policies (Janse, 2019).
3
Although this thesis will focus only on principal-agent and collective action theories, it should be noted
here that there are some other alternative perspectives in addition to these theories. The first is the sociological
approach which sees corruption and corruption control in developing countries from the “logic of
appropriateness” (March and Olsen, 2004). Lambsdorff, Taube, and Schramm (2004, ch.1) provide more
detailed theoretical discussion on corrupt transactions, and Onyango (2017) and David-Barrett et al., (2020)
are useful references for applications of a sociological approach to several cases regarding anti-corruption
reforms in developing countries. The other approach is the historical approach. This approach focuses on
path dependency and informal institutions of anti-corruption institutions. Libman and Obydenkova (2013)
conduct quantitative analysis in post-communist countries from a path-dependency perspective, and
Hellmann (2017) offers a qualitative investigation of East Asian countries using the concept of path
dependency and informal institutions.
CEU eTD Collection
8
Africa America
Asia MENA
Figure 1: Public Sector Corruption, shown at the original scale: 0 (low) - 1 (high)
Source: Variable Graph Public Sector Corruption Index (V-Dem, n.d)
according to Persson, Rothstein, and Teorell (2013), principals’ pursuits of public interests are
assumed in most (anti-) corruption studies from a principal-agent perspective.
4
They assert this
assumption does not match the real-world situation regarding anti-corruption reforms by calling
principals principled principal(s) as follows:
[S]ince the supposed “principled principal(s)” are ... not acting in the interest of the
society but instead pursuing their own narrow self-interests, anticorruption reforms
based on the principalagent framework will invariably fail (2013, p.451).
4
Although this is a well-known view of principal-agent theory in the study of corruption, it is debatable
whether principal-agent theory in the corruption study assumes principals’ pursuits of public interests. While
Persson, Rothstein, and Teorell (2019) reassert the view as a comment to Marquette and Peiffer (2018),
Marquette and Peiffer (2019) disagree. They point out that relevant literature in political economy does not
entail this assumption and it is just misuse of principal-agent theory by later anti-corruption scholars.
CEU eTD Collection
9
In the case of foreign aid and corruption, this implies principals donor countries or institutions
can impose foreign aid conditionality for their strategic reasons rather than aiming at recipient
countries to combat corruption and downgrade conditionalities regarding anti-corruption
reforms.
As the name indicates, collective action theory in corruption study offers another
important lens to analyze anti-corruption attempts: collective action dilemma. Based on
Ostrom’s idea of second-order collective action problem (free-rider problems; Ostrom, 1998),
Persson, Rothstein, and Teorell (2013) argue that people, especially in countries where
corruption is widespread, face challenges to act against corruption, because it entails a huge
cost to fight against corruption and benefits to keeping corrupt behavior. Referring to this study,
Marquette and Peiffer (2015) successfully provide extensive lists of potential causes for the
failures of anti-corruption reforms and actions such as group size, group heterogeneity, trust/
good reputations, heuristics / norms (p.5). This implies two points in the analyses of the effects
of foreign aid on the level of corruption in recipient countries collective action dilemma
among donor countries and population in developing countries.
Theoretically, donors can face collective action dilemma when some or many donors
are not willing to seek to combat corruption in recipient countries, which affects the quality and
volume of aid conditionalities. In addition, if aid is not effectively enforced among donor
countries and institutions due to collective action dilemma, people in recipient countries are
also likely to face collective action dilemma, because they cannot rely on external pressure to
change the institutions.
5
These collective action dilemmas may allow corrupt politicians and
5
In this section, I introduced two most famous literature of collective action in the corruption study. However,
it is worth mentioning that, although it is certain that these studies had a great influence to the study of
corruption, there are similar studies which approach anti-corruption from collective action theory even before
Persson, Rothstein, and Teorell (2013) and Marquette and Peiffer (2015). For example, Hopkin (2002)
critically reviews anti-corruption literature at that time, pointing Rose-Ackerman (1999) ignores the
collective action problems in her analysis of anti-corruption policies.
CEU eTD Collection
10
public officers to continue their corrupt practices or even extend them by wasting aid budget.
As Marquette and Peiffer (2015) point out, this problem will be exacerbated when countries
have larger populations and greater heterogeneity.
2.2 Quantitative Evidence of the Effect of Foreign Aid on Corruption
There are both positive and negative effects of foreign aid on corruption. In this section, I will
firstly review effects of aid on reducing corruption and then the effects of aid on fueling
corruption.
2.2.1 Positive Effect of Foreign Aid
A series of quantitative studies have been conducted and found foreign aid can reduce the level
of corruption in recipient countries. For example, Charron (2011) examines the effects of
foreign aid on corruption, in comparison of pre-1997 and post-1997 when the World Bank
boosted anti-corruption reforms in developing countries (World Bank, 1997). Through analyses
of around 70 developing countries, Charron (2011) finds the level of corruption will decrease
when countries rely more on foreign aid and this positive effect occurs only allocation of foreign
aid after 1997. In addition, to capture different effects depending on the level of corruption in
recipient countries, Okada and Samreth (2012) develop the quantile regression methodology
and find foreign aid has greater effect to reduce corruption in less corrupt countries than more
corrupt countries. They also try to identify the different effects on corruption depending on
donor countries/institutions and show multilateral aid has effects on reducing corruption, while
most bilateral aid (France, the UK, and the US) does not have significant effects. This is an
interesting result in relation to similar previous studies. For example, Alesina and Weder (2002)
study the effects of corruption on donors’ decision over aid allocation. Through regression
analysis with a limited number of cases, they (2002) suggest the level of corruption does not
CEU eTD Collection
11
have significant impacts on aid allocation from both multilateral and bilateral aid. Okada and
Samreth’s (2012) finding implies more nuanced effects of foreign aid depending on donor
countries and institutions.
Several recent studies use econometric methods to find positive effects of foreign aid in
specific times and places. For example, Öhler, Nunnenkamp, and Dreher (2012) examine the
effect of a specific US aid conditionality scheme (Millennium Challenge Corporation - MCC;
from 2004) on the level of corruption in 62 recipient countries. Investing the effects through a
difference-in-differences (DID) method, they find (i) the US aid scheme had effects to reduce
corruption in recipient countries and (ii) the effects have weakened over time. The results of
the study show that US aid conditionality initially worked well because it demanded recipient
countries satisfy requirements for receiving US aid, yet limited additional evaluation processes
after they passed the criteria weakened the continuous efforts for recipient countries to fight
against corruption (Öhler, Nunnenkamp, and Dreher, 2012).
2.2.2 Negative Effect of Foreign Aid
However, other quantitative analyses have also been conducted and have revealed foreign aid
can fuel corruption in recipient countries. As a pioneering study of the relationship between the
two, Knack (1999) finds the slightly negative effect of aid dependence on corruption in 80
recipient countries. Kangoye (2013), who uses the same corruption indicator (International
Country Risk Guide; ICRG), also conducts cross-section and panel regression analyses.
Focusing on unpredictability of foreign aid, he finds aid unpredictability can fuel corruption in
recipient countries (2013).
A number of scholars find similar results in region-specific investigations of the effects
on African countries. Asongu (2012) focuses on the effects of aid on corruption in 52 African
countries using World Governance Indicators (Control of Corruption) and finds Okada and
CEU eTD Collection
12
Samreth's (2012) policy implication aid can reduce corruption through conditionality is not
applicable to African countries.
6
Asongu and Nwachukwu (2016) also analyze the effects of
foreign aid on different components of good governance political: voice and accountability,
political stability; economic: regulation quality, government effectiveness; institutional: rule of
law, and control of corruption. They (2016) find foreign aid has negative impacts on economic
and institutional governance including corruption, whereas it has no impacts on political
governance.
In sum, the effects of foreign aid on corruption in terms of variables used, methods,
cases and time coverage, and results are illustrated in Table 1. As shown in the table, there are
a lot of studies which use instrumental variables (IVs), yet many of them do justify partially or
do not justify the methodology. In addition, their uses of IV do not succeed in finding any causal
relations in specific donor countries or institutions. It is to these this thesis will now turn.
6
Although this is Asongu's (2012) argument, it should be noted this is an overstated argument. His model is
not the replication of Okada and Samreth (2012). In fact, Mohamed et al. (2015) analysed African countries
through similar models and the same method and find foreign aid can reduce corruption in African countries
as well.
CEU eTD Collection
13
Table 1: Summary of the Foreign Aid Literature
Author(s)
Independent
Variable
Dependent
Variable
Control
Variables
Method
Instruments
Time
(Case)
Results7
Asongu
Aid
dependence
WGI (CC),
CPI
Political regime
Trade
IV,
GMM
Legal-origins
Income-levels
Religious-dominations
1996-
2010
(52)
Asongu and
Nwachukwu
Aid
dependence
WGI (all),
CPI
Political regime
Trade
IV
Legal-origins
Income-levels
Religious-dominations
1996-
2010
(52)
Charron
Aid
dependence
(bilateral,
multilateral)
ICRG
(corruption)
CPI
Political regime
GDP per capita
GMM
(n/a)
1986-
2006
(68-75)
+8
Ear
Aid
dependence
(non-lagged
and lagged)
WGI
(CC)
n/a
IV
Infant mortality
1996-
2004
(155)
/+
Kangoye
Aid
unpredict-
ability
ICRG
(corruption)
Income
Population
Rents
Fractionalization
IV
“instruments constructed from
the information set” (p.132;
original idea from Pagan and
Ullah, 1988)
1984
2004
(80)
7
Here, negative sign means aid increases corruption and positive sign means aid reduce corruption.
