CAPITAL FLIGHT AND INSTITUTIONAL GOVERNANCE IN SUB-
SAHARAN AFRICA: THE ROLE OF CORRUPTION

A THESIS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON

BY
KINGSLEY OSEI DOMFEH
(10441782)

IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE
AWARD OF THE MASTER OF PHILOSOPHY (MPHIL) DEGREE IN
ECONOMICS

JULY, 2015

University of Ghana http://ugspace.ug.edu.gh

DECLARATION

This is to certify that this thesis is the result of research undertaken by Kingsley Osei Domfeh
towards the award of a Master of Philosophy (M.Phil.) degree in Economics at the Department of
Economics, University of Ghana.

……………………………………………………
KINGSLEY OSEI DOMFEH
(10441782)

……………………………………….. ………………………………………
DR. MICHAEL DANQUAH DR. ERIC OSEI-ASSIBEY
(SUPERVISOR) (CO-SUPERVISOR)

………………………………………. ………………………………………
DATE DATE
University of Ghana http://ugspace.ug.edu.gh

iii

ABSTRACT

Empirical evidence indicates that macroeconomic uncertainty, political and institutional
instability, less developed financial system, and higher rate of return differentials in other countries
abroad induces capital flight from Sub-Saharan Africa. This research recognizes corruption as an
aspect of a weak political and institutional system. However, the relationship between corruption
and capital flight has received little emphasis, particularly in Sub-Saharan Africa. The study,
therefore, seeks to examine capital flight and institutional governance in Sub-Saharan Africa, the
role of corruption. Panel data set of thirty two (32) countries in Sub-Saharan Africa is analyzed
over the period 2000-2012 employing three different estimation techniques as Generalized Method
of Moments (GMM), Fixed Effect Regression and the pooled-OLS regression models. The
research work is based on the portfolio choice framework.

The main variable of interest (corruption) entered all eight (8) specifications of the econometric
model tested. The result of the empirical estimation established that corruption has a positive and
statistically significant effect on capital flight in SSA in all the specifications. Moreover, the
interaction between corruption and regime durability resulted in a negative and statistically
significant coefficient implying that an increase in regime durability, which also proxy for
institutional strength would reduce corruption, hence capital flight. In other words, in the midst
of strong institutional governance, the potency of corruption in increasing capital flight is reduced
significantly.

University of Ghana http://ugspace.ug.edu.gh

Moreover, the study established that all the controlled macroeconomic variables does not have any
significant effect on capital flight in Sub-Saharan Africa. These variables were found to be
statistically insignificant to capital flight in all the specifications they appeared.

On the basis of the empirical results, the study recommends measures to strengthen institutional
governance in SSAs. Institutional governance reforms are encouraged to be undertaken in order to
reduce corruption and by extension capital flight. The governments of Sub-Saharan African
countries must put in place measures to improve the control of corruption in SSA. Thus, an
effective mechanism aimed at tracking and prosecuting financial crime should be the utmost
priority of the authorities.

University of Ghana http://ugspace.ug.edu.gh

DEDICATION

This thesis is dedicated to the Almighty God, my family, especially Nana Abena Ohenewaa Osei-
Domfeh and my friends.

University of Ghana http://ugspace.ug.edu.gh

ACKNOWLEDGEMENTS

My utmost thanks go to God Almighty for seeing me through my university education and the
successful completion of this thesis work.

I am particularly indebted to my supervisors namely, Dr. Michael Danquah and Dr. Eric Osei-
Assibey, lecturers of the Department of Economics, University of Ghana, who painstakingly
nurtured and taught me how to plumb the depths of this thesis work.

Mr. Derrick Taylor, Mr. Alex Bamfo, Mr. Prince Baah and Miss Deborah Amartey, God richly
reward you for your support and ever sounding advices.

I am also thankful to my family, especially my wife (Iris Naa Osei-Domfeh) for her moral and
prayer support.

University of Ghana http://ugspace.ug.edu.gh

vii

TABLE OF CONTENTS

CONTENTS PAGES

DECLARATION............................................................................................................................. ii
ABSTRACT ................................................................................................................................... iii
DEDICATION ................................................................................................................................ v
ACKNOWLEDGEMENTS ........................................................................................................... vi
TABLE OF CONTENTS .............................................................................................................. vii
LIST OF FIGURES ........................................................................................................................ x
LIST OF TABLES ......................................................................................................................... xi
LIST OF ABBREVIATIONS ....................................................................................................... xii

CHAPTER ONE
INTRODUCTION

1.1 Background of the study ........................................................................................................... 1
1.2 Statement of the Problem .......................................................................................................... 5
1.3 Research Questions ................................................................................................................... 7
1.4 Objectives ................................................................................................................................. 8
1.5 Significance of the study ........................................................................................................... 8
1.6 Organization of the study .......................................................................................................... 9

CHAPTER TWO
LITERATURE REVIEW

2.1 Introduction ............................................................................................................................. 11
2.2 Review of definitions and measurements of Capital Flight .................................................... 11
2.3 Review of Corruption in Sub-Saharan Africa ......................................................................... 16
2.3.1 Introduction ...................................................................................................................... 16
2.3.2 Definition and Causes of Corruption in Sub-Saharan Africa ........................................... 16
University of Ghana http://ugspace.ug.edu.gh

viii

2.3.3 Measures of corruption .................................................................................................... 20
2.3.4 Effects of corruption in Sub-Saharan Africa .................................................................... 22
2.4 Theoretical literature on capital flight and corruption ............................................................ 25
2.5 Empirical Literature Review ................................................................................................... 29
2.5.1 Determinants of capital flight ........................................................................................... 29

CHAPTER THREE
OVERVIEW OF CAPITAL FLIGHT AND CORRUPTION IN SUB-SAHARAN AFRICA

3.1 Introduction ............................................................................................................................. 40
3.2 Trend analysis of capital flight in Sub-Saharan Africa ........................................................... 40
3.3 Trend analysis of average capital flight for Sub-Saharan African countries .......................... 41
3.4 Average capital flight to GDP in SSA (2000-2012) ............................................................... 43
3.5 Trend analysis of Corruption in SSA (2000-2012) ................................................................. 46
3.6 Average Corruption Estimates for Individual SSA country (2000-2012) .............................. 47

CHAPTER FOUR
METHODOLOGY

4.1 Introduction ............................................................................................................................. 49
4.2 Theoretical framework for estimating capital flight ............................................................... 49
4.2.1 Adjustment for Exchange Rate Fluctuations.................................................................... 50
4.2.2 Estimating Trade Misinvoicing ........................................................................................ 51
4.2.3 Adjustment of Underreporting of Remittances ................................................................ 52
4.2.4 Inflation Adjustment ........................................................................................................ 53
4.3 Model for Empirical Estimation ............................................................................................. 53
4.4 A priori Expectation ................................................................................................................ 56
4.5 The Estimation Technique ...................................................................................................... 59
4.6 Description of variables and data sources ............................................................................... 62
4.6.1 The Dependent variable: Capital flight ............................................................................ 62
4.6.2 Explanatory Variables ...................................................................................................... 63
University of Ghana http://ugspace.ug.edu.gh

CHAPTER FIVE
EMPIRICAL RESULTS AND DISCUSSION

5.1 Introduction ............................................................................................................................. 67
5.2 Descriptive Statistics and Analysis ......................................................................................... 67
5.3 Results of Unit root test for stationarity .................................................................................. 69
5.4 Results of Granger Causality Test between capital flight and corruption .............................. 70
5.5 Empirical estimation and discussions ..................................................................................... 72
5.6 Pooled – OLS Estimation Results ........................................................................................... 72
5.7 Fixed Effects Estimation Results ............................................................................................ 75
5.8 System GMM Estimation Results ........................................................................................... 77
5.9 Synthesis of the Results .......................................................................................................... 81

CHAPTER SIX
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

6.1 Introduction ............................................................................................................................. 83
6.2 Summary and Conclusion ....................................................................................................... 83
6.3 Recommendations and Policy Implications ............................................................................ 85
6.4 Limitation of the study and areas of further research ............................................................. 86
References ..................................................................................................................................... 87
Appendices .................................................................................................................................... 97
Appendix A: Real Capital Flight for 32 Sub-Saharan African Countries ( millions of 2010 US
$) ............................................................................................................................................... 97
Appendix B: Pairwise correlation coefficients of regression model ......................................... 99
Appendix C: Hausman Test .................................................................................................... 100
Appendix D: Variables used in the computation of Capital Flight and data sources ............. 101
Appendix E: Variables used in the estimation and data sources ............................................. 102

University of Ghana http://ugspace.ug.edu.gh

LIST OF FIGURES

Figure 3.1 Sub–Saharan Africa: Trends in Capital Flight (2000-2012)…………………………41
Figure 3.2 Average Capital Flight estimates for Sub-Saharan African Countries (2000-2012)…42
Figure 3.3 Average Capital Flight estimates to GDP ratio (2000-2012)………………………...44
Figure 3.4 Trend of Corruption in Sub-Saharan Africa………………………………………….46
Figure 3.5 Average Corruption estimates for Sub-Saharan African Countries………………….47

University of Ghana http://ugspace.ug.edu.gh

LIST OF TABLES

Table 4.1 Independent variable and expected sign………………………………………………56
Table 5.1: Summary Statistics of Panel data of Sub-Saharan Africa……………………………68
Table 5.2: Augmented Dickey-Fuller tests (System GMM)…………………………………….70
Table 5.3 Pairwise Granger Causality Tests on Capital flight and Corruption………………….71
Table 5.4 Pooled – OLS Regression Results (2000-2012)………………………………………73
Table 5.5 Fixed Effects – GLS Regression Results (2000-2012)………………………………..76
Table 5.6 System GMM dynamic panel estimation result (2000-2012)…………………………78

University of Ghana http://ugspace.ug.edu.gh

xii

LIST OF ABBREVIATIONS

ACF Adjusted Capital Flight

AfDB African Development Bank

AR Autoregressive

CAD Current Account Deficit

COR Corruption

CPI Corruption Perception Index

ED External Debt

FE Fixed Effects

FZ French Zone

GDP Gross Domestic Product

GLS Generalized Least Squares

GMM Generalized Method of Moments

IACF Inflation Adjusted Capital Flight

ICRG International Country Risk Guide

IDS International Debt Statistics

IEA Independence of the Executive Authority

ITBT Income from Tourism and Border Transaction

IMF International Monetary Fund

INF Inflation

University of Ghana http://ugspace.ug.edu.gh

xiii

LDCs Least Developed Countries

NFDI Net Foreign Direct Investment

OLS Ordinary Least Squares

OR Official Reserves

PPI Producer Price Index

RD Regime Durability

RE Random Effects

RII Reinvested Investment Income

RL Rule of Law

SFABS Short-term Foreign Assets of the Banking System

SSA Sub-Saharan Africa

TI Transparency International

TNCs Truly Transnational Corporations

UIC Use of IMF Credit

UNODC United Nations Office on Drugs and Crime

UNDP United Nations Development Programme

WB World Bank

WDI World Development Indicators

University of Ghana http://ugspace.ug.edu.gh

CHAPTER ONE
INTRODUCTION
1.1 Background of the study
Capital flight has been an important issue since early 1980s in developing countries. Although
from the end of the 1980s and early 1990s the debt crisis appeared to be contained and attention
to the capital flight phenomenon waned, capital flight still remains a serious problem in a number
of countries. Many developing countries are concerned with the capital flight phenomenon
because of its deleterious impact on economic growth and welfare, macroeconomic stability,
income distribution, illegal activities and other social development matters (Zheng and Tang,
2009).

Economic Stagnation has been a feature in Sub-Saharan Africa’s (SSA) economic performance in
the past four decades. Consistently, the Region has suffered from balance of payments
disequilibria, dwindling government finances, increasing macroeconomic and political instability
and, as a consequence, a higher incidence of poverty (Artadi and Sala-I-Martin, 2003; Collier,
2006). The Region tends to exhibit a significantly higher preference for foreign assets relative to
domestic assets; hence 40 percent of Africa’s private capital was held abroad in the form of capital
flight, the highest ratio of all developing regions, Collier (2001).

According to the most recent estimates, capital flight—the voluntary exits of private residents’
capital either for a safe haven or for investments made in foreign currency—from a sample of 40
Sub Saharan Africa countries amounts to 420 billion in real US dollars over the three decades
spanning 1970– 2004 (Ndikumana and Boyce, 2008). Including interest earning on past flight
University of Ghana http://ugspace.ug.edu.gh

capital, the cumulative amount reaches a staggering $607 billion. One – third of the total capital
flight from Africa is accounted for by the West African Monetary Zone. Moreover, Ndikumana
and Boyce (2010) estimated the total real capital flight from Africa in 2002 to be 18, 249.20 million
dollars of which 6,276 million is linked with the countries in the West African Monetary Zone.
Specifically Ghana, Nigeria, Sierra Leone and Guinea represent 34.39% of the total. A detailed
description can be found in Ndikumana and Boyce (2008, 2010). Again, it is estimated that
between 1970 and 2010, African countries have lost up to $1.3 trillion dollars (in 2010 constant
US dollars) through capital flight. This figure represents the sum of capital flight from 35 sub-
Saharan African countries and 4 North African countries, namely Algeria, Egypt, Morocco and
Tunisia, Ndikumana and Boyce (2012).

The high levels of capital flight pose serious challenges for domestic resource mobilization in
support of investment and growth in Africa. These challenges are even more important because
the Region is confronted with an acute shortage of capital and is increasingly marginalized in the
global distribution of foreign direct investment, which is skewed heavily in favor of OECD and
emerging market economies (World Bank, 2006). Machlup (1942) acknowledged that capital fight
leads to higher interest rates and lower prices, and possible deflationary situation in the capital
exporting countries, resulting in an automatic process of transfer repatriation. As a result capital
flight shows weak economic policies, among which are mis-management of exchange rates,
interest rates, inflation, budget deficits, increase tax burden as well as excessive public sector
borrowing requirement resulting in crowding out of the private sector. M Pastor, (1990) indicated
that capital flight reduces growth potential, erode the tax base and redistributes income from the
University of Ghana http://ugspace.ug.edu.gh

poor to the rich. Capital flight also triggers financial crisis in emerging markets in Sub-Saharan
African countries.

Theory of capital flight suggests that this phenomenon is driven both by private actors and public
authorities (Boyce and Ndikumana, 2003; Ajayi, 2007; Ndiaye, 2011). These authors showed that
capital flight is driven by private actors due to, political and institutional instability, less developed
financial system, macroeconomic uncertainty, and higher rate of return differentials abroad.

In a
context of portfolio choice, these factors lead to increased risk of losses in the real value of
domestic assets of private agents, forcing a shift of portfolio in favour of foreign assets (Collier et
al., 2004). The consequence is that, private agents hold their savings abroad, leading to a reduction
in domestic private investment and economic growth. Similarly, these resources held overseas by
public authorities’ leads to a decline in public resources, thereby causing public investment to fall
and a decline in growth.

There are many studies on capital flight in developing countries. The study by Weeks (2012) is
among the latest studies on capital flight in SSA. Most of these studies can be classified into two
main strands—determinants and associations (Cerra et al., 2008). The determinants literature
concentrates on identifying variables that are responsible for capital flight in a country or a cross-
section of countries. Primarily, this literature identifies macroeconomic policies and non-macro
variables such as political risk factors as significant determinants of capital flight. The literature
on associations spotlights the significant and often contemporaneous association between capital
flight and other perverse macroeconomic outcomes such as low rates of growth, increased aid
inflows, high external debt among others (see Hermes et al., 2002; Cerra et al., 2008).
University of Ghana http://ugspace.ug.edu.gh

Recent literature places a particular attention on the role of non-economic variables, including
institutional governance and political risk and their influence on capital flight from developing
countries (Gibson & Tsakalotos, 1993; Schineller, 1997). In the broad spectrum of the literature, it
has been noted that political instability and poor governance deters investment and induces capital
flight. Public authorities contribute to capital flight under conditions of poor governance and bad
institutional quality (Ajayi, 1992; WJ Awung, 1996; Loungani and Mauro, 2000; Ndikumana and
Boyce, 2003; Le Q.V 2006; Cerra et al., 2008; Ndiaye, 2009a and 2011). These authors reiterate
that, in such a context, corrupt public authorities take advantage of their favorable position to
amass a personal fortune abroad (Boyce and Ndikumana, 2001).

However, empirical research into the role that institutional governance plays on capital flight is
still limited in SSA particularly. Therefore, this study intends to contribute to the empirical
literature by focusing on corruption as an important institutional governance factor and its
relationship with capital flight in SSA. According to Hope (1997), corruption pandemic in Sub-
Saharan Africa, has become a matter of global concern because it has reached cancerous
proportions. In spite of this endemic, so many authors differ in their explanation of what corruption
is made up of. For instance, Osoba (1996) defined corruption as: ... “a form of antisocial behavior
by an individual or social group which confers un- just or fraudulent benefits on its perpetrators,
is inconsistent with the established legal norms and prevailing moral ethos of the land and is likely
to subvert or diminish the capacity of the legitimate authorities to provide fully for the material
and spiritual well-being of all members of society in a just and equitable manner”.

University of Ghana http://ugspace.ug.edu.gh

Corruption in Sub-Saharan Africa has been identified in the literature to include bribery in the
public sector where people need to pay some amount of money before a service is rendered. There
is also theft of public funds by account clerks and other officials as well as misappropriation of
state funds and properties. An observation by the United Nations (1990) indicated that the behavior
of corrupt leaders is to use public authority, office, or official position with the deliberate intent of
extracting personal/private monetary rewards or other privileges at the expense of public good and
in violation of established rules and ethical considerations.

1.2 Statement of the Problem
Capital flight, in economics, occurs when assets or money rapidly flow out of a country, due to an
event of economic consequence. Events such as increase in taxes on capital or the government of
the country defaulting on its debt that disturbs investors and causes them to lower their valuation
of the assets in that country, and losing the confidence they have on the strength of the economy.

The report, entitled “economic perspectives in Africa”, was released in 2012. In the report, the
African Development Bank (AfDB) asserts that capital flight deprived the continent of over US
$700 billion during the preceding decade. The United Nations Development Programme (2011)
now considers the scale of capital flight to be so great that it represents a major obstacle to
mobilizing domestic resources for national development, and also an obstacle to long-term
economic growth.

The link between capital flight and corruption discussed in the literature is limited, especially in
the Sub-Saharan African region. The greatest concern, according to the Transparency
University of Ghana http://ugspace.ug.edu.gh

International’s (TI) 2010 Corruption Perceptions Index (CPI), indicated that the most corrupt
region in the world is Africa. The CPI report defines corruption as the abuse of entrusted power
for private gain, in public and private sectors. More so, six African countries (Angola, Burundi,
Chad, Equatorial Guinea, Sudan and Somalia) were ranked among the 10 most corrupt countries
of the 173 countries surveyed by the Berlin-based group. It provide scores of countries on a 10-
point scale, with zero being the most corrupt. Accordingly, out of 47 African nations surveyed, 44
scored less than five on the index, showing serious levels of corruption. The report again indicated
that the least corrupt African nation (Botswana) scored 5.8 indicating how severity of corruption
in Africa.

The detriments of corruption and capital flight in Sub-Saharan Africa is enormous as it leads to the
widening of the gap between domestic savings and investment. Agents of the economy lose trust
in the social-political and macro-economic environment of the country. Empirical evidence in the
literature has shown that in Sub-Saharan Africa, corruption impairs political, economic and social
development and hinders administrative development and performance. (Osoba, 1996; Hope,
1997). Mauro (1995) indicated that corruption leads to a reduction in private investment, and
consequently dwindling economic growth. Also, Bardhan (1997) reiterated that corruption induces
the creation of inefficiencies leading to a lack of productivity and poor quality public investments
and services. Corrupt practices leads to an unstable economic environment resulting in high cost
of operation and a greater level of risk and expected lower returns for investors (Johnson et.al
2000).

