Professional Documents
Culture Documents
(Conflict, Inequality and Ethnicity) Yoichi Mine, Frances Stewart, Sakiko Fukuda-Parr, Thandika Mkandawire (eds.) - Preventing Violent Conflict in Africa_ Inequalities, Perceptions and Institutions-Pa
(Conflict, Inequality and Ethnicity) Yoichi Mine, Frances Stewart, Sakiko Fukuda-Parr, Thandika Mkandawire (eds.) - Preventing Violent Conflict in Africa_ Inequalities, Perceptions and Institutions-Pa
(Conflict, Inequality and Ethnicity) Yoichi Mine, Frances Stewart, Sakiko Fukuda-Parr, Thandika Mkandawire (eds.) - Preventing Violent Conflict in Africa_ Inequalities, Perceptions and Institutions-Pa
Titles include:
Edited by
Yoichi Mine
Professor, Graduate School of Global Studies, Doshisha University, Japan
Frances Stewart
Emeritus Professor of Development Economics and Director, Centre for Research on
Inequality, Human Security and Ethnicity (CRISE), University of Oxford, UK
Sakiko Fukuda-Parr
Professor, The New School, USA
and
Thandika Mkandawire
Professor, London School of Economics and Political Science, UK
v
vi Contents
References 275
Author Index 295
Subject Index 299
Figures and Tables
Figures
viii
List of Figures and Tables ix
Tables
xii
Foreword xiii
Sadako Ogata
Former UN High Commissioner for Refugees
Former President of Japan International Cooperation Agency
Preface
This book is the result of close research collaboration between the Centre for
Research on Inequality, Human Security and Ethnicity (CRISE), University of
Oxford, and the JICA Research Institute (JICA-RI). We thank all scholars and
development practitioners at CRISE and JICA who have provided invaluable
academic and practical insights into this project.
We are very grateful for the continuous encouragement given by Madam
Sadako Ogata, who has contributed a gracious Foreword to this book,
Dr. Keiichi Tsunekawa, the founding director of JICA-RI, and Mr. Hiroshi
Kato, its present director. JICA-RI has organised several international confer-
ences and seminars in Tokyo and London to discuss research plans, progress
and outcomes, as well as to disseminate our findings. The discussion in
this book takes the lively feedback from the audience at those events into
account.
The perceptions surveys in seven African countries were made possible
by the active engagement of local contracting agencies: Datadesk Con-
sulting Services (South Africa), Mass Public Opinion Institute (Zimbabwe),
Synovate Tanzania (Tanzania), the Institute of Development Studies, Uni-
versity of Nairobi (Kenya), Makerere University (Uganda), Practical Sampling
International (Nigeria) and the Institute of Statistical, Social and Economic
Research, University of Ghana (Ghana). The surveys were conducted in par-
ticipatory ways; questions were modified and improved at various levels
reflecting local realities. We thank each agency as well as all enumerators
who knocked on the doors from house to house.
Finally, we would like to express our cordial gratitude to Dr. Mari
Katayanagi for her wonderful work in coordinating these efforts.
Yoichi Mine
Frances Stewart
Sakiko Fukuda-Parr
and
Thandika Mkandawire
xiv
Acknowledgements
xv
Contributors
Arnim Langer is Director of the Centre for Research on Peace and Devel-
opment (CRPD) and University Lecturer in International Relations at the
University of Leuven, Belgium. His research areas include causes and con-
sequences of conflict, post-conflict economic reconstruction, post-conflict
integration and nation-building, multidimensional group inequalities,
group behaviour and identity formation. His latest works include the
following co-edited books: Horizontal Inequalities and Post-Conflict Devel-
opment (2011), Affirmative Action in Plural Societies: International Experi-
ences (2012) and The Elgar Handbook of Civil Wars and Fragile States
(2012).
xvi
Notes on Contributors xvii
1.1 Introduction
In the period following the end of the Cold War, violent conflicts have
marred the African continent; since 1980, more than half of the countries
of sub-Saharan Africa1 have experienced armed conflict, sometimes multiple
conflicts taking place simultaneously in different parts of the country and
sometimes lasting for decades (Fukuda-Parr et al., 2008). Over the last decade,
the number of active violent intra-state conflicts has declined from 17 in
2001 to 10 in 2010 according to the Uppsala Conflict Data Program (UCDP).2
Despite this modest progress, political unrest and violent conflicts continue
to wreak havoc and delay (or even reverse) development in countries, includ-
ing, for example, Nigeria, Sudan, the Democratic Republic of the Congo
(DRC), Guinea and Mali.3 Most contemporary violent conflicts are identity-
based and have been labelled ‘ethnic’ (Stewart and Brown, 2007). They tend
to have a history of conflict recurrence, which is hardly surprising given
that countries emerging from conflict typically remain ‘very fragile in the
short-to-medium term, with a high risk of return to conflict or of its trans-
formation into other forms of social violence (such as crime or domestic
violence)’ (Langer et al., eds, 2011: 3).4
Yet all of this is a historical anomaly. Africa has largely been a continent of
peace. As Julius Nyerere commented, the ‘surprising thing is not that there
has been so much political instability in Africa but that there has been so
much stability, although this fact is less publicized internationally’ (Nyerere,
2000: xiv). Indeed, while the history of East Asia and Europe are charac-
terised by successive wars, violent revolutions and aggressive state-building
1
2 Introduction
(Tilly, 1992; Wong, 1997), most African states and empires emerged peace-
fully and vanished quietly in pre-colonial times (Herbst, 2000; Connah,
2001; Iliffe, 2007).
Violent conflicts are not inevitable. Thanks to an increasing amount of
research on the causes of intra-state violent conflicts, in particular by polit-
ical scientists and economists, over the last two decades, we now have a
much better understanding of the origins and risk factors contributing to
the emergence of these conflicts, and the ways in which they can be pre-
vented. An important challenge for most African governments and their
development partners is to design and implement development interven-
tions and policies that enhance the prospects for durable peace instead of
aggravating the risks of violent conflict. It is for this reason that the Japan
International Cooperation Agency (JICA) has commissioned this research
project on horizontal inequalities (HIs), focusing in particular on the role of
political institutions and people’s subjective perceptions of their situation.
The aim of this study is to contribute to an understanding of the forms of
political and economic governance that would assist in the prevention of
political violence that can undermine or reverse development.
This study explores the relationship between group inequalities and polit-
ical institutions. It focuses in particular on the role and interlinkages among
three factors: subjective or perceived inequalities among identity groups,
objective inequalities and the participation of groups in political institu-
tions. A major finding of recent research on sources of violent conflicts
is that one of the most important underlying causes is the presence of
HIs or group-based inequalities in access to political power and economic
resources. Severe HIs can provoke feelings of frustration and discontent that
may instigate violent group mobilisation and conflict along ethnic, religious
or regional lines (Stewart, 2000, 2002). While much has been learned about
the nature of HIs, two critical issues that remain unclear are how HIs are
perceived and what are the main influences on people’s perceptions, and the
way in which political institutions affect the relationship between HIs and the
emergence of violent conflicts. In Section 1.3 we will discuss these issues in
more detail.
The findings of the study have important policy implications in relation to
the strengthening of economic and political governance and the prevention
of violent conflicts. One objective of the project is therefore to translate any
new insights into actionable policy recommendations. It should be noted
in this respect that these policy recommendations are relevant not only for
‘conflict’ countries but also for other countries that want to address their
objective and subjective HIs constructively, thereby reducing the risk that
they may become a cause for group mobilisation, violent or otherwise, at a
later stage. Before discussing our framework of analysis, we will look more
closely at the concept of HIs and the evidence linking them to the emergence
of violent conflicts.
Sakiko Fukuda-Parr, Arnim Langer and Yoichi Mine 3
As mentioned in the previous section, research has shown that the presence
of severe HIs or inequalities between ‘culturally defined’ groups such as
ethnic, religious or regional groups in terms of access to political power and
economic resources is often an important contributing factor or cause for
the emergence of violent intra-state conflicts.5 In contrast to the relative
deprivation theory (see Gurr, 1970), the HI hypothesis points to the fact
that it can be the relatively rich, and not just the relatively deprived, who
initiate conflict (Stewart, 2002). One should not assume that it is always
the socioeconomically disadvantaged groups that cause political instability
and initiate violent conflict. While this appears to have been the case in
many disputes and conflicts around the world, sometimes relatively privi-
leged groups have attacked and violently suppressed underprivileged groups
in order to safeguard their political and economic power.
The concept of HIs and its theorised relationship with political instability
and violent conflict was first developed by the Oxford-based development
economist Frances Stewart in the early 2000s and was subsequently exten-
sively researched by the Centre for Research on Inequality, Human Security
and Ethnicity (CRISE) based at the University of Oxford.6 The concept of
HIs differs from the ‘normal’ definition of inequality, which Stewart (2000)
has termed ‘vertical inequality’, because the latter type lines individuals or
households up vertically and measures inequality over the range of indi-
viduals rather than groups. HIs, in contrast, are inequalities among groups
of people that share a common identity (for example, ethnic, religious and
regional identities). The HI literature broadly takes a constructivist approach
to ethnicity, which considers ethnic groups to be constructed socially and
historically according to various identity markers, such as ancestry, language
and region of origin. Although the constructivist view on ethnicity conjec-
tures that ethnic boundaries are malleable and fluid, it also assumes that
the reshaping of group identities is costly and that group boundaries there-
fore tend to have some continuity over time (see Stewart, 2002; Bates, 2004;
Stewart ed., 2008). And ‘it is because of [these] continuities, which go along
with the limited choices most people have to switch identities, that inequal-
ities among groups [can] become a source of unhappiness and resentment,
and a cause of social instability’ (Stewart, 2002: 8).
HIs are conceived of as inherently multidimensional and encompass
economic, social, cultural status and political dimensions:
likely to experience civil war’. This result appears to hold regardless of the
fact whether the group is richer or poorer than the average. Alternatively,
Brown’s (2008) study of the determinants of separatist conflict has found
that the likelihood of separatist conflict increases the richer or poorer a
region is in terms of GDP per capita, compared to the national average.
Due to the emerging consensus in academic circles regarding the causal
relationship between HIs and violent conflicts, in recent years the concept
and theory have also gained purchase in donor agencies and the broader
international development community, particularly in the context of spe-
cific countries undergoing or having recently emerged from violent conflicts
in which such inequalities appear to have played an important role (Brown
and Langer, 2010). While this is a positive evolution, there remain important
issues yet to be analysed and fully understood.
1. Objective HIs. How severe are the prevailing objective HIs in terms of
access to political power and state institutions, levels of socioeconomic
development and access to economic resources? Have they resulted
in group mobilisation, violent or otherwise, along ethnic, religious or
regional lines?
2. Subjective perceptions of HIs. How do people perceive the prevailing polit-
ical, socioeconomic and cultural status inequalities in their country?
To what extent do these perceptions accurately reflect the ‘objective’
situation?
3. Political institutions. What types of formal and informal political insti-
tutions can be distinguished in our case-study countries? How do these
formal and informal political institutions affect the management of the
prevailing objective and subjective HIs? Are certain types of political insti-
tutions more conflict-prone, or, conversely, conducive to sociopolitical
stability?
Political institutions
Socioeconomic HIs
and policies
(Conflict) Historical
Stability/Instability
settings
This study presents case studies of Rwanda, Burundi, Ghana, Côte d’Ivoire,
South Africa, Zimbabwe, Tanzania, Uganda, Nigeria and Kenya (Figure 1.2).
This group includes both ‘peaceful’ countries that have managed their
HIs in a non-violent way (albeit not necessarily in a very constructive way)
and ‘conflict’ countries that are currently confronted with serious political
tensions or active violent conflicts. Large-scale perceptions surveys were con-
ducted in all but three of the case-study countries. It was not possible to
conduct surveys in Côte d’Ivoire because of the ongoing violent conflict
at the time, and in Burundi and Rwanda because of the extremely sensi-
tive nature of the issues involved and the impermissibility of conducting
a survey into ‘ethnic’ issues. While the surveys were not nationally repre-
sentative, they were sufficiently large to ensure that all the major ethnic
and religious groups were included in our survey samples (see Chapter 9 for
more details on these surveys). Consequently, as Langer and Mikami argue
in Chapter 9, the ‘results are therefore only statistically representative for
the selected survey locations, but we can draw wider inferences based on
the assumption that the surveyed areas are qualitatively representative of a
larger part of society’.
The book is organised as follows. Chapter 2 provides a classification of
Africa’s political institutions and analyses whether certain types of for-
mal political institutions lower the risk of experiencing political instability
and/or violent conflicts. Chapters 3–6 present the findings of the case studies
that contrast pairs of neighbouring countries with similar HI structures but
8 Introduction
TUNISIA
MOROCCO
ALGERIA
WESTERN LIBYA EGYPT
SAHARA
MAURITANIA
CAPE VERDE MALI
NIGER
SUDAN ERITREA
SENEGAL CHAD
GAMBIA DJIBOUTI
BURKINA
GUINEA-BISSAU FASO
GUINEA
GHANA NIGERIA SOUTH
SIERRA LEONE ETHIOPIA
CÔTE CENTRAL SUDAN
LIBERIA D'IVOIRE AFRICAN REPUBLIC
CAMEROON
SOMALIA
BENIN
TOGO UGANDA
CONGO RWANDA
GABON
KENYA
DEMOCRATIC
REPUBLIC SEYCHELLES
EQUATORIAL GUINEA OF THE
CONGO
SAO TOME AND PRINCIPE
BURUNDI TANZANIA
COMOROS
MALAWI
ANGOLA
ZAMBIA
ZIMBABWE MADAGASCAR
NAMIBIA
MOZAMBIQUE MAURITIUS
BOTSWANA
SWAZILAND
SOUTH LESOTHO
AFRICA
Notes
1. Throughout this book, when we use the term ‘Africa’, we refer to the countries in
sub-Saharan Africa if not otherwise stated.
Sakiko Fukuda-Parr, Arnim Langer and Yoichi Mine 9
2. For more information on the Uppsala Conflict Data Program (UCDP), visit http://
www.pcr.uu.se/research/ucdp/program_overview/ (last accessed 1 May 2013).
3. The World Bank has strikingly described the devastating consequences of violent
conflicts as ‘development in reverse’ (see Collier et al., 2003).
4. Illustratively, Collier et al. (2008) estimate that 40 per cent of all post-conflict
countries return to conflict within a decade of the cessation of hostilities.
5. This section draws on Stewart ed. (2008) and Brown and Langer (2010), who
extensively discuss the concept of HIs and critically review the empirical evidence
linking HIs and violent conflicts.
6. The Centre for Research on Inequality, Human Security and Ethnicity (CRISE) at
the University of Oxford was a Development Research Centre (DRC) funded by the
UK Department for International Development. For more information on CRISE,
visit http://www.crise.ox.ac.uk (last accessed 1 May 2013).
7. The cultural status dimension was first developed by Langer and Brown (2008). See
also foundational work on the relationship between cultural identity and exclusion
by Kymlicka (1995), UNDP (2004) and Sen (2007).
2
Comparing Political Institutions:
Institutional Choice and Conflict
Prevention in Africa
Yoichi Mine, Mari Katayanagi and Satoru Mikami
2.1 Introduction
10
Yoichi Mine, Mari Katayanagi and Satoru Mikami 11
latter suits the ‘integration’ of conflicting groups into a single public (and
national) identity group (McGarry et al., 2008).
in today’s Africa, except for Liberia and Zimbabwe, which, as a matter of fact,
have the same electoral systems for both houses. A majoritarian electoral
system is typically a single-member constituency system based on the first-
past-the-post voting method, while proportional representation typically
adopts a party list system.
As Lijphart mentioned, ‘we shall simply have to live with concepts that have
very important theoretical and policy significance but that cannot be mea-
sured precisely’ (Lijphart, 2002: 46). Other than the above-listed variables,
we could conceive of several crucial elements of PD and PC institutions in
the African context, which have been dropped because of the paucity of
data and/or the extremely qualitative nature of the information. Some such
limitations are noted below.
First, certain ethnic groups having a monopoly of military power and
bureaucratic positions can be a source of serious ethnic grievances, but it
is very difficult to discover reliable continent-wide data about this subject.
Second, although the measurement of the relative size of the local budget
may be crucial in assessing the practical extent of decentralisation in a uni-
tary state, such data are only available for a handful of African countries
(see also Chapter 10, Table 10.2). Third, some unique qualities of ‘hybrid’
arrangements may not be captured in a quantitative way. For example,
the geographical voting spread requirements in Nigeria should make the
electoral system of the country more power-dispersing (Chapter 8). Finally,
constitutional recognition and everyday practices of traditions are different
matters. It is commonplace in Africa for people never to sell their communal
land in spite of legal provisions to the effect that they can do so.10
Despite all these qualifications, the classification of formal political insti-
tutions based on available data is of importance as a starting point. It is
misleading to assume that formal institutions are eventually swallowed by a
sea of informal practices, for there can also be a very strong influence in the
opposite direction. Formal political institutions like electoral systems often
delineate contours of group boundaries through ethnic coalition-building,
as has been vividly demonstrated in the study of modern ethnic politics in
Zambia by Posner (2005), and thereby shape individuals’ identity choices.
Although people’s identities are historical constructs, people may kill oth-
ers based on their own identity perceptions at a given time. Identity-based
actions are too real, and can also be framed by emergent political institutions
and policies in the formal sphere (see Chapter 1, Figure 1.1). The interac-
tions between formality and informality are discussed fully in the country
case studies in the following chapters.
Note: The ratings are based on the data as of the end of 2011. The outcome of the elections in Gabon on 17 December 2011 is included, while those in
the DRC on 28 November are excluded due to the serious disputes over the election results at the time. The classification of electoral systems is based on
a detailed dataset provided by the International IDEA (International Institute for Democracy and Electoral Assistance), http://www.idea.int/ (last accessed
1 May 2013). Resource-rich almanacs such as Africa South of the Sahara and Political Handbook of the World: Governments and Intergovernmental Organizations
published in various years have been consulted as major references for general features of political institutions and political processes. As for the election
outcomes and the forms of legislature, we have used the data compiled by Nohlen et al. eds (1999) and various contemporary information released by local
electoral commissions and the Inter-Parliamentary Union, http://www.ipu.org (last accessed 1 May 2013). Various articles in journals such as Africa Research
Bulletin: Political, Social and Cultural Series and Jeune Afrique, as well as the constitutional documents of African countries, have also been consulted for the
rating procedure. Eritrea, Somalia and Swaziland are excluded in the following figures due to the lack of relevant data. It must be noted that the ratings in
the centralised/decentralised dimension are based principally on constitutional provisions. In Cameroon and Malawi, however, the upper house has yet to
be established, despite the provisions in their respective constitutions. In such a case, the reality counts. As the standardised scores of each dimension are
rounded off to the second decimal place, the totals on the right-hand side do not necessarily match the simple sums of the two values.
21
22
6 PD-DEC
ETH ZAF
4 COM SDN
NGA
Centralised/Decentralised
NAM LSO
MUS
2 CMR
COD
UGA ZWE
BWA SEN KEN
LSE BDI
ZMB TZA MOZ MRT
0 PC-MJ SYC GHA
CPV PD-PS
BEN MLI
GMB GAB RWA
GIN NER STP
TGO CIV
–2 DJI COG BFA
MDG
MWI
LBR
–4 GNB
AGO CAF TCD GNQ
PC-CEN
–6 –4 –2 0 2 4
Majoritarian/Power-sharing
Electoral system
3
1.5
Division of Composition of the
1
legislature executive
Head of
Local elections
government
Forms of
government
ZAF ETH
5
COM SDN
Centralised/Decentralised
NGA
NAM
LSO
MUS CMR
KEN COD
BWA SWZ
LSE BDI ZWE TZA
MRT SEN UGA
SYC BEN ZMB MOZ
0 GHA
CPV
GMB
GAB RWA MLI
ERI GIN BFA
DJI CIV
STP NER MDG
TGO
COG
LBR MWI
GNB CAF AGO
TCD
GNQ
–5
on the other, except that the top eight most populous countries, Nigeria,
Ethiopia, the Democratic Republic of the Congo (DRC), South Africa, Sudan,
Tanzania, Kenya and Uganda, signify higher degrees of decentralisation
(Figure 2.3). Of these countries, Ethiopia, South Africa, Sudan and Nigeria
present strongly federal features, being classified as PD-DEC (Figure 2.1).
Figure 2.4 shows ethnic fractionalisation based on the dataset developed
in Alesina et al. (2003), indicating that ethnically multipolar societies
tilt towards power-concentration especially in the centralised/decentralised
dimension.12 Although the combination of ethnic diversity and PC insti-
tutions is counterintuitive, it makes sense that multipolar societies can be
stable and thus compatible with democratic centralisation of power. As a
general rule, ethnic conflict is inflamed more easily in bipolar and tripolar
societies where zero-sum situations make it difficult to construct peaceful
ethnic alliances (Bangura, 2006; see also Chapter 6).
Colonial factors also affect institutional choice to some extent. While
Acemoglu et al. (2001) differentiated between extractive and settler insti-
tutions in an attempt to gauge their impacts on economic growth in Africa,
we are more concerned about the degrees to which the patterns of politi-
cal governance originated in colonial centres have been domesticated into
local realities through time. Figure 2.5 indicates that Anglophone Africa
tends to be majoritarian and decentralised, while Francophone Africa tends
24 Comparing Political Institutions
ZAF
5 ETH
COM SDN
NGA
Centralised/Decentralised
LBR
CAF
TCD
GNQ AGO GNB
–5
0 .2 .4 .6 .8 1
Ethnic fractionalisation
4 MDG STP
KEN
COD MRT
NER
Majoritarian/Power-sharing
BFA
2 ZWE
LSO CIV MLI
GIN CPV GNQ
NAM
RWA BDI GNB MOZ
BEN TGO
ZAF CMR SEN
0 LBR TCD
NGA COG
MWI MUS
CAF
TZA SDN
UGA
GHA
LSE
–2 ETH
ZMB
BWA DJI AGO SYC
GAB
COM
–4
GMB
4
Majoritarian/Power-sharing
–2
–4
–6
Anglophone Francophone Lusophone Others
ZAF ETH
5
SDN COM
NGA
Centralised/Decentralised
NAM MUS
LSO
BWA CMR
KEN UGA COD
SWZ MRT
ZWE BDI SEN
LSE TZA
GHA BEN CPV MOZ SYC
0
GMB ZMB MLI
RWA GAB
BFA NER
CIV GIN STP ERI
DJI TGO
MDG COG
MWI
LBR
CAF TCD AGO GNB
GNQ
–5
4
Centralised/Decentralised
–2
–4
A B C D Total E F G H Total
Note: The first row of each country shows the ratings of the period before the ‘major turning points’ specified in the last column, while the second row
indicates the latest ratings as of the end of 2011. In Kenya, although the referendum held on 4 August 2010 resulted in the adoption of a new constitution,
the dates when its provisions enter into force vary, and hence our ratings are based on the previous constitution. The new constitution is meant to limit
executive power and institute a system of checks and balances, thus becoming the culmination of its constitution-making exercise, which had lasted almost
27
two decades (Chitere et al., 2006; Cottrell and Ghai, 2007; see also Chapter 7).
28 Comparing Political Institutions
PD-DEC
ZAF
5
NGA
Centralised/Decentralised
ZWE
UGA
TZA
0 PC-MJ KEN
PD-PS
GHA
RWA
CIV
–5
BDI
PC-CEN
–6 –4 –2 0 2 4
Majoritarian/Power-sharing
20
80
2010 Ibrahim index of African governance
70
60
50
40
30
1 MUS CPV
GHA ZAF BEN
STP
Freedom House sub-category score
BWA NAM
SEN MLI
SYC LSO
total 2010 (standardised)
TZA
0 LSE MWI MOZ
ZMB
COM
BFA
GNB
TGO NER
GMB MDG
CAF GIN
GAB
–1 DJI MRT
COG
RWA
CMR
ZWE
GNQ
–2
PC-MJ PC-CEN PD-DEC PD-PS
1
Freedom House sub-category score
total 2010 (standardised)
–1
–2
PC-MJ PC-CEN PD-DEC PD-PS
Figure 2.8 Freedom House scores (excluding countries recently faced with violent
conflict)
Source: Based on the standardised sums of Freedom House 2010 sub-category scores, http://www.
freedomhouse.org/ (last accessed 1 May 2013).
Yoichi Mine, Mari Katayanagi and Satoru Mikami 33
In the 2010s, dozens of African nations will celebrate 50 years since inde-
pendence, and African countries are continuing their attempts at institu-
tional reforms to accommodate political tensions and multidimensional HIs.
In 2010, Angola, Guinea, Kenya, Madagascar and Niger promulgated new
constitutions. Constitutional referendums were held in Liberia in 2011 and
in Zimbabwe in 2013. Along with such contract-making events, regular elec-
tions are taking root in many African countries. Of course, these attempts all
differ in terms of their content and orientation, and are sometimes tainted
by rioting, suppression and oppositional boycotts, testifying to the fact that
the choice of government and political institutions is a part of the dynamic
political processes in contemporary Africa.
According to Dahl, there are three paths from closed hegemonies
to polyarchies, through competitive oligarchies (liberalisation preced-
ing inclusiveness), through inclusive hegemonies (inclusiveness preceding
liberalisation) and through direct routes (‘the large middle area’), and Dahl
indicated that the first path, by way of competitive oligarchies, could be
the most stable (Dahl, 1971: chapters 1 and 3). These historical sequences
formulated by Dahl imply that there can be plural paths to a univer-
sally desirable state of democratic governance, the transition to which
may not be a simple, unilinear evolution towards modernity. The major-
ity of political institutions in contemporary Africa are, at least formally,
neither oligarchies nor hegemonies but multi-party democracies based on
popular elections, however insecure they may be. PD, PC and interme-
diary or mixed institutions in their ideal forms may be regarded as the
three distinctive pathways towards more stable democracies that would be
geared to contain future violent conflict in Africa (Figure 2.9; see also Dahl,
1971: 7, 34).18
Given such a plurality of pathways to democratisation, let us present
several hypothetical conclusions of this chapter. First, the PD option can be
the best solution for preventing violent conflict as well as putting an end to vio-
lent ruptures, by providing short-term refuge from cumulative tensions in countries
with serious HIs. Despite Dahl’s preference of competitiveness, we endorse the
path of inclusive accommodation of the PD type. On the one hand, a num-
ber of African countries have deliberately opted for ad hoc power-sharing
in the national government to appease worsening political situations. The
term ‘Government of National Unity’, used first in South Africa in 1994,
was subsequently adopted in a number of countries, including the DRC,
Côte d’Ivoire, Burundi, Sudan, Kenya and Zimbabwe (all of which are PD-PS
countries with the exception of Sudan, a PD-DEC country). The attractive-
ness of ad hoc power-sharing in conflict situations is obvious, as the principle
of proportionality is likely to be accepted by all major parties as a second-
best solution. On the other hand, the relatively big countries that have
34 Comparing Political Institutions
PC path
Competitive integration
Hybrid evolution
PD path
One-party
system
Inclusive accommodation
experienced serious conflict in the past, like Nigeria, Tanzania, Ethiopia and
South Africa (all are PD-DEC countries except for Tanzania, which is a PC-MJ
country), have elaborated federal or semi-federal systems and succeeded
in pulling the nation together. In the longer term, these two dimensions,
power-sharing and autonomy, are expected to reinforce each other for the
achievement of sustainable peace in post-conflict situations (Wolff, 2011).
The preference for PD institutions in heterogeneous countries can be natu-
ral, in that such a regime is often an extension of a compromise agreement
among antagonistic parties. PD arrangements can be superior to PC arrange-
ments in terms of conflict resolution, in that a military victory over an
enemy that entails civilian casualties can never be the best way of effecting
regime change, as long as different means to achieve the same ends are avail-
able. PC institutions have sometimes been chosen as an outcome of the end
of war, bringing about the marginalisation of the defeated (see Chapter 3).19
In the long run, however, it would be misleading to endorse the PD option
as having an absolute advantage as a means of conflict prevention. On the
one hand, a limitless continuance of a grand coalition may consolidate the
centralised power of ethnic elites, leaving the masses alienated, as in the case
of the joint rule by Christian and Muslim notables in Lebanon before the
civil war broke out in 1975. On the other hand, the unit autonomy in federal
institutions may exacerbate socioeconomic HIs, unless the regime is supple-
mented by effective mechanisms of resource redistribution.20 At present, this
issue is being treated seriously in many African countries as new discoveries
of oil and minerals in one particular region arouse feelings of jealousy in
other regions.
Second, in order to strengthen national cohesion, it may be of benefit for African
countries in relatively stable stages to take developmental paths, being assisted
Yoichi Mine, Mari Katayanagi and Satoru Mikami 35
(Fjeldstad and Therkildsen, 2008). As people are loath to pay tax to rulers
in remote centres, it is worth noting that decentralised fiscal systems are
required, especially in relatively big countries with PC institutions.
Taking the example of transport development projects in post-
independence Nigeria, A.O. Hirschman distinguished between trait-taking
and trait-making project designs. A trait-taking project accepts local traits,
such as HI-related ethnic cleavages, as given (a highway system open to
many small ethnic trucking firms), while a trait-making project is more
ambitious, presupposing certain changes of traits, such as the swift creation
of national solidarity, for the success of a project (a nationwide railway
system operated from Lagos). The former is feasible but tends to preserve
or even reinforce undesirable traits, while the latter is laudable but may
become ruinous before desirable traits are created (Hirschman, 1967: xi,
128–159). Clearly, the inclusive, accommodative PD path and the competi-
tive, integrationist PC path as two contrastive modalities of nation-building
are trait-taking and trait-making respectively, and the argument of this
chapter is to endorse the relative advantage and practicability of the trait-
taking PD path of gradual nation-building. In order to stabilise a PC regime
with multiple HIs, those inequalities should be rendered visible and recog-
nised as urgent policy issues. In case of a PD regime, it is taken for granted
in its founding contract that group differences should be addressed through
political efforts.
Third, recognising the contrasting features of PD and PC institutions, as well
as the relative advantage of PD institutions for conflict prevention, the diversity of
political institutions in Africa should be fully appreciated. Institutional experi-
ments conditioned by complex relationships between state and society are
to continue in African countries. The discussion of formal institutions in
this chapter has shed light on only the tip of the iceberg, in the sense that
violent conflicts can be triggered by intergroup animosities and then sub-
dued by virtue of a revival of societal norms, irrespective of the types of
institutional gimmicks. However, rather than looking at state and society
separately, it would be more fruitful for us to pay attention to the inter-
actions between the formal and the informal, as incongruity between the
design of formal institutions and the informal configurations of society may
cause serious discontent. In this regard, the past and present attempts at
constitutional reforms in relatively big countries with heterogeneous pop-
ulations, like Nigeria and Kenya, abound with historical lessons on how
to design formal institutions, by trial and error, to make them accommo-
date social cleavages in gradual ways (Chapters 7 and 8). Some countries
have called upon the opposite elements to counteract undesirable effects
of the major elements; for instance, Seychelles, Guinea and Lesotho have
adopted a combination of first-past-the-post and proportionality systems for
parliamentary elections. There is room to consider the merits of this kind of
‘hybrid’ system.