8
This is the effect of multilateral aid. The effect of Bilateral aid after 1997 was weaker but still negative.
CEU eTD Collection
14
Table 1 (cont.)
Author(s)
Independent
Variable
Dependent
Variable
Control
Variables
Method
Instruments
Time
(Case)
Results9
Kangoye
(cont.)
Legal origin
Accountability
Landlock
Africa
Knack
Aid
dependence
10
ICRG
(all)
GDP
Population
IV
Infant mortality rate in
1980s
Initial GDP per capita
Initial population,
French zone dummy
A Central America dummy
1982-
1995
(80)
Mohamed,
Kaliappan,
Ismail, and
Azman-Saini
Aid
dependence
CPI
GDP per capita
Democracy
Rule of law
Political stability
QR
(n/a)
2002-
2010
(42)
+
9
Here, negative sign means aid increases corruption and positive sign means aid reduces corruption.
10
Aid dependence is measured by the ratio of aid/gdp. Other papers which measure aid dependence also use the same measurement.
CEU eTD Collection
15
Table 1 (cont.)
Okada and
Samreth
Aid
dependence
11
WGI
(CC)
GDP per capita
Democracy
Legal origin
QR
(n/a)
1996-
2008
(120)
+12
Öhler,
Nunnenkamp,
and Dreher
MCC
WGI
(CC)
n/a
DID
(n/a)
before
/after
2002
(62)
+
(short-
term)
Tavares
Aid
dependence
ICRG
GDP per capita
Political rights
Fractionalization
Oil export
Population
Expenditure
IV
“geographical and cultural
distance to donor countries”
(p.99)
5-year
avg.
(182)
+
11
They create separate independent aid variables depending on donor countries. They create multilateral aid variables as well.
12
The effects of aid (general) and multilateral aid. Most bilateral countries (except for Japan) do not have significant impacts.
CEU eTD Collection
16
2.3 Summary
This chapter reviewed theoretical and quantitative literature on the effects of foreign aid on the
level of corruption in recipient countries. First, this chapter presented principal-agent theory
and collective action theory in the study of corruption as well as the applications of these
theories in the case of the effects of foreign aid on corruption. The former theory argues foreign
aid helps recipient countries to reduce corruption through conditionalities. Against this, the
latter theory reveals both donors and recipient countries face collective action problems and
argues this allows political executives and public servants to keep or extend their corrupt
practices. Second, this chapter reviewed quantitative studies of the relationship between foreign
aid and the level of corruption through different measurements, methods, and number of cases
and year coverage. While a certain number of studies reveal foreign aid can increase corruption
control in recipient countries, other studies find aid can deteriorate the level of corruption. The
methodological challenges in many of these articles will be discussed in the next chapter.
CEU eTD Collection
17
Chapter3: Methodologies of Corruption Studies: Instrumental
Variables
In this chapter, I will offer the instrumental variable approach in this thesis with a critical review
of previous usages. The present chapter will first critically review the existing instrumental
variable approaches in the study of foreign aid and corruption. Following this review, this thesis
will justify the use of the EU Council presidency as the instrument of analysis.
3.1 Existing Instrumental Variable Approaches
Examining the relationship between foreign aid and corruption in recipient countries involves
many challenges, such as endogeneity and reverse causality. Reflecting such challenges, many
scholars in this field have attempted to examine the effects through instrumental variables (IV).
Many existing relevant literatures recognize “an important issue of endogeneity owing to
reverse causality and omitted variables” (Asongu and Nwachukwu, 2016, p.2176). Instrumental
variables are used in some foreign aid and corruption articles to solve “the problem of
endogeneity of one or more explanatory variables” (Wooldridge, 2016, p.461).
Despite some attempts to examine the effects of foreign aid on corruption using
instrumental variable method (Table 1), most of these existing studies more or less contain
methodological problems: (i) many weak instruments, (ii) unobservable heterogeneity of
instruments, and (iii) limited justifications of the exclusion requirement. As Angrist and Pischke
(2014) review, there are three main requirements for the use of instrumental variables:
(i) The instrument has a causal effect on the variable whose effects we’re trying to capture
[…]
(ii)The instrument is randomly assigned or “as good as randomly assigned” in the sense
of being unrelated to the omitted variables we might like to control for […]. This is known
as the independence assumption.
(iii) Finally, IV logic requires an exclusion restriction. […] [It] describes a single channel
through which the instrument affects outcomes (p.116; italic original).
CEU eTD Collection
18
Most existing studies do not convincingly succeed to pass one or more of these requirements.
First, there are some studies which do not show significant effects of the instrument, which is
against Angrist and Pischk's (2014) first requirement. For example, Knack (1999) uses many
instruments to estimate the effect of foreign aid on corruption.
13
The instruments include infant
mortality rate in the 1980s, initial GDP per capita, initial population, French zone dummy, and
A Central America dummy. According to Angrist and Pischk (2014), this way of instrumental
variable analysis is problematic because using many weak instruments “to generate precise
estimates of a single causal effect” (p.150) cannot detect effectively the causal relationship
between foreign aid and corruption. Since this problem can be solved by using a single
instrument which has significant effects on the independent variable, using many instruments
are considered a risk in instrumental variable analysis.
Second, relating to the independence assumption, most of the instruments in existing
studies have a risk of unobserved heterogeneity, which can weaken the results of foreign aid
measured through the instrumental variable method. Many of the instruments were used in
previous studies, such as “geographical and cultural distance to donor countries” (Tavares,
2003, p.99). Yet, these “instruments may suffer from unobserved heterogeneity” because
scholars cannot provide how exactly assignments proceed (Carnegie and Marinov 2017,
p.671; see also Sekhon and Titiunik, 2012).
Third, some scholars poorly justify the exclusion restriction in their investigation of the
effect of aid on corruption. For instance, while Ear (2007) uses infant mortality similar to Knock
(1999), he provides only limited and intuitive justifications to the exclusion restriction:
Of course, infant mortality is unlikely to be purely exogenous: it is likely affected by poor
governance, though in the long term it is unlikely infant mortality would affect governance.
Infant mortality rates would need to be massive to have a serious impact on a country’s
governance capacity, and save for adult HIV=AIDS prevalence in Africa, this is unlikely to
be the case (Ear, 2007, p.265; emphasis added).
13
The same criticism is also applicable to Asongu, 2012; Asongu and Nwachukwu, 2016.
CEU eTD Collection
19
This limited justification weakens the validity of the findings, as it does not explain how long
infant mortality would no longer have an impact on governance including corruption.
In addition to the limited justifications to three requirements, it should also be noted that
the existing research has another shortcoming, namely lack of consideration over time lag of
the effects of foreign aid on corruption. Except for Ear (2007), none of the articles
14
lagged aid
to examine the effects on corruption in recipient countries. According to statistic literature, the
results of these studies can be less convincing than studies using lagged foreign aid because the
latter can effectively prevent post-treatment bias (Rosenbaum, 1984; Carnegie and Marinov,
2017).
3.2 EU Council Presidency as the Instrument Variable
Reflecting problems in the use of IVs in the previous literature of foreign aid, this thesis will
use EU Council presidency as the instrument to overcome the existing problems. As Carnegie
and Marinov assert, EU Council presidency until 2006 perfectly satisfies the requirements as
the IV: EU Council presidency has a causal effect on EU aid allocation; EU Council presidency
had been rotated based on alphabetical order in their native language (i.e., it is randomly
assigned and the assignment mechanism is quite clear); and EU Council presidency is unlikely
to influence the level of corruption directly and can influence only through foreign EU aid
allocation (2017).
15
In particular, the second point random assignment knowing the
mechanism is quite unique. “[N]atural experiments seldom have a known random assignment
14
Knack (1999), Tavares (2003), Asongu (2012), and Asongu and Nwachukwu (2016)
15
At the beginning of 2007, “the rotation principle changed such that three countries hold the presidency at
a time” (Carnegie and Marinov, 2017), based on “the demographic and geographical range of all the Member
States” (Treaty of Lisbon, 2007, Article 9D-5). EU Council Presidency from 2007 will not be considered as
the ideal instrument because this principle limited the presidency’s power for initiating discussion in the
council (Vaznonytė, 2020) and was not regarded as an as-if randomization in the use of the instrument
variable (Carnegie and Marinov, 2017).
CEU eTD Collection
20
mechanism” (Sekhon and Titiunik, 2012, p.53), yet this is not the case in EU Council
presidency as the instrument. This thesis will utilize this finding by Carnegie and Marinov
(2017) and apply it to the study of corruption, although this instrument limits the scope of the
analysis in this thesis to former European colonies. This is because “former colonizers tend to
have political, cultural, and economic ties to these states” (Carnegie and Marinov, 2017, p.674)
and thus former colonizers are likely to seek to distribute aid more to former colonies.
Nevertheless, considering the huge advantages of the instrument and its potential for
implications (as this thesis noted in the introduction), this paper regards EU Council presidency
as the most suitable instrument.