University of Ghana http://ugspace.ug.edu.gh

Accordingly, research into the corruption – capital flight phenomenon in Sub-Saharan Africa is
worth investigating since it has received little or no attention in particularly SSA, and in attempting
to address this research gap, the study sees corruption as a domestic investment risk function.
The study therefore, uses the portfolio choice framework approach to establish the relationship
between capital flight and corruption, including other objectives for a panel of 32 countries over a
period of thirteen years (2000-2012).

1.3 Research Questions
Empirical literature has indeed identified institutional governance indicators, as well as
macroeconomic indicators as some of the factors that determine capital flight. It is worth noting
that a vast number of institutional governance variables has been outlined in the literature as having
in one way or the other an influence on capital flight. The research work deem it fit to take a look
at some of these important institutional variables and their relationship with capital flight focusing
on corruption as the main variable of interest. In doing this, three (3) specific questions needs to
be asked and answered. These are:
 What is the direction of causality between corruption and capital flight?
 What is the effect of corruption on capital flight in SSA?
 Does the other equally important institutional governance indicators influence the effects
of corruption on capital flight?
University of Ghana http://ugspace.ug.edu.gh

1.4 Objectives
The main focus of this study is to empirically examine the effect of institutional governance on
capital flight with emphasis on corruption in Sub-Saharan Africa. In doing this, the study seeks to
achieve the following specific objectives:
 Examine the effects of corruption on capital flight in Sub-Sahara Africa.
 Investigate the effects of the interaction of corruption and other equally important
institutional governance indicators on capital flight in Sub-Sahara Africa.
 Determine the direction of causality between corruption and capital flight.

1.5 Significance of the study
The essence of this study is as a result of the fact that Sub-Saharan African (SSA) countries, over
the past decades have experienced massive outflows of private capital towards western financial
centers. The private assets surpass the continent’s foreign liabilities, ironically making Sub-
Saharan Africa a ‘net creditor’ to the rest of the world (Boyce and Ndikumana, 2001). Sub-Saharan
Africa shows a significantly higher preference for foreign assets relative to domestic assets when
compared to otehe developing regions; hence Africa has the highest proportion of private assets
held abroad (Collier, 2001). Capital flight from Sub-Saharan Africa economies constitute a serious
development challenge for so many reasons. The phenomenon diverts scarce resources away from
domestic investment and productive activities; and also has a substantial regressive impact on
wealth distribution among others. From this background, the study is critically meant to identify
and bring to light the role of corruption as an institutional governance indicator on capital flight in
Sub Saharan Africa.
University of Ghana http://ugspace.ug.edu.gh

A review of the negative impact of capital flight in the economies of Sub-Saharan African
countries, indicates the need to reduce capital flight in the region to promote healthy economic
environment. Reduction of capital flight represents a crucial avenue to increase the resources
available in Sub-Saharan African countries for both consumption and investment among the public
and private sectors (Fofack and Ndikumana 2009).The study intends to bring to light some possible
ways in this regard. This includes, capital flight repatriation to help raise the level of domestic
investment through the stabilization of the macroeconomic and institutional environment. The
rippling effect is a minimization of uncertainty among agents about the direction of public policies;
low inflation; low taxes and among others.

Finally, this study will add to the existing body of knowledge and can serve as a reference
document for institutions, students, policy makers and other professionals, as well as contribute to
the empirical and theoretical debate.

1.6 Organization of the study
This study is structured into six (6) main chapters: Chapter one is made up of the introduction
consisting of the background of the study, a statement of the problem, the research questions, and
the objectives, significance of the study as well as the organization of the study. Chapter two is
categorized into two, thus, the theoretical and empirical literature of the study. Also chapter three
place emphasis on review and trend analysis of corruption and capital flight in SSA. The fourth
chapter consists of the methodology used for the study with emphasis on model specification,
estimation techniques, source of data, and description of the variables. The remaining Chapters
University of Ghana http://ugspace.ug.edu.gh

five and six are made up of discussion of empirical results and conclusions and recommendations
respectively.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER TWO
LITERATURE REVIEW
2.1 Introduction
This chapter is aimed at reviewing the theoretical and empirical studies on capital flight and
corruption. The chapter is made up of three (3) main themed areas including a review of the
definition and measurement of capital flight. The second section examines the theoretical literature
on capital flight. The last part is an empirical survey of some studies conducted on capital flight
and institutional governance including other determinants of capital flight.

2.2 Review of definitions and measurements of Capital Flight
Capital Flight has been defined differently by different authors indicating that there is no clear cut
uniformity among researchers in agreeing to a common definition. For instance, Cuddington
(1986) defined capital flight as a short-term private capital outflow which occurs in the response
“not only to political crisis but also to economic policy failure”. Deppler and Williamson, 1987
also provided their own version of the definition of capital flight by indicating that it is “acquisition
or retention of a claim of non-residents that is motivated by the owner's concern that the value of
his asset would be subject to discrete losses or impairment if his claims continued to be held
domestically”. More so, Pastor (1990) defines capital flight as resident capital outflow where
capital can be represented by any asset local residents have sent abroad, avoiding national
regulation. According to Cooper and Hardt (2000), capital flight is the flow of financial asset from
one country to another resulted from the holder’s perception that capital is subject to high level of
risk due to hyper-inflation, devaluation, and political turmoil….etc.
University of Ghana http://ugspace.ug.edu.gh

Taking a look at a couple of recent studies where capital flight has been examined, the definition
still differs among researchers. For instance, Schneider (2003) defined Capital Flight as the
outflow of resident capital from a country in response to economic and political risk in the domestic
economy. He reiterated that linking the definition of capital flight to a notion of national welfare
might not be the soundest idea.

Moreover, the differences in the definition of capital have an impact in its measurements and as
such it suffers a similar defect. From empirical literature, measurement of Capital Flight varies
among authors and this leads to differences in its estimates for different countries. Distinguishing
these differences in the literature; Joseph Haregeuin (2012) identified three major measures widely
used in the literature. These are Direct, Indirect, and Dooley measures.

Direct Measures involve the use of direct data for Balance of Payment, where the emphasis is on
short-term outflows (“Hot Money”). Cuddington (1986) established Capital Flight to be the
acquisition of short-term foreign assets by the non-banking private sector. He reiterated that
Capital Flight is the addition of errors and omissions and other short term capital. Capital flight in
this instance is measured by adding up net errors and omissions and non-bank private short term
capital outflows. This measure reflects the view that capital flight goes unrecorded, as a result of
the illegal nature of capital movements. There is the notion that the unrecorded capital movements
appear in the net errors and omissions. More so, by placing emphasis on short-term flows, however
medium and long-term outflows are excluded, which, the author indicated are more normal in
character (Gibson, 1993). C. Kant (1996), emphasized on some criticisms of the direct method of
measurement, among which are the errors and omissions (unrecorded capital flows, measurement
University of Ghana http://ugspace.ug.edu.gh

and rounding errors), registration delays, and unreported imports. Measuring capital flight in this
way is called short-term movements of capital.(Chunchachinda and Sirodom, 2013).

Another measure of capital flight is referred to as indirect measures (also noted as residual method)
and it has been the most widely used measures in the available literature. Among its usage notably
include the World Bank (1985), Morgan Guaranty (1986). The method looks at measurement
under the assumption that capital inflows will be used as a basis of capital outflows. Therefore, the
residual of both types of capital inflows is regarded as the amount of capital flight in the respective
country. Stated in other words, capital flight is measured indirectly from balance of payments
statistics by comparing the sources of capital inflows (Net increases in external debt and the net
inflow of foreign investment) with the use of these inflows (the current account deficit and
additions to foreign reserves). The differences that arise when the sources exceed the uses of capital
inflows, is called capital flight.

The World Bank (1985) suggested the use of indirect measures provided that the increase in
External Debt (ED) and the Net Foreign Direct Investment (NFDI) show the amount of capital
inflows and also the Current Account Deficit (CAD) and the Increase in Official Reserves (OR)
show the amount of capital outflows. Furthermore, the difference between Sources of Funds and
Uses of Funds refers to the increase in Net Foreign Claims by the private sector and is regarded as
the Capital Flight (CF): CFwb = ED + NFDI – CAD – OR. The World Bank used the World Debt
Table data for External Debt increases rather than that of the balance of payment.

University of Ghana http://ugspace.ug.edu.gh

On the other hand, Morgan Guaranty Trust (1986) adjusts the World Bank’s measurement of
capital flight ( 𝐶𝐹𝑤𝑏 ) by omitting Short-term Foreign Assets of the Banking System (SFABS) and
only regarding foreign assets of the non-banking private sector as the Capital Flight (CF): CF
= 𝐶𝐹𝑤𝑏–SFABS. Morgan Guaranty Trust (1986) was unable to show or prove the difference
between the motivations from external claims by the banking system and that of the external claims
by firms and individuals which happens to be the main reason for adjusting the definition of capital
flight introduced by the World Bank.

Furthermore, the way Morgan Guaranty Trust measured capital flight was readjusted by omitting
the Reinvested Investment Income (RII) and the Income from Tourism and Border Transaction
(ITBT) from the current account before measuring the capital flight by Cline (1987). The argument
in this case was that incomes from external assets which do not return to the country should not be
used in capital flight calculations and that the Income from Tourism and Border Transaction
involves transactions in the free market, which is beyond state control. Accordingly, Capital Flight
(CF), can be measured by: 𝐶𝐹𝑐= CFMG - RII – ITBT, Cline (1987).

More so, The IMF working paper (2005) also used the residual measurement approach developed
by various scholars in the World Bank. They explained that the residual approach extracts the best
measured components of the BOP identity (the current account balance, the change in external
debt, net foreign direct investment, and the change in official reserves). The residual measure
approach is constructed as: KF = ΔDebt + DFI + CAS + CHOR; where ΔDebt is the change in
total external debt outstanding, DFI is net foreign direct investment, CAS is a current account
surplus, and CHOR is the net reduction in the stock of the foreign reserves. Using Boyce and
University of Ghana http://ugspace.ug.edu.gh

Ndikumana (2001) methodology, they adjust the change in the long-term debt stock for
fluctuations in the exchange rate of the dollar against other currencies.

The last measure of capital flight is called the Dooley method. This method / approach of capital
flight measurement was developed by M.P Dooley (1986). Conceptually, the method is different
from the residual approach. Dooley estimated capital flight as: CFdo = TCO −ΔES. Where TCO
= FB + FDI + CAS + FR − EO –ΔED. Where TCO, FDI, CAS, FR, FB, EO and ED implies the
total amount of capital outflows, foreign direct investment, current account surplus, a change in
foreign reserves, change in external debt (World Bank data), errors and omissions as well as
foreign borrowing as reported in balance of payments statistics respectively. The next is to
calculate the stock of external assets corresponding to reported interest earnings: ES = IE/rus.

According to Dooley, the approach is a hybrid measure of both direct and indirect methods – thus
defining Capital Flight as the total amount of externally held assets of a private sector that do not
generate income recorded in the Balance Of Payment statistics of a country. He made three
adjustments to capture unrecorded capital flows: Firstly, errors and omissions were added;
Secondly, the difference between the stock of external debt reported in the World Bank data and
those reported in the balance of payments statistics add to the estimate of the increase in private
sector foreign assets. Lastly, Calculation of the stock of external assets needed to give the observed
amount of investment income in the balance of payments at international market rates (eg. US
Treasury bill rate).
University of Ghana http://ugspace.ug.edu.gh

2.3 Review of Corruption in Sub-Saharan Africa
2.3.1 Introduction
Corruption is basically a phenomenon that is difficult to observe, but easy to have opinions about.
It is a complex issue with a vast array of determinants and effects that are often context and country
specific. It is an indisputable fact that corruption has become one of the most persistent and
progressively worsening social problems afflicting virtually all sub-Saharan African countries
today. The practice has indeed found its way into all institutions, both public and private,
governmental and non-governmental. Having reached endemic proportions, corruption has
become not only a way of life, but also a principal method for the accumulation of private property
(Mulinge and Gwen, 1998). Corruption is one of the greatest inhibiting forces to equitable
development and to the combating of poverty and it constitutes the difference between life and
death' - World Bank President James Wolfensohn (2003).

2.3.2 Definition and Causes of Corruption in Sub-Saharan Africa
Corrupt behavior incorporates acts such as the use of public authority, office, or official position
with the deliberate intent of extracting personal or private monetary rewards or other privileges at
the expense of public good and in violation of established rules and ethical considerations (Hope
1987; United Nations 1990). Corruption is a form of antisocial behavior by an individual or social
group which confers unjust or fraudulent benefits on its perpetrators, is inconsistent with the
established legal norms and prevailing moral ethos of the land and is likely to subvert or diminish
the capacity of the legitimate authorities to provide fully for the material and spiritual well-being
of all members of society in a just and equitable manner (Osoba, 1996). The Asian Development
Bank (ADB, 2010) also defines corruption as a behavior on the part of officials in public and
University of Ghana http://ugspace.ug.edu.gh

private sectors, in which they improperly and unlawfully enrich themselves and or those close to
them or induce others to do so, by misusing the position in which they are placed.

Accordingly, the World Bank defined political corruption as the abuse of public power for private
benefit (Bardhan 1997, and Tanzi 1998). The commonly known examples of political corruption
are bribery and embezzlement. This involves the extraction of bribes by Public servants to provide
services they are supposed to offer to citizens. Those with control over spending often abuses that
power by embezzling funds. Buying of votes during elections by political parties and bribery and
favoritism by officials responsible for employment in the public sector are other instances of
political corruption. On the other hand, Jain (2001) defined corruption as “… acts in which public
power is used for personal gains in a manner that contravenes the rules of the game”. In the broad
spectrum of empirical literature, the phenomenon has been established to having a detrimental
impact on domestic investment climate and economic growth.

Empirical literature has indeed established several causes or genesis of corruption in Sub-Saharan
Africa. The literature details the role played by African bureaucracies, private businesses and
individuals, International Actors as well as the historical perspective, thus colonialism. The
industrial revolution of the nineteenth century appears to have been the very initial historical basis
for emergence of corruption. Accordingly, the financial growth which accompanied this historical
event is said to have been directly responsible for the birth of white collar crime of which
corruption is a part (Robb, 1992).

University of Ghana http://ugspace.ug.edu.gh

Three ways have been established by empirical literature as linking colonialism to corruption in
Sub-Saharan Africa. The first link is found in Robb's (1992) argument that the new economic order
resulting from the industrial revolution was responsible for the initial emergence of white collar
crime of which corruption is a part. The argument here is that the new economies, led by the
colonial governments, did not establish the structural groundwork for the origins and sustenance
of corrupt practices. Secondly, the introduction of compulsory cash taxation in the forms of hut
tax and, later, poll tax and the manner in which the tax was collected (by African leaders especially
Chiefs) led to the origin of corrupt practices. The arriving colonialists found no meaningful
monetary economies in the newly acquired territories. Therefore, most colonial governments, and
particularly those of British origin, introduced compulsory taxation payable in cash only with the
objectives of meeting the cost of administration or acquiring a cheap African labour necessary for
the establishment of productive economic activities (Stichter, 1982; Tlou and Campbell, 1984 and
Collier and Lai, 1986;).

The third link between colonialism and corrupt practices arose from the technique of divide and
rule adopted especially by colonies of British descent. Thus, the technique of breaking united front
among the various tribes and indigenes, favoring one tribe over others resulting in the creation of
different groups and paramount’s. In British Uganda and Nigeria, for example, the Baganda and
Ibo, respectively, enjoyed supremacy over other tribes (Roberts, 1962; Hunt and Walker, 1974).
As such, the technique really created immense regional variations in the levels of educational
attainment and economic opportunities and even of independence, those groups favored by the
colonial administration had an edge over those not favored (Leonard, 1991).

University of Ghana http://ugspace.ug.edu.gh

International Actors as a cause of corruption in Sub-Saharan Africa is very crucial and worthy to
note. Globalization has been accompanied by the emergence of uncontrollable market forces and
truly transnational corporations (TNCs) that dominate the basic dynamics of the world global
economy, and by a worldwide spread of manufacturing and sales (Hirst and Thompson, 1996).
According to Johnston (1998), increasing inter- dependence of economies and markets makes it
possible for corrupt agents to extend their dealings across borders. For instance, it allows for the
shifting of illicit profits out of poor countries into numbered bank accounts elsewhere.
Governments, international development organizations, and aid agencies often attach certain
conditions to loans and grants availed to developing countries. Some of these conditions
approximate corrupt practices (Mulinge and Lesetedi, 1999). For instance, the aid or grant may
arrive without considering the technical capacity of the recipient country. There are even situations
where aid organizations and donor agencies provide overpriced, but unqualified, incompetent and
inexperienced technical personnel as part of an aid package to developing countries (Mulinge and
Lesetedi, 1999).

The Structural Adjustment Programs (SAPs) have been established in the literature as a
participatory cause of corruption in sub-Saharan Africa. Where operational, such policies have
been associated with declining social services for the mass of the African population and the
stagnation of wages (Thompson, 1992; Vandermoortele, 1994 ;). As Hope (1997) states, in a
situation of declining incomes, public servants "disavow any sense of civic virtue and attempt to
supplement their incomes by engaging in corrupt acts."

University of Ghana http://ugspace.ug.edu.gh

2.3.3 Measures of corruption
Three measures of corruption have been identified in the empirical literature, namely; internal,
external and the hybrid measures of corruption. The internal measures of corruption captures the
perceptions of firms that operate within a country. It involves a research survey of the perception
of firms about corrupt practices in the country using questionnaires, interviews, among other
approaches. What makes the internal measure an important one is that it help inform firms about
the effect of corruption on the investment climate of a country since firm’s takes investment
decisions on the economic situation of a country. However, internal measures of corruption are
faced with some limitations worth noting. Firstly, firms operate in different countries with different
policy and economic environments. The effect is that the point of reference of these firms is likely
to be different and thus the data may not be easily comparable across countries. For example, firms
that operate in countries where corruption is prevalent may be accustomed to corruption and
therefore have less stringent standards for judging corrupt practices (Cameron et al., 2005).

The second point is the characteristics of firms. For instance, the large firms may have their
corruption ratings different from small firms suggesting that countries with the same level of
corruption, but different composition of firms with may have different internal corruption ratings
as a result the data may not be easily comparable across countries. The last limitation, but not the
least indicates that corruption may be under-reported, as respondents may feel reluctant to provide
sensitive and accurate answers to questions asked by the survey (Azfar and Murrell, 2005)

The external measure of corruption involves the assessment of corruption undertaken by risk
analysts who typically reside outside a country. Generally, private risk rating agencies provide the
University of Ghana http://ugspace.ug.edu.gh

corruption data to a targeted recipient, especially foreign investors. Since various countries are
analyzed and rated by the same entity in this case the risk analyst, outcomes are generally more
consistent and less devoid of measurement errors as compared to the internal measure of
corruption. The drawbacks of this measure of corruption include the analysis and estimates of the
risk analyst which at many times based on media reports and not necessarily personal experience.
This anomaly, often than not affect the levels of corruption reported by these “experts” and it may
not accurately reflect the levels of corruption that prevail in a country. Accordingly, the empirical
results of Haque et al. (2000) found that commercial risk-rating agencies often rate African
countries as riskier than warranted by the fundamentals. This empirical view is also consistent with
the study by Ferri (2004) who found that risk assessments by private rating agencies tend to be
biased against poor countries or smaller countries.

The third corruption measure called the hybrid, actually takes into account a combination of
corrupt data from different sources into a composite index. The combination of all types of
corruption data (including internal and external measures of corruption), helps in the combating of
the problems associated with the other two measures of corruption explained above. However, one
disadvantage of the hybrid measure is that they do not indicate clear distinction among various
forms of corruption, such as bribery, embezzlement of public funds or nepotism. A problem may
arise if different types of corruption have different effects on investment.

Following the available corruption measures analyzed, this thesis employs the hybrid corruption
measure for the empirical analysis. The most widely available hybrid measures of corruption are
the Corruption Perception Index (CPI) compiled by Transparency International.
University of Ghana http://ugspace.ug.edu.gh

2.3.4 Effects of corruption in Sub-Saharan Africa

In recent years, the problem of corruption has gained much attention primarily as a result of the
high level of corruption cases in many countries, an increasing awareness of the cost of corruption
throughout as well as the practical and economic changes in many countries are undergoing.
However, corruption is an issue of development. Staats (1972) noted that corruption is a social
problem found in various “degrees and forms in all but the most primitive societies”.