Yoichi Mine, Mari Katayanagi and Satoru Mikami 37
2.6 Conclusion
Notes
1. See Lewis (1965) as well as Lijphart (1971; 1977: 143–147; 1985: 97–98; 1999: 31,
308; 2002: 45; 2008: 278–279).
2. W. Arthur Lewis, a Caribbean of African descent, was the first ‘non-European’
Nobel laureate in economics (the second was Amartya Sen). Lewis (1965) was
harshly criticised for his unreserved censure of Pan-Africanist leaders, and since
then, in the field of African studies, there has been no serious attempt to reap-
praise this seminal work on constitutional engineering for West Africa (Mine,
2006).
38 Comparing Political Institutions
3. For the recent development in the debate, see the following collections of dis-
cussions: Reynolds ed. (2002), O’Flynn and Russell eds (2005), O’Leary (2005),
Diamond and Plattner eds (2006) and Choudhry ed. (2008).
4. The centripetalist theory of Donald Horowitz is said to be ‘positioned towards
the integrative end’ of the accommodative spectrum in McGarry et al. (2008:
53–56, 69).
5. A similar and thorough attempt to appraise the electoral systems in five democ-
racies in Southern Africa, South Africa, Namibia, Zimbabwe, Zambia and Malawi,
has been made by Reynolds (1999), the criteria of which is based on Lijphart
(1994), an earlier study of 27 countries including no African cases.
6. See also the discussion of power-sharing through the two channels, federal
arrangements and consociation, in Human Development Report 2004 (UNDP, 2004:
51–54). Although this chapter defines power-sharing in a narrow sense, this
concept can be expanded to include vertical power-sharing between the centre
and local units. See the framework of ‘complex power-sharing’ (Wolff, 2009).
7. For comparison, we do not use an average of the election results over time, given
that multi-party elections have begun to take root in many African countries
only recently. Note that our classification only provides a snapshot of changeable
situations in contemporary Africa, rather than a comprehensive evaluation of
historically established institutions over time.
8. A unipolar party system is an outcome of (at least formally) democratic elections
and is different from a de jure one-party system. We consider having plural parties
to be better than having one dominant party, as long as all human societies are
accompanied by conflicts of multiple interests, ideologies and affections. Such
conflicts can be mitigated and transformed effectively through the activities of
political parties not ‘as whole’ but ‘as part’ (Sartori, 1976).
9. According to Watts (2008: 59–61), transitional countries such as the DRC and
Sudan are nominal federations and not yet functioning well as full-fledged federal
countries.
10. Policies towards political HIs should be discussed in terms of the three branches
of government, both in the central and local tiers, paying due attention to the
bureaucracy and the security sector as well as to the nexus between formality and
informality (Stewart, Brown, and Langer, 2008: 306–310). As the present chapter
concentrates on the formal aspects of institutions for cross-country comparison,
many crucial elements are left for the country-specific discussion in the following
chapters.
11. For a given PD country, when the difference after deducting the simple sum of
the majoritarian/power-sharing ratings from that of the centralised/decentralised
ratings is positive, the country can be classified as PD-DEC, because its relatively
high scores due to its decentralised characteristics form the major contributor
to push up this country to the PD position. The same can be applied to other
sub-classifications.
12. Fractionalisation is defined as the probability of two randomly chosen individuals
not being members of the same group. The fractionalisation index of a purely
bipolar society is 0.5. An intrinsic problem with this calculation is that ethnic
boundaries are extremely fluid, as identities are constantly chosen by individuals.
13. These past ratings are weighted using the means and standard deviations of
present ratings.
14. Compared to Figure 2.6, a similar mapping of 26 countries around the world pre-
sented in Lijphart (1999: 255) shows gentler shifts. This difference is reasonable,
Yoichi Mine, Mari Katayanagi and Satoru Mikami 39
as the latter is based on average data over a much longer time period, namely, 50
years. See also note 7.
15. The 2010 Ibrahim Index of African Governance is based on data from 2008 to
2009, http://www.moibrahimfoundation.org/en/media/get/20101020_2010-ibra
him-index.xls (last accessed 1 May 2013). The scores of the 2010 African Gov-
ernance Index for our target countries are as follows: Rwanda, 47.2 and Burundi,
44.7; Ghana, 64.6 and Côte d’Ivoire, 36.8; South Africa, 71.5 and Zimbabwe, 32.7;
Tanzania, 55.0 and Uganda, 50.8; and finally Kenya, 50.5 and Nigeria, 43.3.
16. UCDP Georeferenced Event Dataset v.1.0-2011, Uppsala Conflict Data Program,
Uppsala University, http://www.pcr.uu.se/research/ucdp/datasets/ucdp_ged/ (last
accessed 1 May 2013). The data are the sum total of the deaths caused by wars
and battles, one-sided violence from governments and non-state civil conflicts.
Political instability sometimes looms large with fewer casualties. The number of
conflict-related deaths in Zimbabwe during the same period is 253 in the same
dataset (all these figures of casualties are more or less speculative).
17. The question was: ‘Do you think that the relations between the different ethnic
groups in your country will improve or worsen in the future? Or will there be no
change?’
18. Steven Levitsky and Lucan A. Way argue that, since 1990, competitive author-
itarian regimes have taken three divergent paths, transforming themselves into
democracy, unstable authoritarianism or stable authoritarianism, and they anal-
ysed trajectories followed by 14 African countries according to this perspective.
In their study, the persistent dominance of authoritarianism in Africa is attributed
to the strength of the organisational power of the incumbents, the relative weak-
ness of linkages with the West, as well as the rise of leverage from the West
(Levitsky and Way, 2010: 236–308). While they discuss the paths diverging into
democratic redress and authoritarian regress, our contention is that there are
different paths to a common democratic goal.
19. A PC regime may function as a nominal democracy to consolidate the rule of
a military victor. See also Paul Collier’s mixed praise of President Museveni of
Uganda (Collier, 2009: 182–184), based on an analogy between the political vio-
lence in contemporary Africa and the violent state-making process in European
history as described by Charles Tilly.
20. Discussing the longer-term costs of power-sharing, Rothchild and Roeder (2005)
suggest ‘power-dividing’ through a strategy of civil liberties, multiple majori-
ties, and checks and balances as an alternative. A PD regime may also entrench
patterns of ethnic fragmentation as demonstrated in the cases of Bosnia and
Herzegovina (Slye, 1996; Marko, 2005; Samuels, 2006), post-war Iraq (Williams
and Simpson, 2008; Younis, 2011) and other countries. In the context of Africa,
Ian S. Spears underlines that the external pressures to make political opponents
unite are weak, and that despite prominence of ‘ethnic “arithmetic” and the
creation of multiethnic coalitions’ in African politics, bridging across serious
cleavages is not that easy (Spears, 2002: 123, 125–126).
21. The existing literature on political institutions rarely raises the issues of economic
growth and redistribution, though Lijphart (1991) once argued that proportion-
ality and inclusive government were better suited to stable and steady economic
policymaking than majority government.
3
‘Twin Countries’ with
Contrasting Institutions:
Post-Conflict State-Building
in Rwanda and Burundi
Shinichi Takeuchi
3.1 Introduction
40
Shinichi Takeuchi 41
reflection. The fact that Rwanda and Burundi, often referred to as les pays
jumeaux (the twin countries), share a number of characteristics makes com-
parative methods quite relevant (Durkheim, 1960). This chapter examines
the role played by ethnicity in politics in the two countries, offering a com-
parison of the two countries’ ethnic relations and political institutions in
order to draw out some more general implications for conflict prevention.
Ethnic relations will be analysed through the lens of horizontal inequal-
ities (HIs), which is a useful method for understanding the causal relation-
ships between ethnicity in politics and conflict (Stewart, 2008a). In the case
of Rwanda and Burundi, which have very similar ethnic compositions, the
conclusions that can be drawn from a comparison made using this method
is particularly relevant. Moreover, disaggregating their historical processes
into the political and socioeconomic dimensions of HIs will make any such
comparisons much clearer.
In addition to historical analyses of HIs, this chapter will examine the
political institutions introduced after the recent armed conflicts. In this con-
text institutions can be understood to mean the rules of the game, incentives
and/or norms. The term refers to ‘the humanly devised constraints that
shape human interaction’ (North, 1990: 3), including both formal rules and
informal constraints. In the context of recent post-conflict state-building,
particular attention should be paid to the gaps and interactions of two sorts
of institutions. On the one hand, the international community has tried to
enhance a set of institutions based on the idea of ‘liberal democracy’ – that
is, those that enhance such norms as multi-party democracy, free markets
and human rights, in peace-building processes (Paris, 2004). These efforts
have been reflected in the formal institutions of these countries. On the
other hand, state-building is nothing more than a process of coercive power
accumulation (Tilly, 1992); actors seizing state power thus try to make use
of institutions for their own purposes. The dynamics of post-conflict poli-
tics will therefore be determined through the interactions of these different
motivations.
In comparing the historical processes behind the formation of HIs and
political institutions in Rwanda and Burundi, this chapter aims to clarify
the implications for a general approach to conflict prevention.2 Analyses
of ethnic relations and institutions shed light on the nature of post-conflict
political regimes and the prospects for conflict prevention. In Rwanda, where
revolutionary political change has been a recurrent feature of the state’s
history, authoritarian post-conflict rule led by the ethnic minority casts a
shadow over its peace-building, as the members of the ethnic majority are
likely to resent both political and socioeconomic HIs. While the introduc-
tion of the power-sharing system has drastically reduced the importance of
ethnic politics among the Burundian elites, there is still an imminent danger
of armed conflict breaking out in that country, as the institutionalisation of
the power struggle remains unresolved.
42 Rwanda and Burundi
In both Rwanda and Burundi, the dominant identity groups in politics are
the Tutsi and Hutu, as shown in the repeated outbreaks of ethnically influ-
enced violent conflicts. In the analysis of these groups, two caveats should
be borne in mind. First, the politicisation of the two groups has taken
place largely since the colonial period;4 ethnic antagonism arose between
Tutsis and Hutus as a result of colonial policies in both countries. This
means that the social constructivist vantage point is particularly relevant in
this instance.5 Secondly, the focus should be on the similarities and differ-
ences between the two countries as reflected in their historical experiences.
Although the social characteristics of the Tutsi and Hutu have a number of
similarities, for instance, they share the same language and religions, ethnic
relations in terms of state power have been quite distinct in the two coun-
tries. The necessity and importance of a historical analysis of their group
formation are therefore quite evident.
Pre-colonial Rwanda was the most centralised state in the Great Lakes
Region, a part of the African continent characterised by the presence of
a number of states with stratified societies.6 The centre of Rwandan state
formation, the kingdom of Nyiginya, is thought to have been founded in
the latter half of the seventeenth century. At the end of the nineteenth
Shinichi Takeuchi 43
century, King Rwabugiri exercised control over almost the same territory as
present-day Rwanda. The king could mobilise strong military power through
his control of influential pastoral chiefs. In this context, a relatively unitary
Tutsi group identity had developed among the ruling elites.7 The political
dominance of the Tutsi had been strengthened in particular in the central
area of the kingdom to such an extent that there are records of several revolts
by Hutu agriculturalists at the end of the nineteenth century (Vansina, 2001:
177). Although there is no doubt that the group identities of the Tutsi and
Hutu were clearly shaped during the colonial period, it is clear from the evi-
dence that Rwanda’s social cleavage between these groups can be traced back
to the pre-colonial era.
By contrast, political power in pre-colonial Burundi was more decen-
tralised. Influenced by delays in state formation, Burundian dynastic history
before the nineteenth century is quite obscure, although its regional auton-
omy was very marked even before that period (D. Newbury, 2001). What
characterised pre-colonial Burundi were the diversity of the social cate-
gory ‘Tutsi’ and the existence of serious cleavage among ruling elites. Three
points deserve to be mentioned in this regard. First, the central actors in
Burundian state formation, the dynastic family, had a unique identity as
Ganwa. While the Ganwa are ethnically a part of the Tutsis, it was only
they who had access to the political power of the pre-colonial state, and
thus they had a clearly distinct identity from Tutsis in general. Secondly, the
Ganwa elites were always in conflict internally, due to the rivalry between
different lines of descent, thus hindering the consolidation of central rule.
Especially important in this context was the opposition between the Bezi
and Batare,8 which remained a source of conflict in Burundian politics until
the 1960s. Thirdly, dichotomous relationships between Tutsis and Hutus did
not exist in pre-colonial Burundi, partly because the Tutsis had never been
a monolithic group,9 and partly because the Burundian Hutus had played
more important roles than their Rwandan counterparts in the pre-colonial
kingdom.10
In the period from 1899 to 1962, Rwanda and Burundi were administered
at first by Germany, as a part of German East Africa, then after the First
World War by Belgium, as a mandated territory on behalf of the League
of Nations, and subsequently as a trust territory on behalf of the United
Nations. Under European rule, ethnic tensions were heightened, because
the colonial authorities11 systematically prioritised the Tutsis, and discrim-
inated against the Hutus in the political system. This policy was based on
a Eurocentric ideology, the so-called Hamitic hypothesis, which argued that
the Tutsi were a superior race, with a European origin, whereas the Hutu were
an inferior race of African origin (Sanders, 1969; Chrétien, 2000). Discrim-
inatory policies were justified under the pretext of stabilising ‘traditional
society’; in public schools priority was given to Tutsi children; and Hutu
chiefs were eliminated from the administration.12 As a result of this policy,
44 Rwanda and Burundi
the Hutus generally lagged behind the Tutsis in terms of education as well as
their level of employment in the modern sectors of the economy. The notion
of the Tutsi as ethnically European, based on the Hamitic hypothesis, had an
enormous impact under the unequal power relations of the colonial period,
creating widespread discontent among Hutu elites.
Hutu grievances were more intense and organised in Rwanda, where the
dichotomy between Tutsi and Hutu was much clearer than in Burundi.
In consequence, the political turbulence and ethnic strife, which was
euphemistically termed the ‘social revolution’,13 broke out at the end of the
colonial era. This was Rwanda’s first experience of nationwide ethnic strife,
and resulted in the collapse of the Tutsi-led political system, a massive out-
flow of Tutsi refugees,14 and the Hutu elites gaining a monopoly on political
power after independence. In 1965 their party, the Parti du mouvement de
l’émancipation hutu (PARMEHUTU), won all seats in the national parlia-
ment (Reyntjens, 1985: 445), systematically eliminated Tutsis from political
power and prohibited Tutsi refugees from returning to the country.
The ‘social revolution’ also had a tremendous impact on Burundi, where
people tended to regard it as their possible future, which was desirable for
Hutus but a potential nightmare for Tutsis. The ethnicisation of politics obvi-
ously accelerated after the assassination of the nationalist leader Prince Louis
Rwagasore in October 1961. In spite of King Mwambutsa’s efforts to appease
ethnic tensions, distrust of the king’s political machinations finally resulted
in a coup attempt in October 1965 by a Hutu group in the army and gen-
darmerie. The attempt was severely suppressed by Tutsi groups in the army,
enabling them to seize political power and, subsequently, to overthrow the
monarchy in November 1966. This process of dethroning the king and estab-
lishing the republic resulted in considerable changes in power relations; the
Ganwa elites, who had hitherto occupied the centre of Burundian politics,
were largely replaced by non-Ganwa Tutsi officers in the army, who came
largely from the Province of Bururi. In contrast to the Rwandan Tutsis, the
Burundian (non-Ganwa) Tutsis had not been positioned at the centre of the
traditional kingdom.15
In comparison with Rwanda, where the sudden power shift took place
during the short period of the ‘social revolution’, the post-independence
consolidation of Tutsi hegemony in Burundi advanced only gradually. Their
hegemony, however, was established through bloodshed and mass killing; a
huge number of Hutus were slaughtered following the two aborted coups in
1965 and 1969, and finally in the genocide of 1972.16 It was, in fact, follow-
ing the genocide in 1972 that the Bururi Tutsi monopoly of political power
was confirmed, because the Hutu elites were almost completely eliminated
in the genocide. In 1987, only two seats out of 65 in the Central Committee
of the sole legal party, the Parti de l’union et du progrès national (UPRONA),
were held by Hutus (Lemarchand, 1994: 108).17
Shinichi Takeuchi 45
they have adopted after the serious armed conflicts and explore the back-
grounds of their institutional choices.
3.4.1 Rwanda
Adopted in 2003, nine years after the end of the civil war, the Rwandan
constitution provides political institutions based on a system of multi-party
democracy (Republic of Rwanda, 2003). It can be classified as a semi-
presidential system, in which a president (the head of the state) and a prime
minister (the head of the cabinet) are both active participants in the admin-
istration of the state. As a multi-party democracy, political organisations are
permitted to be formed and to operate freely. However, they are required
not to destabilise national unity (Article 52), and are prohibited from basing
themselves on ‘divisions’ such as race, ethnic group and region (Article 54).
In essence, the Rwandan constitution is shaped by the shadow of the
1994 genocide. Determination not to repeat the genocide and to promote
national unity is expressed repeatedly in its preamble and the text, thereby
emphasising the importance of ‘unity’ and the danger of ‘division’. In fact,
commitments to ‘fighting the ideology of genocide and all its manifesta-
tions’ as well as ‘eradication of ethnic, regional and other divisions and
promotion of national unity’ are stipulated as two of the six fundamen-
tal principles of the state (Article 9). These principles are based on the
idea that the genocide was derived from ethnic discrimination and division
among nationals. The prevention of genocide is pursued further by abolish-
ing a period of limitation for the crime of genocide, and by stipulating that
its ‘revisionism, negationism and trivialisation’ are punishable (Article 13).
In addition, the constitution states not only that ‘discrimination of whatever
kind’ is ‘prohibited and punished by law’ (Article 11), but also that any form
of ‘division’, such as on ethnic, regional and racial lines, is also punishable
(Article 33).
In this context, ethnic power-sharing is unthinkable, because the very
existence of ethnic diversity is formally denied in the constitution. More-
over, as shown in its fundamental principles, the constitution considers
ethnic difference as something to be eradicated; those who emphasise dif-
ference will be regarded as wrongdoers promoting ‘division’. Under the rule
of the RPF, terms such as ‘division’, ‘divisionism’ and ‘genocide ideology’ are
often utilised when the government criticises its opponents. For instance,
before the first post-conflict election, in 2003, the biggest Hutu opposition
party (Mouvement démocratique républicain, MDR) was ordered to dissolve
itself because of its ‘divisive ideology’.33 In 2010 April, a Hutu woman, who
had declared her intention to be a candidate in the presidential election, was
arrested and charged with ‘propagating the Genocide Ideology, Revisionism
and Ethnic Division’ (The New Times, 23 April 2010).
While detailed analysis of the Rwandan constitution has already revealed
a number of problems (Reyntjens, 2003), two points deserve to be men-
tioned here. First, some of its articles constitute institutional arrangements
50 Rwanda and Burundi
that operate to the advantage of RPF rule. One example is the composition
of the legislature, because a limit is placed on the proportion of members
to be elected by universal suffrage (Article 76). In the case of the Cham-
ber of Deputies (the lower house), 53 of its 80 members are elected by
universal suffrage with a secret ballot. The remaining 27 seats are reserved
for representatives of women (24 seats),34 youth (2 seats) and the disabled
(1 seat). These representatives are selected by members of local adminis-
trations and/or related official councils, on which the RPF-led government
can exert a dominant influence (Reyntjens, 2003: 77).35 Considering the fact
that the RPF’s core supporters are an ethnic minority, these measures lim-
iting universal suffrage were likely devised in order to maintain its rule.
Secondly, important civil rights such as freedom of thought, opinion and
religion (Article 33), as well as freedom of the press and information (Arti-
cle 34), are defined as being ‘guaranteed by the State in accordance with
conditions determined by law’. The text indicates the over-presence of the
state, because these fundamental human rights should include freedom from
the state. These clauses demonstrate the intention of the Rwandan state to
control the societal sphere for its own purposes.
The 2003 Rwandan constitution was based on the RPF’s political ide-
ology; against the backdrop of the victory in the civil war, it attempted,
through the constitution, to legitimise the revolutionary change and to insti-
tutionalise its gains, for example, by emphasising national unity, denying
‘division’ and promoting state control over politics and civil society. Since
the RPF’s victory, its elites have virtually monopolised important political
posts, while condemning the ‘bad governance’ of previous regimes as hav-
ing been responsible for the genocide, and have adopted a series of radical
reforms in such fields as local administration, agriculture, education and
health care. In some of these reforms, a tendency towards elitism as well as
radical social engineering has been observable (Ansoms, 2008, 2009). While
it is clear that these political behaviours derive directly from the RPF’s com-
plete victory in the civil war, they have other roots, such as its members’
origins as guerrilla fighters, their ethnic affiliation as minority Tutsis and the
influence of other ‘African new leaders’ (Ottaway, 1999).36 In addition, the
revolutionary ideology has its roots in Rwandan history itself. Monopolies of
power and revolutionary changes in power relations have occurred repeat-
edly in Rwandan history. In this sense the RPF’s victory in 1994 was very
similar to the ‘social revolution’ that had occurred 30 years before.
3.4.2 Burundi
Following the conclusion of the Arusha Peace and Reconciliation Agree-
ment in 2000 and the subsequent transition period, Burundians approved
their constitution in a referendum held in 2005 (République du Burundi,
2005). Like Rwanda, multi-party democracy was chosen as a political system
(Article 75). A president is the head of state, but Burundi does not have a
Shinichi Takeuchi 51
Table 3.1 The ethnic power-sharing mechanisms defined in the Burundian 2005
Constitution
Vice- The two Vice-Presidents shall belong to different ethnic groups and
President different political parties. (Article 124)
Cabinet The cabinet must include a maximum of 60 per cent Hutu Ministers
and Vice-Ministers, and a maximum of 40 per cent Tutsi Ministers
and Vice-Ministers. (Article 129)
Security The Minister in charge of the National Defence Force should not
organisations belong to the same ethnic group as the Minister in charge of the
National Police. (Article 130)
Members of the same ethnic group should not account for more
than 50 per cent of the members of the Defence and Security
organisations. (Article 257)
Public Ethnic representation in public enterprises is assigned as a
enterprises maximum of 60 per cent for Hutus and a maximum of 40 per cent
for Tutsis. (Article 143)
National The National Assembly is composed of at least 100 members, with
Assembly 60 per cent being Hutu and 40 per cent Tutsi. (Article 164)
The election of members of parliament is carried out through a
proportional representation system with closed lists. Party lists must
have a multi-ethnic character, and take gender equality into
account. From every three candidates in order on a list, only two
can belong to the same ethnic group, and at least one of every four
must be female. (Article 168)
Senate The Senate is composed of: (1) two delegates of each Province,
belonging to different ethnic groups; (2) three persons from the Twa
ethnic group; and (3) the former Heads of the State. (Article 180)
Judiciary The composition of members of the Magistrate Upper Council
should take ethnic, regional and gender balance into account.
(Article 217)
Commune Neither of the principal ethnic groups should be represented beyond
67 per cent of the national total of Commune Administrators.
(Article 266)
the compositions of the two main ethnic groups (Tutsi and Hutu) in organ-
isations related to state power. The system, designed along consociational
lines,37 is based on the idea of sharing state power among political elites.
In this power-sharing system, ethnic quotas are calculated on an individ-
ual basis; that is, Tutsi members from Tutsi-led parties (like UPRONA) but
also from Hutu-led parties (like CNDD-FDD) are counted as Tutsis.38 This
arrangement, together with the effect of the electoral rule obliging parties to
put candidates from different ethnic groups on their lists (Article 168), has
shaped the pattern of power struggle in Burundian politics.
The reasons for the particular institutional choice in Burundi deserve some
consideration. There is no doubt that the immediate cause of the power-
sharing agreement was the military stalemate in the civil war. Although
Tutsi-dominated government forces (Forces armées burundaises: FAB) con-
trolled the capital with overwhelming forces and equipment during the civil
war, they could not impose security on rural areas, where the Hutu rebels,
the CNDD-FDD and Forces nationales de libération- Parti pour la libération
du peuple hutu (FNL-Palipehutu), prevailed. The importance of mediating
efforts offered by Julius Nyerere and Nelson Mandela (naturally with the
cooperation of their governments) should be recognised, but the fact that
neither side could win the war by military means was the most fundamental
reason for the local parties to accept a power-sharing arrangement. In addi-
tion to this immediate reason, however, Burundian historical experience
should be considered as another non-negligible factor affecting the choice
of institutions. Looking back on Burundi’s modern history, power-sharing
arrangements were often adopted to appease political tensions; to deal with
political difficulties Burundian political leaders, from Rwagasore to Buyoya,
have resorted to such arrangements. The idea of ethnic equality in political
institutions was a familiar element of the Burundian experience.
2005 CNDD-FDD 43 21 0 64
Frodebu 23 7 0 30
Uprona 0 15 0 15
CNDD 3 1 0 4
MRC 0 2 0 2
Twa 0 0 3 3
Total 69 46 3 118
2010 CNDD-FDD 54 27 0 81
Frodebu-Nyakuri 3 2 0 5
Uprona 5 12 0 17
Twa 0 0 3 3
Total 62 41 3 106
the more the RPF negates ethnicity, the more it tends to be regarded through
the lens of ethnicity; its emphasis on national unity as well as its refusal to
recognise ethnicity have been interpreted from an ethnicity-centred logic, as
the RPF’s agenda for maintaining political power. This logic asserts that the
RPF does not want ethnicity to enter into politics, because it is supported by
the ethnic minority: the Tutsi.
With regard to post-conflict Rwanda, the dangers of social exclusion have
often been pointed out. There are, in fact, three related problems. The first
concerns state patronage; in the circumstances in which the political arena
has been dominated by the RPF, the Tutsis have had greater chances to be
the beneficiaries of state service provision, including opportunities for edu-
cation and employment in the sector, because of their patronage network
with political leaders. Since its seizure of power, the RPF has exerted its influ-
ence not only on the central bureaucracy but also on local administrations.
Today, leaders of local administrations are, almost without exception, ardent
supporters of the RPF.51 While the state patronage network also includes the
Hutus, as both Tutsi and Hutu are far from monolithic groups, it is neverthe-
less beyond doubt that the overwhelming majority of those who have been
excluded from the network are Hutus.
The second problem relates to economic disparities. As we have already
mentioned, rapid economic growth in Rwanda has widened economic dis-
parities, in a setting in which Tutsi returnees have tended to be the winners.
The third danger is related to Gacaca: a popular, participatory transitional
justice process for punishing genocide perpetrators. Its social impact has
been enormous, as the number of delivered judgements has reached around
1.4 million.52 The previous literature disagrees in its assessment of Gacaca,
but researchers who have recently observed the realities on the ground tend
to be critical of the practice;53 one of these studies has argued that, as a conse-
quence of the trials, Rwandan society has suffered from ‘a tension worsening
social cohesion and attitudes towards the “other group” ’ (Ingelaere, 2009).
The reasons for this were mainly twofold; on the one hand, in the huge
number of Gacaca trials, the relationship between accusers and accused has
always been ethnically fixed – the accusers were Tutsi and the accused were
Hutu; on the other hand, although RPF soldiers also committed atrocities
during the civil war, they have rarely been judged or punished (Human
Rights Watch, 2008). In spite of the Gacaca’s stated objectives of reconcil-
iation and national unity, it may in fact have rather widened the ethnic
divide.
It is highly possible that these three layers of exclusion have resulted in the
marginalisation of the Hutus. In addition to the fact that Hutus are in the
majority in the rural population, which makes it generally difficult for them
to benefit from macro-level economic growth, they have a smaller chance
of being included in the patronage network of political leaders. Moreover, it
is rare for a Hutu to have no family members who stood accused in Gacaca
Shinichi Takeuchi 57
trials. In short, the Hutus are quite likely to perceive themselves as suffer-
ing from HIs with regard to both the political as well as the socioeconomic
dimensions.54
In the case of examining perceived HIs in Burundi, it is better to sepa-
rate the perception among political elites from that among ordinary people,
because the ethnic power-sharing system has drastically transformed pat-
terns of political antagonism. The institutional change prohibited politicians
from relying on ethnic logic in organising a political party. As a result, Tutsi
elites chose to enter into Hutu-led parties in pursuing chances to become
MPs, and vice versa. Political elites ceased to use ethnicity as a tool for
mobilising the population in power struggles.
Although the Burundian power-sharing system to date has succeeded in
containing ethnic mobilisation, it fails to institutionalise power struggles.
Since 2010 the country has once again been faced with a rebel movement,
because after the election boycott the FNL rearmed and launched sporadic
attacks. Although they once accepted the ceasefire and the integration, they
defected from the security organisations and returned to the bush to fight.55
Political antagonism between two Hutu-led parties is a new development,
but the outlaw features of the power struggle, such as political violence,
oppression and terrorism, have often been witnessed in modern Burundian
history.
Among ordinary Burundians, CNDD-FDD has remained relatively popu-
lar since its election victory in 2005. The results of the Communal election
in 2010, in which CNDD-FDD gained 64 per cent of the members of the
Communal Council, were an indication of its popularity, especially in rural
areas.56 The main reasons for this popularity seemed to be the improvements
in living standards; in addition to the mere fact of the end of war, policies
ensuring free-of-charge access to primary school and health care have par-
ticularly served to garner popular support.57 It is safe to say that the actual
popularity of the CNDD-FDD is not based solely on ethnic preferences.
Nevertheless, the ethnic problem has remained unsolved among the popu-
lation. The most visible evidence for this is the camps for internally displaced
people (IDPs) that are located throughout the country.58 Burundian IDPs are
mainly Tutsis, who were chased out by their Hutu neighbours during the
civil war in the 1990s. Although the fighting has now ended, the IDPs prefer
to stay in the camps for fear of persecution. In everyday life, the legacies
of the civil war always remain a reality;59 ordinary people continue to hold
memories of fear in terms of ethnicity. It may be possible that such fear could
be manipulated for another mobilisation.60
similar vein, the RPF’s mode of governance reflects not only its total victory
in the war but also Rwanda’s modern history, in which revolutionary polit-
ical change has taken place repeatedly. Institutions for conflict prevention
should therefore be designed taking the historical background into account.
Notes
1. Although no census on the ethnic groups has been carried out in either country,
the approximate proportions of the three groups among their total populations
are said to be almost the same: the Tutsi account for around 15 per cent and the
Hutu 85 per cent, while the proportion of the Twa is less than 1 per cent.
2. As comparative studies between Rwanda and Burundi, Lemarchand’s works (1970,
2006, 2009) are the most notable. Uvin (2010) is also worth mentioning in terms
of research on conflict prevention.
3. The author has continuously conducted field studies since 1999 in Rwanda, and
visited Burundi for research during three months in total, in 2010, 2011 and 2012.
4. The origins of the Tutsi and Hutu have been one of the hottest issues in the
history of the two countries. Although we do not have enough space to fol-
low the debates, it should be stressed that the thesis attributing the origin of
the traditional kingdoms to the migration of and conquest by the Tutsi is seri-
ously questioned today. Recent studies have clarified that the group identities
of Tutsi and Hutu were ambiguous, and their border was blurred in the pre-
colonial period. As representative studies, see C. Newbury (1988), Schoenbrun
(1993, 1998), Chrétien (2000) and D. Newbury (2001).
5. For the importance of the social constructivist view in the analysis of HIs, see
Stewart (2008a: 9–12).