How can EU Council presidency affect foreign aid allocation? Applying theories in
International Relations and American Politics, Tallberg (2004) theoretically argues the
presidency can influence EU policy-making processes through agenda setting. He additionally
provides several case studies in Belgium, France, Finland, Germany, Portugal, and Sweden
from 1999 to 2001. Moreover, as an investigation of specific policies, Warntjen (2007)
examines the effects of EU Council presidency on EU environmental policies for wider periods
(1984-2001) and finds influences using the position of EU Council presidency.
Based on these general findings, some scholars investigate further in which process EU
Council presidency has power to influence EU policies. The most academically agreed
argument is the presidency has an effect at the final stage of policymaking processes, namely
the final revisions and the voting stage. Schalk et al. (2007) find EU Council presidency has
influence most in the final stage from quantitate analysis using Decision making in the
European Union dataset, which include 162 policy issues and 66 proposals. This result is robust
in other studies such as Warntjen (2008). He recognizes the study of Schalk et al. (2007) in
terms of lack of consideration of different policy issues and controls effects caused by different
CEU eTD Collection
21
policy issues, yet Warntjen (2008) still finds the effects of EU Council presidency in EU
policies at the last stage.
3.3 Summary
This chapter reviewed the use of instrumental variables in existing studies and proposed the
instrument this thesis will use. Regarding the literature review on instrumental variables in
studies of foreign aid, this paper points out three limitations of their use of IV: many IVs,
unclear assignment mechanism, and limited justification to exclusion restriction. Reflecting on
such limitations, this paper justified its use of EU Council presidency to analyze the effects of
EU aid on corruption in recipient former colonies as a replication of Carnegie and Marinov
(2017). EU Council presidency is a perfect instrument as it has a causal effect on foreign aid
allocation, the assignment mechanism is known, and it is not likely to affect directly the level
of corruption.
CEU eTD Collection
22
Chapter 4: Analysis
This chapter presents the results of quantitative and qualitative analyses. After outlining the
variables and the model, I will report the effects of aid allocation on four-year average
corruption scores as well as these year-by-year effects. In the following section, this thesis will
critically present qualitative evidence to support quantitative analyses: overall EU attitude
against corruption and case study of DRC.
4.1 Variables and Model
4.1.1 Variables
The present study uses country panel data from the late 1980s to 2006. Due to the
methodological nature of the study, EU aid recepient countries are limited to former European
colonies. Data for the dependent variable, corruption in recepient countries, come from two
indicators: ICRG and v-dem. First, as a measurement of corruption in recipient countries, I use
corruption scores from political risk components of International Country Risk Guide (ICRG)
index, which is available from 1982. Political components of ICRG index is 100-point scale
indicator and measures countries’ political risk for corporate foreign investment by adding the
specific component which can be political risks (The PRS Group, n.d.). Corruption is also
measured (6-point scale) as a component of the indicator. According to The PRS Group (n.d.),
it measures political risks which can be caused by corruption including bribes, patronage, and
nepotism through inefficient economic activities and dark market and is used as one of the
sources of the Corruption Perception Index (CPI). While original scores of ICRG range from
0 (high corruption risk) to 6 (low corruption risk), I inversed and rescale the index from 0 (low
CEU eTD Collection
23
corruption risk) to 10 (high corruption risk) in order to make it comparable to other dependent
variables.
Second, in order to measure specific aspects of corruption, I also use relevant indicators
from the v-dem institute: political corruption, executive corruption, and public corruption. V-
dem is a rich dataset aimed at measuring different components of democracy and provides three
major indicators to measure corruption. Political Corruption score is an aggregated indicator
which covers corruption in executive, judicial, legislative, public sector, judicial, and legislature
corruption. Executive Corruption, Public Sector Corruption, Judicial corruption decision and
Legislature corrupt activities scores measure corruption in the executive, public officers,
judicial branch, and legislative branch respectively. Compared with other corruption indicators
such as World Governance Indicators (WGI) and Corruption Perception Index (CPI), these
five v-dem corruption indicators have a significant advantage in corruption analysis in the sense
that they cover much longer periods than existing corruption indicators. While WGI and CPI
cover only after the late 1990s, v-dem corruption indicators cover from 1789 to the present (V-
Dem Institute, 2019). They are distinguishing indicators because they provide both average
corruption score and each component score for such a long period, which previous corruption
indices except for proxy indicators including ICRG fail to achieve. While original scores of
v-dem corruption indicators range from 0 (low corruption) to 1 (high corruption), I rescaled it
to a 10-point scale (0-10; from low to high).
In terms of independent variable and instrumental variable, as justified in the previous
chapter, the present study will follow Carnegie and Marinov (2017). This thesis will use EU
foreign aid allocation (Carnegie and Marinov, 2017), which originates from the OECD, as the
independent variable. I will also follow the usage of the instrumental variable of EU foreign aid
council until 2006, as Carnegie and Marinov (2017) skilfully find it a useful and effective
instrumental variable. Table 2 presents a first look at descriptive statistics of the data used in
CEU eTD Collection
24
Table 2 Discriptive Statistics
Variable Obs. Mean SD Minimum Maximim
Former colony status 4,027 .0902657 .2858383 0 1
EU aid allocation 4,250 2.154306 1.507431 0 5.784837
ICRG 1,788 5.681581 1.689272 0 10
V-dem
Political corruption 4,762 5.698935 2.544215 .1 9.7
Executive corruption 4,804 5.608364 2.631571 .14 9.75
Public corruption 4,804 5.448416 2.578767 .06 9.74
Judicial corruption 4,754 4.807352 1.425333 1.3525 8.645
Legislature corruption 4,249 5.385051 1.497371 1.61125 9.3
the analysis and Appendix B presents the variables and sources.
4.1.2 Model
Following the models generated by Carnegie and Marinov (2017), the regression model
(second-stage regression) in the present study is the following:
CORit = β0 + β1 log(ODA)i (t1) +  
 +  
 + uit,
where CORitindexes country i’s corruption risk for forging investment (ICRG) or political,
executive, public sector, judicial, and legislative corruption (v-dem) in period t’t, log(ODA)i
(t1) is the logged net EU aid allocation (ODA) in period t 1 which I replicated from Carnegie
and Marinov (2017; originally from OECD data), I reflects country/year fix effect on the model,
and uit is an error term.
As I argued in Chapter 3, the present study will conduct first-stage regression before
analyzing the previous model. The first-stage regression model is
log(ODA)i (t1) = γ0 + γ1Colonyi (t2)2 +   
 +  
 + eit,
where Colonyi (t2)2 is the instrument which shows whether the country i was a former colony of
European countries (UK, France, Spain, Italy, Netherland, Belgium, Portugal). Because of the
different extent of influence through the EU Council Presidency position (as reviewed in
Chapter 3), the present study considers the presidency position only in the later half of the year
CEU eTD Collection
25
(JulyDecember) in period t 2, considering the time lag between the policy decisions and
effects caused by the decisions.
4.2 Results
Table 3 reports the results of 2OLS regressions. Column 1 to 6 show the results of 2OLS
regressions of each independent variable. The first-stage regressions in all models show that
there are statistically significant effects of EU Council Presidency on aid allocation. F-statistics
of excluded instruments in the first-stage regressions show the EU Council Presidency has
strong power as the instrument. Regarding the main models, the second-stage regressions
without covariates (columns 1 to 5) show that EU foreign aid allocation seems to have effects
on reducing corruption, yet the magnitude of the effects is quite small. For example, column 1
shows 100-percent increase of aid causes only 0.6 point reduction (10-point scale)
16
of the level
of corruption in recipient countries. Other models show even smaller effects. The comparison
with the effects of aid allocation on human rights and democracy also shows the scale is small:
one-log increase of EU aid allocation has an impact on improving 1.885 human right score
(scale: 0-14) and 2.031 democracy score (scale: from -10 to 10) (Carnegie and Marinov, 2017).
Compared among components of v-dem political corruption indicators (column 3, 4, 5),
the effects are slightly bigger on executive and public sector corruption (column 3, 4) than on
judicial sector (column 5). This slight difference can be attributed to the different extent of
understanding of corruption in executive and public sector corruption rather than in judicial
corruption. Executive corruption incidents are often revealed in media releases or opposition
16
Sceptical readers may claim this can be substantial effects, yet the effects of ICRG are indeed small.
Although the detailed methodologies of scoring are not published by PSR group, the scoring criteria looks
quite rough in comparison with v-dem indicators and scores tend to jump one point in the 0-6 scale only for
a year (i.e., more than one point on the 10-point scale; see Appendix C for the visualization of score
distribution). Considering this tendency of ICRG scoring, I argue the effects of ICRG are quite small.