The representative of United Nations Office on Drugs and Crime (UNODC) Southern Africa by
name Jonathan Lucas in 2009, labeled corruption as “a crime against development, democracy,
education, prosperity, public health and justice, - what many would consider the pillars of social
well-being." Transparency International’s (TI) 2010 Corruption Perceptions Index (CPI), released
in October 2010, identified Sub-Saharan Africa is also one of the most underdeveloped regions on
earth and Africa as the most corrupt region in the world. In the broad spectrum of empirical
literature, the phenomenon remains a major obstacle to achieving much needed progress and has
a detrimental impact on domestic investment climate and economic growth. Even though
corruption happens to be a world-wide phenomenon, it has been established in the literature to be
costing Africa so much and restricting its development.

The consequences and effects are also on the increasing side. Shleifer and Visny (1993)
acknowledged that corruption precipitates monopolies, and prevent market- based competition and
innovation. Corruption lowers private investment, thereby reducing economic growth, Mauro
(1995). Mauro (1998), concluded that the bulk of the effects of corruption on economic growth,
which operate through private investment accounts for about one third of the total growth effects.
University of Ghana http://ugspace.ug.edu.gh

Rahman et al (1999) in the study of the effects of corruption on economic growth and gross
domestic investment in Bangladesh did establish that corruption is negatively significant to cross-
country differences in economic growth and gross domestic investment (both public and private).
Wei, (2000a) argues that corruption is likely to produce certain compositions of capital flows that
makes a country more vulnerable to shifts in international investors' sentiments and expectations.
He also indicated that other consequences of corruption may possibly include loss of tax revenues
because corruption may encourage people to evade taxes. In addition, by reducing tax revenues
and increasing public expenditure, corruption may lead to adverse budgetary consequences.
Corruption imposes additional costs on growth process as it diverts scarce resources away from
viable investment. It increases the degree of uncertainty and risk associated with investment and
drives away new investment see Fabayo J. et al (2011).

Accordingly, corruption has become a major impediment to political, economic and social
development. It impairs economic efficiency (Gould and Amaro-Reyes 1983). The phenomenon
has indeed led to an increase in poverty and economic growth. This is not to argue that corruption,
however, is peculiar to sub-Saharan Africa. On the contrary, it is a global phenomenon which
"manifests itself with significant similarities and differences in different societies, depending on
the peculiar systems of power distribution and the legal and moral norms operating therein" (Osoba
1996). It also hinders administrative development and performance as well as stifles local initiative
and enterprise (Hope 1997) and intensifies other social problems such as crime, ethnicity and
ethnic conflicts, and family-related problems.

University of Ghana http://ugspace.ug.edu.gh

In conclusion, however, recent empirical literature on capital flight has placed particular emphasis
on the role of non-economic variables (political risk and governance/institutional factors) and their
relationship with capital flight from developing countries (Gibson and Tsakalotos, 1993;
Schineller, 1997). It is understood in the literature that political instability, poor governance and
weaker institutions discourages domestic investment and induces capital flight. The literature has
indeed identified corruption as one of the key measures for evaluating the quality of
governance/institutions along with transparency, the rule of law, among others.

According to Transparency International’s Corruption Perception Index, six of the world’s ten
countries most burdened by corruption are located in the continent. Furthermore, an econometric
analysis suggests that, holding other determinants of capital flight constant, corruption does have
a positive and a significant impact on capital flight. Capital flight and corruption are some of the
main causes of the poverty in Sub-Saharan Africa. (Capital Flight and Corruption Treaty NGO
Alternative Treaties at the 1992 Global Forum). In addition, Le and Rishi, (2008), in the their study
on the role of corruption in impelling capital flight established that corruption is one of the
dimensions of poor governance and has significant positive effect on capital flight. The result
indeed concluded that advocating good governance by combating corruption makes a great deal
of sense for countries aiming to staunch capital flight. The result is consistent with the study of
Lawanson (2006) and Ndiaye (2009). Their findings also established a positive and significant
effect of corruption on capital flight using the system GMM at the 5 % level of significance. The
above empirical evidence certainly showcases that there exist a relationship between corruption
and capital flight.
University of Ghana http://ugspace.ug.edu.gh

2.4 Theoretical literature on capital flight and corruption
Generally, the concept of the causes and determinants of capital flight has been linked to four
main hypotheses by many scholars. The first is the portfolio choice framework that takes into
consideration rate of return and risk differential as the drivers of capital flight (Ajayi, 1992). The
second is the debt-driven flight thesis. This thesis place emphasis on the heavy external debt
burden as the cause of capital flight. Thus, capital flight leads to borrowing in order to promote
growth and further borrowing increases the debt ratio, promoting capital flight and accordingly
leads to poor economic growth. The third is the investment diversion thesis which focuses on
diversion of capital towards a more stable economy due to uncertainties in both the political and
macroeconomic situation of the domestic economy. Lastly, the tax – depressing thesis describes
a situation where domestic government do not have access to funds or wealth held abroad for
taxation purposes and this situation leads to a potential loss of government revenue and
consequently a negative effect on economic growth and development. That is, the direct result of
capital flight is the reduction in the revenue generating capacity of government. (Ajayi, 1992).

Capital movement/ flows have been identified in the literature as having a number of systemic
explanations. However, in the theoretival literature, many researchers have used the portfolio
choice methodology /framework to explain capital flight phenomenon. The theory describe how
capital moves across countries in response to rate of return and risk differentials. Here, emphasis
is placed on the assessment of domestic investment risk and uncertainty that lead individuals to
choose to hold assets abroad instead of investing domestically.

University of Ghana http://ugspace.ug.edu.gh

The original idea of portfolio theory of capital movement can be linked to Williams, (1938). The
principal idea was interest rate differentials as the cause of capital flows. It took MacDougal type
models (determinants of capital flows framed in inter-temporal optimization context), to place
particular attention on risk and not only return differential (see J Tobin, 1958). The literature on
capital flight has built on these earlier theories. Notable among these authors’ included Khan and
Hague (1987) who indicated that capital flows can arise in an instance of where investors face an
asymmetric risk of expropriation. In this instance, investors in the domestic economy will send
their funds abroad when facing a higher risk.

Also, Dooley (1988) placed emphasis on the notion of asymmetric risk by expanding the focus to
a wide range of implicit taxes resulting from either a rise in inflation or exchange rate depreciation.
This led the authorities to depend more on the inflation tax, resultion in the erosion of the value of
financial assets in the domestic economy, hence, capital flown to acquire foreign assets. Alesina
and Tabellini (1989), also reiterated a situation where different governmental regimes with
different ideologies alternate in office, resulting in uncertainty about future policy direction can
lead simultaneously to capital flight, low domestic investment and high external debts.

Accordingly, Ndikumana and Boyce, (2003) viewed capital flight movement as resulting from
investors who in their bid to maximize profits allocate funds between domestic and foreign
investment based on the relative risk-adjusted rate of return at home and abroad. It indicated that
in developing countries with riskier investment environment, will result in lower net risk-adjusted
returns. This phenomenon has invariably been able to explain why capital continues to flow out to
University of Ghana http://ugspace.ug.edu.gh

foreign lands. In addition, foreign investors can be discouraged to invest in the domestic economy
if the situation of risky environment discourages domestic investment.

However, in undertaking a critical look at the link between capital flight and corruption, this thesis
follows tune to the use of Portfolio Choice framework based on Collier et al (2001, 2004); Le and
Zak (2006) and Ali and Walters (2011) to explain how corruption affects capital flight in SSA.

By viewing corruption as a contributor to domestic investment risk, this thesis consider an
economy, say a country in Sub-Saharan Africa with a vast number of infinitely-lived identical
agents. The economic agents optimize their consumption patterns between investment in the
domestic economy or foreign country. This study assume that there exists only one investment in
each country such that agents’ consumption from the return on wealth is allocated to one period
investment in the domestic country or to a single foreign country. Excluding labour and
considering the population as constant, the wealth (W) is is also normalized to unity and the
assumption of a single homogenous commodity produced in both countries.

Investment in the domestic economy is denoted by 𝐼𝑡 at time t. This investment earns a rate of
return 𝑅𝑅𝑡. It is assumed that, in the domestic economy, investment is risky due to poor
governance (e.g in this case is corruption) whiles that of the foreign investment earn a risk-free
rate of return 𝑅𝑅𝑓 when agents invest 𝐼𝑡
𝑓
in say, a US Treasury bill.
With these options, the problem is on the representative Sub-Saharan African agent to choose an
investment portfolio that maximizes utility of his wealth by solving:
𝑀𝑎𝑥
𝑤𝑡
𝐸 ∑ 𝛽𝑡 ∞𝑡=0 𝑈(𝑊𝑡) (1)
University of Ghana http://ugspace.ug.edu.gh

Subject to 𝑊𝑡 = (1 + 𝑅𝑅𝑡) 𝐼𝑡 + (1 + 𝑅𝑅
𝑓) 𝐼𝑡
𝑓 − 𝐼𝑡 + 1 − 𝐼𝑡+1
𝑓 , (2)
The necessary and sufficient conditions below depicts the optimal allocation of the portfolio where
𝑈(𝑤) is strictly increasing, continuous and strictly concave.
𝐼𝑡+1
∗ =
𝐸(𝑅𝑅𝑡+1 − 𝑅𝑅
𝑓)
𝜃𝑉𝐴𝑅(𝑅𝑅𝑡+1)
(3)
The 𝜃𝑉𝐴𝑅(𝑅𝑅𝑡+1) represents the variance of the return on investment in the domestic economy,
whereas the risk aversion is 𝜃 ≡ −𝐸[𝑈′′(𝑤𝑡+1)]/𝐸[𝑈
′(𝑤𝑡+1)].
Assuming the presence of a related problem the individual agents are solving in other countries,
the net capital flight is given by 𝑁𝑡+1
𝑓 = 𝐼𝑡+1
𝑓𝑜 + 𝐼𝑡+1
𝑓∗
Where 𝐼𝑡+1
𝑓𝑜 𝑎𝑛𝑑 𝐼𝑡+1
𝑓∗
are capital outflows and
inflows respectively. Then the average capital invested in the domestic economy from time t to
time t+1 is given by 𝑉𝑡+1 = 𝐼𝑡+1
∗ + 𝑁𝑡+1
𝑓
(4)
The fourth equation indicates a equilibrium of the capital stock which is made up of domestic
investment and net foreign investment. An arrangement of the equation four however, will produce
the equilibrium capital flight as below:
𝑁𝑡+1
𝑓 = 𝑉𝑡+1 −
𝐸(𝑅𝑅𝑡+1 − 𝑅𝑅
𝑓)
𝜃𝑉𝐴𝑅(𝑅𝑅𝑡+1)
(5)
Equation five (5) indicates that a high capital flight is associated with low expected returns on
domestic investment (domestic investment risk is high). So, to obtain capital flight as a ratio of
physical capital stock, equation 5 is divided by 𝑉𝑡, as below:
𝑁𝑡
𝑓
𝑉𝑡
= 1 −
𝐸(𝑅𝑅𝑡 − 𝑅𝑅
𝑓)
𝜃𝑉𝑡𝑉𝐴𝑅(𝑅𝑅𝑡)
(6)
Following the focus of this study as identifying corruption, as an indicator of weaker institutions
and poor governance, and as a contributing factor to domestic investment risk, there is the need to
decompose the variance of domestic investment returns.
University of Ghana http://ugspace.ug.edu.gh

From the above theoretical analysis of portfolio choice theory, a higher capital flight occurs when
expected returns domestically are low and domestic investment is high. That is, corruption-driven
funds moves from a country because corrupt governments are feared with the notion that they will
not provide a stable and conducive environment for investment. This corruption-driven money
explains the earlier statement that corruption is a contributing factor to domestic investment
climate through risk and uncertainty. The approach has been used by some authors in their
empirical works of corruption and investment decisions because of its importance in being able to
explain capital flows from developing countries. Some of these authors are Wedman (1997); Tanzi
and Davoodi, (1997). These authors reiterated that corruption can lead to lowering of the quality
of investment in an economy and also destroy the quality of domestic investment climate through
uncertainty and insecurity.

2.5 Empirical Literature Review
2.5.1 Determinants of capital flight
The empirical literature on the determinants of capital flight has revealed a vast array of factors.
This can be linked to the different definitions, measurements, and the econometric model used.
The most widely mentioned and consistent factors include the macroeconomic factors, capital
inflows, governance and institutional quality, financial development, fiscal policy and rate of
return differentials.

 Institutional and Governance Indicators
Six institutional variables that capture various types of political institutions that the literature has
identified as important drivers of capital flight are political competitiveness as proxied by the level
University of Ghana http://ugspace.ug.edu.gh

of democracy, institutional and political constraints on executive; an index of political
accountability, government fractionalization, and political stability. The six indicators of
governance include the World Bank’s estimates of regulatory quality; the rule of law; government
effectiveness; political risk; corruption; and the World Banks “Ease of Doing Business” rankings.

Quality institutions are generally believed to be a crucial catalyst for domestic investment
climate. Empirical literature has shown a positive relationship between poor governance and
institutions. North, (1990), established that taking a decision to invest in the domestic economy,
depends on whether property rights and other investment-promoting institutions are in place and
that well developed institutions indirectly increase the potential for higher rates of return within
the domestic economy by lowering transaction costs.

Acemoglu et al, (2001, 2003) in their contribution to the literature indicated that institutions
directly influence whether economic agents engage in productive investments or not. Similarly,
strong institutions enhances the domestic investment climate by reducing the likelihood of
distortionary macroeconomic policies. Weak institutions including weak democracy and political
freedom aggravates an illegal outflow of capital from poor countries, diverting scarce resource
from injecting the development pipeline (Lensink et al., 2000). Studies by Hermes and Lensink
(2002) acknowledged that perceived ill institutional variables in any economy may give rise to
capital flight because citizens lose confidence in the domestic economy thereby holding their funds
in abroad. The lack of strong institutional system and good governance expose elites to corrupt the
capital market at the cost of the national interest (Ndikumana and Boyce, 2003).

University of Ghana http://ugspace.ug.edu.gh

Also, effective institutional constraints on executive power, affects capital flight independently.
This result is from a study of testing whether unsound macroeconomic policies or weak institutions
lead to capital flight, using panel data for a large set of developing, emerging market and transition
countries. The results of the research indicated capital flight as a mechanism by which institutional
quality influences volatility, (Cerra; Rishi, and Saxena, 2005).

A study by TB. Pepinsky, (2008), examining the political bases of portfolio investment using a
unique cross-national dataset on net portfolio flows did establish that countries with “better
institution” were no less vulnerable to portfolio outflows than countries with “worse institutions.”
It indicated that Governance quality is the strongest predictor of portfolio capital flows, while
political institutions perform poorly. The existing empirical literature is consistent with the
findings that a good institutional development is associated with a lower incidence of capital flight
(see, for example, Le and Zak, 2006; Cerra, 2008). Ndikumana; Boyce and Ndiaye, (2014), in
using the GMM regressions in their bid to describe the nature of capital flight, the methodologies
used to measure it, and its drivers, they relied on 39 African countries for the period 1970-2010
and found that capital flight is lower in better governed regimes, but that it increases with regime
duration. Their work also indicated that the coefficients are not statistically significant when
omitted country-specific fixed effects are accounted.

Indeed, capital flight can affect economic growth through corruption. According to Transparency
International’s Corruption Perception Index, six of the world’s ten countries most burdened by
corruption are located in the continent. Furthermore, an econometric analysis suggests that,
holding other determinants of capital flight constant, corruption does have a positive and a
University of Ghana http://ugspace.ug.edu.gh

significant impact on capital flight. (Capital Flight and Corruption Treaty NGO Alternative
Treaties at the 1992 Global Forum).

High capital flight is symptomatic of an environment characterized by corruption. This can hurt
economic performance by reducing private investment through adversely affecting the quantity
and quality of public infrastructures, by lowering tax revenues and by declining human capital
accumulation (Ndikumana, 2006, 2011). Baek and Yang (2008), examined the determinants of
capital flight using panel data for 53 developing and 23 developed countries over the period of
1984-2004. Their empirical results showed that political risk and the financial incentive for capital
flows have a statistical robust relationship to capital flight. The results also indicated that
corruption, government stability and law-and-order are common factors affecting Capital flight.

Moreover, Le and Rishi (2008), considered the role of corruption in impelling capital flight in
developed and developing countries using a panel data analysis, they reiterated a positive and
significant effect of corruption on capital flight. In addition, a Dynamic Panel Data Analysis on
the Determinants of capital flight in the Common Market for Eastern and Southern Africa member
countries by Haregewoin (2012), showed a negative but statistically insignificant effect of political
stability and absence of violence on capital flight. This result indeed supports that of Lawanson
(2006), and Ndiaye (2009). Their findings also established a positive and significant effect of
corruption on capital flight using the system GMM at the 5 % level of significance, meaning the
main actors of capital flight from COMESA member countries are corrupted government officials.

University of Ghana http://ugspace.ug.edu.gh

 Macroeconomic Indicators
Variables including external debt, foreign borrowing, the rate of inflation, domestic investment,
budget deficit, real exchange rate, and real GDP, among others, are what the literature has
indicated to be crucial determinants of capital flight.

The government budget is crucial to capital movements. Larger budget deficits motivate domestic
investors to move capital abroad to escape higher future taxation risk through expectations of
higher future inflation (Boyce, 1992, Schineller, 1996, Loungani and Mauro, 2000). Lensink and
Hermes (2000) found positive and significant effect of budget deficit uncertainty on capital flight.
In contrast to the above findings, Ndikumana and Boyce (2003) found that budget deficit and
capital flight from 30 Sub-Saharan African are negatively related. Expectations of domestic
economic agents regarding future tax increases to meet the government debt repayment
obligations, results in capital flight and increased budget deficit (Ndikumana and Boyce (2003).

Also, the degree of currency over-valuation is an indicator that affects the rate of returns for both
domestic and overseas investors. In a study for their African sample indicated that currency over-
valuation helps to explain the occurrence of capital fight in Africa in all specifications. The
coefficient of this variable was found to be positive and highly significant, implying that, on
average, African economies with misaligned exchange rates tend to experience more capital flight,
perhaps reflecting expectations Collier et al. (2001). Similarly, poor exchange rate management,
such as an overvalued currency or a black market premium, may contribute to economic
uncertainty, as they generate incorrect signals to economic agents, Edwards (1989). The existing
evidence suggests that currency overvaluation, in particular, can be harmful since it may result in
University of Ghana http://ugspace.ug.edu.gh

lower economic growth, a higher probability of speculative attacks, increased current account
deficits, shortages of foreign exchange, balance of payments crises and corruption (Frait et al.,
2006; Rodrik, 2008).

Moreover, domestic inflation reduces real returns on domestic capital. More capital tends to flee
abroad to the extent that the government depends on taxing domestic financial assets through
money creation (Dooley, 1988, Pastor, 1990, Loungani and Mauro, 2000). According to Fischer
(1993), high inflation makes domestic asset holders react to the erosion of the real value of their
assets by moving their assets abroad. Also, since inflation is often regarded as an indicator of the
government's overall ability to manage the economy, a rising inflation rate tends to undermine
that ability. Contrary to the above findings, Lensink and Hermes (2000) found an insignificant
effect of the uncertainty relating to inflation on capital flight. Again Capital flight can result in
inflation if domestic sources of revenue generation are eroded and the government resort to
printing money to finance its development activities (Boyce and Ndikumana, 2003).