6. The Great Lakes region, including south-western Uganda, Rwanda, Burundi,
north-western Tanzania and part of eastern DRC, was the site of a number of
pre-colonial kingdoms, among which only Rwanda and Burundi became modern
sovereign states. These kingdoms shared a stratified social composition, partic-
ularly between pastoralists and agriculturalists (d’Hertefelt et al., 1962; Maquet,
1971); in the case of Rwanda and Burundi, pastoralist Tutsis took the domi-
nant position over the agriculturalist Hutus, although it should be noted that
the distinction between the two groups has been ambiguous.
7. The concept was characterised by elitism (Vansina, 2001), connected with state
power, as well as pastoralism, their main socioeconomic activity (Nkurikiyimfura,
1994).
8. Batare refers to family members who were descendants of King Ntare Rugamba
(reigned from 1795 to 1852), while Bezi were descended from Mwezi Gisabo
(1852–1908). The Burundian state expanded considerably during the reign of
Ntare, who, in order to consolidate his territorial gains, appointed his sons to
administer newly acquired provinces. As a consequence of this practice, con-
necting dynastic families systematically with politically important positions, the
Ganwa as a group began to play a decisive role in politics. King Mwezi, on the
other hand, tried to remove his predecessor’s influence and appoint his own sons
to politically important positions, thus exacerbating the rivalry between the two
groups (Lemarchand, 1970: 311).
9. In addition to the Ganwa, the ethnic identity of the Hima seemed to be much
more strongly perceived in Burundi than in Rwanda. While the Hima is a group
60 Rwanda and Burundi
related to the Tutsis, living in Uganda, Rwanda and Burundi, its uniqueness and
independence as a group have been different depending on the political as well
as regional context. Comparing the description of ‘Hima’ in Dorsey (1994) and
that in Eggers (1997), the latter clearly emphasises its uniqueness. For the Hima
in Burundi, see also Lemarchand (1994: 81–82).
10. A good example on this point is the role of the Bashingantahe, elders who are
in charge of justice in local communities (Laely, 1992; Naniwe-Kaburahe, 2008).
‘Although commonly (though not exclusively) Hutu, they were fully recog-
nised within the Burundi political system in a way unknown – even adamantly
opposed – in Rwanda under the Nyiginya dynasty’ (D. Newbury, 2001: 275).
11. Strictly speaking, the Belgian administration on behalf of the League of Nations
and the United Nations is different from colonial rule. In this chapter, how-
ever, the term ‘colonial’ will be used because of the similarity of power relations
between foreigners and natives.
12. The elimination of Hutus from the administration proceeded following the 1920s
in the process of administration reform, which considerably reduced the number
of chiefdoms and sub-chiefdoms under the name of administrative rationalisa-
tion. In Burundi, all 27 posts for Hutu chiefs were struck off between 1929 and
1945, as a result of rationalising chiefdoms from 133 to 35 (Gahama, 2001: 104).
No chiefdom among the 45 that existed in Rwanda as of 1 November 1959 was
headed by a Hutu chief; only 10 sub-chiefdoms among 559 had Hutu sub-chiefs
(Reyntjens, 1985: 269).
13. The term ‘social revolution’ implies a connotation of social progress, moving from
feudalism (rule by the Tutsi minority) to democracy (rule by the Hutu major-
ity). This terminology has therefore been preferred by Hutu elites, who took
power after independence. In contrast, Tutsi elites tend to avoid the term. The
present Tutsi-led government calls the incident ‘the first massacres of Batutsi’,
website of Government of Rwanda, http://www.gov.rw/page.php?id_article=56
(last accessed 1 May 2013). In this chapter, the term is used within quotation
marks.
14. For detailed studies of the ‘social revolution’, see Lemarchand (1970) and
Reynjens (1985).
15. Michel Micombero, the first president after the overthrow of the monarchy,
was of mixed Tutsi–Hima origin and from a family that did not rank high in
traditional prestige (Eggers, 1997: 84). See also Lemarchand (1970).
16. For the Burundian genocide in 1972, see Chrétien and Dupaquier (2007),
Lemarchand (1994).
17. Examples of the extreme ethnic disparity in this period were shown in Nkurunziza
and Ngaruko (2008: 76); the Tutsi monopolised almost all posts in the public sec-
tor, such as ministers, provincial governors, ambassadors, army members (from
officers to the rank and file), policemen, state-owned company directors and
magistrates.
18. Elites from central Rwanda monopolised the core of political power under the first
president (G. Kaybanda), who was originally from Gitarama (Reyntjens, 1985).
Following the coup, Rwandan politics tended to be dominated by political elites
from the north-western part of the country, which was where Habyarimana and
his wife had come from (Prunier, 1995).
19. Data on Rwandan and Burundian political elites are available in series of annuals
‘L’Afrique des Grands Lacs’ issued by the Centre d’étude de la région des Grands
Lacs d’Afrique. Every annual lists the names of members of the political elite
Shinichi Takeuchi 61
such as cabinet members, governors, top officers in the army and ambassadors,
with their political party, ethnic affiliation and regions of origin. The data clearly
show that the Tutsis who had been refugees have occupied important posts in
the government. For instance, top officer positions in the Rwandan armed forces
have always been monopolised by the former Tutsi refugees of the RPF, namely,
the former guerrilla fighters.
20. Despite the deepening Tutsi–Hutu divide during the colonial period, Rwagasore
succeeded in including Tutsis as well as the Hutu population in the UPRONA, thus
making it a nationalistic mass party. In addition to his charismatic leadership,
the principle of incorporating equivalent numbers of Hutu and Tutsi members at
every level of the party organisation was crucial to gaining support from ordinary
Hutus (Lemarchand, 1970: 330).
21. Each of the five governments between 1963 and 1965 comprised an almost even
proportion of Hutu and Tutsi ministers (Lemarchand, 1970: 368).
22. Unlike the case of the mass killing in 1972, the international community fiercely
reacted against the incident in 1988 and demanded that the Buyoya government
improve ethnic relations. The government was thus obliged to show some results
of ‘reconciliation’ for outsiders (Lemarchand, 1994: 128–130).
23. For example, among 48 high-ranking members (ministers and cabinet directors)
in the Buyoya cabinet in 1998, 19 were Hutu, 19 were Tutsi and 10 others were
of ethnically unknown origin. In the same year, all top officers in the Burundian
armed forces and the gendarmerie were Tutsi (Marysse and Reyntjens dir. 1999:
385–389).
24. According to the data from the ‘L’Afrique des Grands Lacs’ annuals, ethnic equiv-
alence in the army seemed to have been achieved around the period of the general
elections in 2005.
25. Before the systematic expulsion, nearly half the students in secondary schools
and universities were Tutsi (Munyarugerero, 2003: 134). For the Tutsi expulsion
in 1973, see also Reyntjens (1985: 501–504).
26. At the beginning of the 1990s, both Rwanda and Burundi were among the least
urbanised countries in sub-Saharan Africa. Burundi was in the same position in
2008, with almost 90 per cent of its total population living in rural areas. How-
ever, rapid urbanisation was seen in Rwanda in the same period, as its urban
population rate increased from 5.4 per cent in 1990 to 18.34 per cent in 2008
(World Development Indicators).
27. A tremendous number of Tutsi former refugees returned after the end of the civil
war. Although there have been various estimates of the number of returnees, it
is safe to say that 600,000 to 700,000 former refugees returned (Huggins, 2009:
69; Bruce, 2009: 112). Ansoms (2009) and Silva-Leander (2012) pointed out the
concentration of Tutsi returnees in urban areas. The RPF-led government, which
is eager to involve the Rwandan Diaspora in the development of the national
economy, has adopted several measures to promote this, such as permission
for dual citizenship (Plaza and Ratha eds, 2011). It is highly probable that the
part of the Rwandan Diaspora that has returned to the homeland since 1994 is
overwhelmingly Tutsi.
28. This fact is clearly shown in the evolution of the percentage share of GDP per
quintile of population. While the share of the top 20 per cent quintile has sharply
enlarged, reaching 51.4 per cent in 2000 from 39.1 per cent in 1985, the bottom
20 per cent shrank from 9.7 per cent to 5.4 per cent in the same period. This
means that ‘almost all the growth generated in the last few years has gone to
62 Rwanda and Burundi
the top quintile’ (UNDP, 2007a: 19), leaving the average income of the other
four quintiles virtually unchanged since 2003. Rwanda’s Gini coefficient in 2000
was 0.468, considerably aggravated from 0.289 in the mid-1980s (UNDP, 2006,
2007b); it further increased to 0.510 in 2006 (Republic of Rwanda, 2007: 13).
29. According to data from the census held in Rwanda in 2000/01, the proportion
of the population whose annual income or expenditure for consumption was
under the national poverty line (64,000 Rwanda Francs, corresponding to roughly
US$150) was over five times higher in rural areas than that in the capital, Kigali
(République rwandaise, 2002: 33).
30. In the Burundian army, members from Bururi had outnumbered those from other
regions since the 1960s. After the coup in 1966, the army established the Conseil
national de la révolution (CNR) as a supreme decision-making body. In 1968,
among seventeen officers included in the CNR, eight were from the Bururi (five
Tutsi, two Hutu and one Hima) and seven were non-Tutsi (three Hutu, two Hima,
one Ganwa and one so-called ‘Swahili’) (Lemarchand, 1994: 79).
31. Compared with the national average, the school enrolment ratio in Bururi was
1.5 times higher, its number of teachers per classroom was twice as large, and
its population per hospital was only 107,000 in comparison with the national
average of 266,000 (Nkurunziza and Ngaruko, 2008: 73).
32. Bururi’s privileged position in public health was no more obvious in recent
statistics; the population per hospital in Bururi province was 129,663 in 2007
(ISTEEBU, 2009: 132); this figure was the fifth-highest rank among 17 provinces.
However, Bururi seemed to maintain its prominence in secondary schools; its
enrolment rate (21.6 per cent) – the second-highest figure, after only Bujumbura
city (37.7 per cent) – was remarkably higher than other rural provinces, among
which the highest (Makamba) was only 13.1 per cent (République du Burundi,
2006: 62–64).
33. IRIN, 23 May 2003. ‘Rwanda, Government to Start Legal Proceedings against
Banned Party’.
34. For the position of women in post-conflict Rwandan politics, see Burnet (2008).
35. In the case of the Senate, all 26 members are either elected from among or
appointed by the Executive Committees of local administrations, the President
of the Republic, the Forum of Political Organisations, and academia.
36. In Ottaway’s analysis, Eritrea, Ethiopia and Uganda are typical countries under
the rule of the ‘African New Leaders’, and Rwanda and the Democratic Repub-
lic of the Congo are their possible followers. Tendencies towards authoritarian
rule and social engineering can be commonly observed among these countries.
In addition, the ideological influence of Museveni’s Uganda on Kagame’s Rwanda
has been well known (Prunier, 1998).
37. Lijphart (1977). For analyses of the Burundian power-sharing system from the
viewpoint of consociationalism, see Sullivan (2005), Lemarchand (2006) and
Vandeginste (2006, 2008, 2009).
38. This arrangement was in accordance with the argument of Hutu parties dur-
ing the negotiations, whereas Tutsi parties had insisted on counting through
the party base, which means that only members from Tutsi parties should be
considered as representatives of the Tutsi ethnic group.
39. The effective party numbers calculated from the results of the latest elections
are 1.53 for Rwanda and 1.49 for Burundi. As a consequence of Rwanda’s leg-
islative election in 2008, the RPF gained 47 of the 53 seats in the Lower House,
PSD (Parti social-démocratique) 7, and PL (Parti libéral) 4. As mentioned earlier,
Shinichi Takeuchi 63
countries), Burundi was in the 181st position among 183 countries, better than
only the Central African Republic and Chad. In addition, during the author’s field
research, a number of interviewees from the civil society pointed out corruption
has intensified under the rule of the CNDD-FDD, and attributed the fact to the
behaviour of new elites, who were busy profiting from their opportunities for
‘eating’.
49. Marysse et al. (2006) contrasts Rwanda’s ‘aid darling’ status with the DRC’s and
Burundi’s ‘aid orphan’ status. Although the contrast between Rwanda and the
DRC is convincing, that between Rwanda and Burundi is not. Comparing the
average net ODA received per capita from 2000 to 2009, Burundi received US$
44.0, which counted for 72 per cent of Rwanda’s US$ 60.7 (data from WDI). Both
of these figures are not far from the average for sub-Saharan countries during the
same period (US$ 58.5).
50. Due to the sensitiveness of questions concerning ethnicity, large-scale social
inquiries were not carried out in Rwanda and Burundi in the research project.
Therefore, unlike other chapters, the perceptions of HIs will be estimated in this
chapter on the basis of analyses made thus far on the history, institutions, and
recent political processes in each country.
51. See Ingelaere (2010). The voting system through queues has confirmed this ten-
dency. In local elections in Rwanda, voters are requested to stand in a queue
behind their preferred candidates.
52. ‘Open Justice: Q&A with Minister of Justice, Tharcisse Karugarama’ Official
Website of the Republic of Rwanda, http://www.gov.rw/OPEN-JUSTICE-Q-A-with-
Minister-of-Justice-Tharcisse-Karugarama (last accessed 1 May 2013). According
to an official document distributed at the closing ceremony (Republic of Rwanda,
2012b), the total number of cases tried by Gacaca court was 1,958,634.
53. See, for example, Waldorf (2006), Drumbl (2007), Rettig (2008), Ingelaere (2009),
Thomson (2011).
54. Ingelaere (2010) supports this point. According to his survey, subjective political
representation rankings are contrasting between Tutsis and Hutus. Tutsis gener-
ally feel that they were more politically represented after the civil war of the 1990s
than before, but Hutus tend to consider themselves to be in the opposite position.
55. A UN expert panel revealed that FNL had entered the DRC for remobilisation. See
UNSC (2010: paras 113–119).
56. Although the main opposition groups such as FNL and Sahwanya–Frodebu
accused the ruling party of carrying out massive fraud and withdrew from
subsequent elections, the general popularity of the CNDD-FDD seemed to be
undeniable except in several provinces, such as Bujumbura Rural, Bururi, and
Bujumbura Marie. The popularity of FNL was particularly strong in the Province
of Bujumbura Rural (CENI, 2010b).
57. Considering that the execution of these policies will result in overloading gov-
ernmental finance, it is uncertain how long these policies and therefore the
popularity of the ruling party will be sustainable. During the author’s field visit in
2010, the diplomatic corps, international agencies, and the civil society were all
worried about the danger to the national finances that the free-of-charge policies
would create.
58. As of 2009, Burundi had 137 IDP camps, with camps in all 17 provinces; the total
number of IDPs amounted to 157,167 (Rwabahungu and Nintunze, 2009: 9).
59. A civil society activist pointed out that, as a result of the civil war, contacts
with different ethnic groups have considerably diminished, even on ceremonial
Shinichi Takeuchi 65
4.1 Introduction
Although the number of violent conflicts and civil wars has been on the
wane since the mid-1990s, today there are still around 30 ongoing violent
intra-state conflicts around the world. The majority of these consist of fight-
ing between groups who are united by a common ethnic or religious identity.
Since 1945, the identity basis of conflicts has become increasingly explicit,
with the proportion of all conflicts that are labelled as ‘ethnic’ increasing
from 15 per cent in 1953 to nearly 60 per cent by 2005 (Stewart and Brown,
2007). In the 1990s, politicians, journalists and intellectuals propagated the
popular view that the breakdown of multiethnic societies and the emergence
of ethnic conflicts, in both Africa and Eastern Europe, was due to the ‘erup-
tion of ancient and irrational tribal antagonisms’ (Turton, 1997: 80). Since
then, and despite the fact that a substantial amount of research has shown
the fallibility of these claims, these views remain prominent in non-scholarly
circles of society.
Indeed, ‘primordial’ ethnic hatred or, for that matter, cultural differences
are generally an insufficient explanation for the emergence of violent con-
flicts. As Abner Cohen aptly argued: ‘Men may and do certainly joke about or
ridicule the strange and bizarre customs of men from other ethnic groups,
because these customs are different from their own. But they do not fight
over such differences alone. When men do, on the other hand, fight across
ethnic lines it is nearly always the case that they fight over some funda-
mental issues concerning the distribution and exercise of power, whether
economic, political, or both’ (Cohen, 1974: 94). In the same vein, Ted Gurr
argued in his well-known book Why Men Rebel (see Gurr, 1970) that a com-
bination of deprivation-induced discontent and a sense of group cultural
66
Arnim Langer 67
Note: The ethnic composition shown here only takes into account
Ivorian nationals.
Source: République de Côte d’Ivoire (2001: 68).
Arnim Langer 71
Population % Population %
Table 4.4 Religion of different ethnic groups in Côte d’Ivoire in 1998 (%)
Note: The data shown here only take into account Ivorian nationals.
Source: République de Côte d’Ivoire (2001: 67).
non-Ivorians are Muslim (about 70 per cent), their presence in Côte d’Ivoire
tilts the religious balance in favour of Muslims at the national level. The
Muslim/Christian balance becomes 39/30 if foreigners are included.
Religious differences appear to reinforce ethno-regional differences.
Table 4.4 shows the religious composition of the five major ethnic groups.
While the Akan and Krou ethnic groups are predominantly Christian, the
northern ethnic groups, Voltaic and Northern Mandé, are mostly Muslim.
Consequently, almost 50 per cent of the people in the northern part of the
country (comprising the Savanes, Vallée du Bandama, Zanzan, Denguélé,
Worodougou and Bafing regions) are Muslim. This percentage increases to
63 per cent if one excludes the most southerly northern region, Vallée
du Bandama. Although the northern population is predominantly Muslim,
about 70 per cent of all Muslims live in the southern regions of the country.
In 1998, if one includes foreign nationals, Muslims were the largest reli-
gious group in the southern part of the country with about 35 per cent of
the population, compared to 34 per cent who are recorded as Christians.
Excluding foreign nationals, however, Christians were the largest religious
group. Although the general perception in Côte d’Ivoire and in the interna-
tional media is that the north is basically Muslim and the south is Christian,
these figures indicate that the actual situation is somewhat more nuanced
and complicated.
Table 4.5 Income per capita in 1965 and 1975 (constant 1965 CFA franc)
4.2.3 The politicisation of ethnicity and the descent into civil war
Although the combination of several years of structural reforms and the
1994 CFA franc devaluation led to a significant recovery in economic
growth, most people did not benefit from it (see, for example, Azam, 2004).
Côte d’Ivoire also continued to experience serious political confrontations
and ethnic tensions, especially in the lead-up to the October 1995 pres-
idential elections. A growing northern consciousness, combined with an
increased dissatisfaction among the northerners towards their (perceived)
position in society, was an important change that contributed to the esca-
lation of ethnic tensions at the beginning of the 1990s. The distribution
of an anonymous document called ‘Le Charte du Grand Nord’ (Charter
of the North) in 1992 illustrated the changed attitudes of the northerners
regarding the sociopolitical system in general and the Baoulé group in par-
ticular. The Charter ‘called for fuller recognition of the Muslim religion [ . . . ],
more efforts to reduce regional inequalities, greater political recognition of
the north’s political loyalty during the upheavals of the 1980s and [ . . . ] an
end to Baoulé nepotism in recruitment to public jobs’ (quoted in Crook,
1997: 226). Northern grievances and dissatisfaction were not limited to the
economic and political sphere, but also had a ‘cultural status’ or religious
dimension. The call for greater recognition of the Muslim religion in Côte
d’Ivoire clearly illustrates this. Côte d’Ivoire’s 1960 constitution was secular
in nature, but as a direct consequence of Houphouët-Boigny’s long stay in
power as head of state, there was a growing blurring in perceptions of the
separation of religion and state. Although the political inclusion of people
from the north in various state institutions helped to counter this impres-
sion, the creation of an immense Basilica in Yamoussoukro by Houphouët-
Boigny in the late 1980s was perceived by many Muslims as a clear indication
of the superior position allocated to Christianity (particularly Catholicism)
in Côte d’Ivoire.
The emergence of a new opposition party, Rassemblement des républi-
cains (RDR), in 1994, reflected a further split in Côte d’Ivoire’s political
elite. The RDR aimed to draw support from people with a northern and/or
Muslim background, predominantly found among the Voltaic and Northern
Mandé ethnic groups. Alassane Ouattara – who in 1994 was again working
in Washington – would soon become their political leader. The emergence of
this new party confronted President Bédié with a serious challenge because
the RDR was likely to reduce PDCI’s electoral support in the northern regions
(Crook, 1997). In response, as Richard Crook notes, ‘Bédié’s initial strategy
was familiar to any student of electoral politics: he stole the opposition’s
clothes, and adopted a policy of Ivorian nationalism, under the slogan of
the promotion of “Ivoirité” (Ivorianness)’ (ibid.: 227).
Although Bédié claimed that the sole aim of the concept of Ivoirité was to
create a sense of cultural unity among all the people living in the territory of
Arnim Langer 77
Côte d’Ivoire, it is widely recognised that it was introduced for specific polit-
ical reasons. First, in contrast to Houphouët-Boigny, Bédié could not count
on the loyalty of the foreign migrants, and he was therefore very concerned
that they might switch their allegiance to the RDR because of their ethno-
cultural and religious affinities with its putative leader, Alassane Ouattara.
Bédié therefore withdrew foreigners’ right to vote once the RDR emerged on
the political scene in October 1994.
Secondly, Bédié also used the concept of Ivoirité to prevent Alassane
Ouattara (the RDR leader with northern origins) from participating in the
presidential elections in 1995. Indeed, in line with the ideology of Ivoirité,
the Electoral Code was amended in December 1994. From then on, individu-
als were only allowed to stand in presidential and parliamentary elections if
both their parents were Ivorian by birth. The revised 1995 Electoral Code fur-
ther stipulated that the candidate himself must have lived in the country in
the five years preceding the elections, and never have used another nation-
ality than that of Côte d’Ivoire. On the basis of the revised Electoral Code,
Bédié and his supporters claimed that Ouattara was ineligible to participate
in the presidential elections because his father was a Burkinabe. Ouattara,
however, countered that he was Ivorian and that he was a descendant of
the royal family of the ancient Kingdom of Kong (situated in the north-
eastern part of Côte d’Ivoire). Ouattara was also accused of having possessed
a Burkinabe passport when he was Governor of the Central Bank for West
African States (BCEAO). In response to this allegation, Ouattara stated that
he only possessed a diplomatic passport and that he never had Burkinabe
nationality. Ouattara was also ineligible according Bédié supporters because
he had lived in the United States since his resignation as Prime Minister
in 1993. However, the Supreme Court (which was responsible for judging
the eligibility of presidential candidates) did not have to proclaim Ouattara’s
non-eligibility officially, as he announced in September 1995 that he would
not participate in the presidential elections.
However, the introduction of the ideology of Ivoirité had an impact far
beyond the political sphere, because it led to a general erosion of northern
Ivorians’ social standing and cultural status, making them de facto second-
class citizens in Côte d’Ivoire.
As a result of Ouattara’s exclusion, the RDR boycotted the October 1995
presidential elections. FPI leader Laurent Gbagbo also decided to boycott the
elections, claiming that the electoral process had been manipulated. Due
to the absence of his main rivals, Bédié achieved a landslide victory in the
October 1995 elections. Until the coup d’état in December 1999, the two
opposition parties, RDR and FPI, together formed the Front Républicain.
In sharp contrast to Houphouët-Boigny, Bédié largely stopped the efforts to
balance the different ethno-regional interests and parties, and increasingly
started to favour people from his own ethnic group, the Baoulé (Dozon,
2000). Importantly, the ‘baoulisation’ of the Ivorian state institutions was
78 Ghana and Côte d’Ivoire
both a cause for and consequence of the increased ethnicisation of the polit-
ical system under Bédié. The ‘baoulisation’ of the Ivorian state in the 1990s
is illustrated in Table 4.6, which shows the ethnic composition of Ivorian
governments in the period 1980–2003. Each group’s relative representation
is calculated by dividing its percentage proportion of government minis-
ters by its percentage size in the entire population. Consequently, unity
means proportional representation; figures higher than one point to over-
representation and less than one to underrepresentation. The data show that
Bédié clearly favoured his own group, the Baoulé, with respect to ministerial
appointments.
In addition to the baoulisation of the political-administrative sector, Bédié
also began to change the ethnic composition of the military forces in favour
of his own ethnic group (Contamin and Losch, 2000). The ethnic tensions
that stemmed from increased favouritism towards the Baoulé were com-
pounded by general discontent in the armed forces due to a gradual decline
in their status during the 1990s, mainly arising from reduced expenditures
due to the precarious financial situation (Kieffer, 2000).
The grievances within the armed forces triggered a coup d’état in
December 1999, initiated by a group of non-commissioned officers who
claimed they were owed financial compensation for their participation in
an international peacekeeping mission in Central Africa (Kieffer, 2000). This
protest movement quickly developed into a large-scale mutiny, at which
stage more senior officers got involved. Although the coup d’état appears
to have initially originated in individual grievances, these grievances and
fears of exclusion cannot be separated from what was happening in the rest
of society.
As Kieffer argues, the opposition parties’ discourse of exclusion and Baoulé
domination of the Ivorian state is likely to have had an important impact on
the attitudes of the young non-commissioned officers involved in the coup
d’état (Kieffer, 2000). Importantly, at the time of the coup d’état in December
1999, both the Baoulé overrepresentation and the underrepresentation of
the northerners in government were the most severe in Côte d’Ivoire’s
postcolonial history up to that point (see Table 4.6; and, Langer, 2005).
Following Bédié’s removal from power, the military forces established the
Comité national de salut public (CNSP), headed by General Gueï, who was
a Yacouba, one of the ethnic groups belonging to the Southern Mandé eth-
nic group. In line with ‘Le modèle Houphouétiste’, Gueï initially promoted
the ideals of national integration and reconciliation, and openly opposed
the ideology of Ivoirité (Akindès, 2004). After negotiations between the dif-
ferent political parties and the military junta, a transitional government was
installed on 4 January 2000. As Table 4.6 shows, this transitional government
had a very inclusive character with a reasonably fair distribution of power
among the major ethno-regional groups. However, after several months in
office, Gueï’s political objectives and strategy changed drastically. In contrast
Table 4.6 Ethnic composition of government, 1980–2003
Nov-80 Jul-86 Oct-89 Nov-91 Dec-93 Jan-96 Aug-98 Jan-00 May-00 Jan-01 Aug-02 Sep-03
% RRa % RR % RR % RR % RR % RR % RR % RR % RR % RR % RR % RR
Akan 0.49 1.15 0.41 0.98 0.47 1.12 0.61 1.46 0.52 1.24 0.52 1.24 0.59 1.40 0.50 1.19 0.30 0.71 0.46 1.09 0.52 1.24 0.40 0.95
Baoulé 0.22 1.32 0.24 1.44 0.20 1.20 0.17 1.02 0.24 1.44 0.28 1.68 0.31 1.86 0.13 0.78 0.04 0.24 0.11 0.66 0.13 0.78 0.09 0.54
Krou 0.19 1.19 0.20 1.37 0.20 1.37 0.17 1.16 0.24 1.89 0.21 1.65 0.16 1.26 0.13 1.02 0.22 1.73 0.29 2.28 0.19 1.50 0.23 1.81
S. Mandé 0.05 0.42 0.10 0.93 0.13 1.21 0.04 0.37 0.04 0.40 0.10 1.00 0.06 0.60 0.08 0.80 0.17 1.70 0.18 1.80 0.16 1.60 0.07 0.70
N. Mandé 0.08 0.59 0.17 1.07 0.13 0.82 0.09 0.57 0.08 0.48 0.07 0.42 0.03 0.18 0.17 1.03 0.17 1.03 0.07 0.42 0.13 0.79 0.19 1.15
Voltaic 0.14 0.91 0.10 0.61 0.03 0.18 0.09 0.55 0.12 0.68 0.10 0.57 0.13 0.74 0.13 0.74 0.08 0.45 0.00 0.00 0.00 0.00 0.12 0.68
Southernersb 1.04 1.06 1.19 1.22 1.23 1.28 1.25 1.10 1.06 1.44 1.34 1.08
Northerners 0.76 0.84 0.50 0.56 0.59 0.50 0.47 0.88 0.73 0.21 0.38 0.91
Northerners, 0.56 0.69 0.41 0.46 0.49 0.42 0.39 0.74 0.62 0.17 0.32 0.76
incl. ‘Foreign
Ivorians’
Ratio N/S 0.73 0.79 0.42 0.46 0.48 0.39 0.38 0.80 0.69 0.14 0.28 0.84
Total N = 37 N = 41 N = 30 N = 23 N = 25 N = 29 N = 32 N = 24 N = 23 N = 28 N = 31 N = 43
Notes: Government: president of the republic, ministers of state and regular ministers. Deputy ministers were not included in the calculations.
a. Relative Representation (RR) is calculated by dividing an ethnic group’s relative proportion in government by its relative size in the entire population.
RR figures with respect to the November 1980 government were based on the ethnic data from the 1975 census. From July 1986 until November 1991, the
ethnic demography data from the 1988 census were used, and from then onwards we have used the 1998 census data to calculate the RR figures.
b. The Northern Mandé and Voltaic ethnic groups together make up the group of ‘northerners’, while the three other groups (the Akan, Krou and Southern
Mandé) constitute the ‘southerners’. In the period from November 1980 to November 1991, ‘northerners’ and ‘foreign Ivorians’ accounted for 4.2 million
people or 39.2 per cent of the population; in the period from November 1991, this was 6.2 million people or 40.6 per cent of the population.
Source: Langer (2008: 173).
79
80 Ghana and Côte d’Ivoire
the northern part of the country and were referring to themselves as the
Mouvement Patriotique pour la Côte d’Ivoire (MPCI). The main grievances
put forward by the insurgents related to the land ownership laws, the cri-
teria of eligibility for presidential elections, the question of identity cards
and the political domination of the northerners by southerners (Dembélé,
2003). Although the vast majority of its forces had a northern background,
the MPCI claimed to have no specific ethnic, regional or religious affiliation.
Although the northern rebellion failed to overturn the Gbagbo regime,
due to the military intervention by France at the end of September 2002,
by then, MPCI firmly controlled the northern part of the country. France’s
intervention led to a situation in which there was no clear winner or loser.
This forced the conflict parties to seek a negotiated solution to their differ-
ences, which, once agreed, proved extremely difficult to implement (Langer,
2010a). After a string of failed peace agreements the conflict parties even-
tually agreed to hold presidential elections in October 2010 in order to
advance the peace process, but the outcome of these elections was chaos
and violence.
The main responsibility for the electoral debacle lay with Laurent Gbagbo,
who refused to accept his defeat in the presidential elections, which inter-
national observers had uniformly declared free and fair. In response to
Gbagbo’s refusal to hand over power to the newly elected President Alassane
Ouattara, the northern insurgents took up arms again and started march-
ing on Abidjan, the largest city in the country and the seat of government.
The forces loyal to Ouattara quickly gained control of the whole country,
apart from Abidjan itself, where a fierce battle between the Ouattara forces
and those forces that remained loyal to Gbagbo emerged and lasted for
several weeks. The fighting eventually ended on 12 April 2011, when the
Ouattara forces (with considerable help from French and UN peacekeeping
forces) were able to enter the Presidential Residence and arrest Gbagbo and
his closest associates. In the wake of Gbagbo’s arrest, Ouattara was officially
inaugurated as president of Côte d’Ivoire.