CEU eTD Collection
26
Table 3: IV Estimates of the Effects of EU Foreign Aid on Corruption in Recipient Countries
Dependent
Variable
(Four-Year
Avarage)
ICRG
V dem
(Political
Corruption)
V dem
(Executive
Corruption)
V dem
(Public
Corruption)
V dem
(Judicial
Corruption)
V-dem
(Legislature
Corruption)
1
2
3
4
5
6
Aid
Allocation
(SE)
-0.631
(0.533)
-0.041
(0.034)
-0.250
(0.151)
-0.262
(0.133)
-0.038
(0.239)
0.136
(0.132)
Countries
65
102
102
102
101
102
Years
18
22
22
22
20
20
Covariates
No
No
No
No
No
No
Fixed Effects
(Year)
Yes
Yes
Yes
Yes
Yes
Yes
Fixed Effects
(Country)
Yes
Yes
Yes
Yes
Yes
Yes
Observations
1538
2002
2017
2017
1994
1945
Note: ICRG and v-dem scores are rescaled as 0-10 scale (10 means high corruption). First-stage coefficient in each model is the folowing:
(1), 0.195 (SE = 0.067, F= 8.46) (2) 0.149 (SE = 0.039, F = 14.64), (3) 0.162 (SE = 0.046, F = 12.31), (4) 0.162 (SE = 0.047, F= 11.55), (5)
0.150 (SE=0.039, F=14.46), (6) 0.158 (SE=0.050, F=10.03).
CEU eTD Collection
27
Figure 2: Year-By-Year Effects of EU Aid Allocation on Corruption (t through t +5)
CEU eTD Collection
28
parties (of course, the extent of influence depends on political environment in each country),
and public sector corruption can be relatively well-known through aid development policy
implementation. In terms of legislature corruption, it is a different result from other models in
the sense that the models (column 6) show that they have small effects to fuel corruption.
Similar to judicial corruption, this may imply that EU policymakers have less ability to monitor
corrupt activities in legislative branches and allow domestic politicians slightly to expand
corrupt activities.
Turning to Figure 2, I test the year-by-year effect of EU aid allocation on corruption in
recipient countries. Overall, year-by-year effects are also quite small in all models over the
period t through t + 5. However, the effects over periods differ slightly among ICRG scores, v-
dem scores for political, executive, and public sector and for judicial and legislature. While the
effects of EU aid on ICRG scores diminished over the period, the effects on v-dem scores
(political, executive, and public sector) strengthened slightly during the later period. This may
imply that foreign investors slightly change the perception of corruption risk immediately
(ICRG), whereas academics take a longer time to change slightly the perception of corruption
in recipient countries. Similar to the findings from Table 3, year-by-year effects on judicial and
legislature corruption show that the size of effects are much smaller than on corruption risk and
specific aspects of corruption regardless of the period after the aid allocation.
4.3 Discussion
4.3.1 Little Interest to Impose Sanctions against Corruption
Why did all models show only small effects of EU aid allocation on corruption in recipient
countries? Previous EU consultation and sanction (aid suspension) implementation illustrates
the different extent of EU’s engagement in anti-corruption until 2009. Table 4 shows the cases
which were under consultation for aid suspension or which EU actually suspended aid
CEU eTD Collection
29
allocation as a sanction. It is clear many countries do not actively impose EU sanctions against
corruption. There is only one country (Liberia in 2001) in which EU imposed a sanction against
corruption and no former European colonies received sanctions against corruption. Limited
interest to impose sanctions against corruption are clearer when it is compared with other
possible reasons for sanctions. EU rarely imposes sanctions against corruption in Africa, the
Caribbean and the Pacific countries, while it imposed sanctions against violations of Human
Rights (13 cases), Rule of Law (23 cases), and Democracy (34 cases).
This trend did not change even after the Cotonou Partnership Agreement in 2000. This
was signed between EU member states and Africa, the Caribbean and the Pacific (ACP) states
in June 2000 and deals with corruption as one of the fundamental political principles. Article 9
of the Agreement refers not only to human rights and democracy promotion but also good
governance as fundamental principles:
2. The Parties refer to their international obligations and commitments concerning respect
for human rights. They reiterate their deep attachment to human dignity and human rights,
which are legitimate aspirations of individuals and peoples. Human rights are universal,
indivisible and inter-related.
3. In the context of a political and institutional environment that upholds human rights,
democratic principles and the rule of law, good governance is the transparent and
accountable management of human, natural, economic and financial resources for the
purposes of equitable and sustainable development. ... Good governance, which underpins
the ACP-EU Partnership, shall underpin the domestic and international policies of the
Parties and constitute a fundamental element of this Agreement. The Parties agree that
only serious cases of corruption, including acts of bribery leading to such corruption, as
defined in Article 97 constitute a violation of that element. (European Union, 2000,
Article 9)
Unfortunately, the comparison of EU sanctions before and after the Agreement shows that while
the number of sanctions against violations of human rights and democracy were increased, the
sanctions against corruption did not. Such little interest to impose sanctions against corruption
was also the case in a DRC aid program designed during the Belgium presidency in 2001, which
this thesis will examine as a typical case study.
CEU eTD Collection
30
Table 4: EU Sanctions on Africa, the Caribbean and the Pacific Countries
Year
Human
Rights
Rule of Law
Democracy
Corruption
Sudan
1990
Haiti
1991
Kenya
1991
DRC
1992
Togo
1992
Malawi
1992
Equatorial Guinea
1992
Guatemala
1993
Nigeria
1993
Gambia
1994
Comoros
1995
Nigeria
1996
Burundi
1996
Sierra Leone
1997
Togo
1998
Niger
1999
Guinea-Bissau
1999
Comoros
1999
Côte d'Ivoire
2000
Haiti
2000
Fiji
2000
Côte d'Ivoire
2001
Liberia
2001
Zimbabwe
2002
Central African Republic
2003
Guinea
2004
Togo
2004
Guinea-Bissau
2004
Mauritania
2005
Fuji
2007
Mauritania
2008
Guinea
2009
Madagascar
2009
Niger
2009
Source: Zimelis, 2011, p.397
CEU eTD Collection
31
4.3.2 Case Study 2001 Belgium Presidency and EU Aid to DRC
In this section, I will examine the case of the 2001 Belgium EU Council Presidency and its
effect on DRC aid programs. This case is categorized as a rare typical case in the sense that it
is typical in the instrument mechanism, independent and dependent variable (Rohlfing, 2012).
17
It is derived from the results that not only did the Belgium presidency succeed in allocating EU
aid to DRC after 2001
18
(following the decisions in EU council meeting 12330/01, 13802/01,
and 15078/01; European Commission, 2001a; 2001b; 2001c) but also most of the indicators of
corruption show that there were no changes to the level of corruption in DRC after 2001 (Figure
3). Before Belgium started to actively engage in development cooperation from 2001, the actual
intervention to DRC from European countries was limited to diplomatic efforts and did not
include much security effort (Kelly, 2007), although EU seemingly increased its intervention
to DRC since the late 90s. As Bourque and Sampson (2001) review, most of the intervention
from European countries to DRC until the 90s were based on domestic and international NGOs.
This trend changed in 2001, after Belgium signed the Agreements with DRC which
promised to offer around 20 million Euros as bilateral development aid. The existing researches
analyze the potential factors which encouraged Belgium to enhance the relationship with DRC.
First, there was a major change of government coalition in Belgium in 1999. The long-ruling
conservative party, which was reluctant to engage in active interventions to DRC stepped down
after the election in 1999 and the liberal parties that support further interventions to DRC joined
in a coalition (Kelly, 2007; Bunse, 2009). For example, the Minister of Foreign Affairs Louis
17
The Belgian case is also valuable in that it is an example of an EU presidency which actually influenced
EU aid policies. This helps scholars to confirm the mechanism between the presidency and aid allocation
which the instrumental variable in this thesis assumes occurred in a real case (Seawright, 2016, ch.7).
18
It should be noted that the particular presidency in Belgium was chosen based on the analysis of Nasra
(2011), which found the 2001 presidency had a significant impact in comparison with later policy influence
from Belgium.
CEU eTD Collection
32
Figure 3: Corruption in DRC (until 2006, 0-10 scale)
Michel actively promoted development cooperation to DRC. Kelly (2007) notes the active
engagement of Louis Michel regarding Belgian intervention to DRC:
When Michel became the Minister of Foreign Affairs at the end of 1999, he called for the
reinforcement of the Lusaka Peace Accords in the DRC. He also used his political power
to surround himself with other individuals who shared his opinion on intervention in the
DRC. In January 2001, he appointed Reginald Moreels his Special Representative for
Humanitarian Affairs to advance the Lusaka process (p.77)
Additionally, Bunse (2009) argues Belgium decided to promote development support to
African countries also in order to “restore Belgium’s international reputation by taking the lead
in peace initiatives, conflict prevention, and humanitarian aid” (p.119).
Reflecting on such political environment, Belgium also promoted EU to engage in
multilateral development cooperation in the Great Lake region (including DRC) during the
Belgian presidency in 2001. During the presidency, Belgium effectively used the presidency’s
agenda-setting role and reported them as council conclusions (Nasra, 2011). Belgium
repeatedly raised development cooperation to DRC as an agenda in weekly meetings (12330/01,
CEU eTD Collection
33
13802/01, and 15078/01: European Commission, 2001a; 2001b; 2001c). Although the first
meeting ended up with just expressing EU’s view that they were in favor of financial support
to DRC (12330/01: European Commission, 2001a), Belgium successfully led the following
meetings to conclude further on the issue. Weekly meeting 13802/01 concluded to confirm the
support (European Commission, 2001b) and weekly meeting 15078/01 discussed further both
signing an aid program in DRC and ways to implement aid programs through National
Indicative Program
19
(NIP) (European Commission, 2001c).