Accordingly, Capital inflows/Foreign Direct Investment influences capital flight, thus increase in
capital inflows provides more resources, thus leading to more capital flight. Ajayi (1995) in his
analysis argued that the simultaneous occurrence of capital inflows and capital outflow are a major
cause of capital flight. Foreign aid is still one of the most important sources of finance for most
African Countries available. Even though the impact of some types of private capital flows such
as FDI on capital flight is ambiguous. In contrast, Kant (1996), in his study suggests that FDI
reduces capital flight through its beneficial effect on the domestic investment climate. In addition,
Lensink et al (2000) in their study of 84 developing countries found an insignificant effect of
University of Ghana http://ugspace.ug.edu.gh

foreign direct investment on capital flight. However, studies by Harrigan et al. (2002) and Cerra
et al. (2008) confirm this empirically. Also, Knack, (2001) argued that aid may be detrimental to
the investment climate of recipient countries as it tends to encourage corruption and rent-seeking.

Moreover, capital flight and the external debt and capital flight causality has many outcomes,
though capital flight is caused by all the possible relationships. Ajayi (1995) and Boyce (1992) as
cited in Sheet (2005) distinguish four possible linkages between the two, (i) debt-driven capital
flight; (ii) debt fueled capital flight; (iii) flight-driven external borrowing; and (iv) flight fuelled
external borrowing. Thus, external borrowings are transformed – sometimes instantaneously from
capital inflow to capital flight, ultimately ending up abroad, usually in a private foreign account.
High indebtedness measured by total debt/GDP can be interpreted as a signal for higher future
taxation, increasing capital flight (Collier et al., 2001, 2003). Empirically, evidence has shown
that capital flight increases significantly the needs for external debt and foreign aid (Boyce, 1992;
Chipalkatti and Rishi, 2001; Cerra, Rishi and Saxena, 2008).

Accordingly, Ndikumana (2009), indicated that capital flight forces the government to increase
its borrowing from abroad, which further increases the debt burden and worsens the fiscal balance.
More so, Ndiaye (2009) finds evidence that short term debt fuels capital fight in his sample of
African countries. Also, Ali and Walters, (2011) in their study, found that higher levels of
indebtedness are associated with increased capital flight; this may reflect the relative riskiness of
African countries.

University of Ghana http://ugspace.ug.edu.gh

Moreover, Capital Flight and the Tax base: Pastor, 1990 did acknowledge that one of the negative
consequences of capital flight importantly, is the tax base erosion channel. More so, capital flight
is noted by Ajayi (1992 and 2007), that it results in the erosion of the tax base, leading to a fall in
government revenue and, consequently, a decline in public investment. This, in turn, can lower
private investment and growth. Accordingly, Cervena, (2006) confirmed the above by reiterating
that the erosion of the tax base occurs by capital flight. On the Otherhand, Lensink and Hermes
(2000) found an insignificant effect of tax payments on capital flight. Ndikumana and Boyce
(2011a) have shown empirically that countries with higher capital flight tend to have lower tax
revenues.

Also, there has been significant interest with regards to the extent to which capital flight has a
detrimental impact on economic development. Collier et atal (2001) in their bid to add to the
debate indicated that private agents hold their savings abroad, which reduces private investment
and consequently a reduction in economic growth due to such factors as macroeconomic
uncertainty, political and institutional instability among others. Cervena (2006) on his part finds
that capital flight has detrimental effects on long-term economic growth for African countries,
Latin American countries, Asian countries and East European countries. An empirical study of
Lan (2009) did find out that capital flight plays a crucial role in influencing economic growth in
the Association of Southeast Asian Nations (ASEAN). In examining the effect of capital flight on
economic growth in the Franc Zone (FZ) for the period 1970 to 2010, the econometric analysis
shows that capital flight significantly reduces economic growth in the FZ. Capital flight, thus
poses a huge threat to high and sustainable economic growth in the FZ. Ndiaye, (2011).

University of Ghana http://ugspace.ug.edu.gh

On the otherhand, Ajayi (1992), in his study on the econometrics analysis of capital flight from
Nigeria found that low economic growth cause capital flight. A study by Ndikumana and Boyce
(2007), showed a negative relationship between the growth rate and capital flight from sub-
Saharan African countries. This, they explained is as a result of high growth performance which
investors interprete as a sign of high overall returns to capital in the country. Moreover, Ndiaye
(2009) in his study involving Franco zone African countries and France found a negative
relationship between the economic growth differential and capital flight, even though the influence
was insignificant. Contrary to these findings, other studies of capital flight from Nigeria, Six Sub-
Saharan African countries including Ajayi (1992), Lensink and Hermes (1992), found the
relationship between economic growth and capital flight insignificant. The differences in the
results is as a result of the different methodologies used, the period of the study and also the
countries involved.

Moreover, using newly available data set, consisting of 139 countries for the period of 2002-2006,
indicated that capital flight has a negative impact on GDP growth. However, its significance is
ambiguous. The results are not robust to specifications, which account for region or year effects,
Valeriia Gusarova (2009). Accordingly, capital flight aggravates resource constraints and
contributes to undermining long-term economic growth (UNDP, 2011).

 Structural features
Structural features are believed to be a catalyst for particular economic shocks which may
adversely affect a country economic performance. An important factor in this case is the
availability of natural resource in a country. Empirical evidence suggests that most African
University of Ghana http://ugspace.ug.edu.gh

countries that are rich in oil and minerals have experienced relatively high levels of capital flight
(Boyce and Ndikumana, 2012). The phenomenon is as a result of poor governance and inadequate
management capabilities.

 Risk and Rate of Return Differentials
Investors are always keen on achieving higher returns on their investment and as such interest rate
differentials between domestic countries and the world market has been a major factor in their
investment decision making. A situation where there is a large differential implies that domestic
agents, in an attempt to maximize their portfolios, may substitute into foreign assets where the
yield on short-term instruments is higher. In this case, capital flight may occur simply because the
returns on assets are higher abroad as compared to assets held domestically.

Numerous studies confirm that capital flight may take place in response to poor returns to domestic
investments. For example, Fedderke and Liu (2002) indicated that the domestic and foreign rates
of return play a crucial role in explaining capital outflows from South Africa. Also, Boyce and
Ndikumana (2002) indicated that the level of risk associated with investment in a country also has
an impact on the level of investment. They reiterated that it determines the level of domestic and
foreign investment that investors allocate after taking into account risk adjusted returns. Collier et
al, (2004) report that returns to capital abroad and domestic economy as well as political
conditions determine capital flight in their sample of countries.

A crucial source to the relative riskiness of developing countries is indebtedness. A high level of
indebtedness increases the country’s vulnerability to external shocks, which heightens uncertainty
University of Ghana http://ugspace.ug.edu.gh

over expected future returns to investments. There is overwhelming evidence in support of the
hypothesis that indebtedness increases risk and thus causes capital flight in LDCs (Lensink et
al.,2000, Collier et al., 2001; Ndikumana and Boyce, 2003; Cerra et al., 2008;). Another source of
increased riskiness in developing countries is the likelihood of economic crises. Some studies
confirm empirically that risk indicators such as low levels of official reserves and weak
government budget positions encourage capital flight (Hermes and Lensink, 2001; Cerra et al.,
2008).

 Financial development
Generally, empirical studies have examined the relationship between financial development and
capital flight. The relationship between capital flight and financial development have been shown
empirically as sensitive to the choice of variable used as a measure of financial development. For
instance, Ndikumana and Boyce (2003) used credit to the private sector as a measure of financial
development and indicated a negative and statistically significant effect on capital flight from Sub-
Saharan Africa. The findings suggest that an increase in the amount of credit for the private sector
is not enough to facilitate investors to invest in the domestic economy, hence a fall in the illegal
outflow of capital from SSA. In contrast however, Ndiaye (2009), established a negative and
significant impact of ratio of deposit to GDP on capital flight. Accordingly, he explained that a
rise in domestic savings will encourage and increase in financing domestic investment, thereby
reducing capital flight.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER THREE
OVERVIEW OF CAPITAL FLIGHT AND CORRUPTION IN SSA
3.1 Introduction
In order to appreciate the state of capital flight and corruption in the Sub-Saharan region, this
chapter provides a trend analysis of the estimates of capital flight and corruption for the period
under consideration. Trend analysis on the average corruption as well as capital flight estimates
for individual Sub-Saharan African country under consideration is also reported.

3.2 Trend analysis of capital flight in Sub-Saharan Africa
The analyses focus on 32 countries in the region for the period 2000-2012. Capital flight is
measured in millions of constant US dollars.

From Figure 3.1 below, capital flight from Sub-Saharan Africa has shown both upward and
downward trend. However, capital flight in the 2000-2012 period increases more than it falls.
Specifically, in the year 2000 it was high, and declined in 2002 but regains its rising mood from
2003 reaching its highest in 2007 and then declined again through to 2011 and upward again in
2012. Total real capital flight in the combined Sub-Saharan African Countries were highest in the
year 2007. This may be due to the productive nature of countries in the Sub region, which have
abundant oil and other natural resources, poor governance and weak institutions, and poor
macroeconomic environment.

University of Ghana http://ugspace.ug.edu.gh

Figure 3.1 Sub–Saharan Africa: Trends in Capital Flight (2000-2012)

Source: Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF.
On the other hand, the lowest capital flight figure is recorded in the year 2002 resulting from the
healthy macroeconomic environment, good governance and quality institutions. Between 2002 and
2003, capital flight in Sub-Saharan Africa recorded the highest percentage change, thus it grew by
227 percent in 2003 followed by a 64 percent increase between 2011 and 2012. This informs us
that capital flight is still an issue that needs particular attention in the Sub-region.

3.3 Trend analysis of average capital flight for Sub-Saharan African countries

Figure 3.2 also shows us an average capital flight for Sub-Saharan African countries in the period
2000-2012. It indicates differences in capital outflows across the countries.
0
10000
20000
30000
40000
50000
60000
70000
1998 2000 2002 2004 2006 2008 2010 2012 2014
C
F
(M
il
li
o
n
s
U
S
D
)
YEARS
University of Ghana http://ugspace.ug.edu.gh

Figure 3.2 Average Capital Flight Estimates for Sub-Saharan African Countries (2000-2012)
Source: Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF.

Capital flight in the Region is positive across all the countries with the exception of Cameroun.
This means that countries in the Region are experiencing net capital outflow. Nigeria, among these
countries has the largest amount and the worse capital flight occurrence among the Sub-Saharan
1190.39
577.01
-43.55
307.47
-185.70
319.55
28.05
61.96
640.26
1832.80
1514.04
1809.02
1361.79
2079.96
74.63
926.56
99.70
145.30
29.62
370.44
1656.88
11115.11
75.980
279.85
462.47
3988.19
19.91
1234.59
541.77
999.55
393.98
463.41
-2000.00 0.00 2000.00 4000.00 6000.00 8000.00 10000.00 12000.00
Angola
Botswana
Burkina Faso
Burundi
Cameroon
Cape Verde
CAR
Chad
Congo Dr.
Congo Rep.
Cote d’Ivoire
Ethiopia
Gabon
Ghana
Guinea
Kenya
Lesotho
Madagascar
Malawi
Mauritania
Mozambique
Nigeria
Rwanda
Seychelles
Sierra Leone
Sudan
Swaziland
Tanzania
Togo
Uganda
Zambia
Zimbabwe
Average C.F
University of Ghana http://ugspace.ug.edu.gh

African Countries. This represent 32 percent of the entire Regions capital outflow. Sudan follows
with a 12 percentage point of the entire Regions capital flight for the period under consideration.

However, Cameroun recorded a negative figure which is the lowest average capital flight from the
Region indicating that the country experienced net capital inflows. This is as a result of healthy
institutional governance. Apart from Nigeria and Sudan, five (5) other countries in the Region
contribute substantially to capital flight. Ghana leads with a remarkable average figure of
$2,079.96 million, followed by Republic of Congo ($1,832.80 million), Ethiopia ($1,809.02
million), Mozambique (($1,656.88 million), and Cote d’voire ($1,514.04) million.

This analysis and findings provide us an insight into where the concentration of capital flight in
the sub-region is positioned and as such the need for the entire Region to help draw policies and
procedures to help curtail the phenomenon in the concentrated area in this case Nigeria and Sudan
especially. This approach will help to reduce the phenomenon of capital flight in the sub-region.

3.4 Average capital flight to GDP in SSA (2000-2012)

Figure 3.3 reviews the ratio of average capital flight to GDP for Sub-Saharan African countries in
the period 2000-2012. It indicates differences in the percentages that each value of capital flight
contributes to GDP across the countries.

Analysis of figure 3.2 and 3.3 reveals that average capital flight to GDP ratio in countries
experiencing high average capital flight estimates in absolute terms differs. For instance, average
University of Ghana http://ugspace.ug.edu.gh

capital flight estimate in Nigeria for the period 2000-2012 was the highest among the samples,
whereas her capital flight for the same period accounted for an average of about 6.0% of the GDP
putting Nigeria 11th position among the samples.
Figure 3.3 Capital Flight Estimates to GDP ratio for Sub-Saharan African Countries (2000-2012)

Source: Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF.
2%
6%
-3%
5%
-15%
2%
2%
1%
4%
24%
8%
9%
13%
11%
2%
3%
7%
4%
0%
15%
20%
6%
2%
32%
24%
11%
1%
6%
22%
8%
3%
6%
-20% -10% 0% 10% 20% 30% 40%
Angola
Botswana
Burundi
Burkina Faso
Cabo Verde
Cameroon
Central African Republic
Chad
Congo, Dem. Rep.
Congo, Rep.
Cote d'Ivoire
Ethiopia
Gabon
Ghana
Guinea
Kenya
Lesotho
Malawi
Madagascar
Mauritania
Mozambique
Nigeria
Rwanda
Seychelles
Sierra Leone
Sudan
Swaziland
Tanzania
Togo
Uganda
Zambia
Zimbabwe
Average Capital Flight Percent of GDP figures
Su
b
-S
ah
ar
an
A
fr
ic
an
C
o
u
n
tr
ie
s
Average C.F % OF GDP
University of Ghana http://ugspace.ug.edu.gh

The same analysis can be made of Sudan and Ghana whose position is 2nd and 3rd respectively in
terms of average capital flight estimates in absolute terms among the samples. With respect to
capital flight to GDP ratio for this two countries, an average of about 11% is recorded for each
country putting them at 7th position among the samples.

Figure 3.3 above shows that Seychelles is experiencing the highest capital flight to GDP ratio from
the sample of Sub-Saharan African countries. For the period 2000-2012, Seychelles capital flight
estimate on average accounted for 32% of her GDP, probably as a result of the poor demand to the
main driver of growth which is Tourism as well as slow down of the manufacturing sector.
Seychelles is followed by Congo Republic and Sierra Leone which represents an average of 24%
for each country, whereas Togo and Mozambigue’s capital flight on average represents 22% and
20% of GDP respectively.

Moreover, for the period 2000-2012, 15 countries out of the sample of 32 Sub-Saharan African
countries under study recorded on average below 5% of capital flight to GDP. Cape Verde and
Burundi among these countries recorded a negative figure, thus, on average -15% and -3% of
capital flight to GDP respectively. Therefore the lowest capital flight to GDP ratio on average is
attained by Cape Verde probably due to an increase in the tourism related foreign investment.
Other Sub-Saharan African countries in the sample with capital flight to GDP ratio below 5%
include Swaziland (1%); Chad (1%); Angola (2%); Camoroon (2%); Guinea (2%); Rwanda (2%);
Central African Republic (2%); Kenya (3%); Zambia (3%), and among others as showed on figure
3.3 above.

University of Ghana http://ugspace.ug.edu.gh

3.5 Trend analysis of Corruption in SSA (2000-2012)

The period (2000-2012) under review depicts an upward trend in the perception of corruption in
Sub-Saharan Africa. The region began the period with a corruption perception score of 12.7 in
2000. Subsequently, the figure shot to 38.2 in 2001 and later fell to 31.9 in 2002. Since then,
corruption perception in SSA has remained on an increasing trend recording the highest score of
95.10 in 2012.
Figure 3.4 Trend of Corruption in Sub-Saharan Africa (2000-2012)

Source: Corruption Perception Index, Transparency International
The increasing rise in corruption in the region is as a result of bribery and embezzlement of funds,
especially by public officials who amass wealth using their official positions. Empirical evidence
has shown that corruption discourages domestic investment thereby leading to increase poverty
and reduction in economic growth.
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Series1 12.7 38.2 31.9 37.6 49.3 64.7 75.6 86 87.7 88.2 88.8 91.19 95.09
0
10
20
30
40
50
60
70
80
90
100
C
O
R
R
U
P
T
IO
N
V
A
L
U
E
S
YEARS
University of Ghana http://ugspace.ug.edu.gh

3.6 Average Corruption Estimates for Individual SSA country (2000-2012)

Figure 3.5 Average corruption estimates in Sub-Saharan Africa (2000-2012)

Source: Corruption Perception Index, Transparency International
1.81
5.91
2.27
1.27
2.21
2.44
1.18
1.23
1.36
1.47
2.04
2.25
1.87
3.38
0.87
1.93
1.80
2.20
2.81
1.21
2.05
1.78
1.82
2.75
1.55
1.28
1.69
2.53
1.15
2.24
2.58
2.23
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00
Angola
Botswana
Burkina Faso
Burundi
Cameroon
Cape Verde
CAR
Chad
Congo Dr.
Congo Rep.
Cote d'voire
Ethiopia
Gabon
Ghana
Guinea
Kenya
Lesotho
Madagascar
Malawi
Mauritania
Mozambique
Nigeria
Rwanda
Seychelles
Sierra Leone
Sudan
Swaziland
Tanzania
Togo
Uganda
Zambia
Zimbabwe
Average Corruption (2000-2012)
S
u
b
-S
a
h
a
ra
n
A
fr
ic
a
n
C
o
u
n
tr
ie
s
University of Ghana http://ugspace.ug.edu.gh

Figure 3.5 above depicts average corruption perception among the various SSA countries under
consideration. The perception of corruption in these countries increases across with differential
magnitudes. For instance, Botswana among the other countries under consideration is regarded as
the most perceived corrupt country as it recorded the highest average score of 5.9. This may be
attributed to poor governance and weak institutions that provide avenues for corrupt individuals in
government and public office to syphon monies from public confers for their own selfish interest.

Accordingly, Ghana happens to be the next perceived corrupt country in the sub-region recording
a 3.38 average score. In addition, Malawi, Seychelles, Zambia, Tanzania, Zimbabwe, Cape Verde,
Burkina Faso, Cameroon, Madagascar, Ethiopia, Cote D’voire and Uganda scored marks between
2.0 and 3.0 whiles the rest of the country scored marks below 2.0. However, Guinea is recording
the lowest perceived corruption score of below 1.0 thus 0.87 specifically in SSA. This is as a result
of prudent macroeconomic policies and strong institutions.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER FOUR
METHODOLOGY
4.1 Introduction
The chapter deals with the model for the empirical estimation, an a priori expectation, estimation
techniques, diagnostic tests, data sources and definitions of the variables.