One important task for President Ouattara in the short term will be to reas-
sure and convince the substantial proportion of the population who voted
for Gbagbo that he will be a president for all Ivorians, regardless of ethnicity,
religion, region of origin or political affiliation. This is a formidable task in
a conflict-ravaged and ethnically polarised society like Côte d’Ivoire. In the
long term, however, a durable peace in Côte d’Ivoire will depend ultimately
on the political elites agreeing on a new set of ‘rules of the (political) game’.
Ghana has had a turbulent postcolonial political history, with long periods
of political instability, repeated non-constitutional regime changes and, at
times, ethnic tensions.3 Despite this, since the introduction of the Fourth
82 Ghana and Côte d’Ivoire
and dances from all major ethnic groups (Lentz and Nugent, 2000); the con-
scious effort to ensure that radio and television programmes are broadcast
in all major languages (ibid.); and the custom that representatives from the
government attend the most important ethnic and/or traditional festivals
and durbars throughout the country on a regular basis. Similarly, examples
of some (symbolic) actions and practices which demonstrate the political
elites’ commitment to promoting and sustaining religious status equality
and inclusiveness include: the practice that representatives from all major
religions are present at official state functions; the state’s active organisa-
tional support for the annual Hajj pilgrimage to the Muslim holy sites in
Saudi Arabia; and the introduction of a new public holiday on the Muslim
festival of Eid-al-Adha in 1996.
Nkrumah also attempted to reduce the prevailing socioeconomic inequal-
ities, in particular between the north and the south. Ghana’s socioeconomic
north–south divide arises from a combination of circumstances and poli-
cies. Three main factors (which are, to a large extent, also applicable in the
case of Côte d’Ivoire) can be identified, the first of which is the geographi-
cal concentration of most agricultural activities and natural resources in the
southern regions; the northern regions predominantly rely on subsistence
farming. A second factor arises from the British colonial policy of investing
more heavily in those regions where exploitable resources such as gold, dia-
monds, timber and cocoa were available, or readily produced, and cheapest
to export. A third factor, which is important for understanding the persis-
tence of socioeconomic inequalities between the north and south, relates to
postcolonial development strategies and investment patterns.
The economic disadvantage of the Northern Region and, to a lesser extent,
the Volta Region compared with the rest of the country (especially with
regard to the Greater Accra Region) at the time of independence is illustrated
in Table 4.7. Despite Nkrumah’s efforts to reduce the north–south divide
and integrate the north into Ghana’s economy, the north remained gener-
ally much poorer in terms of income, infrastructure, education and medical
services at the end of the 1970s. The main reason for this was that the
postcolonial economy had an ‘endogenous’ tendency to favour the south
over the north in terms of the location of economic activities and invest-
ment, much like the situation in Côte d’Ivoire. Indeed, as Charles Anyinam
notes: ‘The economic policies and planning strategies adopted after inde-
pendence failed to alter the colonial legacy of lopsided development. The
development plans implemented between 1957 and 1980 tended to per-
petuate and reinforce the inherited inequalities in Ghana’s space economy’
(Anyinam, 1993: 449).
However, successive Ghanaian regimes did attempt to mitigate the
socioeconomic north–south divide. Thus, for instance, the Rawlings regime
(1983–2001) undertook a range of specific projects in the northern regions,
including the extension of the national electricity grid, the establishment
88 Ghana and Côte d’Ivoire
Table 4.7 Gross value added, per capita and per square mile, by
region, 1960
Notes:
a. UNDP (2007c: 35).
b. Ratio to the national average; calculated by this author.
Source: Data derived from the Ghana Statistical Service. Available at: http://www.ghanainfo.org/
(last accessed October 15, 2008).
% RR % RR % RR % RR % RR % RR % RR % RR
Akan 0.55 1.24 0.62 1.40 0.62 1.40 0.64 1.46 0.33 0.76 0.74 1.67 0.76 1.73 0.50 1.13
Ewe 0.09 0.70 0.08 0.59 0.23 1.78 0.07 0.55 0.33 2.56 0.00 0.00 0.00 0.00 0.29 2.20
Ga/Dangbe 0.09 1.10 0.08 0.93 0.15 1.85 0.14 1.72 0.22 2.68 0.05 0.63 0.06 0.71 0.07 0.86
Southerners 0.73 1.11 0.77 1.18 1.00 1.53 0.86 1.31 0.89 1.36 0.79 1.21 0.82 1.26 0.86 1.31
Northerners 0.18 0.84 0.15 0.71 0.00 0.00 0.14 0.66 0.11 0.51 0.16 0.73 0.18 0.82 0.14 0.66
No. N = 11 N = 13 N = 13 N = 14 N=9 N = 19 N = 17 N = 14
% RR % RR % RR % RR % RR % RR % RR % RR
Akan 0.50 1.13 0.57 1.30 0.43 0.87 0.50 1.02 0.65 1.32 0.53 1.08 0.79 1.61 0.61 1.24
Ewe 0.38 2.88 0.14 1.10 0.14 1.12 0.30 2.36 0.12 0.93 0.12 0.93 0.05 0.42 0.56 0.44
Ga/Dangbe 0.13 1.51 0.14 1.72 0.14 1.79 0.10 1.25 0.06 0.74 0.06 0.74 0.11 1.31 0.22 2.78
Southerners 1.00 1.53 0.86 1.31 0.71 1.02 0.90 1.29 0.83 1.19 0.71 1.02 0.95 1.36 0.89 1.29
Northerners 0.00 0.00 0.14 0.66 0.29 1.23 0.10 0.43 0.18 0.72 0.29 1.21 0.05 0.22 0.11 0.46
No. N=8 N = 14 N=7 N = 10 N = 17 N = 17 N = 19 N = 18
Note: ‘Northerners’ include individuals with a Mole/Dagbani, Gurma or Grusi background. ‘Southerners’ include individuals with an Akan, Ewe and
Ga/Dangbe background. Other smaller ethnic groups (such as the Mandé-Busanga and Guan) are excluded from these calculations. For the period 1954–
1979, the ethnic composition data are based on the 1960 census, while for the period 1981–2005 the 2000 Housing and Population census was used.
Source: Langer (2009: 543).
Arnim Langer 91
In both academic and policy circles, the concept of HIs is increasingly recog-
nised as an important factor for understanding the emergence of violent
conflicts in plural societies. While a substantial amount of empirical research
has found a significant association between the presence of HIs and the
risk of violent conflict, the causal mechanism linking these two phenom-
ena continues to be insufficiently understood. As argued in the introduction
of this chapter, further theoretical refinement is especially needed to explain
why HIs sometimes result in ‘ethnic’ violence, while in other cases similar
inequalities do not translate into violent group mobilisation and conflicts.
The cases of Ghana and Côte d’Ivoire clearly demonstrate that ethno-
regional or horizontal inequalities are in and by themselves insufficient to
produce violent conflict.
In previous work I highlighted the importance of the configuration of
multidimensional HIs as an important factor for understanding when ‘eth-
nic’ inequalities are more likely to escalate into violent conflicts (see Langer
2005). In particular, I argued that ‘ethnic’ violence is more likely to occur in
situations where political, social and economic HIs run in the same direc-
tion, namely, situations in which the same ethnic groups are politically
excluded and socioeconomically disadvantaged. In this chapter I attempt
to deepen our understanding of the process through which HIs may or may
not result in violent conflict. The chapter emphasises the importance of the
intervening variable of ethnic politics in understanding when HIs are more
likely to result in violent conflict; a factor so far ignored in the literature on
HIs. In Chapter 1, I argued that the presence of HIs may directly lead to an
ethnicisation of politics, which in turn may or may not escalate into violent
conflict. The analysis of Ghana and Côte d’Ivoire has shown that both coun-
tries have indeed known episodes where perceptions of HIs were introduced
into the political discourse and this subsequently contributed to the ethni-
cisation of politics and increased ethnic tensions. Yet while both countries
were confronted at different points in time with an ethnicisation of their
respective political systems and discourses, only in Côte d’Ivoire did ethnic
politics escalate into a violent conflict at the national level, which is an issue
we will turn to now.
In Côte d’Ivoire, while ethnicity and ethnic inequalities had an impor-
tant impact on Ivorian politics and occasionally engendered serious ethnic
92 Ghana and Côte d’Ivoire
power by a coup d’état. The failed coup d’état ultimately evolved into a pro-
tracted violent conflict, which arguably only ended in April 2011 with the
arrest of Laurent Gbagbo and the official inauguration of Alassane Ouattara
as President of Côte d’Ivoire.
The case of Ghana is interesting because it shows that successive Ghanaian
regimes have used similar strategies in managing their prevailing HIs (espe-
cially those between the northern and southern regions) and keeping ethnic
politics in check. Illustratively, successive Ghanaian regimes included north-
ern politicians and interests in the main political institutions and undertook
economic redistribution towards the deprived northern regions in order to
diffuse the sharp north–south cleavage. However, as in the case of Côte
d’Ivoire, the actual impact of the economic redistribution measures on
reducing the economic north–south divide was rather limited at best. While
keeping political and economic inequalities between the north and south in
check was an important factor for why Ghana did not experience a north
versus south mobilisation as we saw in Côte d’Ivoire, there are a number of
‘structural’ impediments to the mobilisation of the north as a group which
makes this inherently less likely to occur in the Ghanaian context, includ-
ing ‘the presence of a considerable number of relatively small ethnic groups,
which not only differ greatly regarding their cultural traditions, histories and
institutions but have also been in conflict with one and other at different
points in time’ (Langer, 2009: 544).
The successful management of political HIs throughout nearly the whole
postcolonial period appears to be a very important factor for understanding
why ethnic politics never did escalate into violent conflict in Ghana. More-
over, the ethnicisation of the political system that occurred following Adu
Boahen’s public accusations of ethnic favouritism and overrepresentation of
the Ewes under the Rawlings regime in the late 1980s is particularly useful to
illustrate this point. Indeed, the increased ethnicisation and ethnic tensions
which occurred in the wake of his accusations were basically kept in check
and eventually petered out relatively quickly mainly because the ‘objective’
HIs were rather moderate, as the analysis showed. Thus, in conclusion, it
appears that Ghana was not necessarily better at managing and reducing its
socioeconomic HIs than Côte d’Ivoire, but Ghana appears to have been bet-
ter at managing its political HIs, thereby reducing the appeal and mobilising
‘power’ of ethnicity and ethnic inequalities, which in turn helped to keep
ethnic politics in check.
Notes
1. The labelling of conflicts or violence as being ‘ethnic’ can be misleading, because
it could be seen to suggest that ethnic differences are the cause of these conflicts
(Alexander et al., 2000a). Instead of indicating a causal link between ethnicity and
94 Ghana and Côte d’Ivoire
conflict, in this paper the term ‘ethnic conflict’ solely refers to a conflict where the
different conflict parties have mobilised people to a significant extent along ethnic
lines.
2. This section draws on Langer (2005), (2008) and (2010a).
3. This section draws on Langer (2009) and (2010b).
4. The population data discussed in this section are based on Ghana’s 2000 Housing
and Population Census.
5
Beyond Ad hoc Power-Sharing:
Comparing South Africa and
Zimbabwe
Yoichi Mine
5.1 Introduction
95
96 South Africa and Zimbabwe
Zulu 23.8
Xhosa 17.6
Afrikaans 13.3
Northern Sotho (Pedi) 9.4
Tswana 8.2
English 8.2
Sotho 7.9
Tsonga 4.4
Swati 2.7
Venda 2.3
Ndebele 1.6
Other languages 0.5
Total 100.0
character like the American South, with variegated cultural groups becoming
polarised into two distinct camps: the victims and the beneficiaries of racial
segregation. After 1948, the ruling National Party (NP) intensified institu-
tional HIs between racial groups under the name of apartheid; under this
system Africans, Coloureds and Indians were systematically disenfranchised
and deprived of economic and social entitlements as well as basic human
rights, while European cultures and languages were elevated to a dominant
position. Such institutional retrogression was consolidated in defiance of the
waves of decolonisation in Asia, Africa and the Caribbean in the latter half
of the twentieth century.
In fact, the most striking HIs in South Africa evolved around the land issue.
The conquest of autonomous African polities by white settlers was com-
pleted by the late nineteenth century. Although African peasants actively
responded to the rising demand for foodstuffs in mining cities, their aspi-
rations were crushed by forcible land appropriation (Bundy, 1979). Africans
were supposed to live in homelands, formerly called native reserves, which
constituted only 13 per cent of the South African territory (Figure 5.1). Those
who forged their livelihoods in white areas were vigorously relocated to their
rural ethnic homelands or segregated urban townships, and the total num-
ber of victims of forced removals is estimated to have reached 3.5 million
between 1960 and 1983 (Platzky and Walker, 1985: 10). However, as the
impoverished homelands could not sustain the livelihoods of rural Africans,
migrant workers seeking cash incomes flocked into the white areas, where
98 South Africa and Zimbabwe
ZIMBABWE
MOZAMBIQUE
BOTSWANA
NAMIBIA
Pretoria
Johannesburg SWAZILAND
Bloemfontein
LESOTHO
Durban
Cape Town
Homelands
Kruger National Park
African labour was in much demand in the mining, manufacturing and agri-
cultural industries, especially in the large cities. The proportion of the urban
population in South Africa grew from 32 per cent in 1960 to 43.3 per cent in
1996 in spite of strict measures against urbanisation of Africans (Thompson,
2001: 298).1
The injustice of the apartheid regime functioned as a great leveller of the
oppressed. The Black Consciousness movement strived patiently to bridge
over African, Coloured and Indian politics to form a unified agency of the
‘Black people’. The exile leadership of the African National Congress (ANC)
and the South African Communist Party (SACP), including radical white
activists, made efforts to transcend racial and ethnic boundaries despite occa-
sional tensions (Ellis and Sechaba, 1992). In the popular protest in urban
townships, potential ethnic divides among Africans did not become salient,
and progressive Christians, Muslims and Jews took to the streets hand in
hand. The urban political dynamism as well as the successful containment
of cultural and ideological rivalry appears to have been the most significant
feature of the liberation movement in South Africa.
Such unity was further consolidated as a reaction to the divisive ‘reforms’
of the National Party (NP) government, which experimented with top-down
social engineering to divide and conquer the people. First, in 1983 the gov-
ernment tried to set up two houses of parliament with limited power for
Yoichi Mine 99
100%
90%
80%
50 48
70% 55
60%
50% 4 5
4 8 8
40% 8
30%
20% 38 39
33
10%
0%
1995 2000 2004
African Coloured Indian White
100%
13 11 10
90% 3 2
3 9
80% 9 9
70%
60%
50%
40% 76 78 79
30%
20%
10%
0%
1995 2000 2004
the former Southern Rhodesia, had been weaker than that of their South
African counterparts, even though it was more powerful than in other British
colonies in the tropics. Given that the proportion of the white population
was more or less 5 per cent at its peak in the 1970s,3 ‘it was the size of
Rhodesia’s White population – too small for grand apartheid but too large for
an easy exit – that accounted for the peculiar twists and particular tragedies
in the country’s history’ (Herbst, 1990: 14).
The majority of Zimbabweans are black Africans, and their present-day
ethnic configuration was also shaped by the migration processes associ-
ated with the Mfecane in the early nineteenth century. The Ndebele people,
originally a small Zulu faction lead by King Mzilikazi, settled in the South
African Highveld around today’s Pretoria; subsequently, after being attacked
by Afrikaners, they crossed the Limpopo River and started to consolidate
their permanent polity by assimilating the locals in the late 1830s. In the
north, by the late nineteenth century, the Shona people emerged from an
assemblage of diverse groups, some of which had built up historic kingdoms
and empires on the plateau (Beach, 1994; Mazarire, 2009; Ndlovu-Gatsheni,
2009a). The Shona are concentrated in the northeastern half of the coun-
try, with the Ndebele in the southwestern half. At present, the Shona are
said to account for about 80 per cent of the population, consisting of eth-
nic clusters such as Korekore, Zezuru, Manyika, Ndau, Karanga, Kalanga
and Rozvi, while the Ndebele make up about 10 per cent. This particular
ethnic configuration of majority Shona and minority Ndebele, in con-
trast with the more fragmented, multipolar ethnic configuration in South
Africa (Table 5.1), was to foment serious antagonism within the liberation
movement in Zimbabwe.
The prototypical imperialist, Cecil Rhodes, established the British South
Africa Company (BSAC) in 1889, seeking mining interests and territorial
expansion north of the Limpopo River. While serving as Prime Minister
of the Cape Colony, Rhodes deployed armed forces to crush the peasant
rebellion, the so-called First Chimurenga, in Matabeleland and Mashonaland
102 South Africa and Zimbabwe
MOZAMBIQUE
ZAMBIA
NAMIBIA
Harare
Bulawayo
BOTSWANA
MOZAMBIQUE
Lands for Africans
SOUTH AFRICA
Forests and Parks
0.8
0.7 South Africa
0.6
0.5 Zimbabwe
0.4
0.3
0.2
0.1
0
or a pe
F C g
ric e S e
M Av tate
aZ ala ge
st -N a
N n C al
th pe
el Ave uth
e
as na dl h
m t
aw o
m ela a o
na nd ds
M d W st
ic est
la sv d
C go
l
tra
M ab nd are
Li es
rn n
Af re ap
Ea ulu ng
an rag
M ho Mi ort
ul op
Zi el H ay
na Ma lan
er at
n a
ho la an
a
he te
Kw um era
or a
nd in
la E
W
at we So
en
C
a
u
rn
d
G
an
te
B
a
el
es
p
t
W
ab
ab
b
as
ho
h
at
ut
as
M
So
M
Figure 5.5 Provincial HDI in South Africa and Zimbabwe, 2001
Source: UNDP (2003a: 282; 2003b: 28).
Provinces with a high HDI are urbanised places where there are relatively
high proportions of whites: 19.9 per cent for Gauteng, which accommo-
dates Johannesburg and Pretoria, and 18.4 per cent for Western Cape, which
accommodates Cape Town (Statistics South Africa, 2003). Uprooting poverty
in the rural areas was, and still is, an urgent issue in both countries as wealth
concentrates in cities.
Table 5.3 Shifts in formal political systems in South Africa and Zimbabwe
Country Sequence
Figure 2.6). Let us trace the processes in these countries a little more
closely.
With the demise of the Cold War, President F.W. de Klerk released Nelson
Mandela in 1990 after 27 years of imprisonment, and political foes entered
multi-party negotiations in South Africa. The talks were often derailed by
political violence, including bombings and killings performed by both the
right and the left, and yet the most serious threat to the process was posed
by Zulu nationalists. Soon after the negotiations commenced, IFP support-
ers began to attack ANC supporters in townships and villages and incited
retaliation, pushing up the casualties of political violence to 16,022 from
1990 to 1994 (SAIRR, 1996: 52). In urban townships, ethnically integrated
African communities were assaulted by militant Zulu migrants, against the
backdrop of the bifurcation between urban civil society and the world of
customary law preserved in rural homelands (Mamdani, 1996: 218–284).
Unlike the Gukurahundi in Zimbabwe, the ‘black-on-black’ bloodshed in
South Africa took the form not of military repression but of seemingly vol-
untary ethnic mobilisation. However, the government commission chaired
by Judge Richard Goldstone later established that the IFP violence had been
orchestrated by the ‘Third Force’: a covert network of the apartheid security
forces.
As political fissures loomed large both inside and outside political parties,
South Africans were terrified of their country being on the brink of civil war,
as had occurred in Yugoslavia in the same period. It was at this critical junc-
ture that Mandela perfectly played the unifying role required by all groups
that wished to avoid the horrible consequences of each group promoting
its own self-interest. Mandela embodied the characters of both a seasoned,
dignified African chief and a sophisticated, telegenic urban politician, and
made energetic efforts to bridge racial and ethnic divides, having frequent
talks with leaders of rival political forces. In addition, his long imprisonment
had bolstered his even-handed position within the liberation movement.
The role he played in the transitional period can be understood as a modern
version of that played by the ‘leopard-skin chief’ depicted in the classic of
Evans-Pritchard, The Nuer (1940).
Yoichi Mine 107
Votes (%) Seats Votes (%) Seats Votes (%) Seats Votes (%) Seats
ANC: African National 62.7 252 66.4 266 69.7 279 65.9 264
Congress
NP: National Party (NNP: 20.4 82 6.9 28 1.6 7 — —
New National Party)
IFP: Inkatha Freedom Party 10.5 43 8.6 34 7.0 28 4.6 18
Freedom Front (Freedom 2.2 9 0.8 3 0.9 4 0.8 4
Front Plus)
DP: Democratic Party (DA: 1.7 7 9.6 38 12.4 50 16.7 67
Democratic Alliance)
PAC: Pan Africanist 1.3 5 0.7 3 0.7 3 0.3 1
Congress
ACDP: African Christian 0.5 2 1.4 6 1.6 7 0.8 3
Democratic Party
UDM: United Democratic — — 3.4 14 2.3 9 0.9 4
Movement
ID: Independent Democrats — — — — 1.7 7 0.9 4
COPE: Congress of the — — — — — — 7.4 30
People
Other Small Parties 0.8 0 2.3 8 2.1 6 1.8 5
Total 100.0 400 100.0 400 100.0 400 100.0 400
With the establishment of the GNU, the racial HIs in political representa-
tion and cultural status consolidated during the apartheid era were radically
dismantled. Past injustices committed by the apartheid government and
the liberation movement were the focus of the Truth and Reconciliation
Commission (TRC). On the economic front, the ANC’s policy circle had
worked out a Keynesian-oriented policy prescription of a mixed economy,
the Reconstruction and Development Programme (RDP), which was used as
the ANC’s manifesto for the 1994 elections (MERG, 1993). However, Thabo
Mbeki, who took over the presidency from Mandela in 1999, set aside the
RDP line and instead began to implement a more market-oriented policy
package called Growth, Employment and Redistribution (GEAR). The eco-
nomic power of the white population was largely preserved as a result of
the ANC’s conversion to economic realism. The adherence of the ANC to
the principle of the market economy seems to be the essential reason why
the vote-based political dominance of the ANC was practically supported
by the capitalist sector and grudgingly accepted by the white population
at large.
Meanwhile, the NP withdrew from the GNU in 1996, and the IFP dis-
engaged from the government in 2004. The ANC, as the ruling majority
party, was thus placed in a position to formulate national policy single-
handedly, accommodating critical voices raised by its historical partners in
the anti-apartheid struggle, COSATU and SACP. Reflecting the dissolution
of the grand coalition, the rating of variable C (Composition of the Exec-
utive) in South Africa changed from 3 to 1 (Chapter 2, Table 2.2). As the
influence of the NP and IFP waned, the Democratic Alliance (DA) began to
advocate the value of liberal democracy and emerged as the leading oppo-
sition party. The ANC was placed in a hegemonic position, adopting policy
elements of the left and right in a selective way, but its governing practice
was like walking on a tightrope.
The fiscal discipline espoused by the Mbeki government contradicted
the high expectations of the disadvantaged masses, mainly black South
Africans.7 The growth of employment in the formal sector remained very
slow, and the delay of social service delivery has induced violent protest in
several townships since the mid-2000s. Under the presidency of Mbeki, the
Africanisation of bureaucracy was accelerated, and wealthy Africans, dubbed
‘Black Diamonds’, bolstered their presence, but inequalities within Africans
continued to widen. Counterbalancing such realities, Mbeki often resorted
to racial discourse, for example, blaming white prejudice for the prevalence
of HIV/AIDS, to which his government was reluctant to react with effective
policy measures. As described aptly in MacDonald (2006: 133), ‘emphasizing
racial identities represents poor Africans symbolically, while putting them
off economically’. In the 2004 general elections, the ANC obtained more
than a two-thirds majority, though the voter turnout was relatively low
(Schulz-Herzenberg, 2009).
110 South Africa and Zimbabwe
COSATU and SACP perceived that they were sidelined by the neoliberal
Mbeki mainstream faction in the ruling alliance. The left-wing faction then
began to fight back within the party structure of ANC, enabling Jacob Zuma
to be selected as the party president at the national conference in 2007,
and succeeded in forcing Mbeki to step down. As a possible prelude to a two-
party system, Mbeki loyalists broke away from the ANC to form the Congress
of the People (COPE), which challenged the ANC in the 2009 general elec-
tions. The ANC eventually won a majority with 65.9 per cent of votes, while
DA and COPE secured the position of the loyal opposition.
While politics in South Africa is thus tilting towards ‘normal’ majoritarian
politics, the practice of informal power-sharing still seems to be alive.
After the dissolution of the GNU, the ANC government has always tried
to recruit a few cabinet members from minority parties, including the
Afrikaner nationalist Freedom Front. Moreover, Zuma’s ethnic identifica-
tion as a ‘100% Zulu Boy’ (in the words on the T-shirts of his supporters)
has enabled the ANC to rally Zulu voters with no relapse of violent con-
flict between the ANC and IFP. All in all, the 2009 elections marked a shift
of the power bloc, but the change occurred through peaceful popular vot-
ing, a practice that has firmly taken root in the political life of South Africa
since 1994 (Friedman, 2009). In addition, the multicultural principle of the
Constitution is accepted by the general public, the corporate leadership is
increasingly integrated in terms of race, and the National Economic Devel-
opment and Labour Council (NEDLAC) provides a corporatist framework for
business and labour. Beneath the surface of formal majoritarian politics, the
spirit of informal power-sharing has become an accepted norm in the South
African political life.
2000 2005
Note: In the 2005 elections, an independent candidate won one seat. This is why the sum of the
real seats (FPTP) in 2005 is one seat fewer than the alternative calculation (PR).
Source: Zimbabwe Election Support Network (2009: 26–27).
2008
Note: Regarding the FPTP seats in 2008, see the note to Table 5.5.
Source: Zimbabwe Election Support Network (2009: 28).
positions of foreign farm workers during and after the land invasion sharply
deteriorated (Rutherford, 2003).
were expressed; 89.8 per cent of Africans, 64.5 per cent of Coloureds and
20.1 per cent of whites agreed that affirmative action should continue for
a considerable period. Africans believe that whites benefit most from the
current economic system; 85.9 per cent of Africans and 24.0 per cent of
Coloureds thought that the economic situation of the white population had
become much better than it was ten years ago, while only 1.0 per cent of
whites thought this was the case.
In reality, a majority of the white population lives in relatively rich
suburbs. Their lives are generally well protected from threats from the out-
side world, and even daily contact with neighbours seems to be limited.12
Although the media outside South Africa tend to depict affluent whites as
falling victim to violent crimes committed by thugs and rogues, the main
sufferers are in fact Africans; 60.5 per cent of them answered that they
had personally experienced violent crime in the past year, compared with
26.5 per cent of Coloureds and 11.5 per cent of whites. At the same time, new
forms of inequalities were clearly recognised, especially by African respon-
dents; 82.0 per cent of Africans, 71.0 per cent of Coloureds and 23.4 per cent
of whites strongly agreed with the statement that the class division between
the rich and the poor was becoming more serious than racial discrimination
(50.5 per cent of whites agreed, but not strongly). In a similar vein, 84.9 per
cent of Africans, 73.5 per cent of Coloureds and 40.1 per cent of whites
agreed strongly that corruption rather than racism is the principal problem
in this country.
Power-sharing as a means of political representation is highly valued
across racial groups, to the extent that 81.5 per cent of all respondents felt
positive about an inclusive government like the GNU of 1994; 75.7 per
cent of them added that they would continue supporting their political
parties even after the parties entered into a coalition. Of all respondents,
71.7 per cent of Africans supported the ANC, the ruling party in the central
government, while 68.0 per cent of Coloureds and 85.4 per cent of whites
supported DA, the main opposition in national politics but the ruling party
in the provincial government. As a way to solve the issues in the govern-
ment, no less than 73.7 per cent of Africans, the majority of whom must be
loyal ANC supporters, preferred having a strong opposition. Though it is not
clear whether they were thinking of the central or the provincial govern-
ment, this nonetheless indicates a mature commitment of majority Africans
to multi-party democracy.
People in Cape Town seem to be optimistic about the future of their
nation. Of the total, 84.9 per cent of Africans, 78.1 per cent of whites and
71.5 per cent of Coloureds thought that South Africans were now thinking of
themselves as South Africans more than they did ten years ago. Furthermore
it seems that less impoverished Africans tend to prioritise racial identity
over national consciousness. Cultural events can be immensely instrumen-
tal in nation-building; 80.4 per cent of the total respondents answered that
Yoichi Mine 117
the 2010 FIFA World Cup had strengthened the unity of South Africans.
Defying global trends after the incidents of ‘September 11’, the relation-
ship between Muslim Coloureds and Christians of all racial groups is quite
amicable. More than 80 per cent of Christians and Muslims felt they were
comfortable working with each other.
As for the role of South Africa in conflict mediation, Africans tended to
express more pronounced views; 54.6 per cent of Africans strongly agreed
that their government should take a leadership role in settling political prob-
lems in other countries like Zimbabwe, but 25.9 per cent strongly disagreed
with such initiatives. On the other hand, 16.1 per cent of white respondents
strongly agreed, while 14.6 per cent of them strongly disagreed. Africans
were more pessimistic about xenophobia; 44.4 per cent of Africans (49.5 per
cent for the Africans from Eastern Cape and 40.8 per cent for the Africans
born in Cape Town and its surroundings), 27.0 per cent of Coloureds and
22.9 per cent of whites thought that the relations between South Africans
and foreigners from other African countries would worsen in the future.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Table 5.7 National and ethnic identification of Zimbabwean city dwellers (%)
Group I feel only I feel more I feel I feel more I feel only I don’t know Total
Zimbabwean Zimbabwean equally [ethnic [ethnic
than [ethnic Zimbabwean identity] identity]
identity] and [ethnic than
identity] Zimbabwean
Several practical lessons can be drawn from the discussion in this chapter.
First, power-sharing by all major political parties appears to be useful as
a short-term refuge from cumulative violence, especially when there is a
pressing need to accommodate losing parties still capable of using vio-
lence to destabilise the regime. The security apparatuses developed by the
white minorities in Zimbabwe and South Africa were neutralised effectively
122 South Africa and Zimbabwe
through the emergence of the Lancaster House regime in 1980 and the for-
mation of the South African GNU in 1994 in the respective countries. The
important feature of the South African GNU would be that the coalition
government was formed after a consensus was hammered out through a
period of informal and formal negotiations among local stakeholders. The
1994 elections in South Africa were thus realised as the occasion of the con-
summation of sharing power, rather than as a fighting arena that would
necessitate power-sharing as an ex post solution. The modalities of informal
power-sharing also seem to have taken root in South African politics and
institutions, and contributed greatly to the political stability of the country
after apartheid. One important defect of a grand coalition, however, is that
political processes tend to be dominated by elite talks behind closed doors
to the exclusion of popular participation. In this regard, it is important to
recognise the temporal nature of a grand coalition and to specify its time
boundary. Indefinite extension of a GNU presided over by Robert Mugabe
would be a nightmare for Zimbabwe.
The second lesson is that a ‘winner-takes-all’ principle that is contingent
on a small constituency system could be dangerous in culturally divided
societies, as demonstrated in the successive FPTP elections in Zimbabwe.