Furthermore, the Belgian presidency succeeded in concluding multiple EU development
programs in DRC for later years.
20
Starting from the decision in 2001 to offer 1.9 million Euros
for DRC conflict resolution, EU boosted its amount of development cooperation: 5.5 million
Euros in 2002 and 106 million Euros in 2003 (Bobineau and Gieg, 2016). Considering the aid
suspension from 1992 to 2001 (Arnould and Vlassenroot, 2016), these increases of aid
allocation to DRC from 2001 (particularly from 2003) are indeed significant in terms of DRC’s
development cooperation with EU.
The EU aid programs to DRC after the Belgian presidency were a comprehensive
package of development cooperation: it consists in democracy and human rights promotion,
judicial sector reforms, and humanitarian supports (Bunse, 2009; Arnould and Vlassenroot,
2016). Examples include civilian protection from being attacked, “capacity-building of judicial
officials”, allocation of more resources for judicial officials to expand law enforcement, and
support for and observations of local and national elections (Hoebeke, Carette, and Vlassenroot,
19
NIP “defines the sectors and fields which will receive the aid, explains how the aid will fulfil its objectives,
gives a timetable for implementation, and specifies how NGOs will be involved in the programme (if
relevant)” (European Commission, 2002, p.22).
20
There are several analyses which examine why the Belgian presidency succeeded in not only discussing
in the European Council but also reaching multiple conclusions and agreements. For example, Nasra (2011)
and Bunse (2009) similarly point that it was effective that the Belgian presidency emphasized EU’s role of
aid allocation to African countries due to “historical responsibility” (Bunse, 2009, p.129) rather than stating
Belgian political and economic interests. Bunse (2009) noted Belgium encouraged the EU to allocate aid to
DRC prior to the Belgian presidency through, for example, initiating a dialogue with African countries in
2000.
CEU eTD Collection
34
2007; Arnould and Vlassenroot, 2016, p.12). However, despite these programs being
implemented after the Cotonou Agreement, relevant evaluation reports reveal that EU did not
actively engage in anti-corruption through the aid programs (Davis, 2015). The case in point is
the lack of documentation of corruption risks in DRC:
The programme documents do not mention a number of major risks notably the lack
of political will, fraud and corruption which are a serious matter for concern in the
DRC. (European Court of Auditors, 2013, p.22)
In addition, not only did EU not document any corruption risks in the implementation of aid
programs, but also it arguably increased the risk of corruption in aid programs. As Bourque and
Sampson (2001) note, the budget for the programs were flexible, which permits locals to use
the budget without any prior justification. Although this can be effective to respond
immediately, this may allow locals to use it as a source of corruption. Hoebeke, Carette, and
Vlassenroot (2007), indeed, reveal an aid program which was launched to support salary
payments to soldiers, which allowed corrupt executive soldiers to embezzle most of the
payment, as well as to gain additional payment for their personal use via registering ghost
soldiers.
Furthermore, according to existing reviews, the implementations of aid programs on
judicial reforms, which can reduce corruption indirectly,
21
were not strong enough to reduce
corruption. As Rubbers and Gallez (2012) conclude, judicial reforms “did little or nothing to
stem corruption” (p.5) and corruption by judicial officers remained regardless of the reforms.
Although it is challenging to find the reasons why it was not effective to contribute to tackling
corruption through judicial reforms, there seems to be two main factors which prevented deeper
reforms: idealistic discourse and fragmented reforms. First, the judicial reforms in DRC were
21
Anti-corruption through judicial reforms can be achieved through increasing both resources and
impartiality of judicial decisions (Mendelski 2012). Examples include establishment of anti-corruption
specific courts in Indonesia from 2002 to 2010 when its resource was significantly diminished by separation
of its power into provincial courts (Butt and Schütte, 2014)
CEU eTD Collection
35
implemented with an idealistic discourse that judicial reforms such as increasing resources to
judicial offices themselves could reduce corruption without any further detailed reforms.
Rubbers and Gallez (2012) clearly reveal this point from an interview with a local officer:
We are always confident that justice will win out, except that there are some people who
have come there to make a mockery of justice through corruption. But the truth is, the
truth will always find a way to make itself known, or to win out. (p.101)
Second, the reforms were so fragmented that these did not lead to fundamental changes
including corruption control. The scope of reforms was limited to just increasing judicial
capacity (i.e., not increasing impartiality as well) and not connected well with other relevant
programs such as security aid programs (Arnould and Vlassenroot, 2016).
4.4 Summary
This chapter analyzed the effects of EU aid allocation on corruption in recipient countries. The
results show that while EU aid allocation until 2006 reduced most types of corruption, the
magnitudes of these effects were quite small. It was also revealed that the effects were different
depending on the measurements of corruption. The reduction effects on corruption risk (ICRG)
as well as political (aggregated indicator), executive, and public sector corruption were slightly
bigger than judicial corruption. It was also reported that EU aid allocation had potential to fuel
legislature corruption during the analyzed period. In addition, this chapter tested year-by-year
effects of EU aid allocation. Whereas there were slight effects on ICRG scores immediately,
there were small effects on other types of corruption over a longer period. Similar to the results
of four-year average effects, the effects of political (aggregated indicator), executive, and public
sector corruption were slightly bigger than judicial corruption and there was a slight opposite
effect on legislature corruption. Furthermore, I confirmed the quantitative evidence by
analyzing qualitatively the EU sanctions against violations of human rights, democracy, the
CEU eTD Collection
36
rule of law, and good governance, as well as the rare case of DRC, to which the 2001 Belgian
presidency had a concrete influence on aid allocation.
CEU eTD Collection
37
Chapter 5: Conclusion
The effects of foreign aid on the level of corruption in recipient countries were one of the
popular topics among anti-corruption and political economy scholars. While some researchers
find foreign aid can reduce corruption through mechanisms such as imposing aid conditionality,
other researchers find foreign aid can fuel the level of corruption. Considering the endogeneity
issue on the relationship between the two, they tend to analyze the effects through instrumental
variable approaches or Generalized Method of Moments.
However, such works often poorly justify their theoretical and methodological
foundation. Most of the studies do not provide possible justifications of their analyses and
findings from theories of corruption principal-agent theory and collective action theory. These
studies also often fail to justify their use of methodology with methodological literature. This
is particularly the case in the use of instrumental variables (IV), which just replicate the previous
usages of IVs. Additionally, previous studies examine the effects with proxy measurements of
corruption, such as International Country Risk Guide (ICRG), or aggregated indicator which
measure the overall level of corruption, such as World Governance Indicators (WGI) and
Corruption Perception Index (CPI).
Reflecting the innovative study by Carnegie and Marinov (2017) and recent
developments of new corruption indicators, the present study applied their approach to the study
of corruption with multiple new corruption indicators. In other words, this thesis addressed the
issues of theoretical justification, measurement, and methodologies, using EU council
presidency rotation as an as-if randomization and incorporating multiple corruption indicators,
including ones measuring particular aspects of corruption (such as executive and public sector
corruption) with an extensive theoretical review of aid and corruption. Not only analyzing
average effects of EU aid allocation, this thesis also examined the year-by-year effects of EU
aid allocation on each aspect of corruption.
CEU eTD Collection
38
The presented analyses until 2006 showed that the size of effects of EU aid allocation
on corruption risk as well as on political, executive, public sector, and judicial corruption were
quite small. The comparison among each v-dem corruption indicator (executive, public sector,
and judicial corruption) also revealed the former two types of corruption were more likely to
be affected by EU aid allocation than judicial corruption. Regarding legislative corruption, the
result even shows EU aid allocation can slightly fuel corruption. Additionally, the present study
also tested year-by-year effects of EU aid. While the effects on corruption risk diminished over
the period, the effects on political, executive, and public sector corruption strengthened slightly
in the later period.
In addition to these quantitative analyses, this thesis also conducted qualitative analysis
and found the EU did not impose any sanctions against corruption in former European colonies
that received EU aid until 2006. EU rarely imposed sanctions against corruption, whereas it
actively imposed sanctions against other violations (i.e., democracy, human rights, and the rule
of law) over the period. This was revealed not only from general records of EU sanctions but
also from a case study of the effects of aid on corruption in DRC after the 2001 Belgian
presidency a rare typical case which gave a concrete influence over EU aid policies using the
Council presidency position. Despite the Cotonou Agreement, the present study found reducing
corruption has not been paid enough attention as one of the conditions of aid distribution by the
EU.
The findings, overall, suggest EU aid conditionality does not have substantial effects on
reducing corruption in recipient countries in comparison with effects on other development
issues such as democracy promotion and human rights improvement due to little interest to
impose sanctions against corruption. Although the present study limits the scope and period of
the analysis, the findings will also be applicable to EU aid allocation in recent years, as the EU
still rarely imposes sanctions against corruption in aid recipient countries (Crawford and
CEU eTD Collection
39
Kacarska, 2019). Anti-corruption scholars and practitioners need to work further on how to
ensure reduction of corruption in recipient countries through aid conditionality (i.e., imposing
sanctions). Otherwise, aid allocation would not be convincing for citizens who oppose foreign
aid allocation and would not effectively help combat corruption in recipient countries.