4.2 Theoretical framework for estimating capital flight
This study estimates capital flight employing the methodology outlined by Boyce and Ndikumana
(2001) which is a variant of the World Bank (1985) residual method. The method calculates capital
flight as the residual difference between capital inflows and recorded foreign-exchange outflows.
Adjustments made include; trade mis-invoicing, under-reporting of remittances, inflation and
exchange rate. Capital flight is estimated for country 𝒊 in year 𝒕 using the following equation:
𝐶𝐹𝑖𝑡=𝛥𝐷𝑒𝑏𝑡𝐴𝑑𝑗𝑡𝑖𝑡+𝐷𝐹𝐼𝑖𝑡–(𝐶𝐴𝑖𝑡+𝛥𝑅𝐸𝑆𝑖𝑡) +𝑀𝐼𝑆𝐼𝑁𝑉𝑖𝑡) (4.1)
where:
𝐶𝐹𝑖𝑡 is Capital Flight
𝛥𝐷𝑒𝑏𝑡𝐴𝑑𝑗𝑡𝑖𝑡 is the change in the country’s stock of external debt adjusted for exchange rate
fluctuations
𝐷𝐹𝐼𝑖𝑡 is net direct foreign investment;
𝐶𝐴𝑖𝑡 is the current account deficit
𝛥𝑅𝐸𝑆𝑖𝑡 is the change in the net stock of foreign reserves
𝑀𝐼𝑆𝐼𝑁𝑉𝑖𝑡 is net trade mis-invoicing.
University of Ghana http://ugspace.ug.edu.gh

4.2.1 Adjustment for Exchange Rate Fluctuations
The adjustment is made to the change in the long-term debt stock for fluctuations in the exchange
rate of the dollar against other currencies. This is meant to correct for potential discrepancies. The
estimate for country 𝒊, the US dollar value of the beginning-of-year stock of debt at the end-of-
year exchange rate is obtained as follows:
𝑁𝐸𝑊𝐷𝐸𝐵𝑇𝑖,𝑡−1=∑ (
7
𝑗=1 𝛽𝑖𝑗,𝑡−1 ∗ 𝐿𝑇𝐷𝐸𝐵𝑇𝑖,𝑡−1)/(𝐸𝑋𝑗𝑡 /𝐸𝑋𝑗,𝑡−1) +
𝐼𝑀𝐹𝐶𝑅𝑖,𝑡−1 /(𝐸𝑋𝑆𝐷𝑅,𝑡 /𝐸𝑋𝑆𝐷𝑅,𝑡−1) + 𝐿𝑇𝑂𝑇𝐻𝐸𝑅𝑖,𝑡−1 + 𝐿𝑇𝑀𝑈𝐿𝑇𝑖,𝑡−1 + 𝐿𝑇𝑈𝑆𝐷𝑖,𝑡−1 + 𝑆𝑇𝐷𝐸𝐵𝑇𝑖,𝑡−1
where:
LTDEBT is the total long-term debt
𝛽𝑖𝑗 is the proportion of long-term debt held in currency j, for each of the seven non-US currencies
EX is the end-of-year exchange rate of the currency of denomination against the dollar (expressed
as units of currency per US dollar)
IMFCR is the use of IMF credit
LTOTHER is long-term debt denominated in other unspecified currencies
LTMULT is long-term debt denominated in multiple currencies
LTUSD is long-term debt denominated in US dollars
STDEBT is short-term debt
DEBT is the total debt stock as reported by the World Bank.
The exchange rate adjustment is given as: 𝐸𝑅𝐴𝐷𝐽𝑡=𝑁𝐸𝑊𝐷𝐸𝐵𝑇𝑡−1− 𝐷𝐸𝐵𝑇𝑡−1 (4.2)
Finally, the adjusted change in debt is given as: 𝛥𝐷𝑒𝑏𝑡𝐴𝑑𝑗𝑡𝑡 = 𝐷𝐸𝐵𝑇𝑡 ̶ 𝑁𝐸𝑊𝐷𝐸𝐵𝑇𝑡−1 (4.3)
University of Ghana http://ugspace.ug.edu.gh

4.2.2 Estimating Trade Misinvoicing
Trade mis-invoicing is calculated by comparing a country’s declared import and export statistics
to those of its trading partners considering the addition of the cost of freight and insurance. The
assumption is that trade data from advanced countries are deemed to be accurate as compared to
that of the African country, hence, a measure of the discrepancy between them. For instance, for
the purpose of this study, the researcher calculates trade mis-invoicing of Sub Saharan African
countries relative to that of Advanced/Industrialized countries, and use this as a benchmark to
compute overall trade mis-invoicing.

The method for calculating trade mis-invoicing is as follows:
𝐾𝑋𝐼𝐶𝑖𝑡=𝐼𝑀𝐴𝐶𝑖𝑡− (𝐸𝑋𝐴𝐶𝑖𝑡 * 𝐶𝐼𝐹𝑡) (4.4)
𝐾𝑀𝐼𝐶𝑖𝑡 = 𝑀𝐼𝐶𝑖𝑡− (𝐸𝑋𝐼𝐶𝑖𝑡 * 𝐶𝐼𝐹𝑡) (4.5)
where:
𝐾𝑋𝐼𝐶𝑖𝑡 is the SSA Country 𝒊 in time 𝒕 export discrepancies with industrialized countries
𝐼𝑀𝐴𝐶𝑖𝑡 is the value of imports from the African country as reported by the industrialized trading
partners

𝐸𝑋𝐴𝐶𝑖𝑡 is the African country’s exports to industrialized countries as reported by the African
country
𝐾𝑀𝐼𝐶𝑖𝑡 is the SSA Country 𝒊 in time 𝒕 import discrepancies with industrialized countries
𝑀𝐼𝐶𝑖𝑡 is the African country’s imports from industrialized countries as reported by the African
country
𝐸𝑋𝐼𝐶𝑖𝑡 is the industrialized countries’ exports to the African country as reported by the
industrialized trading partners.
CIF is the c.i.f/f.o.b factor, representing the costs of freight and insurance.
University of Ghana http://ugspace.ug.edu.gh

TTM = Total trade mis-invoicing
TEX = Country’s total exports
TIM = Country’s total imports

Therefore, Total trade mis-invoicing (TTM) is the sum of mis-invoicing of exports and mis-
invoicing of imports. A positive sign on export mis-invoicing indicates a net outflow (export
under-invoicing) – increasing net capital flight, whiles a negative sign indicates a net inflow
(export over-invoicing) – reducing net capital flight.
Thus, 𝑇𝑇𝑀𝑖𝑡 =
𝐾𝑋𝐼𝐶𝑖𝑡
𝑇𝐸𝑋𝑖
+
𝐾𝑀𝐼𝐶𝑖𝑡
𝑇𝐼𝑀𝑖
(4.6)

4.2.3 Adjustment of Underreporting of Remittances
Sub-Saharan African countries receive enormous inflows of remittances from their citizens
working outside the region and these inflows are under-reported in the official BOP statistics.
According to the World Bank (2006, p. 92), under-reporting in the BOP statistics is the largest in
Africa accounting for more than half of total remittance flows. These under-reporting of remittance
have an effect on capital flight estimates because the quantum of foreign exchange that enters
African countries is greater than that captured in the BOP statistics. So omitting these inflows from
the residual – based estimates of capital flight would lead to underestimation of its true magnitude.

Accordingly, we estimate the quantum of unreported remittances comparing estimated inflows
from industrialized countries to the total inflows recorded in the official BOP statistics. The
University of Ghana http://ugspace.ug.edu.gh

discrepancy would be calculated based on 2006 data (the year for which the alternative estimates
are available), and extrapolate from this to estimate discrepancies for earlier years;
𝑅𝐼𝐷𝑖𝑡 = (𝐴𝑅𝐼𝑖, 2006 - 𝐵𝑃𝑅𝐼𝑖 , 2006)* 𝐵𝑃𝑅𝐼𝑡 / 𝐵𝑃𝑅𝐼 2006
Where:
𝑅𝐼𝐷𝑖𝑡 is the remittance inflow discrepancy in country 𝒊 in year 𝒕;
𝐴𝑅𝐼𝑖, 2006 is the alternative measure of remittance inflows to African countries as a whole in years
𝑡 and 2006
𝐵𝑃𝑅𝐼𝑖, 2006 is the BOP measure of remittance inflows in country 𝑖 in year 2006
𝐵𝑃𝑅𝐼𝑡 and 𝐵𝑃𝑅𝐼 2006 are the BOP measures of remittance inflows to African countries as a whole
in years 𝑡 and 2006, respectively.

4.2.4 Inflation Adjustment
Inflation is adjusted so as to make annual capital flight comparable over an extended period of
time. The US producer price index with base year of 2010 is employed to convert nominal flows
to constant dollars. Empirically, other studies have used this approach and notable among them are
Boyce (1992), Ajayi (1997) and also Chipalkitti and Rishi (2001). The method for calculating
inflation adjusted real capital flight is calculated as IACF = ACF/PPI
Where, IACF is the inflation adjusted capital flight; ACF is the adjusted capital flight; and PPI is
the US producer price index.

4.3 Model for Empirical Estimation
This study employs different estimation techniques including the Pooled OLS regression, Static
panel model of fixed effects and random effect and the Dynamic panel regression to analyze the
University of Ghana http://ugspace.ug.edu.gh

effect of corruption on capital flight in Sub-Saharan Africa. The Pooled OLS regression involves
running a simple Ordinary Least Squares (OLS) on the whole panel data to assess the effects of
the explanatory variables on the dependent variable without taking cognizance of the panel
structure of the data. The Random Effect model is used when the variations across countries is
assumed to be random and uncorrelated with the explanatory variables; On the other hand, the
Fixed Effect models are used to assume that countries possess certain individual characteristics
which are unique to them and are time-invariant. The Dynamic Panel regression model is seen to
be more appropriate for the empirical study as it allows the inclusion of lagged dependent variables
as explanatory variables in the model to capture its effects on the dependent variable.
The dynamic panel model is specified as: 𝑪𝑭𝒊𝒕 = 𝑿𝒊𝒕𝜷 +𝜺𝒊𝒕 (4.7)
Where 𝐶𝐹𝑖𝑡 represents capital flight measured in millions of constant US dollars for each country
at time t; 𝑋𝑖𝑡 is the matrix of all explanatory variables; and 𝜀𝑖𝑡 is the error term which is made up
of two components, thus unobserved country-specific effects 𝜇𝑖 and the idiosyncratic error
term 𝑣𝑖𝑡, thus; εit = μi+vit. Where i represent each country under study, t denotes the number of
years under consideration.

Along the lines of the theoretical framework of capital flight equation postulated in equation (6)
above and using the Dynamic Panel model in equation (4.7), the actual econometric model to be
estimated is written as: 𝐶𝐹𝑖𝑡=𝛽0+𝛽1𝐶𝑂𝑅𝑖𝑡+𝛽2𝑅𝐿𝑖𝑡+𝛽3𝑅𝐷𝑖𝑡+𝛽4𝐼𝐸𝐴𝑖𝑡+𝛽5𝑋𝑖𝑡 + 𝜀𝑖𝑡 (4.8)
𝐶𝐹𝑖𝑡=𝛽0+𝛽1𝐶𝑂𝑅𝑖𝑡+𝛽2𝑅𝐿𝑖𝑡+𝛽3𝑅𝐷𝑖𝑡+𝛽4𝐼𝐸𝐴𝑖𝑡+𝛽5𝑋𝑖𝑡+𝛽6𝐶𝑂𝑅𝑖𝑡 ∗ 𝑅𝐷𝑖𝑡+𝜀𝑖𝑡 (4.9)
𝐶𝐹𝑖𝑡=𝛽0+𝛽1𝐶𝑂𝑅𝑖𝑡+𝛽2𝑅𝐿𝑖𝑡+𝛽3𝑅𝐷𝑖𝑡+𝛽4𝐼𝐸𝐴𝑖𝑡+𝛽5𝑋𝑖𝑡+𝛽6𝐶𝑂𝑅𝑖𝑡 ∗ 𝑅𝐿𝑖𝑡+𝜀𝑖𝑡 (4.10)
𝐶𝐹𝑖𝑡=𝛽0+𝛽1𝐶𝑂𝑅𝑖𝑡+𝛽2𝑅𝐿𝑖𝑡+𝛽3𝑅𝐷𝑖𝑡+𝛽4𝐼𝐸𝐴𝑖𝑡+𝛽5𝑋𝑖𝑡+𝛽6𝐶𝑂𝑅𝑖𝑡 ∗ 𝐼𝐸𝐴𝑖𝑡+𝜀𝑖𝑡 (4.11)
University of Ghana http://ugspace.ug.edu.gh

However, equation (4.8) above is estimated to test the first objective of this study whereas
equations (4.9), (4.10), and (4.11) is estimated to test the second objective of the study.
Where;
𝐶𝐹𝒊𝒕 = Capital Flight for each country at time t
𝐶𝑂𝑅𝑖𝑡 = Corruption for each country at time t
𝑅𝐿𝑖𝑡 = Rule of Law for each country at time t
𝑅𝐷𝑖𝑡 = Regime Durability for each country at time t
𝐼𝐸𝐴𝑖𝑡 = Independence of the Executive Authority for each country at time t
𝐶𝑂𝑅𝑖𝑡 ∗ 𝑅𝐷𝑖𝑡 = Interaction between corruption and regime durability for each country at time t
𝐶𝑂𝑅𝑖𝑡 ∗ 𝑅𝐿𝑖𝑡 = Interaction between corruption and rule of law for each country at time t
𝐶𝑂𝑅𝑖𝑡 ∗ 𝐼𝐸𝐴𝑖𝑡= Interaction between corruption and independence of executive authority for each
country at time t
𝑋𝑖𝑡 = Control Variables for each country at time t
𝜺𝒊𝒕 = Error term
𝛽0, 𝛽1, 𝛽2, 𝛽3, 𝛽4, 𝛽5, 𝛽6 represent the parameters to be estimated. Independence of the
Executive Authority is proxied by executive constraints. Control variables capture the
macroeconomic environment proxied by; GDP growth, Inflation and use of IMF credit as a
financial development variable.

University of Ghana http://ugspace.ug.edu.gh

4.4 A priori Expectation
The table 4.1 indicates independent variables and their expected sign. A brief empirical link for
their expected sign is reported. A vivid description and source of variables are captured in section
4.7 below.

Table 4.1 Independent variable and expected sign
Independent
Variable
Variable Description Expected
sign
Data Source
Corruption Defined as the perceived levels
of public sector corruption
Positive (+) CPI of
Transparency
International
Rule of Law Captures perceptions of the
extent to which agents have
confidence in and abide by the
rules of society
Negative (-) ICRG & World
Bank data base
Regime
Durability
End of transition period defined
by the lack of stable political
institutions
Positve (+) /
Negative(-)
Polity IV database
Independence Of
Executive
Authority
Indicates the extent to which the
chief executive takes into
account preferences of others
when making decisions
Negative (-) Polity IV database
Use of IMF
credit
Denotes member's drawings on
the IMF other than amounts
drawn against the country's
tranche position
Negative (-) World Bank, IDS
Inflation (CPI) Reflects the annual percentage
change in the cost to the average
consumer of acquiring a basket
of goods and services
Positive (+) World Bank, WDI
GDP growth At purchaser's price is the sum of
gross value added by all resident
producers in the economy
Negative (-) World Bank,
national accounts
data
Source: Author’s Computation, 2015

University of Ghana http://ugspace.ug.edu.gh

 The coefficient of corruption is expected to be consistent with the findings of Le and Rishi
(2008), thus positively related to capital flight. This implies that a rise in the perception of
corruption among public authorities as in bribery, kickbacks in public procurement,
embezzlement of public funds and among others facilitates an increase in illegal outflow of
capital from Sub-Saharan Africa. Hence, (β1 > 0).

 The variable rule of law is used as a measure of good governance and institutional quality.
Consistent with existing empirical evidence that a good institutional development is associated
with a lower incidence of capital flight (see, for example, Lensink et al., 2000; Collier et al.,
2004; Le and Zak, 2006; Cerra et al., 2008). The study expects rule of law to be negatively
related to capital flight. Thus, (β2 < 0).

 Regime durability has been established in the empirical literature as having varied outcomes.
Regime durability (i.e. less frequent regime changes) may reduce capital flight since it is a
source of political stability (Collier et al. 2004). On the other hand, Ali and Walters, (2011)
found that regime durability is associated with higher capital flight since it may be linked to
weaker governance, high corruption, and poorer domestic investment climate. Hence, the
research expects a positive/negative relationship between capital flight and corruption. i.e
(β3 > 𝑜𝑟 < 0).

 Independence of the Executive Authority is proxied by executive constraints. Empirical
literature indicates a negative relationship to capital flight, (Cerra et al 2005).The results
suggest that strong constraints on the executive power lead to an exacerbation of capital flight.
University of Ghana http://ugspace.ug.edu.gh

This research work expects a negative relationship between the independence of executive
authority and capital flight. Thus, (β4 < 0)

 GDP growth is one of the variables used as an indicator of the macroeconomic environment.
Empirical literature has established that the quality of macroeconomic environment is
associated with lower capital flight. Higher economic growth is a signal of higher expected
returns on domestic investment, which induces further domestic investment and thus reduces
capital flight (Ndikumana and Boyce 2008). This research work expects a negative
relationship between GDP growth and Capital flight.

 Inflation is also an indicator of the macroeconomic situation of an economy. Capital flight
can result in inflation if domestic sources of revenue generation are eroded and the
government resort to printing money to finance its development activities (Boyce and
Ndikumana, 2002). According to Fischer (1993), high inflation make domestic asset holders
react to the erosion of the real value of their assets by moving their assets abroad and also
reiterated that since inflation is often regarded as an indicator of the government overall ability
to manage the economy, a rising inflation rate tends to undermine that ability. A positive
relationship between inflation and capital flight is expected.

 Use of IMF credit is used as a factor of financial development. Financial development has
been shown in the literature as one of the determinants of capital flight and sensitive to the
choice of variable used as a measure. A rise in domestic savings resulting from an increase in
the ratio of deposits to GDP will encourage and increase in financing domestic investment,
University of Ghana http://ugspace.ug.edu.gh

thereby reducing capital flight (Ndiaye, 2009). This thesis expects a negative relationship
between the use of IMF credit and capital flight because these credits are expected to be
efficiently invested in the local economy.

4.5 The Estimation Technique
The study employs three different panel data approaches to ensure robustness of the results across
various econometric techniques. The initial technique used is the pooled-OLS technique. The
pooled OLS estimation is often used as starting point in applied analyses despite its potential biases
resulting from the presence of individual heterogeneity and endogeneity problems. The connection
of potential endogeneity with the model for empirical estimation renders estimates using the
Ordinary Least Squares (OLS) biased and inconsistent. Due to the presence of heterogeneity and
endogeneity, this thesis adopts two other approaches, namely the Static panel model of fixed
effects/random effects and the System-Generalised Method of Moments (Arellano and Bover,
1995; Blundell and Bond, 1998) to find out the effects of corruption on capital flight in SSA.

Two assumptions pertaining in the econometric literature pertaining to correlation between the
time –invariant error term (𝜇𝑖) and the explanatory variables account for the Fixed Effects (FE)
and the Random Effects (RE) models. The random effect model assumes that the unobserved
country-specific, time-invariant effects are uncorrelated with the regressors. The model is used
when the variations across countries is assumed to be random and uncorrelated with the
explanatory variables; Thus, (𝜇𝑖𝑋𝑖𝑡) = 0 . In contrast to the random effects, the fixed effects
(FE) model assumes that the country-specific, time-invariant effects correlate with the explanatory
variables, and thus controls for them in the model. The FE models are therefore used to assume
University of Ghana http://ugspace.ug.edu.gh

that countries possess certain individual characteristics which are unique to them and are time-
invariant. The presence of these country- specific, time-invariant effects lead to the problem of
endogeneity and subsequently biases the estimates. The FE model eliminates the time-invariant
effects from the estimation. Both the fixed effect and random effect estimator are models that
handle the specific structure of longitudinal or panel data. That is, unobservable individual
heterogeneity is taken into account by both models. The Hausman test is used in choosing between
the RE and FE.

However, the System GMM estimation technique would yield more reliable estimates when the
data fit for fixed effect. Then, because of the more reliable estimates the GMM estimation
technique produces due to its ability to estimate the Equation in the presence of the endogeneity,
this thesis adopts the GMM estimation approach. The GMM procedure is best because it resolves
the Dynamic panel bias problem resulting from the endogeneity associated with such models. The
GMM technique is more preferable as it uses the lags of the endogenous variables as instruments;
in which case, the endogenous variables are predetermined and therefore not correlated with the
stochastic error term. The use of GMM allows correcting for insufficiencies related to problems
of simultaneity bias, inverse causality and omitted variables (Kpodar, 2005).