When the choice between ‘yes’ and ‘no’ is a close game, ‘wasted’ votes may
amount to nearly half of the votes cast, and the electorate tends to remain
deeply divided even after the campaign is over. The same applies to national
referendums and presidential elections, which are fought in a nationwide
single constituency. This is why elections based on a PR system, in which
political opinions are represented proportionally, are considered superior in
countries in volatile transition. The PR system may also contribute to a pro-
gressive change in the rules of national politics beyond vested interests. For
example, after the 2009 elections in South Africa, 172 of the 400 seats in the
national assembly were held by women due to the gender quota introduced
by the ANC to its party list of candidates, pushing South Africa up to the
fourth in the global ranking of women’s equitable representation in parlia-
ment. This kind of quota system can more easily be adopted in a PR system
than under a FPTP system.
The negative side of a PR system is that the accountability of representa-
tives to the electorate tends to be weak and indirect, as the MPs are inclined
to turn their faces not to the voters but to the party leaders. In our sur-
veys in Zimbabwe where the FPTP system is adopted, 77.6 per cent of all
respondents answered that consideration of the personal qualities of candi-
dates was important when they cast their vote, and 65.7 per cent answered
that the political party was important, while 22.6 per cent said that ethnic
considerations were important. A PR system reduces such opportunities for
the electorate to examine the individual qualities of representatives. This
weakness of the PR system is the very point emphasised by Mandela in
1999 in his farewell speech to the South African Parliament on the day of
Yoichi Mine 123
Notes
1. The urbanisation rate of the total South African population reached 61.7 per cent
in 2010, much higher than 38.3 per cent in Zimbabwe and the average figure for
sub-Saharan Africa, 37.0 per cent (UNDP, 2009a).
2. For the general trend, see Seekings and Nattrass (2005: 300–339). Even the official
unemployment rates excluding the chronically unemployed are near 30 per cent.
3. This proportion was less than half that of the white population in South Africa.
In addition, in Zimbabwe, a sizeable ‘buffer’ population like the Coloureds and
Indians in South Africa is absent. The proportion of white Zimbabweans has
probably declined to less than 0.5 per cent by now, although exact population
statistics are non-existent for the moment.
4. For a detailed description of the complex ethnic power struggles within the lib-
eration movement in Zimbabwe, see Sithole (1999). The vibrant urban politics
during the period up to the 1960s was overshadowed by the sectarian liberation
politics (Scarnecchia, 2008).
5. In Zimbabwe, the Lancaster House Agreement in 1979 was mediated by the
British government, while the formation of the GNU in 2009 was mediated by
the Southern African Development Community (SADC). In South Africa, theo-
retical blueprints for a new political dispensation proposed by American political
scientists, Lijphart (1985) and Horowitz (1991), did play positive roles. Although
the 1994 regime in South Africa was closer to the consociational prescriptions of
Lijphart, one of the major four components of his proposal, mutual veto, was
rejected in the course of negotiations.
6. The head of the state of South Africa is called the president, but is elected in
parliament like a prime minister in other countries. Therefore, variable D (Head
of Government) in South Africa is rated as 2 instead of 1 (Chapter 2, Table 2.1).
7. Respected Afrikaner economic historian Sampie Terreblanche (2002: 95–124) and
leading Canadian anti-globalist Naomi Klein (2007: 245–274) expressed the same
view that the ANC’s compromise with the business sector during the backroom
negotiations in the early 1990s was the critical turning point at which the former
became trapped in free-market orthodoxy.
8. The draft constitution was a cocktail of strong poison and potentially effective
medicine. In addition to the clause on land reform, in which compensation to
white farmers was to be paid out by the British government, the document con-
tained a Bill of Rights, electoral reform including the partial introduction of a PR
system, the establishment of a senate, and the inauguration of the post of prime
minister. The presidency was to be restricted to two terms, though this would
be applied only after the enforcement of a new constitution, enabling Mugabe
to stay in power. For the tension between the top-down approach of the gov-
ernment and the bottom-up approach of civil society in the aborted process of
constitutional reform in the 1990s, see Sachikonye (2004). The new constitution
approved in the referendum of March 2013 is downloadable from: http://www.
copac.org.zw/ (last accessed 1 May 2013).
9. The war veterans have always been relatively autonomous from the ruling ZANU
elite as demonstrated in Kriger (2003) and Sadomba (2011) from different angles.
10. Although the introduction of a quota system and bicameralism contributes to
making Zimbabwe fall within the category of a power-dispersing (PD) regime
(Chapter 2, Table 2.2), the reforms initiated by the ruling party were clearly meant
to concentrate the power base of ZANU(PF). The credibility of the election results
Yoichi Mine 125
has also been questionable, but non-partisan observers agreed that the May 2008
elections (Table 5.6) were conducted in a relatively free and fair environment
(Masunungure ed., 2009: 73–75), to the effect that the result could be a good ‘cen-
sus’ of people’s preferences of political parties. Terrible violence occurred before
the subsequent presidential run-off in the same year.
11. For a detailed record of racial incidents in South Africa since 1994 as well as a
collection of interviews with opinion leaders on the topic of racism, see Holborn
(2010).
12. Of the respondents, 73.9 per cent of Coloureds, 73.2 per cent of Africans and
27.1 per cent of whites answered that their neighbourhood was a very important
part of their identity. In this survey, the refusal rate was exceptionally high for the
residents of white suburbs, 65 per cent, laying a heavy burden on enumerators.
13. In Harare, 45.2 per cent of respondents identified themselves as Shona, 28.2 as
Zezuru and 10.5 as Karanga. In Bulawayo, 54.6 per cent identified themselves as
Ndebele, 26.9 per cent as Shona and 7.4 per cent as Karanga. Although Shona-
speaking peoples like the Zezuru and Karanga are often classified as being part
of the Shona, the ethnic classification in this survey is based on their own self-
identification.
14. The proposed multiple-seat system with a single non-transferrable vote is similar
to the former electoral system of the Japanese lower house, which has now been
converted to a combination of a small constituency system and a PR system.
The same kind of medium-sized constituency system is currently in use in a few
countries such as post-conflict Afghanistan.
15. For the prime importance of state legitimacy, see Moyo (2010). Given what
Mugabe symbolises, it seems to be difficult for the government of Zimbabwe to
restore full legitimacy in the eyes of the Western countries as long as Mugabe
himself is part of the leadership (Ndlovu-Gatsheni, 2009b: 343). In any case, the
‘Mugabe factor’ is only a part of the history, and ‘there is blood on many hands’
(Bourne, 2011: 236).
6
Is Ethnic Autonomy Compatible
with a Unitary State? The Case
of Uganda and Tanzania
Yuichi Sasaoka and Julius E. Nyang’oro
6.1 Introduction
Uganda and Tanzania, two countries in East Africa, have experienced sharply
contrasting conditions of political instability and stability. On the one hand,
it is estimated that more than half a million Ugandans lost their lives
between 1971 and 1986, during the period of Idi Amin’s dictatorship and
of the subsequent ‘liberation war’ waged by the National Resistance Army
(NRA) led by the current president, Yoweri Museveni. After the victory of
Museveni’s army, a serious armed resistance continued in the country’s
Northern region, and Uganda is prone to potentially violent ethno-regional
conflict even today. On the other hand, Tanzania has been largely peaceful
for many years, going back to the days of the late president, Julius Nyerere.
As evidenced in our perceptions surveys, ethnic identity and national iden-
tity compete on even terms in Uganda, while common identity as a single
nation has firmly taken root in Tanzania.
Despite such contrasting situations, Uganda and Tanzania share several
common traits. Both countries have had to deal with influential regional
separatist movements that sought political and cultural autonomy: the
Buganda Kingdom in Uganda and the islands of Zanzibar in Tanzania. These
movements posed serious challenges to the post-independence legitimacy
of the governments of these countries. Furthermore, prominent leaders in
these countries, President Nyerere of Tanzania and President Museveni of
Uganda, were ardent proponents of African socialism and integrationist
nation-building, which were meant to go beyond ethnic particularism
and secessionism. More recently, these countries have accepted multi-party
democracy and started to implement decentralisation policies as a form of
governance reform.
The objective of this chapter is to explain the reasons why these coun-
tries, neighbours on opposite sides of Lake Victoria, have taken remarkably
126
Yuichi Sasaoka and Julius E. Nyang’oro 127
1991 2002
Uganda’s major cities and towns are inhabited largely by single dominant
ethnic groups. For example, about 80 per cent of the inhabitants of Gulu
in Northern Uganda are thought to be Acholi, and the same percentage of
the inhabitants of Mbarara in the Western region to be Banyankole (UBOS,
1995; Moses 2007).
The Buganda Kingdom, having hierarchical authority and a sophisticated
social order, was established as early as the fourteenth century, reached its
zenith in the eighteenth and nineteenth centuries, and then entered a period
of ‘modernising autocracy’ (Apter, 1961). At the time of Uganda’s indepen-
dence, ethnic cleavage was palpable between the subjects of the Buganda
Kingdom and the rest of the country. During both the colonial period
and immediately after independence, Buganda was granted a ‘special status’
within Uganda. It is worth recalling that soon after independence (when
Queen Elizabeth of the UK relinquished the post of head of state), King
Mutesa II of Buganda, supported by the Kabaka Yekka (KY: King Only) party,
became the first (non-executive) president of the country and the region
was offered semi-autonomous status under the 1962 Constitution. Mean-
while, a conflict between the Baganda elite and the central government led
by Prime Minister Milton Obote, a politician from the Langi ethnic group in
the North, emerged over various constitutional issues, especially on the posi-
tion of Buganda within the now independent, unitary state of Uganda. The
conflict escalated into violence when two counties of Buganda were trans-
ferred to Bunyoro after a referendum in 1964. The ‘Buganda Crisis’ of 1966
followed, and eventually Obote, who represented the ruling Uganda People’s
Congress (UPC), gave an order to arrest Mutesa and to abolish all traditional
kingdoms in Uganda including Buganda, Ankole, Toro, and Bunyoro. The
Busoga chieftaincy was also abolished in 1967.
The series of political struggles that took place in the 1960s was a harbinger
of subsequent conflicts between the people of socioeconomically advanced
Yuichi Sasaoka and Julius E. Nyang’oro 129
in the West, the home region of President Museveni, are favoured by the
government. Thus, the potential conflict between the Baganda and the West-
erners, as well as the question of the relative deprivation of the Northerners,
is of great significance in present-day ethnic relations in Uganda.
As the Buganda Kingdom has historically been the centre of political and
economic power in Uganda, many Baganda aspire for the political auton-
omy of Buganda within the country. When President Museveni acquiesced
to the reintroduction of monarchies in the early 1990s, there was strong dis-
satisfaction within Buganda at the limited nature of the authority granted to
these monarchies. Baganda royalists had hoped that the new dispensation
would essentially be a return to the status quo ante of the pre-1966 period,
when Buganda was almost a country within a country. The call for a federal
system in Uganda is thus driven mainly by Baganda royalists who seek the
restoration of the political autonomy of the Buganda Kingdom.
Gulu
UGANDA
SOMALIA
Hoima
Buganda Kampala
Kingdom KENYA
Mbarara
Lake
Victoria
RWANDA
BURUNDI
INDIAN
OCEAN
ZAMBIA
MALAWI
MOZAMBIQUE
borders and ensuring minority rights, but also has something to do with the
national identity of the Republic of Tanzania. When we discuss the issue
of Zanzibar, the political division between the two major islands is also of
great significance, because the pros and cons of autonomy have been hotly
debated within Zanzibar itself, sometimes with polarised opinions. Broadly
speaking, the CCM (Chama cha Mapinduzi: Party of the Revolution), which
is the ruling party in Mainland Tanzania, has taken root in Unguja, while
the Civic United Front (CUF), the leading opposition in Mainland, has its
stronghold in Pemba (Bakari, 2001; Killian, 2008; Burgess, 2009). In a sense,
the ethnic politics within Zanzibar is a microcosm of Tanzanian politics as a
whole, and this justifies our looking into what has been happening in that
compact space in some detail.
Mainland Tanzania gained independence from Britain as the Republic
of Tanganyika in December 1961, while Zanzibar gained independence as
a constitutional monarchy in December 1963. However, soon after inde-
pendence, the Sultan of Zanzibar was deposed in a violent revolution in
January 1964 (Yeager, 1989; Petterson, 2002), in which about 5,000 ‘Arabs
and Indians’ on the islands were massacred by ‘African revolutionaries’. The
massacre, which was a precursor to the deportation of Indian Ugandans by
Idi Amin, was based on a perception of HIs between the rich Asian minority
and the poor African majority. April 1964 saw the formation of the United
Republic of Tanganyika and Zanzibar, which was then renamed the United
Republic of Tanzania. Given the political sensitivity of prior sovereignty, the
Mainland and Zanzibar formed a power-sharing arrangement. The arrange-
ment was that if the president of Tanzania were to be elected from the
Mainland, then the vice president would come from Zanzibar, and vice versa.
This has been the political arrangement for almost five decades. Zanzibar
was also given a disproportionate number of seats in the National Assembly.
In more recent times, a power-sharing arrangement has been instituted in
Zanzibar where the ruling party shares power with the opposition, as we will
see shortly.
Uganda and Tanzania are somewhat similar with regard to religious
identification in that Christians (Catholics, Protestants) and Muslims are
competing over new converts. In Tanzania, the Muslim population (just
below 40 per cent) is larger than it is in Uganda (12 per cent), and it is
predominant in Zanzibar (98 per cent) and the coastal areas.1 African Tra-
ditional Religions (ATRs) are also practised in both countries. For example,
many Baganda practice traditional rituals, as does almost every other eth-
nic group. In terms of language, there is a major contrast between the two
countries: virtually everyone in Tanzania speaks Kiswahili as the official and
national language while in Uganda, Kiswahili is mostly spoken in the East
(towards Kenya) and in the North (as the language of trade and the military).
In Uganda, the teaching language at the secondary school level and beyond
is English.
134 Uganda and Tanzania
Table 6.4 National and regional poverty rates in Uganda, 1999/2000–2009/2010 (%)
Note: The data for 1999/2000 and 2002/2003 exclude some districts.
Source: UBOS (2003, 2006, 2010).
Girls Boys
domination at the elite level, however, is not groundless. In 2008, 74 per cent
of Ugandan military commanders were from the West, 17 per cent from the
Central region, and 9 per cent from the North (Tumushabe, 2009: 19). Fol-
lowing the overthrow of the Amin and the Obote regimes, Northerners who
had occupied high military positions were replaced by Westerners, especially
the Banyankole. The Baganda continue to be excluded from the military, as
has always been the case since colonial times.
Male/Female Total
1999 2005
the secret talks held in 2009 between President Amani Abeid Karume, the
son of the Zanzibar’s first president, and Seif Shariff Hamad, the CUF sec-
retary general, were informally arranged but instrumental in helping bring
about the agreement of a coalition government in Zanzibar. The considera-
tion of mutual trust between political leaders thus leads us to the next topic
of our discussion, the question of leadership.
The sites of the interview surveys were selected on the basis of their regional
representativeness. In Uganda, major towns in the Northern, Central and
Western regions were chosen. In Tanzania, the locations were Dar es Salaam,
the capital on the Mainland, and Unguja and Pemba, the two Zanzibar
islands. Ethnic concentration in each town is evident in Uganda. In our
survey, the Acholi constitute 89 per cent of the population of Gulu in the
Yuichi Sasaoka and Julius E. Nyang’oro 145
Northern region, the Banyoro are 70 per cent of the population of Hoima in
the Western region, the Banyankole account for 58 per cent of the popula-
tion of Mbarara in the Western/Southern region and the Baganda account
for 44 per cent in Kampala, the Central region. The ethnic groups in Dar es
Salaam are diverse, with the largest ethnic group, the Zaramo, accounting
for only 13 per cent, while the Shirazi form 86 per cent of the residents of
Pemba and 55 per cent of Unguja residents.
Our Ugandan survey consisted of 500 residents. The regional breakdown is
as follows: Kampala, 200 respondents; Hoima, 100; Mbarara, 100; and Gulu,
100. Kampala is the largest city, attracting many immigrants from the sur-
rounding rural areas. Kampala represents the opinion of the Central region,
but the sample size had to be doubled in order to obtain an appropriate
weight of ethnic groups. The interview survey was conducted just before
the presidential election that took place in January and February 2011. In
Tanzania, the number of respondents was 600. The breakdown is as fol-
lows: Dar es Salaam, 300 respondents; Unguja, 148; and Pemba, 152. The
perceptions survey was conducted in March and April 2011. In both coun-
tries, a stratified sampling method was applied to obtain a collective sample
of the same quality from each group.
6.5.1 Uganda
According to our survey, the living standards of the people in the Western
region appeared to have improved and to be approaching those in the
Central region. This trend was consistent with the results of recent house-
hold surveys. In our surveys, in Kampala and Mbarara, the ratio of those
who had completed secondary education or above was high (85 per cent
and 49 per cent, respectively), and many households possessed refrigera-
tors (35 per cent and 36 per cent), although there were many people who
had experienced problems of not having enough water, food and money
in the previous 12 months. The North represents the poorest region, and
Northerners recognised their disadvantaged status (71 per cent replied that
their status was ‘worse’ or ‘much worse’ than other ethnic groups), and
tended to perceive that their situations had been worse five years ago than 10
or 20 years ago. Ninety per cent of respondents in all three places shared the
view that the North had been the least favoured by the government. There
was also a dominant view, even in remote Gulu, that the Banyankole of the
West were the richest group.6 This must reflect the discovery of a promising
oil field in the Western region in 2009, as well as the ethnicity of President
Museveni. Our survey town in the West, Hoima, is close to the oil field in
the Albertine Basin, bordering the DRC.
The recognition of their own poor status by Northern people seems to
have made them feel alienated. The support rate for the opposition FDC
was 28 per cent in Gulu, which was the highest in all places, even though
33 per cent in the same place supported the ruling NRM. Northern people
146 Uganda and Tanzania
used to have an aversion solely to Central people, but they now seem to have
the same level of discomfort in working with Western people. The answer
that the ethnic factor was important when they voted for the national par-
liament and local councils was preponderant in Gulu and Kampala where
people thought that politicians of other groups utilised ‘tribalism’ to garner
political support (43 per cent of respondents in Gulu and 40 per cent in
Kampala replied that politicians did this ‘often’ or ‘always’). In addition, the
survey results indicated that the Baganda and Banyankole did not always
like each other. In the question asking if they would accept a daughter’s or
a sister’s marriage with a member of another group, a significant number
of people in Mbarara in the West (23 per cent) opposed marriage with a
Muganda husband.
Northern people perceived their underrepresentation in the government,
police and other institutions and felt that the Banyankole were the most
prosperous. In Gulu, 52 per cent considered Banyankole to be the most
prosperous, while 38 per cent replied that the richest were the Baganda.
In contrast, Western people do not share such perceptions. In Mbarara, in
the West, the willingness to give extra economic assistance to poorer groups
was thinner than in other places: those people who ‘strongly agree’ with this
accounted for only 49 per cent in Mbarara. Concurrence with the opinion
that some portion of government ministers should be allocated to disadvan-
taged groups was also weak in Mbarara. A possible explanation for this is that
the Banyankole, the influential group in the West, have become relatively
rich only recently and may still think that their group should be prioritised
by the government.
The ruling NRM was supported by 69 per cent of the respondents in
Hoima, 57 per cent in Mbarara, 48 per cent in Kampala and 33 per cent
in Gulu. This is consistent with the results of the election conducted in
February 2011, signifying the end of the long-term decline of the NRM.
The view that the president should not have too much power was agreed
to by 86 per cent of people in Gulu, 68 per cent in Kampala, 39 per cent in
Hoima and 36 per cent in Mbarara. Even though the living standards in the
North and Central regions were very different, people in Gulu and Kampala
share a much more cautious view on presidential power than do people in
the West.
The statement that ‘giving semi-autonomous status to Buganda is fair and
justified’ was supported by a majority of the Baganda (54 per cent ‘strongly
agree’ or ‘agree’), while only 23 per cent of Northern people supported this
view. The approval of the statement that ‘the former kingdom should be
given more political power’ showed similar results. There was discontent
regarding regional resource distribution in Gulu and Kampala. Regarding the
question asking if ‘oil revenue will be distributed fairly among the different
regions in future’, the respondents who ‘strongly disagree’ were 74 per cent
in Gulu, 58 per cent in Kampala, 36 per cent in Mbarara and 32 per cent
Yuichi Sasaoka and Julius E. Nyang’oro 147
in Hoima. In the North and Central regions, people tended to perceive that
the current decentralisation policy exacerbated ethnic tension at the local
level (68 per cent ‘strongly agree’ or ‘agree’ in Gulu, 56 per cent in Kampala,
35 per cent in Hoima and 28 per cent in Mbarara).
6.5.2 Tanzania
The living standards were generally higher in Dar es Salaam than in Unguja
and Pemba. As Dar es Salaam was the only place surveyed on the Mainland,
the survey result must contain an urban bias. In Zanzibar, Unguja was gen-
erally richer than Pemba, but it should also be noted that there were more
people who replied that they did not have enough clean water and food over
the past year in Unguja than in Pemba. Forty-eight per cent of the respon-
dents in Dar es Salaam answered that the living conditions of their own
ethnic group were considered to be the same as other groups. On the other
hand, in Unguja (Pemba), 46 per cent (58 per cent) answered ‘much better’
to the same question. This high rate of satisfaction indicates that grievances
related to economic HIs were not very strong in Zanzibar.
With regard to the question asking whether or not ‘you feel uncom-
fortable to work with other groups’, 20 per cent of all respondents said
‘yes’. In Zanzibar, the proportion of respondents who felt uncomfortable
working with Christians was higher than in Dar es Salaam. Regarding the
question about marriage, reluctance was more evident in case of different
religious groups than in cases of different ethnic groups.7 While the num-
bers of Muslims and Christians were almost even in Dar es Salaam, Muslims
were dominant in Zanzibar. As for social trust, people of Pemba tended to
attach the highest value to relatives, neighbours and co-religionists, but
this was less so in Unguja and Dar es Salaam.8 Regarding future prospects,
the respondents showed general optimism that ethnic (racial) relation-
ships would improve rather than worsen in the future, with the people of
Pemba the most positive about such a view.9 In Dar es Salaam, the rela-
tionship between different religious groups seems to have worsened, as the
respondents who answered ‘worsen a lot’ accounted for 26 per cent, while
23 per cent answered ‘stay the same’.
The idea that a specific identity group was favoured in government jobs
was shared by a majority of the people in Pemba, followed by Unguja, and
least in Dar es Salaam. In Pemba, 55 per cent of people replied that eth-
nic (racial) background became ‘much more important’ or ‘somewhat more
important’ after the introduction of the multi-party system in 1992. About
the statement that ‘politicians tend to use ethnicity for mobilising political
support’, 48 per cent of people in Dar es Salaam and 43 per cent in Pemba
answered ‘sometimes’, while 42 per cent in Unguja said ‘never’. The opti-
mism of people in Pemba towards the future of Tanzania turned out to be
stronger than in other places, indicating their satisfaction with the newly
established power-sharing arrangement on the island.
148 Uganda and Tanzania
The support rates for the ruling CCM were 56 per cent in Dar es Salaam
and 60 per cent in Unguja, while in Pemba, CCM and the opposition CUF
seemed to be in close competition with each other, and 28 per cent refused to
answer, reflecting the sensitivity of party support in Pemba. Regarding voting
in presidential elections, 62 per cent of people in Unguja and 64 per cent
in Pemba answered that religion was more important than other elements,
while only 28 per cent in Dar es Salaam stated this to be the case. As it
happens, the current president of the country, Jakaya Mrisho Kikwete, is a
Muslim who previously served as chief political instructor in the military.
In contrast, 72 per cent of respondents in Dar es Salaam thought that the
‘policy program and proposed actions are much more important’ at the time
of elections. The ethnic origin of a president was not the main issue in all
three places, where the view that a political party should not be formed in
line with ethnicity was widely shared in Mainland (agreed to by 90 per cent
in Dar es Salaam, 44 per cent in Unguja and 61 per cent in Pemba).
People were generally happy with the formation of the unity govern-
ment in Zanzibar. The respondents who answered either ‘strongly agree’ or
‘agree’ regarding the formation of the unity government were 85 per cent in
Unguja and no less than 94 per cent in Pemba. In Pemba, expectations were
especially high for the positive effects of the unity government on poverty
reduction. The majority of the respondents in the three places expressed the
view that the president had too much power and agreed that the current
decentralisation policy that devolved power to the district and community
levels was useful for the improvement of public service provision (agreed
to by 81 per cent in Dar es Salaam, 85 per cent in Pemba and 53 per cent
in Unguja). Moreover, the statement that ‘more autonomy and indepen-
dence given to Zanzibar is desirable’ was supported by many in both Unguja
(82 per cent) and Pemba (93 per cent).
(74 per cent) and Dar es Salaam (60 per cent). In Uganda, the equilibrium of
perceptions seems to lie in dual identity: ‘I feel equally Ugandan and [own
ethnicity]’ (61 per cent on average).11
There are also contrasting perceptions about decentralisation policy.
Respondents in Uganda tended to think that resource allocation is biased
towards a certain region, while such a perception was not strong in Tanzania.
People in Gulu and Kampala who think they are disadvantaged in terms of
several or all dimensions of HIs find many problems in the present decentral-
isation policy and do not regard it as a useful means of resource allocation.
In contrast, in Tanzania, people are generally satisfied with the decentrali-
sation policy in their own country. People in Zanzibar share the view that
social services in Pemba is inferior to that of Unguja, and such a perception
of socioeconomic HIs should form part of the expectations held by local
islanders in respect of what the government of national unity in Zanzibar
can achieve in the future. In Uganda, there is a widely shared perception
that the Northern region is still the poorest in the country. Tackling the
150 Uganda and Tanzania
6.6 Conclusion
in the north and the discovery of oil in the Western region. Some have
argued that the arena of conflict in Uganda was shifting from the national to
the local level due to the progress of decentralisation (Green, 2008), but this
proposition is only speculative as long as the central government holds the
reins on the process. As power concentrates, counteracting poles are likely
to emerge. The failure of President Museveni to promote a reliable successor
also clouds the vision of the country’s political future. The Ugandan people
are to deal with these uncertainties to reach their own solution.
Notes
1. Muslims make up 35 per cent of the population of Mainland Tanzania and more
than 99 per cent in Zanzibar, and 12 per cent of the total population in Uganda.
(The 2002 Uganda Population and Housing Census. UBOS, 2005.)
2. World Bank Country Results Profile, http://data.worldbank.org/country/uganda
(last accessed 1 May 2013).
3. ‘Bukenya lauds Buganda crackdown, CBS closure’, The Observer (Kampala), 13
September 2009, http://www.observer.ug/index.php?option=com_content&task=
view&id=5119&Itemid=59 (last accessed 1 May 2013).
4. In Tanzania as a whole, in 2010, Chadema (Party for Democracy and Progress)
was the largest opposition party, and the CUF was the second opposition party.
5. BBC News, ‘Tanzania PM to Resign over Graft’, 7 February 2008, http://news.bbc.
co.uk/2/hi/africa/7232141.stm (last accessed 1 May 2013).
6. Twenty-nine per cent pointed to the Banyankole, while 28 per cent indicated the
Baganda. In Mbarara, 31 per cent of the respondents replied that the Bahima,
a subgroup of the Banyankole, were the richest and 22 per cent said that the
Baganda were the richest.
7. Fifty-two per cent of respondents in Unguja, 55 per cent in Pemba and 21 per cent
in Dar es Salaam revealed their reluctance to work with Catholic people.
8. Ninety-six per cent of respondents in Pemba and 91 per cent in Unguja said that
they trust their relatives, while 76 per cent in Dar es Salaam said they did so.
9. Sixty-three per cent of respondents in Pemba, 45 per cent in Unguja and
42 per cent in Dar es Salaam said that the ethnic relationship would improve
in the future.
10. Eighty-eight per cent of respondents in Dar es Salaam answered that national-
ity was very important, while 76 per cent in Unguja and 64 per cent in Pemba
answered thus. On the other hand, 86 per cent in Dar es Salaam answered that
religion was very important, whereas this ratio was 85 per cent in Unguja and
95 per cent in Pemba.
11. People in Uganda, including Baganda, often seem to strategically select a more
advantageous identity in a given situation.
7
The Politics of Identity, Horizontal
Inequalities and Conflict in Kenya
Mwangi S. Kimenyi
7.1 Introduction
153
154 Kenya
Kikuyu 20.1 22
Luhya 13.3 14
Luo 13.9 13
Kalenjin 10.9 12
Kamba n.a. 11
Kisii n.a. 6
Meru n.a. 6
Other African n.a. 15
Non-African n.a. 1
UGANDA
SOMALIA
5
4
2
Lake
Lake 3 1
Victoria
Victoria
6
7
TANZANIA
investment than the rest of the country. These inequalities persisted after
independence.
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Nairobi Coast Central Rift Valley Nyanza North Eastern Western
Eastern
Lowest Highest
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
rn
bi
st
za
rn
rn
tra
lle
ro
oa
e
e
te
n
st
en
st
Va
ya
ai
es
C
Ea
Ea
N
W
ift
th
R
or
N
160
Deaths per 1,000 live births
140
120
100
80
60
40
20
0
Central Eastern Rift Nairobi North Coast Western Nyanza
Valley Eastern
areas, and especially in Nairobi and Mombasa and the adjoining areas. This
has resulted in wide differences in access to water, electricity and sanita-
tion. Figure 7.5 shows the regional inequalities in terms of access to different
types of infrastructure. Nyanza, North Eastern and Western provinces have
the lowest access to running water and sanitation. Similarly, Nyanza, Eastern
and Western provinces have the lowest access to electricity.
Mwangi S. Kimenyi 161
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Nairobi Coast Central Rift Nyanza North Eastern Western
Valley Eastern
Water supply Sanitation Electricity
Figure 7.5 Access to piped water/public tap, sanitation and electricity by province
Note: Access to sanitation means access to flush toilet or flush pit latrine.
Source: Kenya Demographic and Health Surveys, 2008–2009, http://www.measuredhs.com/data/
dataset/Kenya_Standard-DHS_2008.cfm?flag=0 (last accessed 1 May 2013).
80%
70%
60%
50%
40%
30%
20%
10%
0%
Nairobi Coast Central Rift Nyanza North Eastern Western
Valley Eastern
Refrigerator TV
In summary, the data provided above show that Kenya is a country that
is characterised by significant regional inequalities in terms of quality of
life and access to public goods and services. We have observed that these
regional inequalities do in fact reflect ethnic inequalities due to the fact that
ethnic groups occupy distinct areas of the country. Although the political
inequalities may not necessarily explain socioeconomic inequalities, they
do exacerbate the inequalities. Nonetheless, the origin of the inequalities is
not important in terms of the potential for the inequalities to trigger con-
flict. Of importance is how inequalities interact with political processes and
institutions.
structures that could have served to disperse power among several points
were systematically eliminated, and power was eventually concentrated in
the presidency. Not only was the bicameral legislature done away with, but
the limited powers of local governments were eventually transferred to the
central government. Furthermore, and probably most important, the coun-
try slowly gravitated to a de facto single-party state, then de jure single-party
rule, before returning to multi-party system in the early 1990s. It has only
been since the enactment of the new constitution in 2010 that meaning-
ful constitutional dispersion of political power has now come to define the
political landscape. Of course, the implementation of the new constitution
is still in the very early stages. Below I discuss the various phases that seem
to define Kenyan political institutions and processes, and the relationship
with ethnicity and HIs.
banned in 1969 and its leader, Odinga, was jailed. With the banning of KPU,
Kenya became in essence a de facto single-party state. Over time, several other
important power-dispersing institutions were all but extinguished. For exam-
ple, a series of constitutional amendments allowed the president to appoint
high court judges, a law to detain opponents without trial was enacted,
the two legislative houses were merged to form the National Assembly and
independent candidates were barred from contesting seats in the National
Assembly. Finally, in 1982, following an attempted coup, Section 2A of the
constitution was introduced, effectively turning the country into a de jure
one-party state. Essentially, by 1982, the country was mainly in a civilian
dictatorship.