Furthermore, this thesis suggests scholars of anti-corruption need to re-consider the
effects of foreign aid based on different corruption indicators. The results of IV models in the
thesis showed different effects of EU aid on corruption depending on measurements of
corruption corruption risk (ICRG), political corruption (v-dem), and each component score
(executive, public sector, judicial, and legislative corruption; v-dem). These findings imply it
is important to examine effects of aid on specific aspects of corruption in addition to general
level of corruption using different multiple corruption indicators.
The present study limits the scope of analysis in terms of time (until 2006; due to
methodological reasons) and places (former European colonies), which limits the
generalizability of the research. However, the findings should be considered in future research.
First, the findings do not indicate any effects of other multilateral and bilateral aid on corruption.
Since the aid allocation strategies differ from each donor institution, such effects should be
examined with rich theoretical and methodological foundations, as well as a variety of
measurements. Second, the thesis neither includes any covariates in the models nor directly
covers effects of EU aid allocation on corruption after 2007. Future research should include
covariates in the models with theoretical justification for variable selection in order to estimate
the models more precisely and should examine whether the recent policy revisions
22
by EU
policymakers have had any changes on the effects.
22
For example, from 2008, the “governance incentive tranche” instrument was introduced to aid programs
to Africa, the Caribbean and the Pacific Countries (Mungiu-Pippidi 2019, p.75). This policy is different from
traditional negative aid conditionality in that “governance incentive tranche” is categorized as an example
policy which is actually based on the idea of positive aid conditionality.
CEU eTD Collection
40
Third, further researches are required to find the reasons why the EU is not keen to
impose aid conditionality against corruption in recipient countries. While collective action
theories in corruption studies suggest the lack of interest is caused by collective action dilemma,
I could find no convincing data to prove the point. This should be analyzed with more primary
sources (e.g., interviews). Finally, future studies should extend the present study by addressing
the heterogeneity effects of the aid allocation on the level of corruption. Qualitative and
quantitative comparisons between countries with low and high institutional capacity for anti-
corruption (from the perspectives of monopoly, discretion, and accountability; Klitgaard, 1988)
deserve attention as a further explanation of possible differences of the effects of aid allocation.
CEU eTD Collection
40
Appendix
Appendix A. List of States
1. Former Spanish Colonies
Argentina
Barbados
Bolivia
Chile
Colombia
Costa Rica
Cuba
Dominican Republic
Ecuador
El Salvador
Equatorial Guinea
Guatemala
Guyana
Haiti
Jamaica
Mexico
Honduras
Nicaragua
Panama
Paraguay
Peru
Suriname
Trinidad and Tobago
Uruguay
Venezuela
2. Former Portuguese Colonies
Angola
Brazil
Cape Verde
Guinea-Bissau
Mozambique
Sao Tome and Principe
3.Former British Colonies
Bahamas
Bahrain
Bhutan
Botswana
Brunei
Burma/Myanmar
Cyprus
Egypt
Eritrea
Eswatini
Fiji
Ghana
India
Iraq
Jordan
Kenya
Lesotho
Malawi
CEU eTD Collection
41
Malaysia
Maldives
Mauritius
Namibia
Nigeria
Oman
Pakistan
Qatar
Seychelles
Sierra Leone
Singapore
Solomon Islands
South Africa
Sri Lanka
Sudan
Tanzania
The Gambia
Uganda
Vanuatu
Yemen
Zambia
Zimbabwe
4. Former French Colonies
Algeria
Benin
Burkina Faso
Cambodia
Cameroon
Central African Republic
Chad
Comoros
Côte d'Ivoire
Djibouti
Gabon
Guinea
Laos
Lebanon
Madagascar
Mali
Mauritania
Morocco
Niger
Republic of the Congo
Senegal
Syria
Togo
Tunisia
Vietnam
5.Former Belgian Colonies
Burundi, Democratic Republic of the
Congo, Rwanda
6. Former Italian Colonies
Libya, Somalia
7. Former Dutch Colonies
Indonesia
CEU eTD Collection
42
Appendix B: List of Variables
1. Instrumental Variable
l2CPcol2: Former colony status (twice lagged in the second half of the year)
(Source: Replication data of Carnegie and Marinov, 2017)
2. Independent Variable
EV: aid allocation (EU net ODA)
(Source: Replication data of Carnegie and Marinov, 2017)
3. Dependent Variables (Rescaled 0-10 scale; 10 means high corruption)
icrgavg: ICRG corruption score (averaged over four years)
(Source: International Country Risk Guide - The Corruption Index)
v2x_corravg_rs10: Political corruption score (averaged over four years)
v2x_execorravg_rs10: Executive corruption score (averaged over four years)
v2x_pubcorravg_rs10: Public sector corruption score (averaged over four years)
v2jucorrdc_rs10: Judicial corruption decision (averaged over four years)
v2lgccpt_rs10: Legislature corrupt activities (averaged over four years)
(Source: V-Dem Institute, 2019)
CEU eTD Collection
43
Appendix C: Distributions of ICRG and V-dem (Political Corruption) score
Figure 4: Distributions of ICRG and V-dem (Political Corruption) score
Appendix D: Replication Data
Replication data (data files, codes for merge and analysis) is stored in Harvard Database
https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/9L2LLQ.
All data is available upon request.
CEU eTD Collection
44
Bibliography
Angrist, Joshua D., and Jörn-Steffen Pischke. 2014. Mastering’metrics: The Path from Cause
to Effect. Princeton: Princeton University Press.
Arnould, Valerie, and Koen Vlassenroot. 2016. “EU Policies in the Democratic Republic of
Congo: Try and Fail?” Security in Transition SiT/WP/06/16. Accessed June 18, 2020.
https://biblio.ugent.be/publication/7155430/file/7155436
Asongu, Simplice. 2012. “On the Effect of Foreign Aid on Corruption.” Economics Bulletin
32 (3): 21742180.
Asongu, Simplice A., and Jacinta C. Nwachukwu. 2016. “Foreign Aid and Governance in
Africa.” International Review of Applied Economics 30 (1): 6988.
Bobineau, Julien, and Philipp Gieg. 2016. The Democratic Republic of the Congo. La
République Démocratique Du Congo: Problems, Progress and Prospects. Problèmes,
Progrès et Perspectives. Münster: LIT Verlag Münster.
Bourque, André, and Peter Sampson. 2001. The European Union’s Political and Development
Response to the Democratic Republic of the Congo. Maastricht: European Centre for
Development Policy Management.
Bullough, Oliver. 2018. Moneyland: Why Thieves And Crooks Now Rule The World And How
To Take It Back. London: Profile Books.
Bunse, Simone. 2009. “The 2001 Belgian Presidency: Internal Market and Foreign Policy
Priorities and Achievements.” In Small States and EU Governance, 114158. Berlin:
Springer.
Butt, Simon, and Sofie Arjon Schütte. 2014. “Assessing Judicial Performance in Indonesia:
The Court for Corruption Crimes.” Crime, Law and Social Change 62 (5): 603619.
Carnegie, Allison, and Nikolay Marinov. 2017. “Foreign Aid, Human Rights, and Democracy
Promotion: Evidence from a Natural Experiment.” American Journal of Political
Science 61 (3): 671683.
Charron, Nicholas. 2011. “Exploring the Impact of Foreign Aid on Corruption: Has the ‘Anti-
Corruption Movement’ Been Effective?” The Developing Economies 49 (1): 6688.
Crawford, Gordon, and Simonida Kacarska. 2019. “Aid Sanctions and Political
Conditionality: Continuity and Change.” Journal of International Relations and
Development 22 (1): 184214.
David-Barrett, Elizabeth, Aoife Murray, Johanna Polvi, and Richard Burge. 2020.
“Evaluating Anti-Corruption Agencies: Learning from the Caribbean.” Journal of
Development Effectiveness 12 (1): 74-88.
Davis, Laura. 2015. “The EU and Peacebuilding in the Democratic Republic of Congo.” Civil
Society Dialogue Network Discussion Paper 9. Accessed May 23, 2020.
http://eplo.org/wp-content/uploads/2017/03/EPLO_CSDN_Discussion_Paper_The-
EU-and-peacebuilding-in-the-DRC.pdf.
European Commission. 2001a. “2372nd Council Meeting - GENERAL AFFAIRS.” Text.
European Commission - European Commission. Accessed May 25, 2020.
https://ec.europa.eu/commission/presscorner/detail/en/PRES_01_337.
———. 2001b. “2386th Council Meeting - GENERAL AFFAIRS -.” Accessed May 25,
2020.
https://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/gena/DOC.6862
0.pdf?fbclid=IwAR0ykdYwl7d2bJzHQM9cKCExNXxPOu-14CL8Ix8_DSO3UCI-
LLXGwqwkdJ8.
CEU eTD Collection
45
———. 2001c. “2397th Council Meeting - GENERAL AFFAIRS -.” Accessed May 25,
2020. http://www.sussex.ac.uk/Units/spru/hsp/documents/2001-
1210%20GAERC%20Concs.pdf.
———. 2002. “The European Development Fund in a Few Words.” Accessed May 25, 2020.
http://aei.pitt.edu/40696/1/DE_112.pdf.
European Court of Auditors. 2013. “EU Support for Governance in the Democratic Republic
of the Congo: (Pursuant to Article 287(4), Second Subparagraph, TFEU).” Accessed
May 26, 2020.