The literature, however establishes two (2) types of GMM estimation procedures. The difference
GMM introduced by Arellano and Bond (1991) and the System GMM introduced by Arellano and
Bover (1995) and Blundell and Bond (1998). The difference GMM estimator procedure resolves
the inconsistency problem resulting from the endogeneity associated with of some of the regressors
in equation (4.8). This procedure eliminates sources of the inconsistency in the estimation by
University of Ghana http://ugspace.ug.edu.gh

applying the first difference operator to the estimation. After differencing, the equation is
subsequently estimated by the difference GMM by including the lags as regressors. The process
eliminates the unobserved country–fixed effects by taking the first difference of the equation and
in the process deal with the inconsistency and biases resulting from endogeneity of the explanatory
variable by using lagged values of the endogenous explanatory variables as instruments. The
difference GMM estimator is based on the momentary condition under the assumption that the
regressor is weakly exogenous and not serially correlated. The standard GMM estimation which
eliminates the unobserved country-specific effects have been found to produce unsatisfactory
results because the process may pose serious biases when the dependent variable is highly
persistent, and there is a weak correlation between the instruments and the endogenous variable.
The weak instruments actually increase the variance of the coefficients and bias the coefficients in
the small sample.

The System GMM estimator designed by Arellano and Bover (1995) and Blundell and Bond
(1998) rectifies this problem of weak instruments associated with the differenced GMM estimator
by a system of two equations namely the level equation and the difference equation. The System
GMM has been shown to perform much better (less bias and more precision) especially when the
dependent variable is persistent. Hence, the system GMM estimator is preferred to the differenced
GMM when the dependent variable is persistent.

University of Ghana http://ugspace.ug.edu.gh

4.6 Description of variables and data sources

4.6.1 The Dependent variable: Capital flight
Capital flight is expressed in millions of constant US dollars. The data of capital flight are taken
from the database of Ndikumana & Boyce (2012). The capital flight data is updated (see procedure
outlined in chapter four, subsection 4.2 above) by the researcher using data obtained from World
Development Indicators of the World Bank. Three different methods have been identified as a
measure of capital flight. These are the direct measures, indirect measures, and the Dooley
measure.

This study employs the indirect method which is also called the residual method, the most widely
used measures in the available literature. It is considered to be the broadest estimate of capital
flight in order to minimize potential biases in narrower measures. The method looks at
measurement under the assumption that capital inflows will be used as a basis of capital outflows.
In other words, capital flight is measured indirectly from balance of payments statistics by
comparing the sources of capital inflows (Net increases in external debt and the net inflow of
foreign investment) with the use of these inflows (the current account deficit and additions to
foreign reserves). See Chapter two for a vivid approach to the measurement of capital flight.

University of Ghana http://ugspace.ug.edu.gh

4.6.2 Explanatory Variables

 Corruption
Corruption the main variable of interest is defined as the perceived levels of public sector
corruption. It captures bribery of public officials, kickbacks in public procurement, embezzlement
of public funds, and questions that probe the strength and effectiveness of public-sector anti-
corruption efforts. Due to the difficulty in measuring corruption, the researcher employed the
Corruption Perceptions Index (CPI) developed by Transparency International as the basic measure
of corruption. The CPI ranks countries by the degree to which corruption is perceived to exist
among public officials and politicians. The scores range between 10 (highly clean) and 0 (highly
corrupt).

Corruption is, thus, a threat to foreign investment because it distorts the environment in both
economics and finance; it also leads to a reduction in the efficiency of government and business
as well as introducing an inherent instability in the political process. Corruption poses a particularly
serious challenge in Sub-Saharan Africa. According to Transparency International’s Corruption
Perception Index, six of the world’s ten countries most burdened by corruption are located in the
Sub Saharan Africa.

 Rule of Law
Rule of Law captures perceptions of the extent to which agents have confidence in and abide by
the rules of society, and in particular the quality of contract enforcement, property rights, the
police, and the courts, as well as the likelihood of crime and violence. The estimate gives the
country's score on the aggregate indicator, in units of a standard normal distribution, i.e. ranging
University of Ghana http://ugspace.ug.edu.gh

from approximately -2.5 to 2.5 (International Country Risk Group). The data are taken from the
International Country Risk Guide published by the Political Risk Services (PRS) Group.

The phenomenon requires the government to exercise its power in line with established and clearly
written rules, regulations, and legal principles. It primarily refers to the influence and authority of
law within society, particularly as a constraint upon behavior, including behavior of government
officials. For the UN, the Secretary-General defines the rule of law as “a principle of governance
in which all persons, institutions and entities, public and private, including the State itself, are
accountable to laws that are publicly promulgated, equally enforced and independently
adjudicated, and which are consistent with international human rights norms and standards. It
requires, as well, measures to ensure adherence to the principles of supremacy of law, equality
before the law, accountability to the law, fairness in the application of the law, separation of
powers, participation in decision-making, legal certainty, avoidance of arbitrariness and procedural
and legal transparency.” (Report of the Secretary-General; the rule of law and transitional justice
in conflict and post-conflict societies” (2004).

 Regime Durability

Regime Durability talks about the number of years since the most recent regime change (defined
by a three point change in the Polity score over a period of three years or less) or the end of the
transition period defined by the lack of stable political institutions (denoted by a standardized
authority score). In calculating the durable value, the first year during which a new (post-change)
polity is established is coded as the baseline “year zero” (value = 0) and each subsequent year adds
University of Ghana http://ugspace.ug.edu.gh

one to the value of the durable variable consecutively until a new regime change or transition
period occurs( Polity IV Database). The data are obtained from the Polity IV data base.

 Independence of the Executive Authority
This is proxied by Executive Constraints (Decision Rules). It indicates the extent to which the
chief executive takes into account the preferences of others when making decisions. The polity IV
dataset considers the extent to which decision rules constrain the executive actions. A seven
category scale is used. Thus, these constraints take values ranging from 1 – strongest constraints,
then the worst institutional quality to 7 – smallest constraints, then the best institutional quality.
The data set is obtained from the Polity IV data base.

 GDP Growth
GDP growth (annual percent) at purchaser's prices is the sum of gross value added by all resident
producers in the economy plus any product taxes and minus any subsidies not included in the value
of the products. The dataset is obtained from the World Bank (WDI). The variable is used as an
indicator of the macroeconomic environment. Capital flight aggravates resource constraints and
contributes to undermining long-term economic growth (UNDP, 2011).

 Inflation
Inflation (annual %) as measured by the consumer price index reflects the annual percentage
change in the cost to the average consumer of acquiring a basket of goods and services that may
be fixed or changed at specified intervals, such as yearly. The dataset is derived from World
University of Ghana http://ugspace.ug.edu.gh

development indicators (2014) of the World Bank. The variable inflation is also used as a measure
of the macroeconomic environment. Inflation is often regarded as an indicator of the government's
overall ability to manage the economy, a rising inflation rate tends to undermine that ability.

 Use of IMF credit
Use of IMF credit denotes members' drawings on the IMF, other than amounts drawn against the
country's reserve tranche position. Use of IMF credit includes purchases and drawings under
Stand-By, Extended, Structural Adjustment, Enhanced Structural Adjustment, and Systemic
Transformation Facility Arrangements as well as Trust Fund loans. SDR allocations are also
included in this category. Data are measured in current U.S. dollars and it is obtained from the
World Bank, International Debt Statistics (2014). IMF loans/credits are meant to help member
countries tackle balance of payments problems, stabilize their economies, and restore sustainable
economic growth. The IMF however, provides consultation and advice to all countries that draw
upon their credit.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER FIVE
ESTIMATION RESULTS AND DISCUSSION
5.1 Introduction
Focusing on the idea of static panel data model with different estimation methods, this chapter
presents the main parts of the empirical results such as the descriptive analysis and econometric
analysis.

5.2 Descriptive Statistics and Analysis
Table 5.1 below provides a summary descriptive statistics relating to thirty two (32) countries in
SSA for the period 2000-2012. The table indicates the summary descriptive statistics of central
tendency and measure of variability. The mean values indicate the average value of the variables
used in the overall model. The standard deviation also captures the distribution of data around the
average value. It also shows the closeness of data to the mean value over the period under
consideration.

More so, the spread of the data is indicated by the range and this is measured by the maximum and
minimum values in each different model. The range is an indicator of the level of variations in the
variables. The larger the range values, the higher the level of variations in a variable and vice versa.
The mean of capital flight of the sample is 1.086 and ranges in value between 37.99 and -25.67.

University of Ghana http://ugspace.ug.edu.gh

Table 5.1: Summary Statistics of Panel data of Sub-Saharan Africa
Variables Mean Std. Dev. Min Max Obs.
Capital Flight Overall 1.085903 4.575747 -25.67407 37.991 N=410
Between 2.032335 -0.1856977 11.11511
Within 4.108826 -35.70327 27.9618

Corruption Overall 2.897595 2.135918 1 37 N=330
Between 1.261311 1.782274 7.372722
Within 1.780682 -1.975126 32.52487

Regime Durability Overall 9.562189 9.623193 0 46 N=402
Between 8.382298 0.1538462 40
Within 4.936809 -6.283965 35.71604

Rule of Law Overall -0.77926 0.6008635 -1.94745 0.66833 N=384
Between 0.5905893 -1.705126 0.6198181
Within 0.1491942 -6.283965 -0.3209315

Ind. Of Exec. Authority Overall 0.7385445 12.43596 -91 95 N=371
Between 2.165533 -0.1666667 7.916667
Within 12.25118 -90.51146 94.48854

Use of IMF Credit Overall 2.568045 17.05824 -54.4619 238.4408 N=384
Between 4.265024 -0.8407249 19.75132
Within 16.53226 -52.81894 221.2575

Inflation Overall 12.35024 36.94877 -9.616154 513.9069 N=393
Between 16.42681 1.938055 78.90683
Within 33.0274 -63.75658 447.3503

GDP growth Overall 4.866867 5.071729 -17.66895 33.73578 N=416
Between 2.569378 -1.844626 9.806402
Within 4.394499 -10.95745 30.4258
Source: Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF; CPI of TI, WB-IDS, ICRG, and Polity IV data.

Over the period 2000-2012, the average capital flight for the thirty two (32) SSA countries under
study averaged 1.086 million constant US dollars ranging from a maximum score of 37.99 and -
25.67 showing high level of variations. The range of the corruption values (the main value of
University of Ghana http://ugspace.ug.edu.gh

interest) indicates that between countries observations in the region attain as low as 1.78 and as
high as 7.37 scores within the period under consideration, whereas within countries observation
shows a wider variation ( low = -1.97 and high= 32.52). The corruption index averaged 2.89.
Empirical studies on capital flight in SSA, is determined by the value of the corruption perception
index. From the table (5.1) above, the maximum value of this variable observed is 37 with
minimum of 1.

5.3 Results of Unit root test for stationarity
Although the Unit root test is usually regarded as a time series problem, conducting unit root tests
in panel datasets could also be appropriate in order to ensure that the variables under study are
stationary. Gujarati (2003 pp. 713) stated that, “ a stochastic process is said to be stationary if its
Mean, and Variance are constant over time and the value of Covariance between two time periods
depends only on the distance between the two time periods and not on the actual time at which the
Covariance is reported”.

The research carried out the panel unit-root test before proceeding to the estimation. This is
intended to prevent any distortions in estimated regression relations as well as spurious regression
due to time-series process (Greene, 2012). The researcher used the Augmented Dickey-Fuller
(ADF) test for unit-root on all the variables. The unit-root test is actually undertaken on the Null
Hypothesis, which states that all panels contain unit-root (meaning all panels are non-stationary).

University of Ghana http://ugspace.ug.edu.gh

Table 5.2: Augmented Dickey-Fuller tests (System GMM)
Variable
Statistic
Probability
Value Lag Structure Value
Capital Flight Level 7.6549 0.0000
Corruption First Diff 12.5794 0.0000
Rule of Law Level 9.9873 0.0000
Regime Durability Level 3.4162 0.0003
Ind. of Exec. Auth. Level 9.7493 0.0000
GDP Growth Level 17.89 0.0000
Inflation Level 58.2851 0.0000
Use of IMF Credit First Diff. 2.0148 0.0220
NB: Statistical values reported are based on the Modified inv. chi-squared Pm
Source: Author’s Computation 2015

The results of the Unit root test shown in table 5.2 depicts the results of the Augmented Dickey-
Fuller (ADF) test for unit-root, which indicates that we reject the null hypothesis for tests in all the
variables implying that all variables were found stationary.

5.4 Results of Granger Causality Test between capital flight and corruption
The basic concept of Granger causality can be traced back to Wiener. Granger proposed the idea
in his 1969 paper to describe the ‘causal relationships’ between variables in econometric models.
The idea of Granger – causality is that a variable X Granger-causes variable Y if variable Y can
be better predicted using the histories of both X and Y than it can be predicted using the history of
Y alone. Thus, in line with most of the literature in Econometrics, one variable is said to Granger-
University of Ghana http://ugspace.ug.edu.gh

cause the other if it helps to make a more accurate prediction of the other variable than had we
only used the past of the latter as a predictor.

Another element to define is that of the feedback effect. This occurs if variable X Granger-causes
variable Y, and Y Granger-causes X, denoted X↔Y. In testing, we indicate the null hypothesis
that Xt does not Granger-cause Yt and Yt does not Granger-cause Xt. In each case, a rejection of
the null hypothesis implies there is Granger causality between the variables. All variables in the
model have been tested for stationarity.

Table 5.3 Pairwise Granger Causality Tests on Capital flight and Corruption
Sample: 1416
Lags: 2
Null Hypothesis: Observations F-Statistic Probability
Corruption does not Granger Cause Capital flight 106 3.1263 0.0482

Capital flight does not Granger Cause Corruption 2.80521 0.0652
Source: Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI, and GDF; CPI of TI

The table 5.3 depicts the results of the Granger Causality tests conducted to understand the
interrelationships between the respective variables under consideration. The results indicate a uni
- directional causal relationship running from corruption to capital flight at the 5% significance
level.

University of Ghana http://ugspace.ug.edu.gh

5.5 Empirical estimation and discussions
This section presents the estimation results of the explanatory variables using the Pooled Ordinary
Least Squares regression, Fixed Effects – GLS regression and the System GMM estimation. Tables
(5.4), (5.5), and (5.6) display the regression results for the OLS, FE, and GMM respectively.
Diagnostic test results are also reported. Eight (8) specifications of the econometric model were
tested. The main variable of interest in this case corruption entered the model in specification (1)
and it is included in all the other specifications. Specification (2) captures corruption and regime
durability and their interaction. Also, specification (3) captures corruption and rule of law and their
interactions whereas specification (4) is made up of corruption and the independence of the
executive authority and their interaction.

Moreover, specifications (5), includes the full model and the interaction between corruption and
regime durability. Specification (6) on the other hand, comprises the full model (thus, all the
explanatory variables) and the interaction between corruption and rule of law whereas
specification (7) involves the interactions between corruption and independence of the executive
authority and the full model. Finally, the last column indicates specification (8) which consists of
the full model that includes all the variables with the exception of the interactive variables to
determine their combined significance.

5.6 Pooled – OLS Estimation Results
Estimated results reported in table 5.4 indicate that, although the main variable of interest, thus,
corruption is statistically significant to capital flight in most of the specifications it appeared, it did
not retain the expected sign in the first two specifications.
University of Ghana http://ugspace.ug.edu.gh

Table 5.4 Pooled – OLS Regression Results (2000-2012)

Variables (1) (2) (3) (4) (5) (6) (7) (8)
COR -0.1794* -0.376* 0.385** -0.0973 0.532* 0.827*** 0.853*** 0.837***
(0.1048) (0.206) (0.201) (0.1151) (0.311) (0.242) (0.243) (0.2393)

RL -1.498*** -1.46*** -1.72** -1.62*** -1.59***
(0.525) (0.385) (0.5612) (0.386) (0.3777)

COR*RL 0.126 0.066
(0.174) (0.1975)

RD -.0805** -0.703* -0.0202 -0.1899 -0.0183
(0.364) (0.374) (0.1688) (0.1595) (0.0158)

COR*RD 0.0191** 0.138
(0.0086) (0.009)

I.E.A 0.0252 -0.033 -0.0024 -0.0374 -0.002
(0.0911) (0.009) (0.092) (0.088) (0.0091)

COR*IEA -0.1062 0.0153
(0.396) (0.038)

UIC 0.012** 0.0109* 0.011* 0.0107*
(0.057) (0.0573) (0.057) (0.0057)

Inflation 0.5933* 0.673* 0.6996** 0.699**
(0.3375) (0.3411) (0.332) (0.3315)

GDP growth 0.0422* 0.0385 0.0374 0.0386
(0.253) (0.0253) (0.0255) (0.0252)
Note: The dependent variable is capital flight; the figures in parenthesis are the Standard Errors of the
estimates, where ***, **, * represent the statistical significance of the estimates at 1%, 5% and 10% level
of observations.
Source:Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF; CPI of TI, WB-IDS, ICRG, and Polity IV data.

The research can confidently say that, corruption and capital flight are related positively.
University of Ghana http://ugspace.ug.edu.gh

One significant feature worth noting in the above results is that an important institutional variable,
thus rule of law is shown to be negative and statistically significant at the 1 % level of significance
of capital flight in all the specifications it is included. This implies that in a context of
institutional/governance parameters, good governance and strong institutions facilitate a fall in
capital flight in Sub-Saharan Africa.

Accordingly, independence of the executive authority and regime durability did retain their
expected sign, but the former is statistically insignificant to capital flight in all the specifications
it occurred. Also, the interaction between corruption and regime durability is positive and
statistically significant at the 5 % level, implying that an increase in regime durability will
necessitate a fall in corruption and a repelling effect to declining capital flight in SSA.The reason
is that regime durability is identified in the literature as a source of institutional quality. However,
the interaction between corruption and the rule of law, as well as between corruption and the
independence of the executive authority is statistically insignificant in all specifications they
occurred.

Finally. The macroeconomic indicators, controlled in this thesis did not retain their expected signs
with the exception of inflation, which is positive and statistically significant to capital flight in all
the specifications it appeared. The study employed the fixed effects model due to the fact that the
pooled-OLS model is believed not to be efficient in doing panel data analysis.

University of Ghana http://ugspace.ug.edu.gh

5.7 Fixed Effects Estimation Results
The table 5.5 depicts the fixed effects regression results. Three specifications of the econometric
model are tested using the fixed-effects model to control for the country-specific characteristics.
The appropriateness of the fixed-effects model for all regression specifications is confirmed by the
Hausman test after running it on the null hypothesis that the preferred model is Random Effect to
the alternative the Fixed Effects. The test is to find out whether the unique errors are correlated
with the regressors, the null hypothesis is they are not. The results of the Hausman test are
displayed in Appendix C. In addition, the study finds that heteroscedasticity and autocorrelation
are a severe problem associated with the panel model, hence; the study employs the robust model.
This approach has been used by Eshete (2005) and Daniel Hoechle (2007).

Results from the fixed effect model suggest that, in all the specifications, the variable of interest,
in this case corruption, was found to be explaining the occurrence of capital flight in Sub-Saharan
Africa. The coefficient is positive and statistically significant at predominantly 1% level of
significance, indicating that economies in Sub-Saharan Africa region with high corruption tend to
experience more capital flight. The result conforms to empirical findings reported by Le and Rishi
(2008), who considered the role of corruption in impelling capital flight in developed and
developing countries using a panel data analysis, they reiterated a positive and significant effect of
corruption on capital flight.