In brief, as noted in Kibara (2003: 1), after independence, the ruling party
gradually undermined the checks on executive power to create a one-party
state. The ruling party harassed and co-opted political opposition, removed
effective powers of the judiciary and parliament and tampered with the
constitution to consolidate presidential power. From 1964 to 1990, thirty
amendments were made to the constitution, depriving the original values
and principles of the constitutions. As a result, highly personalised rules
emerged from the wreckage of the dreams of Kenya’s founding fathers.
In addition, the period was also marked by a persistent neutralisation of
non-state actors. The civil society was weakened through intimidation, the
use of force and detention without trial, denial of permits and outright
de-registration of non-governmental organisations. Thus the period was not
only one marked by the negation of constitutionalism, but was also marked
by widespread and far-reaching abuses of human rights.
What is particularly significant to note about the era of single-party
rule was that the president was extremely powerful and was able to solely
determine who held positions of authority, which groups were rewarded
and which were punished. It was through these wide discretionary pow-
ers that both Presidents Kenyatta and Moi were able to discriminate in
favour of members of their communities for government appointments,
contracts, land and the concentration of public services. As already observed,
the Kikuyu benefited disproportionately during Kenyatta’s rule while the
Kalenjin were favoured by Moi. The allocation of land to Kikuyus in Kenya’s
Rift Valley remains one of the key issues that have divided the Kikuyus and
Kalenjins, and is the primary source of grievances that have triggered recent
conflicts between these two groups. The patronage system perfected during
the long rule of President Moi, who discriminated against the Kikuyu and
other ethnic groups and awarded preferences to the Kalenjin, had the effect
of widening the divisions in Kenyan society.
I feel equally
Kenyan and my
own ethnic group,
289 (32%)
120 111
100
100
80
67 64
60 53
39
40 31 29
25 23
19 16 15
20 13 12 12 10
0
M a
R ria
ba
so
bu
a
Em a
Ba e
Ka ru
o
yu
in
Tu kot
sa
ni
Lu ii
Ka ali
an
hy
l
Lu
di
nj
Te
ju
Ku
m
ra
ku
Ki
aa
m
Po
M
en
rk
le
Bo
So
Ki
Figure 7.8 Would object to marriage with a specific ethnic group (N = 321)
Source: JICA Surveys.
321 respondents: 35 per cent), Kikuyu (100 out of 321: 31 per cent), Somali
(67 out of 321: 21 per cent) and Kalenjin (64 out of 321: 20 per cent)
were the most cited groups that respondents would object to having their
sister/daughter marry.
Respondents were not asked why they objected to these specific groups.
Was the objection based on a past unpleasant encounter or just a general dis-
like of the other group? In conducting future research it might be of interest
to find out why respondents objected to these groups.
170 Kenya
80 72
70 66
60
50 46
40 36
30 22 19 18
20 14 14
11 7
10 6 6 5 5 3
4
0
yu
na
Bo o
ba
Tu ria
o
le
u
bu
in
R a
M ot
Ba i
Ka ali
ni
Ka ii
sa
s
hy
Lu
er
an
s
di
nj
ku
ra
Te
ju
m
Ku
m
Em
Ki
aa
Po
M
en
Lu
le
rk
Ki
So
Figure 7.9 Would feel uncomfortable working with a specific ethnic group (N = 253)
Source: JICA Surveys.
Mwangi S. Kimenyi 171
Relatives
Neighbours
%
%
%
%
%
%
0%
10
20
30
40
50
60
70
80
90
10
I trust them a lot I trust them somewhat Just a little
Not at all Don't know/refused to answer
a little’ – were people from other ethnic groups (63 per cent), people from
other religions (60 per cent) and people from other countries (60 per cent).
The high level of distrust towards other ethnic groups is somewhat surprising
given the respondents’ responses to previous questions about marriage and
working together. These results suggest that other factors, such as religion
and proximity (that is, being neighbours), are also important factors in how
Kenyans perceive other groups.
The majority of respondents, 63 per cent, believe that inter-ethnic rela-
tions in Kenya are either positive or very positive (Figure 7.11). Only a small
number of respondents, 4 per cent, believe that inter-ethnic relations are
negative or very negative. However, it appears that Kenyans have somewhat
polarised opinions about future inter-ethnic relations (Figure 7.12); about
14 per cent of respondents expect things to worsen somewhat or a lot, while
13 per cent expect a lot of improvement.
The study then sought to examine respondents’ perceptions of whether
ethnicity affects access to public goods and services (Figure 7.13). The over-
whelming majority of respondents believed that ethnicity affects a person’s
chances of getting a government job (80 per cent) or a government contract
(79 per cent). Ethnicity plays a much larger role in access to government jobs
than access to private sector jobs (66 per cent), though it plays a large role
in getting access to a job in both sectors. Historically, land ownership has
172 Kenya
Public housing
Land ownership
University education
Pre-university education
Private loans
Don't know/
Much less
refused to
important, 27
answer, 53
(3%)
(6%)
Somewhat less
important, 54
(6%) Much more
important, 342
No change, (38%)
130 (14%)
Somewhat
more
important, 301
(33%)
Figure 7.14 Has ethnicity become more important in Kenyan politics since the
introduction of multi-party elections in 1992? (N = 907)
Source: JICA Surveys.
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Today compared Today Compared Today Compared
to Kibaki to Moi from to KANU
before 2007 1992 to 2002 before 1992
Figure 7.15 Perceptions about today’s socioeconomic conditions of own ethnic group
compared to the past (N = 907)
Source: JICA Surveys.
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
a
Ka o
in
a
ba
R na
le
so
hy
u
Bo ia
yu
So u
t
M ii
Tu ali
ni
Em i
ko
sa
Lu
an
er
b
nj
di
r
ju
m
Te
ra
Ki
Lu
ku
Ku
m
M
Po
aa
le
en
rk
Ba
Ka
Ki
Figure 7.16 Which ethnic group do you feel is most favoured by the government?
(N = 907)
Source: JICA Surveys.
Table 7.5 Compared to other ethnic groups, do you think your ethnic group
currently gets its fair share? (% of group)
less than their fair share. Indeed, feeling excluded from access to government
resources with no opportunity to redress this issue could fuel violent action
against the government and against those seen to be garnering the lion’s
share of benefits (Table 7.5).
resource allocation, because resource allocation decisions will not rest solely
on the presidency.
Although the new constitution retains a presidential system, it has sev-
eral features that disperse political power to many centres and positions and
also restrains the power of the executive. In this new constitution, the pres-
ident is still powerful, but there are various constraints on the powers of
the executive as compared to the previous system. For example, although
the president will still be able to appoint the cabinet and senior govern-
ment secretaries, the appointees must be vetted by various organs of the
legislature and other independent commissions. Furthermore, there are sev-
eral requirements regarding the composition of the cabinet and other senior
appointments that guarantee broad representation of the Kenyan people.
Thus, while the new constitution retains a presidential system, it is no longer
the same as the ‘imperial’ presidency that has dominated Kenyan politics.
Another feature of the new constitution that will serve to disperse power
is the establishment of devolved structures of governance. The constitution
calls for the establishment of 47 devolved governments called ‘counties’.
These counties will be the focal points for the delivery of services to the
people. Each country will have its own government headed by a governor
and a legislature. Although the counties will receive resource transfers from
the central government, counties will have the power to determine their
own policies and development programmes, including raising some forms of
revenue. The devolved governments will serve to increase the participation
of the people in determining their own destinies and will ensure broader
inclusion in decision-making than has been the case under the centralised
system.
A key power-dispersing feature of the new constitution brings back a
bicameral legislature with a parliament and a senate. Members of parliament
will be elected as they were in the past, based on electoral constituencies
much like they are now. The members of parliament will be responsi-
ble for making broad policies at the central government level. Members
of the senate will be elected to represent county interests in the central
government.
Looking ahead
As noted, for most of the post-independence era, Kenyan politics has been
dominated by an ‘imperial’ presidency. The new constitution radically alters
Kenyan politics by placing checks on the executive and dispersing power to
many centres. Although a presidential system was retained, there are various
checks on executive power. However, it is still not clear how the estab-
lishment of county governments will impact ethnic relations and also HIs.
Finally, the commencement of cases by the International Criminal Court
(ICC) against six suspects related to the post-election violence in Kenya is
Mwangi S. Kimenyi 177
also creating uncertainty about the future and will most likely redefine the
politics and ethnic alliances in a major way.
The 2008 post-election violence in Kenya revealed how easily the coun-
try could disintegrate into chaos. Although violence was triggered by
perceptions that the election had been stolen by President Kibaki, it is
apparent that there are underlying grievances that have divided the Kenyan
people. In this chapter, I have focused on HIs and have suggested that
these inequalities could be important triggers for conflict. I have also sug-
gested that, to a certain extent, power-concentrating institutions have in
part contributed to the existing HIs. The new constitution radically changed
the institutions of governance to provide for wide dispersion of power and
effective checks on the executive.
However, the information reported concerning perceptions of ethnicity
and inequality reveals that negative perceptions about other ethnic groups
are widespread. To the extent that such perceptions do influence the way
people act, the probability of inter-ethnic conflict in Kenya remains high.
Just because a new constitution has been adopted may not resolve the
differences unless serious steps are taken to address these differences.
Note
1. See also Alwy and Schech (2004).
8
Managing Horizontal Inequalities
and Violent Conflicts in Nigeria
Ukoha Ukiwo
8.1 Introduction
178
Ukoha Ukiwo 179
mobilisations and conflicts. Sections 8.4 and 8.5 will examine the political
institutions developed to manage conflicts and how they have functioned
over time. In Section 8.6, I will present the findings of a popular perceptions
survey conducted to gauge public views on the nature of identities, HIs,
and political institutions. Section 8.7 presents the concluding remarks of
the study.
Historically, ethnicity, region, and religion have been the major social cleav-
ages in Nigeria as well as the fault lines for most political mobilisations and
violent conflicts in the country. The probability of violent conflict tends to
increase where one social cleavage reinforces another. For instance, when
ethnic identity is coterminous with religious identity, as we have in some
parts of the North Central zone (the Middle Belt), mobilisations along emer-
gent ethno-religious identities have tended to result in frequent violent
conflicts.
Nigeria has about 350 ethnic groups (Otite, 1990). Since the colonial
period, however, some of these ethnic groups have played dominant roles in
politics. These groups, known as the ‘major ethnic groups’, historically dom-
inated the three regions formally established as centres of government in
1946. In fact, the placement of the peoples with similar cultures and related
languages in colonially drawn administrative boundaries promoted the rise
of ethnic consciousness, cohesion, and mobilisation (Nnoli, 1978). Thus,
Hausa and Fulani ethnic groups dominated the Northern Region, while
the Yoruba and Igbo were dominant in the Western Region and Eastern
Region, respectively. The other ethnic groups in the regions constituted
the minorities and were known as Northern minorities, Western minori-
ties, and Eastern minorities. The larger minority groups in the North are
the Kanuri, Angae, Berom Gwari, Kataf, Nupe, Tiv, Idoma, Ebira, Chamba
and Igala. These groups are found mostly in the North Central and North
East geopolitical zones. The Western minorities include the Edo, Urhobo,
Ijaw, Itsekiri, Ishan, Isoko, Ika Igbo, and Etsako among others, while the
Eastern minorities include the Ijaw, Efik, Ibibio, Ikwerre, Anang, Ogoni, and
Ejagham among others. Both the Eastern and Western minorities currently
belong to the South South geopolitical zone. A key feature of Nigeria’s eth-
nic configuration is the fluidity of both identities and boundaries. Against
this background, groups can adopt entirely new identities or fuse with other
groups. For example, because Igbo identity was considered a political lia-
bility after the collapse of the Igbo-led Biafran secessionist attempt, some
groups on the Igbo frontiers have privileged their sub-group identities over
Igbo identity (Ndu, 2011).
It is almost impossible to estimate the current population size of the eth-
nic groups. This is because of the disputes that threatened the fabric of the
180 Nigeria
nation over the controversial census exercises of the 1950s and 1960s. Since
the end of the civil war, successive governments have omitted ethnic and
religious identities from the censuses. Although ethnic leaders have tried
to use the state populations of successor states and provinces as proxies
of the populations of the different ethnic groups, the figures they present
are evidently exaggerated as the non-ethnic resident populations are hardly
factored into such computations.
Ethnic mobilisations in the late colonial period and early years of indepen-
dence were marked by the quest for minorities in each of the three regions
to pull out of the regions through the creation of new states. Often the three
major ethnic groups that tussled for supremacy at the federal tier of gov-
ernment were drawn into the conflicts due to the alliances they forged with
minorities in neighbouring regions in order to undermine their rival parties.
Because ethnic identity almost coincided with party affiliation, such political
alliances often crystallised as ethnic alliances (Coleman, 1958; Sklar, 1963;
Vickers 2000).
Successive state-creation exercises have, however, aggravated the politics
of marginalisation and ethnic domination. They have led to the emergence
of new major ethnic groups and new ethnic minorities in the new states.
This has resulted in endless agitations for the creation of new states. As a
result of the dynamic changes in the structure of the federation as shown
in Table 8.1, the configuration of ethnic cleavages is fluid. Throughout the
colonial and early postcolonial periods, the structure was tripolar, with the
Hausa, Igbo, and Yoruba competing for supremacy. However, this configura-
tion has evolved into a multipolar structure with the ascendancy of minority
ethnic groups following the successive creation of states, the recognition of
six geopolitical zones, and the discovery of oil and gas deposits among the
Southern minorities. The geopolitical zones have become particularly salient
since 1999 when they became the template for distribution of principal fed-
eral posts such as the president, vice president, Senate president, speaker
of the House of Representatives, and secretary to the federal government
(Figure 8.1).
Furthermore, there are situations where the most salient cleavage is
between the North and South, leading to a bipolar structure. In such
NIGER
CHAD
Sokoto
Katsina
Jigawa
Zamfara
Kano Yobe Borno
Kebbi
Figure 8.1 Political map of Nigeria, showing states and geopolitical zones
circumstance, the minorities in one region align with the major ethnic
group(s) in the region against the major ethnic group(s) and minorities in
the other region(s). The situation is also dynamic in the 36 states of the
federation, depending on their ethnic composition. For instance, sub-ethnic
identities become more salient and are mobilised for political contestations
in states where there is only one indigenous ethnic group. Such situa-
tions often result in multipolarity. Multipolarity also characterises the ethnic
structure of states with a number of fairly equally sized ethnic groups.
However, in cases, where there is a dominant ethnic group, a bipolar struc-
ture often emerges, as the minority ethnic groups coalesce into a bloc to
neutralise the dominant group.
Religious cleavages are also salient in Nigeria, with Islam and Christianity
competing for influence and dominance. The historical dominance of the
religions in different regions has generated conflicts as dominant groups
react to protect their turf in response to the quest of rival religions to
make inroads. Although other religions exist in Nigeria, the configuration
182 Nigeria
Zone Military Head Prime Senate Speaker of Chief Total Rank Percentage Rank in
of State Minister/ President House of Justice of number in of total population
President Representatives the of posts posts population
Federation 2006a
Note: a. As this column excludes FCT, the figures do not add up to 100.
Source: Author’s compilation.
183
184 Nigeria
service positions. The application of such quotas has helped to bridge the
gap between the North and the South. This explains why the North West
and North East zones have a disproportionately high percentage of directors
relative to their share of staff in the service (Table 8.6).
It should be noted that consolidation of staff by geopolitical zones and the
entire public service occludes certain salient disparities. There is a tendency
for particular units of government to be dominated by persons from particu-
lar zones and states. This suggests that, though very salient, the North–South
dichotomy hides more than it reveals. For instance, while the South South
and South West zones are generally well represented in the federal service,
this cumulative representation masks the overrepresentation of states such
as Ogun in the South West and underrepresentation of states such as Bayelsa
in the South South.
Moreover, some groups tend to dominate different government depart-
ments. For instance, the staff profile of the National Assembly (Table 8.7)
186 Nigeria
shows the dominance of personnel from the North Central, South East and
South South zones. The fact that two of the three zones (North Central and
South East) have produced more Senate presidents and speakers of the House
of Representatives, especially in the early years of the institution, lends some
credence to suspicions that heads of governmental institutions influence
recruitment of persons from their ethnic groups.
These disparities within zones and regions have moderated the potential
of HIs triggering violent conflict, as they stifle the emergence of an undif-
ferentiated community of the deprived and, ipso facto, the possibility of
mobilisations of one group against the other. A clear case in point was the
undisguised lukewarm attitude of key leaders of the North Central zone to
mobilisations for the election of another president from the North after the
death of Alhaji Umar Musa Yar’Adua in March 2010. Another dimension
of HIs in Nigeria that has helped to stymie violent conflict between the
North and South is the fact that political HIs and socioeconomic HIs are not
mutually reinforcing. As shall be seen in the following discussion of socioe-
conomic HIs in Nigeria, there is an emergent division of powers where each
Ukoha Ukiwo 187
discrimination when they lose out in public sector employment and pro-
motions due to the application of quotas in consonance with the Federal
Character Principle.
Generally, Nigerians in the Southern zones enjoy a better standard of liv-
ing than their Northern compatriots. As Table 8.12 shows, the Northern
zones lead in the percentage rate of Nigerians classified as ‘core poor’. In the
same vein, the Northern zones perform dismally relative to the Southern
zones in most of the human development indicators (Table 8.13). Further
disaggregation of standard of living measures indicates that the Southern
zones also fare relatively better in access to improved sources of water
and electricity. However, there is a mixed record with respect to access to
improved and not-shared toilet facilities, as the North West (47.6 per cent)
and South East (39.6 per cent) zones fare better than the South West (17.8
per cent) and North Central (22.3 per cent) zones (Table 8.14). Nevertheless,
the Southerners appear to be wealthier than the Northerners. As Table 8.15
shows, a greater percentage of Southerners belong to the highest wealth
quintile, while the Northerners have a greater percentage of their population
in the lowest wealth quintile.
The pattern of socioeconomic HIs derives from a number of historical,
economic, and ecological factors, notably the level of educational attain-
ments as well as the type and volume of economic activities. Socioeconomic
HIs are not driven or sustained by any specific state policies aimed at priv-
ileging the South. On the contrary, the North appears to be favoured in
state redistribution initiatives. For instance, Table 8.16 shows that more than
9000 public health facilities are located in the North as against 5000 in the
South. However, this does not imply that the Northern zones fare better than
190 Nigeria
Source: NPC of Nigeria and ICF Macro (2009: 321, 324, 327).
events. For instance, it has become fashionable for Nigerian presidents (and
first ladies) to dress like their hosts during their tours of different states
around the country.
Moreover, the presence of state government-owned media allows full
expression to cultures of minorities in the states in which they live. The
prominence given to the major ethnic groups in the federal government-
owned media is also being neutralised by the emergence of private media
organisations. Some of the media houses, such as African Independent
Television (AIT), which is owned by a business tycoon from the South
South zone, have consciously promoted the visibility of minority cultures
in the public sphere. Finally, different states and local government areas
have distinct policies aimed at supporting the development of the lan-
guages and cultures of their communities. All these direct and indirect
forms of representation of most Nigerian cultures in the public sphere have
reduced the risks of conflict associated with cultural alienation in plural
societies.
It was against this background that the military officers who seized power
sought to portray themselves as nationalists. Although the political class
had infiltrated the Nigerian military during the First Republic, the military
was one of the institutions that still had a national orientation at the time
the first coup took place (Luckham, 1971). However, certain actions and
inactions of the early military regimes, which culminated in the civil war,
eroded the autonomy of the military from social forces and its credibility as
a national institution. Successive post-war regimes have subsequently grap-
pled with the challenge of repositioning the military as a national institution
that would be instrumental in promoting national integration.
The prevailing perception that the virtual independence of the regions was
the bane of the Nigerian federation influenced the military regimes to intro-
duce institutions aimed at transforming the country from a federation with
a historically weak centre and strong federating units to a federation with a
strong centre and weak federating units. This centralising tendency was also
a product of the inherently hierarchical nature of the military as an insti-
tution. The policy instruments for the centralisation of the federation and
the enfeeblement of the federating units included dividing big regions into
smaller states and transferring more powers and resources to the federal gov-
ernment. The modality of military-engineered state-creation exercises also
factored in the need to discourage ethnic mobilisation. Consequently, the
boundaries of new states were drawn in such a way that they do not neatly
coincide with ethnic boundaries; also important was the fact that no state
was named after an ethnic group. As shown above in Table 8.1, the feder-
ating units increased from four regions to 12 states in 1967 and 19 states
in 1976.
Furthermore, the military jettisoned the Westminster cabinet system char-
acterised by shared leadership and collective responsibility in favour of a
presidential system where power is concentrated in the chief executive.
The introduction of a powerful executive at the federal and state levels
endowed with enormous powers and resources was intended to avert the
kinds of political crises that had been associated with jostling for power and
influence between the prime minister and the president in the First Repub-
lic. As midwives of Nigeria’s Second Republic (1979–1983), the military
regimes reckoned that the status of the prime minister as first-among-
equals was ill suited to a federation with strong, assertive, and centrifugal
ethno-regional-based groups.
The persistence and exacerbation of fears and allegations of ethnic dom-
ination also contributed to the attempts to further entrench power-sharing
and affirmative action policies in the 1979 Second Republic Constitution.
For instance, the Federal Character Principle, which stipulated that the com-
position of government institutions should reflect the diversity of groups
within the country, was enshrined in the 1979 Constitution. The Federal
Character Principle was strengthened further with the establishment, in
194 Nigeria
1997, of the Federal Character Commission (FCC), which was charged with
monitoring and enforcing application of the Federal Character Principle, and
the subsequent entrenchment of the Commission in the Fourth Republic’s
1999 Constitution.
Moreover, in its desire to promote national unity, the military government
introduced a number of institutions to discourage the creation of ethnic-
based parties. Since the Second Republic, successive Nigerian political parties
have been required to have a national outlook (Horowitz 1985). Successive
constitutions have stipulated that, to qualify for registration, parties should
have offices in a minimum of two-thirds of the federation’s states and that
they should also reflect the Federal Character in the selection of their offi-
cers. This stringent requirement for registration has been relaxed since 2007
when the law courts ruled that it infringed upon the fundamental rights of
Nigerian citizens to freedom of association. However, the electoral system
stipulates that in addition to securing a majority of votes, candidates for
presidential and governorship elections should also secure a quarter of the
votes in two-thirds of the states in the federation or of the local government
areas in the state, respectively, which continues to be a major incentive for
the emergence of inter-ethnic and cross-regional alliances in political party
formation.
The attrition rate from minority parties is, however, mitigated by the ten-
dency for parties that lose national elections to win elections in vital states
or zones. Control of state power in some states enhances prospects for the
survival of the party if the persons elected under its platform remain com-
mitted to the party and provide it with patronage resources. The survival
of the All Nigeria Peoples Party (ANPP) and the Action Congress of Nigeria
(ACN) has stemmed from their ability to retain control in some strategic
states, where loyal party members were elected as governors. Such governors
have in some cases been able to lure members of other parties, including
the dominant PDP, to their parties. The PDP has also lost members in situa-
tions where there is a conflict of interest between state-based party stalwarts
and the national leadership. For instance, the quest of some prominent Igbo
politicians to contest for the presidency contributed to the PDP’s loss of two
states in the South East zone in 2007.
remained the need to preserve unity in the midst of diversity. The fed-
eral structure was considered to be the ‘natural’ form of government in a
country with over 300 ethnic groups who want to preserve their individual
identities while seeking common protection and benefits under the Nigerian
federation.
Federalism is the overarching power-dispersing institution in Nigeria.
Nigeria’s federalism has been calibrated to address the twin fears of ethnic
domination and national disintegration. As indicated earlier, this has been
accomplished by creating centres of countervailing power at the federal,
state and local levels. Federalism, in its three-tiered format, provides a sphere
of autonomy for most identity-based groups. Groups that are demographi-
cally insignificant in the context of national politics can become salient at
the level of either state governments or local councils.
Nigerian federalism was adopted not just in recognition of the diverse
ethnic topography of the country, but also as an institutional mechanism
to address HIs that contribute to violent group mobilisation. It is apposite
at this juncture to examine the extent to which the complex federal archi-
tecture has impacted on the nature of HIs. This will be examined through
an analysis of the outcomes of some strategies designed over the years to
operationalise Nigeria’s federalism.
principle was aimed at ensuring the spread of opportunities across the coun-
try and preventing ethnic domination. More precisely, the objectives of the
Federal Character Principle are:
Forty five per cent of the total number of management positions in the
recent promotion exercise conducted by the Nigerian National Petroleum
Corporation (NNPC) went to officers from the South South geopoliti-
cal zone . . . The home region of both President Goodluck Jonathan and
Minister of Petroleum Resources Diezani Allison-Madueke grabbed 17
new management staff positions out of the 38 on offer. The docu-
ments indicated that six out of the top eight management positions
of Group General Manager (GGMs) are occupied by South South indi-
genes . . . . In the promotion list, Southeast came second with seven posts
equivalent to 18 per cent while Northwest came third with 5 new posts
or 13 per cent. South West and North Central zones occupied the fourth
position with 4 new posts each which or 10.5 per cent, while the North
East was left holding the hat with only one post . . . In all, the North has
only 10 new posts out of 38 and the remaining 28 went to the southern
part of the country.
(Muhammad, 2011)
Another area where the application of the Federal Character Principle has
contributed to tackling HIs is in the location of federal institutions. Since
the 1970s, when it started to proactively promote university education, the
Federal Government has pursued a policy aimed at establishing federal uni-
versities in all states of the federation (Ukiwo, 2007). Although economic
recession forestalled realisation of this objective in the 1980s and 1990s,
improvements in the revenue profile of the federal government have made
the government respond to public pressure for the establishment of federal
universities. In 2011, President Goodluck Jonathan approved the establish-
ment of nine more federal universities. The result of such initiatives is that
most states now have educated elements to work in the state civil service
and fill the state’s quota in the federal civil service, thereby reducing ten-
sions that arose from the dependence of some states/regions on educated
elements from other states for the basic functioning of public services. Simi-
lar initiatives have been adopted by the federal government in the location
of tertiary health care institutions and airports, as well as in the construction
of federal highways.
Another intrinsic mechanism of the country’s federalism that tackles HIs is
the Federation Account. The 1999 Constitution allows for federally collected
revenues to be pooled into an account known as the Federation Account and
empowers the National Assembly to legislate on the formula for distribution
of the revenues. These include vertical allocation between the three tiers of
government (federal, state and local government) as well as horizontal allo-
cation among the states. Among the criteria applied for revenue allocation
are derivation, population, landmass, equality of states and ecological chal-
lenges. The criterion of equality of states has led to the provision of block
grants that have enhanced the capacity of less-endowed and indigent states
to provide basic state services. Although such equalisation transfers have not
bridged historical gaps, given disparities in resource endowment and tax col-
lection capacities, they have ensured the survival of many non-viable states,
and, ipso facto, the stability of the federation. Nevertheless, there is still no
consensus about a revenue allocation formula, and contestations over the
revenue allocation formula remain a perennial threat to national cohesion.
The contentious issues remain the share of the federal government, and the
demands of oil-producing states for increment of the proportion allocated
on the basis of derivation.
Hausa 40 90 26 38
Igbo 28 72 33 61
Yoruba 42 65 19 55
Group I feel only I feel more I feel equally I feel more I feel only I don’t Total
Nigerian Nigerian Nigerian and [ethnic [ethnic know
than [ethnic identity] identity]
[ethnic identity] than
identity] Nigerian
Hausa 25 8 51 8 7 100.0
Igbo 3 13 60 17 7 100.0
Yoruba 10 14 60 12 2 100.0
All respondents 11 14 58 12 5 100.0
between groups (Langer and Ukiwo, 2008). It is against this background that
a perceptions survey was conducted for the present study. As a result of
resource constraints, the study was conducted in only one location: Lagos.
The choice of Lagos – Nigeria’s capital between 1906 and 1991 – was predi-
cated on its position as the most ethnically and religiously diverse city in
the country. This was based on the need to ensure that Nigerians from
all social backgrounds had an equal opportunity of being selected for the
survey.
The study sampled 12 enumeration areas with concentrations of diverse
populations. A total of 412 respondents based on a 50/50 gender quota
were selected for the study. Based on the random sampling method adopted
for the study, as expected, most of the respondents in the survey were
from the major ethnic groups, as follows: Yoruba, 45 per cent; Igbo, 22 per
cent; Hausa, 17.5 per cent; Southern minorities, 9.7 per cent and Northern
minorities, 6 per cent.
The study confirmed the perception that Nigerians attach importance to
their religion. However, it did not confirm the perception that Nigerians con-
sider their ethnic identity to be more important than their national identity,
as seen in the responses of persons from the major ethnic groups shown in
Tables 8.19 and 8.20.
Ukoha Ukiwo 205
The study confirmed the perception and official statistics that indicate that
Southern Nigerians generally fare better than Northern Nigerians in social
and economic indicators.
There was no evidence of ethnic animosity, as there was no strong objec-
tion to intermarriage on the basis of ethnicity. Most of the respondents
who would object to their sister or daughter being married to someone
from another group would do so on religious grounds. In fact, most of the
respondents (95 per cent) indicated their willingness to work with persons
from other ethnic groups, and only a cumulative 6 per cent of respondents
claimed that their perception of people from any other ethnic group was
negative. Positive attitudes were also generally expressed about the current
and future state of ethnic relations in the country.
The respondents were, however, polarised in their perception of the
importance of ethnic identity in access to state resources such as government
jobs, government contracts and admissions to public universities. Gener-
ally, respondents from Southern ethnic groups felt ethnicity mattered a great
deal for access, while their counterparts from the North felt ethnic identity
was not important. However, most of the respondents felt that ethnicity has
become more important in the public sphere since 1999.
Interestingly, most of the respondents felt that the socioeconomic status
of their ethnic group was better than it had been 20 years ago. Signifi-
cantly, though, the Igbo felt they were underrepresented in government
and the armed forces. There was strong support for affirmative action for
marginalised groups, and most of the respondents felt the federal govern-
ment accorded recognition to their culture, even though most of them felt
the Hausa were given the greatest visibility in the public sphere.
The survey confirmed the perception among the Igbo that they are
marginalised in the Nigerian federation. Perception of marginalisation
tended to influence voting behaviour with the Igbo respondents more than
any other group, indicating they would attach importance to the ethnic
identity of candidates during presidential elections. This probably explains
the strong clamour for an Igbo presidency, as this group is the only major
ethnic group that has not produced a democratically elected president.