Https://Www.Eca.Europa.Eu/Lists/ECADocuments/SR13_09/SR13_09_EN.Pdf.
2013.
European Union. 2000. “Cotonou Agreement.” Accessed May 28, 2020.
https://www.ohchr.org/EN/Issues/RuleOfLaw/CompilationDemocracy/Pages/Cotonou
Agreement.aspx.
Fazekas, Mihály, and Lawrence Peter King. 2019. “Perils of Development Funding? The Tale
of EU Funds and Grand Corruption in Central and Eastern Europe.” Regulation &
Governance 13 (3): 405430.
Gerring, John, and Strom C. Thacker. 2004. “Political Institutions and Corruption: The Role
of Unitarism and Parliamentarism.” British Journal of Political Science 34 (2): 295
330.
Hellmann, Olli. 2017. “The Historical Origins of Corruption in the Developing World: A
Comparative Analysis of East Asia.” Crime, Law and Social Change 68 (12): 145
165.
Hoebeke, Hans, Stéphanie Carette, and Koen Vlassenroot. 2007. “EU SUPPORT TO THE
DEMOCRATIC REPUBLIC OF CONGO.” Paris: Centre d’analyse Stratégique.
2007. Accessed May 25, 2020.
http://www.egmontinstitute.be/content/uploads/2013/10/EU_support_to_the_DRC1.p
df?type=pdf.
Hopkin, Jonathan. 2002. “States, Markets and Corruption: A Review of Some Recent
Literature.” Review of International Political Economy 9 (3): 574590.
Huntington, Samuel P. 1969. Political Order in Changing Societies.(Second Printing.). New
Haven: Yale University Press.
IMF. 2005. “The IMF’s Approach to Promoting Good Governance and Combating
Corruption -- A Guide.” 2005. Accessed April 24, 2020.
https://www.imf.org/external/np/gov/guide/eng/index.htm.
Janse, Ronald. 2019. “Is the European Commission a Credible Guardian of the Values? A
Revisionist Account of the Copenhagen Political Criteria during the Big Bang
Enlargement.” International Journal of Constitutional Law 17 (1): 4365.
Kangoye, Thierry. 2013. “Does Aid Unpredictability Weaken Governance? Evidence from
Developing Countries.” The Developing Economies 51 (2): 121144.
Kaufmann, Daniel. 1997. “Corruption: The Facts.” Foreign Policy 107, 114131.
Kelly, Catherine. 2007. “Belgian Intervention Policy in the DRC: Causes and Consequences
of the Reorientation, 1999-2006.” Les Cahiers Du RMES 4 (2): 20072008.
Klitgaard, Robert. 1988. Controlling Corruption. Berkeley: Univ of California Press.
Knack, Stephen. 1999. "Aid Dependence and the Quality of Governance: A Cross-Country
Empirical Analysis". The World Bank WPS 2396: 1-44.
Kunicova, Jana, and Susan Rose-Ackerman. 2005. “Electoral Rules and Constitutional
Structures as Constraints on Corruption.” British Journal of Political Science 35 (4):
573606.
CEU eTD Collection
46
Lambsdorff, Johann Graf, Markus Taube, and Matthias Schramm. 2004. The New
Institutional Economics of Corruption. Vol. 64. London: Routledge.
Leff, Nathaniel H. 1964. “Economic Development Through Bureaucratic Corruption.”
American Behavioral Scientist 8 (3): 8-14.
Libman, Alexander, and Anastassia Obydenkova. 2013. “Communism or Communists?
Soviet Legacies and Corruption in Transition Economies.” Economics Letters 119 (1):
101103.
Lovitt, Jeff. 2004. “The Information Challenge: Transparency International and Combating
Corruption.” Accessed April 25, 2020. https://ec.europa.eu/anti-
fraud/sites/antifraud/files/docs/body/lovitt_en.pdf.
March, James G., and Johan P. Olsen. 2004. “The Logic of Appropriateness.” In The Oxford
Handbook of Political Science, edited by Robert E. Goodin. Oxford: Oxford
University Press.
Marquette, Heather, and Caryn Peiffer. 2015. “Corruption and Collective Action.”
Developmental Leadership Program Working Paper 32.
———. 2018. “Grappling with the ‘Real Politics’ of Systemic Corruption: Theoretical
Debates versus ‘Real-World’ Functions.” Governance 31 (3): 499514.
https://doi.org/10.1111/gove.12311.
———. 2019. “Thinking Politically about Corruption as Problem-Solving: A Reply to
Persson, Rothstein, and Teorell.” Governance 32 (4): 81120.
https://doi.org/10.1111/gove.12444.
McGovern, Feancis E. 1908. “Legal Repression of Political Corruption.” In Proceedings of
the American Political Science Association, 4:266276. Cambridge University Press.
Mendelski, Martin. 2012. “EU-Driven Judicial Reforms in Romania: A Success Story?” East
European Politics 28 (1): 2342.
Mohamed, Masoud Rashid, Shivee Ranjanee Kaliappan, Normaz Wana Ismail, and W. N. W.
Azman-Saini. 2015. “Effect of Foreign Aid on Corruption: Evidence from Sub-
Saharan African Countries.” International Journal of Social Economics.
Mungiu-Pippidi, Alina. 2019. Europe’s Burden: Promoting Good Governance Across
Borders. Cambridge: Cambridge University Press.
Nasra, Skander. 2011. “Governance in EU Foreign Policy: Exploring Small State Influence.”
Journal of European Public Policy 18 (2): 164180.
Öhler, Hannes, Peter Nunnenkamp, and Axel Dreher. 2012. “Does Conditionality Work? A
Test for an Innovative US Aid Scheme.” European Economic Review 56 (1): 138153.
Okada, Keisuke, and Sovannroeun Samreth. 2012. “The Effect of Foreign Aid on Corruption:
A Quantile Regression Approach.” Economics Letters 115 (2): 240243.
Onyango, Gedion. 2017. “Collectivism and Reporting of Organizational Wrongdoing in
Public Organizations: The Case of County Administration in Kenya.” International
Review of Sociology 27 (2): 35372. https://doi.org/10.1080/03906701.2017.1298429.
Ostrom, Elinor. 1998. “A Behavioral Approach to the Rational Choice Theory of Collective
Action: Presidential Address, American Political Science Association, 1997.”
American Political Science Review 92 (1): 122.
Pagan, Adrian, and Aman Ullah. 1988. “The Econometric Analysis of Models with Risk
Terms.” Journal of Applied Econometrics 3 (2): 87105.
Persson, Anna, Bo Rothstein, and Jan Teorell. 2013. “Why Anticorruption Reforms Fail—
Systemic Corruption as a Collective Action Problem.” Governance 26 (3): 449471.
———. 2019. “Getting the Basic Nature of Systemic Corruption Right: A Reply to Marquette
and Peiffer.” Governance 32 (4): 799810. https://doi.org/10.1111/gove.12403.
CEU eTD Collection
47
Peters, B. Guy. 2010. “Institutional Design and Good Governance.” In The Good Cause,
edited by Gjalt de Graaf, Patrick von Maravić, and Pieter Wagenaar, 1st ed., 8397.
Theoretical Perspectives on Corruption. Leverkusen: Verlag Barbara Budrich.
Pratt, JW, and RJ Zeckhause. 1985. “Principals and Agents: An Overview.” In Principals and
Agents: The Structure of Business. MA: Harvard Business School Press.
Rohlfing, Ingo. 2012. Case Studies and Causal Inference: An Integrative Framework.
London: Palgrave Macmillan.
Rose-Ackerman, Susan. 1978. Corruption: A Study in Political Economy. Nova Iorque:
Academic Press.
———. 1999. Corruption and Government: Causes, Consequences, and Reform. Cambridge:
Cambridge University Press.
Rose-Ackerman, Susan, and Bonnie J. Palifka. 2016. Corruption and Government: Causes,
Consequences, and Reform. Cambridge: Cambridge university press.
Rosenbaum, Paul R. 1984. “The Consequences of Adjustment for a Concomitant Variable
That Has Been Affected by the Treatment.” Journal of the Royal Statistical Society:
Series A (General) 147 (5): 656666.
Rubbers, Benjamin, and Emilie Gallez. 2012. “Why Do Congolese People Go to Court? A
Qualitative Study of Litigants’ Experiences in Two Justice of the Peace Courts in
Lubumbashi.” The Journal of Legal Pluralism and Unofficial Law 44 (66): 79108.
Schalk, Jelmer, René Torenvlied, Jeroen Weesie, and Frans Stokman. 2007. “The Power of
the Presidency in EU Council Decision-Making.” European Union Politics 8 (2): 229
250.
Seawright, Jason. 2016. Multi-Method Social Science: Combining Qualitative and
Quantitative Tools. Cambridge: Cambridge University Press.
Sekhon, Jasjeet S., and Rocio Titiunik. 2012. “When Natural Experiments Are Neither
Natural nor Experiments.” American Political Science Review 106 (1): 3557.
Szarek-Mason, Patrycja. 2010. The European Union’s Fight Against Corruption: The
Evolving Policy Towards Member States and Candidate Countries. Cambridge:
Cambridge University Press. https://doi.org/10.1017/CBO9780511676086.