However, the empirical results show a statistically insignificant interactions between corruption
and rule of law, between corruption and regime durability, and between corruption and the
independence of the executive authority.
University of Ghana http://ugspace.ug.edu.gh

Table 5.5 Fixed Effects – GLS Regression Results (2000-2012)
Variables (1) (2) (3) (4) (5) (6) (7) (8)
COR 1.2662** 1.194*** 1.158*** 1.369*** 1.619*** 1.245*** 1.485*** 1.466**
(0.2221) (0.2535) (0.3303) (0.213) (0.233) (0.2192) (0.202) (0.2454)

RL 0.672 -0.8395 0.113 -0.8265 -0.8236
(1.185) (1.015) (1.0412) (1.0314) (0.7339)

COR*RL -0.434 -0.388
(0.397) (0.3684)

RD 0.481 0.979 0.644* 0.0599 0.0614*
(0.069) (0.064) (0.361) (0.3699) (0.0274)

COR*RD 0.0035 -0.1234
(0.0165) (0.1455)

I.E.A -0.045 0.0047 0.0048 -0.4324 0.0045
(0.3176) (0.0055) (0.0057) 0.0299 (0.0063)

COR*IEA 0.0212 0.207*
(0.0125) (0.012)

UIC -0.0036 -0.004 -0.0032 -0.0033
(0.003) (0.003) (0.0028) (0.0041)

Inflation 0.1927 0.217 0.178 0.1711
(0.2465) (0.236) (0.2557) (0.3113)

GDP growth -0.0156 -0.0134 -0.158 -0.0147
(0.0223) (0.21) (0.218) (0.0200)
No. of Observations: 223 215 215 204 189 189 189 189
Number of Groups: 32 31 32 31 31 31 31 31
R-squared 0.1317 0.1794 0.1733

0.1596 0.2671 0.2690 0.2677 0.2648
Hausman Spec. Test 0.0000 0.0000 0.0032 0.0000 0.0000 0.0000 0.0000 0.0000

Note: The figures in parenthesis are the Robust Standard Errors of the estimates, whereas ***, **, *
represent the statistical significance of the estimates at 1%, 5% and 10% levels of observations.
Source:Ndikumana and Boyce (2012), and data sample expanded using data from IMF, IFS,
DOTS, WDI and GDF; CPI of TI, WB-IDS, ICRG, and Polity IV data.
University of Ghana http://ugspace.ug.edu.gh

This implies that institutional and governance interactions do not play any crucial role in
explaining the influence of corruption on capital flight using the fixed effect model.

Moreover, the controlled macroeconomic variables of inflation, GDP growth and use of IMF credit
showed a statistically insignificant effect in all the specifications they were included. This result
indeed explains that the macroeconomic indicators used in this analysis play no important role in
explaining capital flight from SSA. Due to the presence of potential endogeneity, the thesis work
did replicate the above econometric exercise using the System GMM as presented in table 5.6.

5.8 System GMM Estimation Results
Following the empirical estimation and discussion of the pooled-OLS and the fixed effect
regression results, the study suspected the presence of autocorrelation, heteroscedasticity and the
potential endogeneity problem. As part of the robustness check, the system GMM is used to
undertake these diagnostic tests to ensure that the data fits the model and that the results from the
system GMM estimation are valid and reliable.

The test for autocorrelation for this study is carried out using the Arellano-Bond test for
autocorrelation in the first difference errors. The test fails to reject the null hypothesis of no
autocorrelation in the second order AR (2) for all the regressions indicating that the error terms are
not correlated with each other i.e. MA (1) process. For this condition of no autocorrelation,
employing the System GMM estimation is satisfied. The test results are reported together with the
estimation results in a table (5.6)

University of Ghana http://ugspace.ug.edu.gh

Table 5.6 System GMM dynamic panel estimation result (2000-2012)
Variables (1) (2) (3) (4) (5) (6) (7) (8)
COR 0.7543** 1.561*** 0.992*** 1.207*** 2.9083** 0.8663** 0.9687*** 0.9007**
(-0.3666) (0.509) (0.278) (0.194) (1.344) (0.4284) (0.371) (-0.3879)

RL -0.362 -6.083** -1.273 -1.645** -1.514**
(0.890) (2.704) (1.027) (0.6825) (-0.6373)

COR*RL 0.127 -0.3026
(0.195) (0.422)

RD 0.410 0.3393** -0.0125 -0.0864** -0.0452
(0.961) (0.1666) (0.498) (0.0479) (-0.0499)

COR*RD -0.0248 -0.098**
(0.338) (0.046)

I.E.A 0.327 -0.012 0.0123** -0.8789 0.0127*
(0.2006) (0.0077) (0.0062) 0.8062 (-0.0072)

COR*IEA -0.136 0.3922
(0.888)) (0.355)

UIC -0.0009 -0.0008 -0.0013 -0.0013
(0.0024) (0.0014) (0.001) (-0.0012)

Inflation 0.1531 -0.673 -0.7466 -0.70**
(0.73) (0.419) (0.4199) (-0.3537)

GDP growth 0.0065 -0.014 -0.0157 -0.0092
(0.0169) (0.0109) (0.0124) (-0.0106)
No. of Observations: 100 97 96 97 72 138 138 138
Number of Groups: 25 24 25 24 18 31 31 31
No. of Instruments: 64 25 65 65 62 80 80 79
AR1 0.3455 0.0193 0.0547 0.0085 0.0201 0.0287 0.0244
AR2 0.1415 0.2686 0.1174 0.1465 0.1687 0.1133 0.1438
Sargan 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
Note: The dependent variable is capital flight; the figures in parenthesis are the Standard Errors of the
estimates, whereas ***, **, * represent the statistical significance of the estimates at 1%, 5% and 10%
levels of significance. Source:Ndikumana and Boyce (2012), and data sample expanded using data
from IMF, IFS, DOTS, WDI and GDF; CPI of TI, WB-IDS, ICRG, and Polity IV data.
University of Ghana http://ugspace.ug.edu.gh

Accordingly, the System GMM estimation technique requires that the instruments used must be
valid in order to obtain consistent and efficient estimates. As a result, the Sargan test of the over-
identifying restrictions, which tests the null hypothesis that the over-identifying restrictions are
valid is employed. The test as reported in Table 5.6 fails to reject the null hypothesis and this
provides strong evidence that the instruments are valid for the regression model.

Table 5.6 reports the results of the system GMM on the relationship between capital flight and
corruption. The research found that corruption is positively related to capital flight and statistically
significant across all the specifications from Sub-Saharan African countries. This actually means
that a higher perception of corruption among public authorities as in bribery of public officials,
kickbacks in public procurement, embezzlement of public funds and among others facilitates an
increase in illegal outflow of capital from Sub-Saharan African countries.

The results confirm that of Le and Rishi (2008); and Ndiaye (2009), who reported a positive and
significant effect of corruption on capital flight. The coefficient of corruption in specification (8)
thus, the full model in particular implies that a rise in corruption by one standard deviation is
associated with a rise in capital flight in Sub-Saharan Africa by 0.9007 million of constant US
dollars. This result indicates that the threat corruption poses to capital flight is very huge and severe
in SSA.

Moreover, the results indicated that the interaction between corruption and regime durability in
specification (5) depicts a negative and statistically significant coefficient. This result shows that
University of Ghana http://ugspace.ug.edu.gh

a reduction in the number of years since the most recent regime change or the end of a transition
period, will lead to a decline in corruption and subsequently a decline in capital flight. This is
consistent with the empirical results of Ali and Walters, (2011) who in their study of the causes of
capital flight in SSA using a panel data set of 37 countries between 1980 and 2005, found that
regime durability is associated with higher capital flight since it may be linked to weaker
governance, high corruption, and poorer domestic investment climate. It is therefore evident that
regime durability is crucial in explaining the effect of corruption on capital flight. However, the
interaction between corruption and the rule of law and between corruption and the independence
of the executive authority in the various specifications they appeared revealed a statistically
insignificant effect.

In addition, the other equally institutional variables (regime durability, independence of the
executive authority) especially the rule of law in most of the specifications it appeared indicate a
negative and statistically significant coefficient at 5% level of significance. This is consistent with
existing empirical evidence that good institutional development is associated with a lower
incidence of capital flight (see, for example, Lensink et al., 2000; Collier et al., 2004). On the other
hand, the controlled macroeconomic variables in all the specifications they were included depicted
statistically insignificant effect on capital flight with the exception of inflation but even that it
failed to retain the expected positive coefficient. This result tells us that the macroeconomic
environment does not have any significant impact on capital flight in Sub-Saharan Africa.

University of Ghana http://ugspace.ug.edu.gh

5.9 Synthesis of the Results

In order to appreciate better the idea of the various techniques used to estimate the model on the
effects of corruption on capital flight, this section compares the results of the system GMM and
the results of the fixed effects-GLS and the Pooled-OLS. The results from these three estimation
techniques shows both similarities and differentials in all the specifications produced. An
important feature worth noting is that the results from the system GMM and that of the fixed effects
regression shows that corruption, which happens to be the main variable of interest was positive
and predominantly statistically significant at the 1 % level of significance in all the specifications
it appeared. Even in the pooled-OLS regression, corruption depicted a positive, statistically
significant effect on capital flight in most of the specifications it occurred. As already indicated
this results tell us that corruption plays an essential role in explaining the phenomenon of capital
flight in SSA.

Another similarity is that, all the controlled macroeconomic indicators except inflation (use of
IMF credit, and GDP growth) used in this study did not retain their expected signs and was also
statistically insignificant especially in the system GMM model and fixed effect – GLS regression
approach. However, inflation is statistically significant and positively related to capital flight in
the pooled-OLS regression. This is the case because inflation leads to the lowering of the returns
to non-indexed assets and increases the opportunity cost of holding money. This is consistent with
the Portfolio-choice theory which suggests that capital flight is driven by relative risk-adjusted
expected return.

University of Ghana http://ugspace.ug.edu.gh

Accordingly, the various interactions between corruption and the other equally important
institutional and governance variables used in the study reveals some similarities among the
estimation techniques used in the estimation. Specifically, the interaction between regime
durability and corruption showed statistically significant results in both the system GMM and the
Pooled OLS. Again, the interaction between corruption and the rule of law is found to be
statistically insignificant in all the estimation techniques used and across all specifications of the
models. More so, the interaction between corruption and the independence of the executive
authority indicated an insignificant statistical result in both the system GMM and the pooled-OLS
estimation techniques.

However, even though the similarities of the results seem to be predominant across all the three
estimation techniques, the system GMM estimator is preferred to the other two techniques since it
allows correcting for insufficiencies related to problems of simultaneity bias, inverse causality and
omitted variables (Kpodar, 2005).

University of Ghana http://ugspace.ug.edu.gh

CHAPTER SIX
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
6.1 Introduction

The chapter provides a summary of the study. Based on the findings of the study, this chapter also
draws conclusions and make recommendations for policy analysis. The study aimed at finding the
effect of institutional governance on capital flight in Sub-Saharan Africa, focusing on the role of
corruption. Panel data set of thirty two (32) countries in Sub-Saharan Africa is analyzed over the
period 2000-2012. The portfolio choice frame work is used for the theoretical model. Also, three
different estimation techniques as Generalized Method of Moments (GMM), Fixed Effect
Regression and the pooled-OLS regression models are used.

6.2 Summary and Conclusion
Sub-Saharan Africa is home to a young and growing population, which is experiencing the fastest
population growth in the world and the region as a whole is expected to be the second fastest-
growing globally (behind Asia Pacific) achieving its fastest rate of growth since 2010. However,
the region has experienced massive outflows of private capital towards western financial centers.
These private assets surpass the continent’s foreign liabilities, ironically making Sub-Saharan
Africa a ‘net creditor’ to the rest of the world (Boyce and Ndikumana, 2001). For instance, the
report, entitled “economic perspectives in Africa”, was released in 2012. In it, the African
Development Bank (AfDB) asserts that capital flight deprived the continent of over US $700
billion during the preceding decade. The high levels of capital flight pose serious challenges for
domestic resource mobilization in support of investment and growth in Africa.
University of Ghana http://ugspace.ug.edu.gh

Most studies on capital flight as observed in the empirical literature is skewed towards the cause
and determinants with limited studies emphasizing on corruption as a key factor particularly in
Sub-Saharan Africa. Hence, the main objective of this study is to establish the relationship between
capital flight and corruption in Sub-Saharan Africa. The study was based on the Portfolio Choice
framework and engaged three different estimation techniques, thus, the system GMM, fixed effects
regression, and the pooled-OLS regression based on data of thirty two (32) Sub-Saharan African
Countries for the period 2000 – 2012. The GMM estimation technique is employed in the empirical
estimation procedure in order to achieve the set objectives of the study. The technique is preferred
to other techniques (pooled - OLS and Fixed Effect regression) used because it is perfect for a
Dynamic model that exhibit a number of sensitive characteristics such as endogeneity, fixed
effects, heteroscedasticity, serial correlation and other data challenges.

An estimation of capital flight on corruption as well as other controlled macroeconomic and
equally important institutional variables of the model was undertaken. Accordingly, the variable
of interest, thus corruption regressed on capital flight retain its expected positive sign and
statistically significant across all specifications of the model. The relationship remains robust even
when macroeconomic variables and equally other institutional variables that have been used in
empirical literature and significantly affect capital flight are taken into account. The result was
consistent with empirical evidence. However, most controlled macroeconomic variables entered
the model insignificantly. This provides us enough evidence to conclude that poor governance and
weak institutions, in this case an increase in corruption strongly leads to a rise in capital flight in
Sub-Saharan Africa.

University of Ghana http://ugspace.ug.edu.gh

Moreover, interactive variables regressed also revealed interesting results, including a statistically
significant positive coefficient between corruption and regime durability indicating that regime
durability is an important institutional variable that play a critical role in explaining the effect of
corruption on capital flight in SSA. However, the interaction between corruption and the rule of
law, as well as corruption and independence of the executive authority show statistically
insignificant result across all the specifications in the model. Finally, from aforementioned
discussions, this research work has enough evidence to conclude that corruption has a positive
effect on capital flight. Also, this thesis indicate that there is the need for stronger institutions and
good governance to help reduce capital flight in SSA.

6.3 Recommendations and Policy Implications
With regards to the estimation results of the study, the following measures are recommended.

 Institutional and macroeconomic reforms are encouraged to be undertaken in order to reduce
corruption and by extension capital flight. The governments of Sub-Saharan African countries
must put in place measures to improve the control of corruption in SSA. Thus, an effective
mechanism aimed at tracking and prosecuting financial crime should be the utmost priority of
the authorities.

 Also, leaders in Sub-Saharan Africa should establish better political environments that will
make them desist from an oppressive apparatus that suppressed demand for political opening.
They should create a democratic environment linked to stronger governance, decline in
corruption and better domestic investment climate.
University of Ghana http://ugspace.ug.edu.gh

 Last but not the least, a common development agenda needs to be instituted among member
countries to streamline accountability and development projects. This will help reduce the level
of corruption in the economy and its effect on capital flight.

6.4 Limitation of the study and areas of further research
This research is conducted to identify the effects of corruption on capital flight in Sub Saharan
Africa for the period 2000-2012.

As a matter of fact, this study cannot exhaust all the issues pertaining to the topic under study, as
no single study is exhaustive enough. Notwithstanding this, however, it will be enough for the
academic purpose for which it is being carried out. The number of Sub-Saharan African Countries
considered were 32 due to the problem of data availability. Time constraints are a limiting factor
during the study since the researcher has a short period at his disposal because the study is also
carried out alongside other equally important academic commitment.

Future research should be geared towards the use of other estimation techniques, for instance the
Structural Vector Autoregressive model to test for the effects of institutions and governance
parameters on capital flight in SSA.

University of Ghana http://ugspace.ug.edu.gh

References
Abdilahi Ali and Bernard Walters (2011), On the causes of Capital Flight from Sub-Saharan
Africa, University of Manchester,
Acemoglu Daron, Simon J., James R., and Yunyong T., (2003), Institutional causes,
macroeconomic symptoms: volatility, crises and growth, Journal of Monetary Economics
Ajayi ibi (1992), “An Econometric Analysis of Capital Flight from Nigeria”. World Bank
Working Paper number 993.
Ajayi, I.S. (2007). “Dynamics of capital flight: Causes and determinants”. Paper prepared for
the Senior Policy Seminar on Capital Flight from Sub-Saharan Africa: Implications for
Macroeconomic Management and Growth, Pretoria, South Africa, 30 October to 2
November 2007.
AK Jain (2001), Corruption: A review- Journal of economic surveys, - Wiley Online Library
Alesina, A. and G. Tebellini (1989), external debt, capital flight and political risk, journal of
international economic 27 (3/4), 16-44.
Artadi, E. V. and Sala-I-Martin, X. (2003). The economic tragedy of the XXth century: growth
in Africa. NBER Working Paper Series, Nr. 9865.
Azfar, O., & Murrell, P. (2005). Identifying Reticent Respondents: Assessing the Quality of
Survey Data on Corruption and Values. Working Paper, University of Maryland.
Bardhan, P. (1997). Corruption and development: A review of issues. Journal of Economic
Literature, vol. 35, p. 1321.
Boyce, J. K. and L. Ndikumana (2001) “Is Africa a Net Creditor? New Estimates of Capital
Flight from Severely Indebted Sub-Saharan African Countries, 1970-1996”, Journal of
Development Studies, 38(2).
University of Ghana http://ugspace.ug.edu.gh

Cameron, L., Chaudhuri, A., Erkal, N., & Gangadharan, L. (2005). Do Attitudes Towards
Corruption Differ Across Cultures? Experimental Evidence from Australia, India,
Indonesia, and Singapore? Working Paper No. 943. University of Melbourne.
Capital Flight and Corruption Treat, Non- Governmental Organization Alternative Treaties at the
1992 Global Forum.
Cerra Valerie, Meenakshi Rishi,and Sweta Saxena, (2008), Robbing the Riches: Capital Flight,
Institutions, and Instability, IMF Working Paper 05/199
Cervena, M. (2006). The Measurement of Capital Flight and Its Impact on Long-term Economic
Growth: Empirical Evidence from a Cross-section of Countries. Unpublished Master’s
thesis, Comenius University Bratislava.
C Kant (1996), Foreign direct investment and capital flight- princeton.edu
C Iversen (1935), Aspects of the theory of international capital movements.
Cho, Wonbin and Matthew F. Kirwin. “A Vicious Circle of Corruption and Mistrust in
Institutions in sub-Saharan Africa: A Micro-level Analysis.” 2007.
Collier, P. and D. Lai (1986), Labour and Poverty in Kenya 1900-1980, Oxford: Clarendon
Press.
Collier, Paul, Anke Hoeffler and Catherine Pattillo (2001) "Flight capital as a portfolio choice,"
World Bank Economic Review, 15, 1,55-80
Collier, P., A. Hoeffler et C. Pattillo (2004a) “Africa’s Exodus: Capital Flight and the Brain
Drain as Portfolio Decisions”, Journal of African Economies, 13(1), ii15-ii54(1).
Collier, P. (2006). African growth: why a .big push.? Journal of African Economies, (15),
Supplement 2, 188.211.
University of Ghana http://ugspace.ug.edu.gh

Cuddington, John T. (1986). Capital flight: Estimates, issues, and explanations. Princeton
University, N. J., Department of Economics, International Finance Section. Princeton
Studies in International Finance, no. 58.
Daniel Kaufmann, Aart Kraay, and Pablo Zoido-Lobatón (1999 a), “Aggregating Governance
Indicators,” World Bank Policy Research Working Paper No. 2195 (Washington),
www.worldbank.org/wbi/ governance. —— (1999 b), “Governance Matters,” World
Bank Policy Research Working Paper No. 2196 (Washington),
www.worldbank.org/wbi/- governance.
D Green, S Branford –(2012), Faces of Latin America: (Revised)- books.google.com
D Rodrik (2008), The real exchange rate and economic growth- Brookings papers on economic
activity, - muse.jhu.edu
Driscoll, J. C., and A. C. Kraay. (1998). Consistent Covariance Matrix Estimation with Spatially
Dependent Panel Data. Review of Economics and Statistics 80: 549–560.
Dooley, M. P. (1986) “Country-Specific Risk Premiums, Capital Flight and Net Investment
Income Payments in Selected Developing Countries”,Washington DC: IMF
Dooley, M. P., Kletzer, K. M. (1994). Capital Flight, External Debt and Domestic Policies.
National Bureau of Economic Research. (http : //www.nber.org/papers/w4793.pdf)
Economist Intelligence Unit (1999) The Russian Federation, Country Report (London EIU)
Edwards, S. (1989), ‘Capital Flows, Foreign Direct Investment, and Debt-Equity-Swaps in
Developing Countries’, NBER Working Paper No. 3497, National Bureau of Economic
Research.
EV Artadi, X Sala-i-Martin (2003), The economic tragedy of the XXth century: growth in
Africa- nber.org
University of Ghana http://ugspace.ug.edu.gh