Since the end of the civil war in 1970, Nigeria has continued the nation-
building project by institutionalising mechanisms that would simultane-
ously protect national stability and prevent ethnic domination. The fact
that the country has not experienced any major secessionist conflict dur-
ing this 40-year period testifies to the modest progress recorded in national
integration. The country has so far defied the thesis that countries that
have experienced civil wars are likely to experience repeated implosions (see
Collier and Hoeffler, 2004). This has largely resulted from the ability of the
206 Nigeria
9.1 Introduction
208
Arnim Langer and Satoru Mikami 209
In this section we will examine the main reasons why people’s perceptions of
the prevailing HIs in a country may differ sharply from more objective mea-
surements or assessments of these inequalities. One issue that complicates
matters in this respect is that there may be sharp differences in the perceived
inequalities across ethnic groups. Thus, for instance, in societies with sharp
objective HIs (possibly resulting from past and/or ongoing discriminatory
practices by the state), it is not unlikely that the deprived groups will ‘cor-
rectly’ perceive that they are in a relatively disadvantaged position compared
to other groups, but their perceptions may nonetheless reflect a considerably
worse or better picture than the one that emerges from the analysis of objec-
tive data. Relatively advantaged groups in horizontally unequal societies
may also ‘correctly’ perceive their relatively privileged position, although
they may have very different views about the level of inequality compared
with disadvantaged groups and also of the causes of the prevailing inequal-
ities. Moreover, even in cases where the objective HIs are not very severe,
there may still be substantial differences in the perceived levels of HI across
different ethnic groups.
There are a number of reasons why there can be a mismatch between the
objective and subjective HIs in a particular society, which we outline below.
ethnic group with different levels of income and welfare should in prin-
ciple have the same perceptions about their group’s situation and relative
position. However, it is not unlikely that people’s individual socioeco-
nomic background and situation may colour their perceptions of the
prevailing group inequalities.
• Manipulation of perceptions by elites or group leaders. In order to gain polit-
ical support (or pre-empt losing it), the leaders or elites of a particular
group may decide to manipulate their constituents’ perceptions of the
prevailing HIs. While elites occasionally attempt to mitigate perceptions
of inequality (for example, to pre-empt criticism that they have not done
enough to improve their group’s socioeconomic situation and relative
position), it appears to be more common that they try to exacerbate
the existing perceptions of inequality among their group members or
constituents in order to gain or maintain political support. Leaders of
relatively advantaged groups, in turn, may play down the severity of the
prevailing inequalities and concomitantly stress that the deprived groups
are themselves to blame for their relatively disadvantaged situation.
• Inaccurate media reporting. The media can play an important role in
bringing objective HIs to the attention of the population at large. Yet,
inaccurate reporting on the part of the media due to sloppy reporting, a
lack of sufficiently qualified and experienced journalists, or for political
reasons can clearly have a major impact on people’s perceptions of the
existing HIs and possibly their perceived causes.
• Lack of objective data on horizontal inequalities. Ethnically segregated
socioeconomic data are usually not readily available. Sometimes – as, for
example, in Nigeria – ethno-cultural variables are not included in surveys
because of their political sensitivity (Okolo, 1999). While language and
region can sometimes be used as proxies for ethnic groups, in a substan-
tial number of countries this might not be possible or might not provide a
sufficiently accurate picture of the prevailing HIs. The absence of accurate,
comprehensive and independent data on HIs in many multiethnic coun-
tries increases the risk that people’s perceptions might instead be based
on personal experiences, opinions and stories of friends, family and peo-
ple in positions of ‘power’ (such as politicians, community leaders and
church leaders), or even on rumours and hearsay.
• Insufficient access to information. Another reason why objective and sub-
jective inequalities may differ is because certain groups may lack access
to the necessary information and data to form a reasonably accurate pic-
ture of the prevailing HIs in their country. Thus, for instance, groups in
rural and geographically remote areas may have insufficient access to the
media or other sources of information, which in turn makes it difficult for
them to compare their own situation to that of other groups. Low mobil-
ity among the people living in rural and remote areas is another obstacle
for assessing the relative position of their own group.
212 Objective and Subjective Horizontal Inequalities
To what extent these factors are at play in our five case studies, in which
specific combinations and to what effect are issues that go beyond the scope
of this chapter. However, in cases where we observe a mismatch between
objective and subjective HIs (see Section 9.5), we will examine the impact of
some of these factors in more detail.
Igbo (n = 358)
Yoruba (n = 491)
Other Nigerians
Hausa/Fulani
Igbo (n = 366)
Yoruba (n = 486)
Other Nigerians
(n = 603)
(n = 814)
(n = 578)
(n = 815)
Ghana: BHNs (not adjusted) Ghana: assets (not adjusted)
1 1.2
0 1
0.8
–1
0.6
–2 0.4
–3 0.2
–4 0
Akan (n = 588)
Ewe (n = 156)
Ga/Dangbe
MoleDagbani
Other
Ghananians
Akan (n = 581)
Ewe (n = 157)
Ga/Dangbe
Mole/Dagbani
Other
Ghananians
(n = 123)
(n = 100)
(n = 123)
(n = 103)
(n = 192)
(n = 193)
Ndebele
Other
Zimbabweans
Shona (n = 897)
Ndebele
Other
Zimbabweans
(n = 147)
(n = 147)
(n = 122)
(n = 121)
Figure 9.1 Ethnic inequalities according to the BHN-index and the Assets-index
Source: Authors’ calculations based on the Afrobarometer R4 surveys.
214 Objective and Subjective Horizontal Inequalities
Luo (n = 135)
Luhya
(n = 135)
Somali (n = 95)
Kamba
(n = 115)
Kalenjin
(n = 128)
Kisii (n = 66)
Other
Kenyans…
(n = 206)
Kikuyu
(n = 202)
Luo (n = 129)
Luhya
Somali (n = 93)
Kamba
Kalenjin
Kisii (n = 66)
Other
Kenyans…
(n = 193)
(n = 130)
(n = 109)
(n = 127)
Uganda: BHNs (not adjusted) Uganda: assets (not adjusted)
0 0.9
–1 0.8
0.7
–2 0.6
–3 0.5
–4 0.4
0.3
–5 0.2
Baganda
Banyoro
Banyankole
Acholi
Other
Ugandans
Baganda
(n = 523)
Banyoro
(n = 117)
Banyankole
(n = 273)
Acholi
(n = 179)
Other
Ugandans
(n = 1,289)
(n = 522)
(n = 117)
(n = 276)
(n = 177)
(n = 1,320)
to both indices. The figures depict the linear predictions of the point esti-
mates as well as the 95 per cent confidence intervals that were calculated on
the basis of the Afrobarometer surveys.
As can be seen in each plot, all countries covered here contain consider-
able gaps between the main ethnic groups according to one or both welfare
indices. In Nigeria, for instance, the Hausa/Fulani are poorer than the other
two main ethnic groups (the Yoruba and the Igbo) as well as the com-
bined group of other ethnic minorities; in Ghana, the Ga/Dangbe seem to
be significantly wealthier than other groups regardless of how we measure
living standards; in Zimbabwe, we find a difference between the Ndebele
and other ethnic minorities in terms of household assets; in Kenya, rela-
tionships are more complicated due to the greater number of major ethnic
groups, but the results still indicate that there is a significant gap between
the Kikuyu and the Somali, with rough parity between the remaining ethnic
groups; in Uganda, the Acholi and residual ethnic minorities exhibit con-
sistently lower levels of welfare compared to the three main ethnic groups
(the Buganda, Banyoro and Banyankole). Moreover, the picture that emerges
from the point estimates is much in line with other data and information
that are available on the relative socioeconomic situations of these differ-
ent ethnic groups (see, for example, for Ghana: Gyimah-Boadi and Asante,
2006; Langer, 2008; for Nigeria: Mustapha, 2006; Langer and Ukiwo, 2008;
for Kenya: Kanyinga, 2006: Stewart, 2010a).
It should be noted that the causes and origins of the prevailing socioeco-
nomic inequalities between different ethnic groups and/or regions in most
Arnim Langer and Satoru Mikami 215
The ethnic inequalities observed in the previous section do not take into
account differences in factors that are conventionally used to explain dif-
ferences in standard of living across individuals. If we assume that the
factors conventionally associated with lower levels of individual socioeco-
nomic welfare are unrelated to ethnicity, we should be able to explain a
large proportion of the observed ethnic inequalities by means of these indi-
vidual ‘risk’ factors. In order to test this, we aim to explain as much as
possible the variances of (a) assets and (b) BHNs with the help of several
commonly used risk factors and then see if the included ethnic dummies
are significantly different from zero or not. The risk factors we include in
our regression analysis are: (1) level of educational attainment; (2) people’s
employment status; (3) level of infrastructural development of people’s liv-
ing environment; and (4) people’s experience of physical insecurity due to
violence and crime. Also incorporated are individual demographic variables,
such as sex, age, region of birth (whether a respondent is indigenous to a par-
ticular place of living) and ethnicity as well as a range of context variables,
such as the urban/rural distinction, and a series of dummies for each of the
administrative distinctions made in a country (province, region or state).
It is important to note here that not all variables are available in both
survey datasets. Thus, while Afrobarometer lacks information on people’s
region of birth, an infrastructure variable is not included in our survey.
Also, variables for insecurity and employment are differently measured (see
Appendix 1 for more details on these issues). While we can use an OLS regres-
sion for our survey, this is not appropriate for the Afrobarometer surveys
216 Objective and Subjective Horizontal Inequalities
Nigeria
Ghana
Zimbabwe
Kenya
Uganda
In this section we will analyse whether the risk factors associated with indi-
vidual standards of living are distributed equally among different ethnic
groups. If the probability of being exposed to these risks does not differ from
ethnic group to ethnic group, only then can these factors be considered to
contribute to individual well-beings independently of ethnic affiliations. If,
however, the probability of facing one of these risk factors varies across eth-
nic groups, the root cause of inequality is ethnicity, whose effect is mediated
through these factors.
It should be further noted that the risk factors that were analysed in the
previous section are related to each other. For example, someone’s risk of
living in a rural area depends on the region in which he/she lives, because
the proportions of rural areas differ from region to region. Moreover, the
region of residence is itself a risk factor that could affect individuals’ stan-
dard of living due to a region’s climatological and ecological characteristics.
Likewise, the risk of facing infrastructural underdevelopment depends on,
among other things, the place of residence (that is, in an urban or rural
area) as well as the region of residence. Risk of quitting school after primary
education depends on factors such as the extent to which infrastructure is
available (such as secondary schools), access to school, degree of urbani-
sation and region of residence, as well as on gender and generation. Risk
of being threatened or actually victimised by violent crime is determined
largely by similar factors. Finally, risk of being unemployed depends on edu-
cational attainment, gender, age and infrastructure. Therefore, in assessing
ethnic gaps in risk factors, we need to control for these ‘interdependencies’
accordingly.
Let us start by examining whether the infrastructural development people
enjoy systematically differs according to the group to which they belong. Of
course, infrastructure is supposed to be ‘public’, meaning everyone, at least
among the citizens of a country, should be able to use it without being dis-
criminated against. In some cases such as paved roads, even non-citizens can
benefit from infrastructural development. Indeed, one of the defining char-
acteristics of public goods is their non-excludability. However, if different
Arnim Langer and Satoru Mikami 227
groups are segregated from each other even within the same region and if
they have very low levels of mobility, inherently ‘public’ infrastructure can
become ‘private’ (or a ‘club’ good). The individuals who benefit from the
infrastructural development of a particular area tend to be members of a
particular ethnic group because the group tend to live in that area. In such
a context, infrastructure development site or its priority might be contested
among ethnic groups.
In order to detect possible group differences resulting from ethnic com-
petition or discrimination in infrastructural development, we regress our
infrastructural development index on ethnic affiliation, controlling for
regions of residence as well as the urban/rural distinction. Figure 9.2 presents
the 95 per cent confidence intervals of the average scores of infrastructure
development for each ethnic group. In Nigeria and Zimbabwe, no signifi-
cant ethnic gap is found in terms of infrastructure development, although
average scores for the Yoruba and the Ndebele are higher than the respec-
tive reference groups. In Ghana, however, the Ga/Dangbe have statistically
significant advantages in terms of infrastructure development over the Ewe
and other residual ethnic minorities. In Kenya, too, we find significant eth-
nic gaps: groups with relative advantages are the Luhya, the Kalenjin and
the Kisii, while the Luo and the Kamba face higher risks of infrastructure
underdevelopment. Finally, in Uganda, we find a significant gap between
the Baganda and other residual ethnic minorities on the one hand and the
Banyoro on the other hand. It should be noted that the Banyankole, like
the Kikuyu in Kenya, do not necessarily benefit most from infrastructural
development, although they do not suffer most either.
Let us now turn to the ethnic gaps that might exist in education, espe-
cially in terms of the risk that one quits schooling early. Because educational
attainment usually varies across generations and gender as well as environ-
mental factors like the availability of infrastructure just examined, we need
to control for these variables when estimating ethnic influences on the risk
of ‘quitting’ education before or at completion of primary school. Therefore,
we applied a mixed-effect logit model, assigning 1 to those whose highest
level of educational achievement is at most completion of primary educa-
tion and 0 otherwise. We present the adjusted predicted probabilities (based
on fixed parts) of lower educational attainment for each ethnic group (that
is, quitting school before or at completion of primary schooling) in our case-
study countries for the case of a 35-year-old male living in an urban area
with an average infrastructure in the capital city (Figure 9.3).
The results show that in Nigeria, the Hausa/Fulani face a significantly
higher risk of having lower educational attainment compared to the Igbo
and the Yoruba, while the educational difference with the residual ethnic
minorities is not statistically significant. In Ghana, the Mole/Dagbani face
a higher risk of not having post-primary schooling when compared to the
Ewe and the Ga/Dangbe. However, these differences do not have statistical
228 Objective and Subjective Horizontal Inequalities
Nigeria Ghana
1.6 3.2
1.5 3
1.4 2.8
1.3 2.6
2.4
1.2
2.2
1.1
2
1 1.8
Hausa/ Igbo Yoruba Other
an
ni
an r
ni he
Ew
gb
ba
Fulani Nigerians
Ak
s
na Ot
an
ag
D
D
a/
e/
G
ol
ha
M
G
Zimbabwe Kenya
4.1 5
4.8
4 4.6
4.4
3.9 4.2
4
3.8 3.8
3.7 3.6
3.4
3.6 3.2
3
3.5 2.8
Shona Ndebele Other
in
Ka ba
So a
yu
Lu o
ny er
Ka ali
i
hy
si
s
Lu
nj
Ke Oth
ku
an
m
Zimbabweans
Ki
le
Ki
Uganda
4.2
4
3.8
3.6
3.4
3.2
3
a
le
li
ho
nd
or
nd er
ko
s
an
ny
ga th
Ac
ga
an
O
Ba
Ba
ny
Ba
Nigeria Ghana
35% 60%
30%
50%
25%
40%
20%
15% 30%
10% 20%
5% 10%
0% 0%
Hausa/ Igbo Yoruba Other
an
ni
an r
ni he
Ew
gb
ba
Ak
s
na Ot
Fulani Nigerians
an
ag
D
D
a/
e/
G
ol
ha
M
G
Zimbabwe Kenya
60% 100%
50%
80%
40%
60%
30%
20% 40%
10% 20%
0% 0%
Shona Ndebele Other
in
So a
Ka ba
yu
Lu o
ny er
Ka li
i
si
hy
Lu
nj
s
a
ku
Ke Oth
m
an
Ki
m
Zimbabweans
le
Ki
Uganda
40%
35%
30%
25%
20%
15%
10%
5%
0%
da
le
ro
nd er
li
s
o
ho
an
n
ga h
nk
ga
ny
U Ot
Ac
ya
Ba
Ba
n
Ba
Total Total
Shona Shona
Ndebele
Ndebele
Other
Zimbabweans Other
Zimbabweans
Think about the condition of your ethnic group. Are their economic
conditions worse, the same as, or better than other groups in this
country?
The Afrobarometer surveys included the same question. People were asked
to respond to this question on the following five-point ordinal scale: 1,
much better; 2, better; 3, same; 4, worse; and 5, much worse. We sub-
sequently consolidated people’s answers into a three-point ordinal scale:
‘superior’ (much better/better); ‘same’ (same); and ‘inferior’ (worse/much
worse). Figure 9.4 depicts the distributions of responses according to our
three-point ordinal scale for each case-study country, where the results based
on the Afrobarometer surveys are displayed on the left-hand side and those
based on our surveys on the right-hand side.
It is interesting to see that the distribution patterns of both surveys are
quite similar despite the underlying differences in the scope and timing of
both surveys. For example, in both surveys, it emerges that the most frus-
trated groups in each country are as follows: the Igbo in Nigeria, the Ewe
in Ghana, the Ndebele in Zimbabwe, the Somali in Kenya and the Acholi
in Uganda. Except for Nigeria, there also appears to be consistency between
both surveys regarding the most ‘satisfied’ ethnic groups: the Mole/Dagbani
in Ghana, the Shona in Zimbabwe, the Kikuyu and the Kisii in Kenya and
the Banyankole in Uganda.
232 Objective and Subjective Horizontal Inequalities
Comparing the results of Figure 9.4 with our previous analysis, it appears
that there are some notable discrepancies between the perceived and
objective HIs in our case-study countries. Thus, for instance, in Nigeria,
while the Hausa/Fulani were ‘objectively’ the poorest ethnic group, it was
among the Igbo respondents that the highest proportion of people felt
that they had an inferior level of socioeconomic development. In Ghana,
it was the Mole/Dagbani group that was objectively most disadvantaged, yet
the Ewe respondents were the ones who held this perception most widely.
In Zimbabwe, while there was no sharp objective divide between the Shona
and the Ndebele, the latter group clearly perceived themselves to be seriously
disadvantaged. In Uganda, while the Baganda and the Banyoro were at least
objectively at the same level as the Banyankole, they do not seem to perceive
things that way. Moreover, some ethnic groups had a better than expected
view of their own socioeconomic situation compared to other ethnic groups.
Thus, for instance, the confidence in their superior status displayed by the
Kisii respondents in Kenya was largely ungrounded on the basis of the objec-
tive analysis. For some ethnic groups, the perceptions of HIs appear to offer
a relatively good reflection of the actual situation. Thus, for instance, the
Somali people in Kenya and the Acholi people in Uganda perceived them-
selves correctly to be at a disadvantage compared to other ethnic groups.
Similarly, the Kikuyu in Kenya and the Banyankole in Uganda correctly
perceive themselves to be in a relatively advantaged position.
Returning to the factors that could induce a mismatch between objec-
tive and subjective HIs discussed in Section 9.2, we will now examine the
extent to which people’s individual socioeconomic situation influenced or
blurred people’s perceptions of the prevailing objective HIs. In order to do
this, we have to control for the effects of individual attributes as well as
environmental factors affecting respondents’ lives and see if the adjusted
distributions of perceived HIs converge to the expected distributions based
on the objective HIs. It is worthwhile recalling here that if perceptions reflect
objective group differences, people from the same ethnic group should
respond with the same answer regardless of their personal socioeconomic
status. Thus, two respondents from the same ethnic group – one being
rich and the other being poor – should choose ‘inferior’, if their group is
relatively disadvantaged compared to other ethnic groups. Similarly, two
equally wealthy respondents, with one of them belonging to a relatively rich
group and the other belonging to a relatively poor group, should choose
different answers regardless of the fact that they are in the same personal
socioeconomic situation. Complicating matters in this respect is the fact
that if there is an objective difference between ethnic groups, it is possible
that individual wealth correlates with perceiving one’s group to be supe-
rior in terms of socioeconomic development, as richer individuals are more
likely to belong to the relatively richer group, and vice versa. Yet once we
include ethnic dummy variables in our model, the variables that measure
respondents’ individual wealth situation should no longer have explanatory
Arnim Langer and Satoru Mikami 233
power, because the coefficients are adjusted to the situation where the
respondent’s ethnic group is the same.
Infrastructural development can also affect respondents’ perceptions of
the prevailing socioeconomic HIs. This is because it may have a direct
demonstration effect via its visual impact upon perceptions; but it can also
influence perceptions indirectly through the improvement of objective indi-
vidual living standards, which therefore also have to be controlled for by
individual wealth indicators. Therefore, in estimating the effects of ethnic
affiliation on respondent’s choice from our three possibilities (‘inferior’, ‘the
same’ and ‘superior’), it is necessary to control for the effects of differences
in infrastructural development as well as individual socioeconomic charac-
teristics, including education and employment. We also control for group
size by means of a dummy variable, which captures whether or not groups
are larger than 10 per cent of the total population. The logic for including
this variable is that larger groups are in a better position to ‘claim’ socioe-
conomic benefits and resources from the state than smaller groups, thereby
possibly improving their group’s position, which might be reflected in their
perceptions of the prevailing HIs.
Given that there is no mixed effect multinomial logit model, we decided
to divide the three answering options (‘inferior’, ‘the same’ and ‘superior’)
into two sets of binary variables, and used mixed-effect logit instead. When
estimating the probability of feeling inferiority, the dependent variable takes
the value 1 if a respondent chooses ‘inferior’ and 0 if the answer is ‘the same’,
excluding those who choose ‘superior’ from the analysis. When estimating
the probability of perceiving group ‘superiority’, respondents who choose
‘inferior’ are excluded from the analysis. Respondents who chose ‘superior’
are assigned the value 1, while respondents who answered ‘the same’ were
assigned 0. The results of this analysis are shown in Table 9.3, where each
case-study country has four models. The first column from the left lists the
effects of the above variables on the probability that respondents perceived
their own group to be inferior to other groups. Conversely, the third col-
umn lists the effects on the probability that respondents perceived their own
group to be superior to other groups.
As expected, most variables measuring individual socioeconomic status
have no significant effect. The exception is the BHN indices. In four of our
five case studies, it emerges that the respondents who are better able to sat-
isfy their basic human needs are also less likely to perceive their own group
to be inferior to other groups. This suggests that people’s individual socioe-
conomic situation to some extent affects their assessment of the prevailing
socioeconomic HIs, which was argued to be a possible factor for the exis-
tence of a mismatch between objective and subjective HIs in Section 9.2.
Conversely, there is little evidence to support the idea of a ‘visual’ effect
of infrastructural development. Although respondents in areas with more
infrastructural development are less likely to think their group is inferior
to other groups in Uganda, unexpectedly, the opposite effect is found in
Table 9.3 Determinants of perceived socioeconomic HIs
234
coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value
Nigeria
Politically inferior 2.176 0.000
(0.162)
Politically superior 1.767 0.000
(0.142)
Assets 0.132 0.234 0.079 0.514 0.278 0.005 0.229 0.029
(0.111) (0.120) (0.100) (0.105)
BHNs −0.048 0.005 −0.041 0.027 0.032 0.035 0.018 0.258
(0.017) (0.018) (0.015) (0.016)
Lower education 0.005 0.973 −0.076 0.664 −0.007 0.966 0.022 0.888
(0.163) (0.176) (0.152) (0.160)
Part-time 0.140 0.389 0.222 0.205 −0.263 0.083 −0.290 0.069
(0.163) (0.175) (0.152) (0.160)
Full-time 0.357 0.028 0.363 0.039 −0.200 0.187 −0.150 0.351
(0.162) (0.176) (0.151) (0.161)
Rural 0.140 0.460 0.229 0.229 −0.159 0.335 −0.282 0.090
(0.190) (0.191) (0.165) (0.166)
Infrastructure 0.016 0.868 0.117 0.240 0.030 0.723 −0.024 0.777
(0.099) (0.100) (0.084) (0.085)
Groupsize10%+ −0.474 0.220 −0.226 0.570 0.181 0.589 −0.305 0.385
(0.387) (0.397) (0.335) (0.351)
Yoruba −0.114 0.653 −0.212 0.407 −0.298 0.148 0.157 0.463
(0.254) (0.256) (0.206) (0.214)
Igbo 0.509 0.059 0.029 0.918 −0.538 0.027 0.108 0.668
(0.269) (0.277) (0.242) (0.252)
Other Nigerians −0.228 0.556 −0.289 0.466 0.082 0.808 0.139 0.691
(0.387) (0.396) (0.339) (0.351)
Intercept −0.556 0.213 −1.487 0.001 −0.462 0.237 −0.878 0.030
(0.447) (0.461) (0.391) (0.404)
sd(_cons) 0.893 0.769 0.579 0.475
(0.121) (0.132) (0.104) (0.117)
Number of observation 1305 1286 1336 1319
Number of psu 235 234 231 231
Wald ch2 24.200 0.012 198.560 0.000 22.950 0.018 168.810 0.000
Log likelihood −841.692 −717.444 −885.569 −788.459
Ghana
coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value
Zimbabwe
coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value
Kenya
Uganda
Politically inferior 2.367 0.000
(0.142)
Politically superior 2.093 0.000
(0.190)
Assets 0.045 0.660 0.076 0.504 0.247 0.030 0.151 0.219
(0.102) (0.114) (0.114) (0.123)
BHNs −0.091 0.000 −0.055 0.002 0.074 0.000 0.062 0.005
(0.015) (0.017) (0.020) (0.022)
Lower education −0.285 0.025 −0.216 0.134 0.006 0.967 −0.075 0.657
(0.127) (0.144) (0.155) (0.169)
Part-time −0.173 0.202 −0.238 0.121 −0.479 0.005 −0.470 0.010
(0.136) (0.154) (0.169) (0.183)
239
Table 9.3 (Continued)
240
coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value coefficient (SE) p-value
In the same vein as for the perceived socioeconomic HIs in Figure 9.4,
Figure 9.5 shows the distribution of responses to this question for both
surveys for our five case-study countries. The results show that the
Hausa/Fulani respondents perceive themselves to be in a relatively advan-
taged position, while the Igbo feel relatively marginalised politically.
In Ghana, it is the Akan who feel relatively advantaged, while a substan-
tial proportion of Ewe respondents feel politically inferior according to
Afrobarometer. In Zimbabwe, the Ndebele feel largely excluded from polit-
ical power. The Kenyan situation is very complex; yet, it appears that the
Luo are quite satisfied with their degree of political power perhaps due to
Odinga’s rise to power, while more than 70 per cent of the Somali people
felt they were politically inferior. Lastly, in Uganda, Banyankole respondents
appear relatively satisfied with their political situation, while the Banyoro
and Acholi feel considerably disadvantaged in terms of political power.
Without going into detail as to the extent to which people’s perceptions
of the prevailing political HIs are in line with the objective situation, the
Shona Shona
Ndebele Ndebele
Other Other
Zimbabweans Zimbabweans
Hausa/Fulani Igbo
26.34
28.36
30 22.15 30
15.94
25 25 16.96
10.23
20 20 9.06
8.72 6.43
15 11.91 15 11.99
5.7 2.85 4.97
7.6 4.39
10 10
4.19 Politically superior Politically superior
5 2.18 5
Politically equal Politically equal
Politically inferior Politically inferior
0 0
Economi- Economi- Economi- Economi- Economi- Economi-
cally cally cally cally cally cally
superior equal inferior superior equal inferior
36.53
25.32
40 30
35
25 14.56
30 16.49
17.97
20
25
8.86 6.71
6.89
20 9.39 15 6.96 4.68
12.32
15 7.31 2.92 8.99
10 5.95
10 Politically superior Politically superior
3.97 4.18
Politically equal 5 Politically equal
5
0 Politically inferior 0 Politically inferior
Economi- Economi- Economi- Economi- Economi- Economi-
cally cally cally cally cally cally
superior equal inferior superior equal inferior
deprivation is not controlled for. Likewise, when we control for the feeling of
political superiority, all ethnic dummies in Nigeria, Ghana, and Zimbabwe
lose statistical significance.
In Kenya and Uganda, we find a considerable reduction of the positive
influence of ethnic dummies on the feeling of economic inferiority in the
models that control for political group deprivation, except for the case of the
Luo who somewhat surprisingly show a strengthened tendency to exhibit
feelings of group economic deprivation. With regard to the probability of
feeling group economic superiority, we find that the Kamba and residual
minorities in Kenya, who are less likely to feel group economic superiority
compared to the Kikuyu when political feeling is not controlled for, are still
less likely to do so, but the size of the negative impacts themselves are con-
siderably reduced. The Kisii, on the other hand, become more likely to feel
economic superiority, which means that their economic confidence is sup-
pressed to a degree by their political feelings; this could be explained by the
fact that the Kisii are one of the most economically active communities in
Kenya. In the same vein, the Baganda in Uganda, who are not economically
inferior to the ruling Banyankole in objective terms, become more likely to
feel economic group superiority after political feelings are controlled for.
This chapter examined the extent to which objective and subjective HIs dif-
fered in five African countries. So far, this issue has been largely ignored in
the literature on HIs. It was established that there were significant socioe-
conomic HIs in each of our five case-study countries. We also examined
the extent to which these seemingly ethno-based gaps could be explained
by personal characteristics and/or other environmental factors commonly
used to explain differences in standards of living between individuals. While
these individual risk factors explained a significant portion of the observed
inequalities, ethnic affiliations remained important as well. It also emerged
that different ethnic groups faced different chances of experiencing these
risk factors. In other words, there were significant ethnic differences in the
distribution of these risks.
Having established the presence of objective socioeconomic HIs, we
analysed people’s perceptions of these inequalities and found a surpris-
ingly large discrepancy between our objective and subjective measures
of inequality, especially in Nigeria, Ghana and Zimbabwe. Further anal-
ysis of the mismatch between objective and subjective HIs showed, first,
that people’s individual socioeconomic situation tended to ‘distort’ their
perceptions of their group’s situation and relative position, and, secondly,
that there appeared to be cross-dimensional contamination, whereby peo-
ple’s perceptions of political HI had a significant effect on their perceptions
of the prevailing socioeconomic HIs. Moreover, given the finding that
246 Objective and Subjective Horizontal Inequalities
Appendix 1: Operationalisation
A. Radio
B. Television
C. Motor vehicle, car or motorcycle
Where is your main source of water for household use located? (1 = Inside the
house, 0 = Inside the compound or outside of the compound)
How often do you use:
A. A mobile phone?
B. A computer?
C. The internet?
248 Objective and Subjective Horizontal Inequalities
1 = Every day, A few times a week, A few times a month or, Less than once a
month, 0 = Never.
BHNs (both Afrobarometer and JICA): sum of standardised (within the
capital) scores of responses to the following five questions:
Over the past year, how often, if ever, have you or anyone in your family gone
without:
JICA Survey
Variable Country Obs Mean Std. Dev. Min Max Variable Country Obs Mean Std. Dev. Min Max.