Tallberg, Jonas. 2004. “The Agenda-Shaping Powers of the Council Presidency.” In
European Union Council Presidencies, 2746. London: Routledge.
The PRS Group. n.d. “ICRG Methodology.” Accessed May 22, 2020.
https://www.prsgroup.com/wp-content/uploads/2018/01/icrgmethodology.pdf.
Thomson, Robert. 2008. “The Council Presidency in the European Union: Responsibility with
Power.” JCMS: Journal of Common Market Studies 46 (3): 593617.
Treaty of Lisbon. 2007. Official Journal of the European Union, 50. Accessed June 19, 2020.
https://eur-lex.europa.eu/legal-
content/EN/TXT/HTML/?uri=OJ:C:2007:306:FULL&from=EN
achudova, Milada Anna. 2009. “Corruption and Compliance in the EU’s Post-Communist
Members and Candidates.” J. Common Mkt. Stud. 47 (9): 43-62.
V-Dem Institute. 2019. “V-Dem Codebook (V9).” Accessed February 25, 2020..
https://www.v-dem.net/media/filer_public/e6/d2/e6d27595-9d69-4312-b09f-
63d2a0a65df2/v-dem_codebook_v9.pdf.
Vaznonytė, Auste. 2020. The rotating Presidency of the Council of the EUStill an agenda-
setter?. European Union Politics, 1-22.
Warntjen, Andreas. 2007. “Steering the Union. The Impact of the EU Presidency on
Legislative Activity in the Council.” JCMS: Journal of Common Market Studies 45
(5): 11351157.
CEU eTD Collection
48
———. 2008. “The Council Presidency: Power Broker or Burden? An Empirical Analysis.”
European Union Politics 9 (3): 315338.
Wooldridge, Jeffrey M. 2016. Introductory Econometrics: A Modern Approach. Tronto:
Nelson Education.
World Bank. 1997. Helping Countries Combat Corruption: The Role of the World Bank.
Washington: World Bank
———. n.d. “James David Wolfensohn.” Text/HTML. World Bank. Accessed April 24,
2020. https://www.worldbank.org/en/about/archives/history/past-presidents/james-
david-wolfensohn.
Zimelis, Andris. 2011. “Conditionality and the EU–ACP Partnership: A Misguided Approach
to Development?” Australian Journal of Political Science 46 (3): 389406.
CEU eTD Collection
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Efforts to evaluate the impact of anti-corruption programmes face numerous difficulties related to the complexity and hidden nature of corruption, the political sensitivity of the topic, and the ability of corrupt networks to respond flexibly to interventions. This makes it difficult to measure changes in corruption levels and problematic to attribute them to interventions. This article shows how a theory-based evaluation that builds on academic research can elaborate a set of intermediate outcomes for the evaluation of anti-corruption programmes, and showcases several purpose-built tools for evaluating different aspects of capacity building. It also draws on learning from two evaluations of Anti-Corruption programmes in the Caribbean to demonstrate how anti-corruption theory is being translated into law enforcement practice in two ways: (i) through economic models of criminals as rational actors whose behaviour can be changed through incentives and disincentives; (ii) and through social norms models which argue that reducing corruption requires deeper social change that resonates with or adapts local norms.
Article
Full-text available
Given the unprecedented scale of intergovernmental development funding and the importance of institutional quality for human wellbeing, it is imperative to precisely understand the impact of development funds on corruption. In Europe, EU Funds provide a boost to public spending in recipient member states while introducing additional corruption controls. We investigate whether EU Funds increase high-level corruption in the Czech Republic and Hungary in 2009-2012. We analyse newly collected data from over 100,000 public procurement contracts to develop objective corruption risk indicators and link them to agency-level data of the public sector. Propensity score matching estimations suggest that EU funds increase corruption risks by up to 34%. The negative effects are largely attributable to overly formalistic compliance and EU Funds overriding domestic accountability mechanisms in public organisations entirely dependent on external funds. The policy implications are profound: governments should reduce barriers to market entry by lowering red tape and they should prevent excessive concentration of funds.
Article
Full-text available
A growing body of research argues that anticorruption efforts fail because of a flawed theoretical foundation, where collective action theory is said to be a better lens for understanding corruption than the dominant principal–agent theory. We unpack this critique and advance several new arguments. First, the application of collective action theory to the issue of corruption has been, thus far, incomplete. Second, a collective action theory-based approach to corruption is in fact complementary to a principal–agent approach, rather than contradictory as is claimed. Third, applications of both theories have failed to recognize that corruption persists because it functions to provide solutions to problems. We conclude by arguing that anticorruption effectiveness is difficult to achieve because it requires insights from all three perspectives—principal–agent theory, collective action theory, and corruption as serving functions—which allows us to better understand how to harness the political will needed to fight corruption.
Article
Full-text available
This paper explores dimensions of collectivism to explain patterns of reporting organizational wrongdoing in public organizations. The findings herein illustrate that responses to wrongdoing in public organizations in developing countries like Kenya are more rationalized or interpreted within communitarian or supported by parochial social ideologies, which tend to override instrumental accountability norms and structures. Accordingly, responses towards organizational wrongdoing are more informed by the logic of appropriateness, as potential and actual complainants prefer informal channels for addressing organizational wrongdoing over formal reporting mechanisms. These findings present important insights for designing accountability mechanisms or anti-corruption strategies in public organizations. With a focus on unfamiliar settings to the literature on ethical culture, the paper rides on its in-depth analysis while contributing to the current research on organizational behavior and decision-making.
Article
Full-text available
Purpose – This paper investigates the effect of foreign aid on governance in order to extend the debates on foreign aid and to verify common positions from Moyo’s ‘Dead Aid’, Collier’s ‘Bottom Billion’ and Eubank’s ‘Somaliland’. The empirical evidence is based on updated data from 52 African countries for the period 1996-2010. Design/methodology/approach – An endogeneity robust instrumental variable Two-Stage-Least Squares empirical strategy is employed. Findings – The findings reveal that development assistance deteriorates economic (regulation quality and government effectiveness) and institutional (corruption-control and rule of law) governance, but has an insignificant effect on political (political stability, voice and accountability) governance. While, these findings are broadly in accordance with Moyo (2009) and Collier (2007) on weak governance, they neither confirm the Eubank (2012) position on political governance nor the Asongu (2012) stance on the aid-corruption nexus in his debate with Okada & Samreth (2012). Practical implications – The use of foreign aid as an instrument to influence the election and replacement of political leaders in Africa may have insignificant results. It is time to solve the second tragedy of foreign aid and that economists and policy makers start rethinking the models and theories on which foreign aid is used to influence economic, institutional and political governance in recipient countries. Originality/value – The paper extends the debate on foreign aid and institutions in Africa in the light a plethora of recent studies in the aid literature.
Article
What role does the rotating Council Presidency maintain a decade after Lisbon? This article argues that, regardless of institutional changes, the rotating Presidency still shapes the Council agenda to a large extent. Based on an original hand-coded dataset of rotating Presidency programmes between 1997 and 2017, I show that some policies are ‘stickier’ on the Council agenda, while the others exhibit significant changes in salience over time. Since the magnitude of these shifts varies from Presidency to Presidency, the analysis focuses on domestic political factors and the country positioning vis-à-vis the European Union to determine their relationship with agenda volatility. By means of a panel model, the examination demonstrates that the government issue salience can best explain the levels of issue salience in the Presidency programmes.
Book
Cambridge Core - European Government, Politics and Policy - Europe's Burden - by Alina Mungiu-Pippidi
Article
In reply to Marquette and Peiffer's article “Grappling With the ‘Real Politics’ of Systemic Corruption: Theoretical Debates Versus ‘Real‐World’ Functions,” this article employs three criticisms: Marquette and Peiffer's call to grapple with the “real politics of corruption” does not bring much new to the table, is conceptually flawed, and risks serving as an excuse for corrupt elites to pursue “business as usual.” In response, we reaffirm three insights gained from collective action‐based approaches toward corruption. Although corruption might solve individual‐level problems in the short term, it is still a de facto problem at the aggregate level, the tools derived from principal–agent theory will not solve the collective action problem of systemic corruption, and elites will be the least likely to implement reform. We conclude by calling for the continued fight against corruption—a fight informed by empirical and theoretical knowledge.
Article
Political conditionality was first introduced by Western governments into their development aid policy a quarter of a century ago, threatening to invoke aid sanctions in the event of human rights abuses or democratic regression in aid recipient countries. This paper examines how political conditionality has evolved in the subsequent years and analyses what has changed and why. It does so through a review of sanctions cases in the EU and the US aid from 2000 to date, with discussion located within the post-2000 international environment in which foreign policy and aid policy are situated. The paper focuses on three regions: sub-Saharan Africa, the Middle East and North Africa, and Central Asia. Patterns of change and continuity are identified in relation to how political conditionality has been implemented. Our findings are that political conditionality remains a significant policy tool, contrary to the perception that its use has declined. However, while selectivity and inconsistency in policy application continue, security interests have become a more prominent explanatory factor in the post-2000 period. Indeed, the initial normative agenda of political conditionality as a tool for the promotion of democracy and human rights, as stated in policy rhetoric, has been replaced by its use as an instrument to promote Western security interests in line with the securitisation of development.