Fedderke, J. W. and Liu. W. (2002). Modeling the determinants of capital flows and capital
flight: With an application to South African data from 1960 to 1995. Economic Modeling,
19, 419-444.
F Ghazi – (2014) Corruption and Growth- etd.ohiolink.edu
Fofack, Hippolyte, and Leonce Ndikumana (2009) "Potential Gains from Capital Flight
Repatriation for sub-Saharan African Countries," Policy Report Working Paper 5024
(Washington: World Bank)
Frait, J., Komarek, L. and M. Melecky, M. (2006). The real exchange rate misalignment in the
Central European countries. Warwick Economic Research Paper No. 739, University Of
Warwick.
FS Ayadi (2008), Econometric analysis of capital flight in developing countries: A study of
Nigeria- 8th Global Conference on Business & - theglobaljournals.gcbe.us
Granger, C.W.J (1969). Investigating Causal Relations by Econometric Models and Cross
Spectral Methods. Econometrics. 37:424-35.
Grange, C.W.J (1980). Testing for Causality. A personal Viewpoint. Journal of Economic
Dynamic and Control 2(4), 329 – 352
Gibson, H. & Tsakalotos, E. (1993) Testing a flow model of capital flight in five European
countries, Man- chester School of Economic and Social Studies, 61, 144–166
Hermes, N., R. Lensink and V. Murinde (2002) “Flight Capital and its Reversal for Developing
Financing”, UNU/WIDER, Discussion Paper No. 2002/99.
Hirst, P. and G. Thompson (1996), Globalisation in Question: The International Economy and
the Possibilities of Governance, Cambridge, MA: Polity Press.
Hoechle, D. (2007). Robust standard errors for panel regressions with cross-sectional
dependence. Stata Journal, 7(3), 281.
University of Ghana http://ugspace.ug.edu.gh

Hope, K.R. Sr. (1997) African Political Economy: Contemporary Issues in Development,
London: M.E. Sharpe.
Hunt, C.L. and Walker, L. (1974) Ethnic Dynamics: Pattern of Intergroup Relations in Various
Societies, Homewood, Illinois: The Dorsey Press.
Johnston, M. (1998), "Fighting Systematic Corruption: Social Foundations for Institutional
Reform", in M. Robinson (ed.), Corruption and Development, London: Frank Cass.
J Tobin (1958), Liquidity preference as behavior towards risk- The review of economic studies
Johnson and Robinson (2000), Institutions and Geography: Comment on Acemoglu
Kaufmann, D., Kraay, A. & Zoido-Lobato ´n, P. (1999) Governance matters, World Bank Policy
Research Working Paper No. 2195, Washington, DC.
Kaufmann, D. (2003). “Governance Crossroads.” Global Competitiveness Report 2002-2003.
Khan, Mohsin S. and Haque, Nadeem Ul. (1987), Capital flight: An examination of this
phenomenon and the issues it raises. Finance and Development (IMF). Washington D.C.,
Mar., vol. 24, no. 1, pp. 2-5.
Lan, S.K. (2009). Effects of Capital Flight on Economic Growth in Selected ASEAN Economies.
Unpublished PhD thesis, University Putra Malaysia.
Le Quan V, Rishi M (2006) Corruption and capital flight: an empirical assessment. International
Economic Journal 20: 523 - 540.
Le, Q.V. & Zak, P.J. (2006) Political risk and capital flight, Journal of International Money and
Finance, 25, pp. 308–329
Lensink, R. Hermes, N & Murinde, V. (2000) Capital Flight and Political Risk, Journal of
International Money and Finance WIDER, Helsinki
University of Ghana http://ugspace.ug.edu.gh

Lensink, R. & Murinde, V. (2002) Flight capital and its reversal for development financing,
Discussion Paper No. 2002/99
Leonard, D.K. (1991) African Successes: Four Public Managers of Kenyan Rural Development,
USA: University of California Press.
Lessard, D.R., and J. Williamson, eds. (1987). Capital flight and Third World debt. Washington,
DC: Institute for International Economics.
Loungani, P. & Mauro, P. (2000) Capital flight from Russia, World Economy, 24(5), pp. 686–
709.
Machlup, Fritz. (1942). Theory des Kapi- talflucht. Welwirtschaftliches Archiv. Vol. 36, pp. 512-
520.
Mikkelsen, Jan Giehm. (1991), An econometric investigation of capital flight. Applied
Economics. Vol. 23, issue 1, pp. 73- 85
Morgan Guaranty Trust Company (1986) “LDC Capital Flight”, World Financial Markets, 2
M Pastor (1990), Capital flight from Latin America- World development, - Elsevier
Mulinge, M. M. and G. N. Lesetedi (1998), "Interrogating Our Past: Colonialism and the Birth
and Entrenchment of Corruption in sub-Saharan Africa", African Journal of Political
Science, Vol. 3, No.2, 15-28.
Mulinge, M. M. and G. N. Lesetedi (1999), "The Genesis and Entrenchment of Corruption in
sub-Saharan Africa: A Historical and International Contextualization" in K. Frimpong
and G. Jacques (eds.), Corruption, Democracy and Good Governance in. Africa: Essays
on Accountability and Ethical Behavior, Gaborone: Lentswe La Lesedi Publishers.
Mulino, M. (2002) On the determinants of capital flight from Russia, Atlantic Economic Journal,
30(1), pp. 148–169
University of Ghana http://ugspace.ug.edu.gh

Ndiaye, Ameth Saloum (2011) "Capital Flight and its determinants in the franc Zone," SERC
Research Paper 215 (Nairobi: AERC)
Ndikumana and Boyce, (2003), Public Debts and Private Assets: Explaining Capital Flight from
Sub-Saharan African Countries, World Development 31 (1)
Ndikumana, L. and J. K. Boyce (2007) “New Estimates of Capital Flight from Sub-Saharan
African Countries: Linkages with External Borrowing and Policy Options”, Paper
prepared for the Senior Policy Seminar on “Capital Flight from Sub-Saharan Africa:
Implications for Macroeconomic Management and Growth”, October 30-November 2,
2007, Pretoria, South Africa
Ndikumana, L. and H. Fofack (2008) “Capital Flight Repatriation: Investigation of its Potential
Gains for Sub-Saharan African Countries”, Paper presented at the 2008 African
Economic Conference, November 2008, Tunis, Tunisia.
Ndikumana, Léonce and James K Boyce (2010) "Measurement of capital flight: methodology
and results for sub-Saharan African countries," African Development Review 22, 4, 471-
81
Ndikumana, Léonce and James K Boyce (2012) "Measuring Capital Flight," All Africa
(http://allafrica.com)
Nicholson, A. (2005) Russian Capital Flight More than Triples (Associated Press)
http://www.ocnus. net/ cgi-bin/exec/view.cgi?archive¼61&num¼15922.
North, D. C. “Institutions, Institutional Change and Economic Performance.” Washington:
Cambridge University Press, 1990
Osoba, S.O. (1996) "Corruption in Nigeria: Historical Perspectives", Review of African Political
Economy, 69: 371-386.
P Mauro (1995) “Corruption and growth”. The quarterly journal of economics - JSTOR
Robb, G. (1992), White-collar Crime in Modern England, London: Cambridge University Press.
University of Ghana http://ugspace.ug.edu.gh

Roberts, C. (1962) "The Sub-Imperialism of the Baganda", Journal of African History, 3 (3):
435-450.
Rodrik, D. (2008). The real exchange rate and economic growth. Brookings Papers on Economic
Activity, 2, 365.412.
Schineller, L. (1997) A nonlinear econometric analysis of capital flight. International Finance
Discussion Paper, (Washington, DC: Board of Governors of the Federal Reserve
System).
Schneider, B. (2003). Measuring Capital Flight: Estimates and Interpretations. Working Paper.
Overseas Development Institute.
Seid, R. (1996) Corruption and capital flight rob Mexico of needed funds, Christian Science
Monitor
Seung (2010), Institutional Quality, Capital Flight: Estimates And Capital Flows, The Korean
Economic Review, Volume 26, Number 1, Summer 2010
S Edwards (1989), Exchange rate misalignment in developing countries- The World Bank
Research Observer, - World Bank
Sheets, N. (1995). Capital flight from the countries in transition: some theory and empirical
evidence. International Finance Discussion Paper No. 514. Board of Governors of the
Federal Reserve System, Washington, DC.
S Knack (2001), Aid dependence and the quality of governance: cross-country empirical tests-
Southern Economic Journal, - JSTOR
Staats, S. (1972). Corruption in the Soviet system. Problems of Communism, 21 (1), 40-47
Stichter, S. (1982) Migrant Labour in Kenya: Capitalism and African Response, 1895-1975,
Harlow: Longman Group Ltd.
University of Ghana http://ugspace.ug.edu.gh

Tanzi, V. & Davoodi, H. (1997) Corruption, public investment, and growth, Working Paper 139,
(Washington, DC: IMF).
Thomas (2010), Institutions and Capital Flight in the Global Economic Crisis, Depatrment of
Government, Cornell University
Thompson, C. (1992), Harvests Under Fire, London: Zed Books.
Tlou, T. and Campbell, A. (1984) History of Botswana, Gaborone: Macmillan Botswana
Publishing Company.
TB Pepinsky (2008), Institutions, economic recovery, and macroeconomic vulnerability in
Indonesia and Malaysia - Crisis as Catalyst: Asia's Dynamic Political, -
crawford.anu.edu.au
United Nations (1990), Practical Measures Against Corruption, New York: United Nations.
UNDP (2011), Illicit financial flows from the Least Developed Countries: 1990–2008. New York
Vandemoortele, J. (1994), "Austerity: Does it have a Human Face? A UNDP Note on Budgetary
Expenditure and Recent Development in Malawi", mimeo
Weeks, John (2012b) "Economic Impact of Capital Flows in sub-Saharan Countries, 1980-
2008,“ PERI Working Paper 290, Amherst, Massachusetts 33 2012c
WH Cooper, JP Hardt (2000), Russian capital flight, economic reforms, and US interests: An
analysis
WJ Awung (1996), An economic analysis of capital flight in Cameroon- African Economic
Research Consortium
World Bank (1985). World Development Report, Washington DC: World Bank.
University of Ghana http://ugspace.ug.edu.gh

World Bank (1997), The State in a Changing World: World Development Report 1997, Oxford:
Oxford University Press.
World Bank (2006), Wolfowitz’s corruption agenda. World Bank Press Review. Feb. 20, pp.1-2.
World Bank (2013), The Worldwide Governance Indicators, www.govindicators.org
Y Zheng, KK Tang (2009), Rethinking the measurement of capital flight: an application to Asian
economies- Journal of the Asia Pacific Economy
ZS Eshete (2014), Public spending composition and efficiency: It's simplifications for growth,
structural change and household welfare in Ethiopia (using recursive dynamic CGE -
International Academic Journals, - iajournals.org

University of Ghana http://ugspace.ug.edu.gh

Appendices
Appendix A: Real Capital Flight for 32 Sub-Saharan African Countries (millions of 2010 US $)
COUNTRY 2000 2001 2002 2003 2004 2005
Angola 1329.6 1342.3 1342 2502 3442 2519
Botswana 286.7 1571.8 -3756.8 1397.9 1290 765.1
Burkina Faso -98.8 -49.2 -126.2 461 -201.3 -310.8
Burundi 181.4 238.5 372.2 367.9 196.4 404.3
Cameroon -26.3 6088.6 -762 -216 -1488 -2620
Cape Verde 39.5 138.3 130.5 63.3 249.2 195.3
CAR -27.9 -25.3 288.9 -15.6 77.1 -62.2
Chad -62.6 88.6 213 797 -206 3
Congo Dr. 2952.5 -897.8 805.5 1659.6 392.4 -80.7
Congo Rep. 3549.2 914.6 -49 2120 5319 1215
Cote d’Ivoire 3629.7 -680.2 1342.2 3271.5 -63.2 3280
Ethiopia 406.2 2518 3073.4 1598.4 1400.8 -319.9
Gabon 2939.5 -78.3 144 1039 1437 2198
Ghana 209.6 199.5 1041 397.8 499.2 -908
Guinea -56.3 -232.9 54.2 -49.6 -325.4 -725.5
Kenya 62.7 518.3 1826.4 1735.9 1087.6 -929.3
Lesotho -84 -134 303.8 220.6 244.2 -87.2
Madagascar -72.9 -1010.4 78.6 45.6 21.4 -117.1
Malawi -63.5 19.5 -63.5 19.5 45.9 61.5
Mauritania -33.8 332.6 312.2 218.7 -30 -15.5
Mozambique -45.2 1305.2 384.6 -1564.1 -15.5 -657.9
Nigeria 518 3357 2723 13107 9812 29263
Rwanda -37.9 -121 21.3 17.3 -141.9 -161.5
Seychelles -12 141.9 46.7 84 -33.6 399.2
Sierra Leone 102.2 -74.7 291.7 183.4 348.1 264.9
Sudan 1061 -370 -1811 2031 6414 5898
Swaziland 16.8 -69.7 285.6 227.1 151.5 -193.9
Tanzania 545.8 -319.8 758 624.2 1020.4 332.5
Togo 669.6 118.2 -412.6 -699.1 -716.5 -227.2
Uganda 190.1 515 563.6 1047 -2488.9 263
Zambia 475.1 -156.1 -32.4 183.9 1543.8 2077.2
Zimbabwe 1512.2 60.7 -65.4 -2375 48.2 186.8

University of Ghana http://ugspace.ug.edu.gh

COUNTRY 2006 2007 2008 2009 2010 2011 2012
Angola 5763 4205 10164 7155 2149 -13943 -12495
Botswana 1200 1010.6 631.8 1205.9 358.8 -482.59 2022.37
Burkina Faso -324.5 -52.8 151.3 181.7 -196.5 705.04 959.71
Burundi 464.6 318.4 169.5 839 -6.4 281.64 169.62
Cameroon -939 -669 -660 -458 -730 3012.67 2829.59
Cape Verde 234.2 220.9 238.7 247.7 846.16 846.16 704.18
CAR 74.2 -8.6 -36.4 -91 59.9 26.9 104.67
Chad -139 -199 -521 0 0 395.97 435.57
Congo Dr. 784.7 3012.2 1715.3 -411.6 1813.6 -6913.6 3491.29
Congo Rep. 2133 535 2283 3091.6 2714.95
Cote d’Ivoire 491.8 979.3 -1123 -1995.6 101.5 7343.46 3105.01
Ethiopia 104 1696.6 -263.8 1875.1 3407.7 3284.95 4735.83
Gabon 2146 927 2191 699 1433 1361.42 1266.68
Ghana 748.3 733.2 1445 678 1184 9667.27 11144.6
Guinea -10.9 -127.2 26.6 -538 -154.2 1825.5 1283.86
Kenya -973.5 -303.4 -1370 -1952.8 -2219 8906.37 5656.16
Lesotho -59.4 -46.3 -0.8 140.1 -64.5 433.21 430.35
Madagascar 321.2 415.3 -440.2 1308.68 1338.73
Malawi 53.3 -762.3 -192.7 -707.9 -311 1303.16 983.05
Mauritania 226.1 -170.4 197.7 -245.9 284.4 931.55 2808.13
Mozambique 1803 241.8 72 132 733.6 7058.38 12091.5
Nigeria 24307 26908 37991 29029 18455 25300 25674
Rwanda -130.1 50.9 -192.3 -274.3 -317.7 713.08 1561.86
Seychelles 323.5 4.1 137.1 -160.4 1002.1 1002.09 703.34
Sierra Leone -23.8 121.2 298.6 484.2 369 2591.52 1055.74
Sudan -5628 21061 1966 346 3298 6548 11032.5
Swaziland 45 335.7 -381 -299.1 181.5 312.35 -353.06
Tanzania -329.6 -505.7 -1063 -216.7 -151.2 7904.57 7449.84
Togo 47.9 1155.7 2636.1 3081.3 1508 -358.7 240.35
Uganda 4123 980.4 913.9 141.4 -163 3731.99 3176.71
Zambia 2203 135.3 -1780 -2185.8 -2359 2213.9 2803.14
Zimbabwe 3145 601.5 -495.9 -1696.4 -801.7 2652.99 3251.13
Source: Ndikumana and Boyce (2012), and sample expanded using data from: IMF, International
Financial Statistics, IMF, Balance of Payments Statistics, IMF, Direction of Trade Statistics,
World Bank, Global Development Finance

University of Ghana http://ugspace.ug.edu.gh

Appendix B: Pairwise correlation coefficients of regression model

VARIABLES C.F COR RD RL IEA UIC INF GDP
C.F

1.0000

COR -0.1144

1.0000

RD -0.0912 0.4323

1.0000

RL -0.2345 0.4159 0.4543

1.0000

IEA 0.0009 -0.026 -0.0647 -0.0259

1.0000

UIC 0.147 -0.0219 -0.0096 -0.0577 0.0216

1.000

INF 0.1384 -0.0314 -0.1031 -0.0974 0.0692 0.024

1.0000

GDP 0.1369 -0.0425 -0.0257 0.056 0.0692 -0.004 -0.0748 1.0000

Appendix B above report pairwise correlation of the variables under consideration. The correlation
coefficients of these series are found to be less than 5 percent which ruled out the issue of
multicollinearity in the estimations.

University of Ghana http://ugspace.ug.edu.gh

100

Appendix C: Hausman Test

Capital Flight
(b)
fixed
(B)
random
(b-B)
Difference
Sqrt (diag(V_b-
V_B)) S.E
Corruption 1.466 1.2369 0.2291 0.0797
Rule of Law -0.8236 -1.9507 1.1271 0.5679
Regime Durability 0.0614 0.0112 0.5021 0.1619
Ind. Of Exec.
Authority 0.0045 0.0032 0.0013
Use of IMF credit -0.0033 0.0003 -0.0036
Inflation 0.1711 0.2721 -0.101 0.0371
GDP growth -0.0147 -0.0073 -0.0074

b = consistent under HO and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from
xtreg
Test: Ho: difference in coefficients not systematic
chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) =
39.80
Prob>chi2 = 0.0000
(V_b-V_B is not positive
definite)
Source: Hausman Test Result

University of Ghana http://ugspace.ug.edu.gh

101

Appendix D: Variables used in the computation of Capital Flight and data sources

VARIABLE DEFINITION SOURCE
DEBT Total external debt stock World Debt Tables

CA Current account balance Balance of Payment Statistics, WDI

DFI Direct foreign investment Balance of Payment Statistics, WDI

RES
Change in reserves & related
items
Balance of Payment Statistics, WDI

TEX Total exports to the world Direction of Trade Statistics

EXAC
Exports to industrialized
countries as reported by the
African country
Direction of Trade Statistics

TIM Total imports from the world Direction of Trade Statistics

MIC
African country’s imports from
industrialized countries reported
by the African country
Direction of Trade Statistics

IMAC
The African country’s imports
from industrialized countries as
reported by industrialized
countries
Direction of Trade Statistics

EXIC
Exports to industrialized
countries as reported by
industrialized countries
Direction of Trade Statistics
CIF/FOB CIF/FOB factor Assumed to be 1.10
NB: Exchange rates of the French franc, Deutsche mark, Swiss franc, Pound sterling, Yen, and
SDR against the dollar: SOURCE; International Financial Statistics

University of Ghana http://ugspace.ug.edu.gh

102

Appendix E: Variables used in the estimation and data sources

VARIABLES DEFINITION SOURCE
CF
Capital Flight (millions 2010
US$)
Ndikumana & Boyce (2012), World Bank Data
(WDI)
RD Regime Durability Polity IV database
I.E.A
Independence of the Executive
Authority
Polity IV database
COR Corruption
Corruption perception index, Transparency
International.
RL Rule of Law
International Country Risk Guide, and World
Bank data base
GDP GDP growth (annual %) World Bank national accounts data
UIC
Use of IMF Credit (current
US$)
World Bank, International Debt Statistics.
INF Inflation (CPI annual %) World Bank (WDI)

University of Ghana http://ugspace.ug.edu.gh

103

University of Ghana http://ugspace.ug.edu.gh