Assets Nigeria 406 0.869 0.534 0.000 2.586 Lower education Nigeria 411 0.129 0.336 0 1
Ghana 324 1.033 0.767 0.000 2.889 Ghana 324 0.395 0.490 0 1
Zimbabwe 400 1.406 0.802 0.000 3.127 Zimbabwe 402 0.142 0.349 0 1
Kenya 904 0.885 0.663 0.000 2.213 Kenya 902 0.213 0.410 0 1
Uganda 498 1.257 0.905 0.000 3.875 Uganda 499 0.297 0.457 0 1
BHNs Nigeria 402 0.029 3.950 −13.929 4.276 Insecurity Nigeria 408 0.593 0.492 0 1
Ghana 324 0.000 3.395 −13.432 3.386 Ghana 324 0.309 0.463 0 1
Zimbabwe 397 0.107 3.376 −7.947 7.720 Zimbabwe 397 0.587 0.493 0 1
Kenya 892 −1.313 4.611 −19.482 2.980 Kenya 902 0.220 0.414 0 1
Uganda 491 −0.572 4.045 −14.819 3.712 Uganda 494 0.287 0.453 0 1
Female Nigeria 412 0.500 0.501 0 1 Job Nigeria 410 0.688 0.464 0 1
Ghana 324 0.491 0.501 0 1 Ghana 324 0.858 0.350 0 1
Zimbabwe 402 0.498 0.501 0 1 Zimbabwe 402 0.729 0.445 0 1
Kenya 907 0.502 0.500 0 1 Kenya 906 0.670 0.470 0 1
Uganda 500 0.566 0.496 0 1 Uganda 500 0.652 0.477 0 1
Age Nigeria 412 33.049 13.013 18 84 PSEHI Nigeria 404 2.698 0.862 1 5
Ghana 323 38.774 15.688 18 99 Ghana 313 3.080 1.002 1 5
Zimbabwe 402 33.458 12.665 18 84 Zimbabwe 385 2.917 0.809 1 5
Kenya 907 32.871 11.077 18 99 Kenya 880 2.880 0.849 1 5
Uganda 499 30.826 11.639 18 100 Uganda 484 3.045 1.064 1 5
Indigene Nigeria 412 0.427 0.495 0 1 PPHI Nigeria 403 2.918 0.939 1 5
Ghana 324 0.241 0.428 0 1 Ghana 303 2.865 1.076 1 5
Zimbabwe 402 0.343 0.475 0 1 Zimbabwe 354 2.672 0.981 1 4
Kenya 905 0.259 0.438 0 1 Kenya 875 3.170 1.388 1 5
Uganda 483 0.290 0.454 0 1 Uganda 484 3.279 1.099 1 5
249
Afrobarometer
250
Variable Country Obs Mean Std. Dev. Min Max Variable Country Obs Mean Std. Dev. Min Max
Assets Nigeria 2245 0.812 0.684 0.000 2.909 Rural Nigeria 2324 0.507 0.500 0 1
Ghana 1177 0.688 0.732 0.000 3.480 Ghana 1200 0.547 0.498 0 1
Zimbabwe 1186 0.825 0.888 0.000 3.607 Zimbabwe 1200 0.633 0.482 0 1
Kenya 1062 0.640 0.661 0.000 2.770 Kenya 1104 0.775 0.418 0 1
Uganda 2381 0.525 0.676 0.000 3.646 Uganda 2431 0.799 0.401 0 1
BHNs Nigeria 2266 −0.588 4.390 −13.556 4.635 Small Nigeria 2324 0.129 0.335 0 1
Ghana 1174 −1.281 5.020 −18.695 3.258 urban Ghana
Zimbabwe 1181 0.601 2.839 −6.204 7.385 Zimbabwe
Kenya 1097 −1.188 3.774 −13.758 4.380 Kenya
Uganda 2412 −2.432 4.182 −14.787 3.690 Uganda
Female Nigeria 2324 0.499 0.500 0 1 Infra- Nigeria 2051 1.648 0.942 0.000 3.640
Ghana 1200 0.500 0.500 0 1 structure Ghana 1096 2.046 1.328 0.000 3.915
Zimbabwe 1200 0.500 0.500 0 1 Zimbabwe 1192 1.985 1.418 0.013 3.816
Kenya 1104 0.502 0.500 0 1 Kenya 1104 1.808 1.399 0.000 4.863
Uganda 2431 0.499 0.500 0 1 Uganda 2423 1.838 1.570 0.000 6.020
Age Nigeria 2316 31.303 11.410 18 86 Employ- Nigeria 2299 0.773 0.841 0 2
Ghana 1184 39.017 16.459 18 110 ment Ghana 1191 0.959 0.894 0 2
Zimbabwe 1189 36.563 15.294 18 94 status Zimbabwe 1199 0.362 0.700 0 2
Kenya 1100 35.215 12.918 18 95 Kenya 1100 0.595 0.786 0 2
Uganda 2421 33.709 12.286 18 81 Uganda 2430 0.538 0.729 0 2
Lower Nigeria 2319 0.266 0.442 0 1 PSEHI Nigeria 2245 3.017 1.005 1 5
education Ghana 1195 0.623 0.485 0 1 Ghana 1121 3.093 1.114 1 5
Zimbabwe 1200 0.333 0.472 0 1 Zimbabwe 1142 2.948 0.901 1 5
Kenya 1100 0.459 0.499 0 1 Kenya 1078 3.412 1.053 1 5
Uganda 2431 0.504 0.500 0 1 Uganda 2334 3.586 1.059 1 5
Insecurity Nigeria 2293 −0.009 2.334 −1.80298 12.15934 PPHI Nigeria 2239 2.926 1.062 1 5
Ghana 1194 0.001 2.046 −1.36252 14.01731 Ghana 1087 2.959 1.211 1 5
Zimbabwe 1196 0.003 2.126 −1.91515 11.36737 Zimbabwe 1097 3.027 1.067 1 5
Kenya 1096 0.000 2.195 −1.81166 12.48099 Kenya 1073 3.185 1.119 1 5
Uganda 2421 −0.001 2.236 −1.84514 10.24266 Uganda 2290 3.559 1.044 1 5
Arnim Langer and Satoru Mikami 251
Notes
1. Langer (2005) provides a theoretical foundation for this empirical finding
by emphasising that the simultaneous presence of severe political horizontal
inequalities and socioeconomic horizontal inequalities forms an extremely explo-
sive sociopolitical situation, because in these situations the excluded political
elites not only have strong incentives to mobilise their supporters for violent
conflict along ethnic lines but also are likely to gain support among their ethnic
constituencies quite easily.
2. Please note that the word objective was put in inverted commas to indicate that
any indicator can only be an approximation of the objective reality. Furthermore,
the selection of entities, variables or indicators used to quantify objective horizon-
tal inequalities at a particular point in time is clearly to some extent an arbitrary
choice by the researchers involved.
3. For more information on the Minorities at Risk project, visit http://www.cidcm.
umd.edu/mar/ (last accessed 1 May 2013).
4. These perceptions surveys were conducted as part of the JICA-RI project ‘Preven-
tion of violent conflicts in Africa’.
5. For the Afrobarometer surveys, visit http://www.afrobarometer.org/ (last accessed
1 May 2013).
6. Appendix 1 provides a detailed description of the operationalisation of the
variables used in our analysis.
7. Please note that scores were ‘normalised’ vis-à-vis the capital of the country,
expect for Nigeria, where Lagos was used as a base.
8. Stewart and Langer (2008) propose a formal framework for understanding the
persistence of group inequalities based on the following factors: unequal rates of
accumulation; dependence of returns of one type of capital on the availability of
other types of capital; and asymmetries in social capital.
9. We used OLS for the model that examined determinants of ‘Assets’ in Zimbabwe
because the mixed-level model failed to converge.
10
Findings and Implications: The Role
of Development Cooperation
Frances Stewart, Thandika Mkandawire and Mari Katayanagi
10.1 Introduction
252
Frances Stewart, Thandika Mkandawire and Mari Katayanagi 253
The authors used the available data on socioeconomic HIs but a daunting
problem was the paucity of appropriate statistics in conflict-prone countries
in Africa. The authors combined statistical evidence with a description of
the historical evolution of boundaries and inequalities between groups.
In conflict processes, people’s perceptions and views play critical roles.
Where available, therefore, case studies incorporate the results of JICA
perceptions surveys, which were designed to examine how people them-
selves see their own identities, those of others and their views of the
prevailing HIs and their country’s political system, as well as their attitudes
towards violence. Chapter 9 provides a cross-country analysis of the survey
results using polimetric methods.
In the following section, we give a brief summary of each chapter. Next, we
formulate general findings from the study as a whole, presenting some tables
that elucidate the relationships between different dimensions of HIs and
classify the political institutions of each country based on both their for-
mal and informal features. We then touch on contemporary processes in
conflicts in Africa, involving multiple actors both inside and outside the
continent. The final section concludes with a set of policy recommenda-
tions for effective conflict prevention, directed to development agencies and
other stakeholders in Africa.
The backbone of this book consists of case studies. Most of the country cases
consist of paired comparisons: Rwanda and Burundi (Chapter 3), Ghana
and Côte d’Ivoire (Chapter 4), South Africa and Zimbabwe (Chapter 5) and
Tanzania and Uganda (Chapter 6). Each of the countries in the paired com-
parisons exhibits differences in the way they manage HIs through political
institutions and policies, with contrasting political outcomes in terms of
stability and instability. In addition to these four pairs, there are two single-
country cases of large ethnically heterogeneous countries, Kenya (Chapter 7)
and Nigeria (Chapter 8), which exemplify unique approaches to managing
diversity through their political systems, as well as in changes over time.
These case studies are framed by two chapters that contain quantitative anal-
ysis on institutions (Chapter 2) and perceptions (Chapter 9). The integration
of quantitative analysis and qualitative case studies is a major feature of this
book. The following are summaries of all these chapters.
Despite the strong potential for peace in Africa following the end of the
Cold War which had instigated and sustained a number of conflicts, the
continent faced a spate of violent conflicts in the 1990s against the back-
drop of post-Cold War globalisation. Many of these conflicts are identity
based. Previous studies, both case studies and econometric analysis, have
shown a strong relationship between HIs and violent conflict, and, con-
sequently, this study gives weight to HIs. We focus in particular on two
issues that have previously been dealt with insufficiently: the relationship
between objective and subjective HI; and the interaction between political
institutions and HIs.
Based on the quantitative analysis and qualitative case studies just sum-
marised, this section sets out our general findings. Tracing the flow of logic
in the case studies, we start by presenting a series of findings about eth-
nic configurations and HIs in these countries. Next, findings about popular
perceptions on political and socioeconomic HIs and their implications are
formulated. Then the advantage of different types of political institutions in
addressing overall HIs is discussed in the light of the lessons learnt from case
studies.
there was also political exclusion. This points to the importance of political
inclusion – and consequently of the political system – as well as the ways
in which people perceive their positions in such a system, for stability. It is
important to note that when it is difficult to introduce or sustain redistribu-
tive policies, possibly due to economic recession, and political exclusion
continues, the conflict risk may sharply increase (for example, Côte d’Ivoire,
Zimbabwe).
a Consistent if same group is in higher (or lower) position in both socioeconomic and political
HIs.
b Balanced if one group is dominant in politics, another in socioeconomics, or conversely.
the population who are Muslims in Ghana and Côte d’Ivoire, or the different
degrees of urbanisation in South Africa and Zimbabwe.
Our case studies confirm the importance of political systems that lead to
power dispersal in supporting stability. Such power-dispersal is of two types:
first, including all major groups in central government, or power-sharing at
the centre, through cooperative or coalition government (PS); and secondly,
decentralising power to regions and/or districts (DEC). Both appear to be
important. PS may be formal or informal. In Ghana, informal convention
ensures that in general all major groups participate in power. In Nigeria,
formal arrangements were introduced, as a response to the Biafran war, via
the Federal Character Principle, as well as informal arrangements – with the
convention that the presidency should alternate between North and South.
Formal PS is a feature of the Tanzanian, Burundian and South African con-
stitutions. In Kenya and Zimbabwe, coalition government was introduced
in the late 2000s in response to conflict, but has not been enshrined in
the countries’ constitutions. Rwanda, Uganda and Côte d’Ivoire currently
lack PS.
PS may reduce the likelihood of political conflict in the short run, but
if large socioeconomic HIs remain, there is a potential for group grievance
which could give rise to violent conflict at some point. That is especially the
case when PS is ad hoc, and comes to an end. Consequently, redistributive
socioeconomic policies are an important aspect of conflict prevention. Con-
versely, we should bear in mind that prolongation of ad hoc PS, especially
involving a coalition government of all major parties, may negatively affect
democratisation (as may be the case in Burundi and Zimbabwe), if such a
formal union is not supplemented by mutual trust and informal PS. It is pos-
sible, however, to combine democratic processes with PS arrangements, as in
Nigeria, Tanzania and South Africa. On the other hand, concentrating power
at the centre, based on majoritarian systems without PS, may be supported
in the belief that this will contribute to national unity (as has been argued,
for example, in Uganda and Rwanda), but this can then lead to grievances
among groups excluded from the central government, which may then have
negative effects on national unity. In such situations, close attention should
be paid to the nature of grievances which can simmer below the surface, as
well as to the democratic quality of regimes. Problems of this kind appear
in a number of our cases, including Rwanda, Côte d’Ivoire, Zimbabwe and
Uganda.
DEC, an important aspect of power-sharing in a broad sense, can be
achieved through a federal system, as in Nigeria, or by decentralisation
in a unitary state, as in Ghana and Uganda. In each case, the reality of
decentralisation – whether the arrangements involve genuine devolution of
decision-making and finance, or the centre has retained power – should be
investigated. For example, it is argued that in Uganda the centre has retained
power, despite the fact that there has been a large amount (30 per cent of
266 Findings and Implications
% of locally raised
revenue in local
government revenueb
% of locally raised
revenue in local
government revenueb
Notes: + indicates that the characteristic is present and X indicates that characteristic is not present; number in parenthesis, e and n.a. indicate the year,
the estimate and the unavailability of the data respectively.
a. Difficult to assess.
b. Based on Commonwealth Local Government Forum Country Profiles, http://www.clgf.org.uk/country-profiles/ (last accessed 1 May 2013). For Zimbabwe, the
estimation is obtained from Dennys et al. (2000: 120, Table D9). Comparison of the sizes of local government expenditure is also of significance, but the
cross-country data is difficult to obtain.
Frances Stewart, Thandika Mkandawire and Mari Katayanagi 269
Centralisation Decentralisation
emphasised that leadership styles matter for conflict prevention in the short
term. If such styles take root as long-term practices, they will contribute
greatly to bringing sustained stability to their countries. The different
personalities of Nelson Mandela and Robert Mugabe, for example, were
responsible for the differences in the PS and majoritarian paths taken by
South Africa and Zimbabwe.
10.4 Conclusion
In closing, the major messages of this book may be put as follows. First
of all, we emphasise the necessity of mainstreaming the HI perspective in
development policy-making in Africa. As confirmed in the whole of our
study, serious HIs, political, socioeconomic and cultural, do exist in African
countries, and the HI perspective provides one (although not the only) sig-
nificant explanation of the cause of violent conflict between groups. The
recent growth of African economies is improving the absolute situation
for many people but we still need to monitor the quality of develop-
ment closely, paying due attention to shifting patterns and intensities of
multidimensional HIs. Secondly, ‘subjectivity’ matters, as demonstrated by
the quantitative survey analysis. People take actions not because of objec-
tive data on structural inequalities but because of their subjective, sometimes
emotional perceptions about such inequalities, the nature of which must be
studied regularly. Thirdly, the political dimension of HIs is of prime impor-
tance, especially given the post-Cold War fluidity of politics and political
institutions in Africa. It is interesting to note that our research suggests that
people’s perceptions of political HIs may override those of socioeconomic
HIs. Fourthly, the case studies of paired countries as well as the cross-country
comparison of political institutions have confirmed the general advantage of
power-dispersing institutions, namely, a combination of sustainable PS and
DEC institutions, in conflict resolution and prevention.
Fifthly, the significance of informal institutions also emerged in the
case studies. Inclusive leadership and informal power-sharing have played
important roles in preventing violent conflicts in all the stable countries
investigated. The case studies have also shown that, depending on historical
contexts, informal values and practices strengthen, neutralise or sometimes
betray the intentions of formal reforms. Sixthly, all case studies indicate that
Frances Stewart, Thandika Mkandawire and Mari Katayanagi 273
[1] HIs are major root causes of violent conflict. The perspective of
multidimensional HIs should be mainstreamed in development policy-
making in Africa.
[2] Objective and subjective HIs are not always in accord. Development
practitioners and government officials should take care to avoid pro-
moting misperceptions concerning HIs and their own fairness and
impartiality.
[3] Inclusive institutions, both formal and informal, are the foundation
of peace. Outside actors should respect local initiatives for sustainable
power-sharing and effective decentralisation.
1. In the face of escalation of violence, warring parties are encouraged
to form power-sharing arrangements with the support of regional
organisations and thereby alleviate inter-group animosities.
2. Development agencies should continue to assist efforts towards
decentralisation so as to enhance power-sharing at the local level,
while ensuring that there are redistributive measures to support local
units across regions and localities.
3. Leadership quality is of prime importance. Opportunities for mutual
learning should be provided for African young leaders to embed in
them a culture of conflict prevention and inclusive nation-building.
References
Acemoglu, D., Johnson, S., and Robinson, J.A. 2001. The colonial origins of
comparative development: An empirical investigation. The American Economic
Review 91(5): 1369–1401.
Adebajo, A. 2008. Hegemony on a shoestring: Nigeria’s post-cold war foreign policy.
In Adebayo, A., and Mustapha, A.R., eds, Gulliver’s Troubles: Nigeria’s Foreign Policy
after the Cold War. Scottsville: University of KwaZulu-Natal Press.
Agyeman, D.K. 1998. Ethnic conflicts and politics in Ghana. In Nnoli, O., ed., Ethnic
Conflicts in Africa. Dakar: CODESRIA.
Akindès, F. 2004. Racines des crises socio-politiques en Côte d’Ivoire et sens de
l’histoire. Identity, Security and the Negotiation of National Belonging in West Africa:
Reflections on the Crisis in Côte d’Ivoire. CODESRIA and Nordiska Afrika Institut
Conference, Dakar, May 15–16.
Alesina, A., Devleeschauwer, A., Easterly, W., Kurlat, S., and Wacziarg, R. 2003.
Fractionalization. Journal of Economic Growth 8(2): 155–194.
Alexander, J., McGregor, J.A., and Ranger, T. 2000a. Ethnicity and the politics of con-
flict: The case of Matabeleland. In Nafziger, W.E., and Stewart, F., eds, War, Hunger,
and Displacement: The Origins of Humanitarian Emergencies, Vol. 1: Analysis. Oxford:
Oxford University Press.
Alexander, J., McGregor, J.A., and Ranger, T. 2000b. Violence and Memory: One Hundred
Years in the ‘Dark Forests’ of Matabeleland. Oxford: James Currey.
Alston, P. 2009. Press statement by Prof. Philip Alston. UN Special Rapporteur on
Extrajudicial, Arbitrary or Summary Executions, Mission to Kenya. New York: United
Nations.
Alwy, A., and Schech, S. 2004. Ethnic inequalities in education in Kenya. International
Education Journal 5(2): 266–274.
Amnesty International. 1997. Rwanda: Ending the Silence, No. AFR 47/032/1997, http://
www.amnesty.org/en/library/info/AFR47/032/1997 (last accessed 1 May 2013).
Anderson, B. 1983. Imagined Communities: Reflections on the Origin and Spread of
Nationalism. London: Verso.
Ansoms, A. 2008. Striving for growth, bypassing the poor? A critical review of
Rwanda’s rural sector policies. Journal of Modern African Studies 46(1): 1–32.
Ansoms, A. 2009. Re-engineering rural society: The visions and ambitions of the
Rwandan elite. African Affairs 108(431): 289–309.
Anyinam, C. 1993. Spatial implications of structural adjustment programmes in
Ghana. Tijdschrift voor Economische and Sociale Geografie 85(5): 446–460.
Apter, D.E. 1961. The Political Kingdom in Uganda: A Study in Bureaucratic Nationalism.
Princeton, NJ: Princeton University Press.
Azam, J.-P. 2004. Poverty and growth in the WAEMU after the 1994 devaluation.
Journal of African Economies 13(4): 536–562.
Bach, D.C. 2004. Inching towards a country without a state: Prebendalism, violence
and state betrayal in Nigeria. In Clapham, C., Herbst, J., and Mills, G., eds, Big
African States. Johannesburg: Witwatersrand University Press.
Bakari, M.A. 2001. Democratization Process in Zanzibar: A Retarded Transition, Hamburg
African Studies 11. Hamburg African Studies 11. Hamburg: Hamburg University Press.
275
276 References
Bakary, T. 1984. Elite transformation and political succession. In Zartman, W.I., and
Delgado, C., eds, The Political Economy of Ivory Coast. New York: Praeger Publishers.
Banégas, R., and Losch, B. 2002. La Côte d’Ivoire au bord de l’implosion. Politique
Africaine 87: 139–161.
Bangura, Y. 2006. Introduction: Ethnic inequalities and public sector governance.
In Bangura, Y., ed., Ethnic Inequalities and Public Sector Governance. Basingstoke:
Palgrave Macmillan.
Baskin, J. 1991. Striking Back: A History of COSATU. Johannesburg: Ravan Press.
Bates, R. 2004. Ethnicity. In Clark, D., ed., The Elgar Companion to Development Studies.
Cheltenham: Edward Elgar Publishing.
Beach, D. 1994. The Shona and Their Neighbours. Oxford: Blackwell.
Behrend, H. 1999. Alice Lakwena and the Holy Spirits: War in Northern Uganda, 1986–
1997. Oxford: James Currey.
Bentley, K.A., and Southall, R. 2005. An African Peace Process: Mandela, South Africa,
and Burundi. Cape Town: HSRC Press.
Beswick, D. 2010. Managing dissent in a post-genocide environment: The challenge
of political space in Rwanda. Development and Change 41(2): 225–251.
Blum, S. 2006. East Africa: Cycles of violence, and the paradox of the peace. ZEF
Discussion Paper 107. Bonn: Center for Development Research, University of Bonn.
Boahen, A.A. 1992. The Ghanaian Sphinx: Reflections on the Contemporary History of
Ghana, 1972–1987. Accra: Sankofa Educational Publishers.
Bond, P., and Manyanya, M. 2002. Zimbabwe’s Plunge: Exhausted Nationalism,
Neoliberalism and the Search for Social Justice. Pietermaritzburg: University of Natal
Press.
Booysen, S., and Toulou, L. 2009. Zimbabwe. In Denis, K., and Booysen, S., eds,
Compendium of Elections in Southern Africa, 1989–2009. Johannesburg: EISA.
Bourne, R. 2011. Catastrophe: What Went Wrong in Zimbabwe? London: Zed Books.
Bresson, Y. 1980. Politique économique et revenus en Côte d’Ivoire. World Employment
Programme Research Working Papers, WEP 2-23/WP 95. Geneva: ILO (International
Labour Office).
Brown, D. 1982. Who are the tribalists? Social pluralism and political ideology in
Ghana. African Affairs 81(322): 37–69.
Brown, G.K. 2005. Balancing the risks of corrective surgery: The political economy of
horizontal inequalities and the end of the new economic policy in Malaysia. CRISE
Working Paper No. 20. Oxford: CRISE (Centre for Research on Inequality, Human
Security and Ethnicity), University of Oxford.
Brown, G.K. 2008. Horizontal inequalities and separatism in Southeast Asia: A compar-
ative perspective. In Stewart, F., ed., Horizontal Inequalities and Conflict: Understanding
Group Violence in Multiethnic Societies. Basingstoke: Palgrave Macmillan.
Brown, G.K., and Langer, A. 2010. Horizontal inequality and conflict: A critical review
and research agenda. Conflict, Security & Development 10(1): 27–55.
Bruce, J.W. 2009. International standards, improvisation and the role of interna-
tional humanitarian organizations in the return of land in post-conflict Rwanda.
In Pantuliano, S., ed., Uncharted Territory: Land, Conflict and Humanitarian Action.
Warwickshire: Practical Action Publishing.
Brukum, N.J.K. 1995. Ethnic conflicts in the northern region of Ghana: A study of the
conflict in the Gonja district, 1980–1994. In Oquaye, M., ed., Democracy and Conflict
Resolution in Ghana. Accra: Gold Type Press.
Bundy, C. 1979. The Rise and Fall of the South African Peasantry. Berkeley, CA: University
of California Press.
References 277
Burgess, T. 2009. Race, Revolution, and the Struggle for Human Rights in Zanzibar. Athens,
OH: Ohio University Press.
Burnet, J.E. 2008. Gender balance and the meanings of women in governance in post-
genocide Rwanda. African Affairs 107(428): 361–386.
CCJP/LRF (Catholic Commission for Justice and Peace in Zimbabwe and the Legal
Resources Foundation). 2008. Gukurahundi in Zimbabwe: A Report on the Disturbances
in Matabeleland and the Midlands, 1980/1988. New York: Columbia University Press.
Cederman, L.A., Gledditsch, K.S., and Weidmann, N.B. 2010. Horizontal inequal-
ities and ethno-nationalist civil war: A global comparison. Paper prepared for
presentation at Yale University, April.
Cederman, L.A., Gledditsch, K.S., and Weidmann, N.B. 2011. Horizontal inequali-
ties and ethno-nationalist civil war: A global comparison. American Political Science
Review 105(3): 478–495.
CENI (Commission Electorale Nationale Indépendante). 2010a. Elections législatives
2010: Liste nominative des élus. Bujumbura: CENI.
CENI (Commission Electorale Nationale Independante). 2010b. Tableau d’affectation
des administrateurs communaux aux fins des équilibres ethnique et de genre. Bujumbura:
CENI.
Chachage, S.L.S. 2000. Environment, Aid and Politics in Zanzibar. Dar es Salaam: Dar es
Salaam University Press.
Chitere, P., Chweya, L., Masya, J., Tostensen, A., and Waiganjo, K. 2006. Kenya
Constitutional Documents: A Comparative Analysis. Berge: Chr. Michelsen Institute.
Choudhry, S., ed., 2008. Constitutional Design for Divided Societies: Integration or
Accommodation? Oxford: Oxford University Press.
Chrétien, J.P. 2000. L’Afrique des Grands Lacs: Deux mille ans d’histoire. Paris: Aubier.
Chrétien, J.P., and Dupaquier, J.F. 2007. Burundi 1972 au bord des génocides. Paris:
Karthala.
Chung, F. 1988. Education: Revolution or reform. In Stoneman, C., ed., Zimbabwe’s
Prospects: Issues of Race, Class, State, and Capital in Southern Africa. London:
Macmillan.
CIA (Central Intelligence Agency). 2011. The CIA World Factbook 2012. New York:
Skyhorse Publishing.
CIPEV (Commission of Inquiry into the Post Election Violence). 2008. Report of the
Commission of Inquiry into the Post-Election Violence in Kenya (Waki Report). Nairobi:
CIPEV.
Cohen, M.A. 1974. Urban Policy and Political Conflict in Africa: The Study of the Ivory
Coast. Chicago, IL: The University of Chicago Press.
Coleman, J. 1958. Nigeria: Background to Nationalism. Berkeley, CA: University of
California Press.
Collier, P. 2009. Wars, Guns, and Votes: Democracy in Dangerous Places. New York:
HarperCollins.
Collier, P., Elliott, V.L., Hegre, H., Hoeffler, A., Reynal-Querol, M., and Sambanis,
N. 2003. Breaking the Conflict Trap: Civil War and Development Policy. Washington,
DC: World Bank.
Collier, P., and Hoeffler, A. 2004. The challenge of reducing the global incidence of
civil war. Copenhagen Consensus Challenge Paper, http://www.copenhagenconsensus.
com/sites/default/files/CP%2B-%2BConflicts%2BFINISHED.pdf (last accessed 1 May
2013).
Collier, P., Hoeffler, A., and Söderbom, M. 2008. Post-conflict risks. Journal of Peace
Research 45(4): 461–478.
278 References
Electoral Task Team. 2003. Report of the Electoral Task Team. Pretoria: Govern-
ment Printer, http://www.elections.org.za/content/DynamicDocs.aspx?id= 329&
BreadCrumbId= 329&LeftMenuId= 251&name= home (last accessed 1 May 2013).
Ellis, S., and Sechaba, T. 1992. Comrades against Apartheid: The ANC and the South
African Communist Party in Exile. London: James Currey.
Englebert, P. 2002. Born-again Buganda or the limits of traditional resurgence in
Africa. Journal of Modern African Studies 40(3): 345–368.
Englebert, P., and Hummel, R. 2005. Institutionalising ethnic representation: How
effective is the Federal Character Commission in Nigeria? African Affairs 104(416):
399–427.
Evans-Pritchard, E.E. 1940. The Nuer: A Description of the Modes of Livelihood and
Political Institutions of a Nilotic People. Oxford: Clarendon Press.
Fearon, J.D., and Laitin, D.D. 1996. Explaining interethnic co-operation. American
Political Science Review 90(4): 715–735.
Figueroa, A. 2006. Education, labour markets and inequality in Peru. Paper presented
at the CRISE Latin America Workshop, Santa Cruz, Bolivia, September 18–20.
Fjeldstad, O.H., and Therkildsen, O. 2008. Mass taxation and state-society relations
in East Africa. In Bräutigam, D., Fjeldstad, O.H., and Moore, M., eds, Taxation and
State-Building in Developing Countries. Cambridge: Cambridge University Press.
Frempong, K.D. 2001. Ghana’s election 2000: The ethnic under current. In Ayee J.,
ed., Deepening Democracy in Ghana: Politics of the 2000 Elections. Accra: Freedom
Publications.
Freund, B. 2007. South Africa: The end of Apartheid and the emergence of the ‘BEE
elite’. Review of African Political Economy 34(114): 661–678.
Friedman, S. 2009. An accidental advance? South Africa’s 2009 elections. Journal of
Democracy 20(4): 108–122.
Fukuda-Parr, S., Ashwill, M., Messineo, C., and Chiappa, E. 2008. Conflict-
development nexus: Survey of armed conflicts in sub-Saharan Africa, 1998–2006.
Journal of Peacebuilding and Development 4(1): 6–19.
Furnivall, J.S. 1948. Colonial Policy and Practice: A Comparative Study of Burma and
Netherlands India. Cambridge: Cambridge University Press.
Gaffney, P.D. 2000. Burundi: The long sombre shadow of ethnic instability.
In Nafziger, E.W., Stewart, F., and Väyrynen, R., eds, War, Hunger, and Displacement:
The Origin of Humanitarian Emergencies, Vol. 2. Oxford: Oxford University Press.
Gahama, J. 2001. Le Burundi sous administration belge: La période du mandat, 1919–1939.
Deuxième édition. Paris: Karthala.
Geschiere, P. 2009. The Perils of Belonging: Autochthony, Citizenship, and Exclusion in
Africa and Europe. Chicago, IL: University of Chicago Press.
Glazer, N., Moynihan, D.P., and Schelling, C.S. 1976. Ethnicity: Theory and Experience.
Cambridge, MA: Harvard University Press.
Glickman, H. 1998. Ethnicity, elections and constitutional democracy in Africa.
In Sisk, T.D., and Reynolds, A., eds, Elections and Conflict Management in Africa.
Washington, DC: United States Institute of Peace Press.
Global Witness. 2011. Congo’s Minerals Trade in the Balance: Opportunities and Obstacles
to Demilitarisation. London: Global Witness.
Gonin, G. 1998. Ethnicity, politics and national awareness in Côte d’Ivoire. In Nnoli,
O., ed., Ethnic Conflicts in Africa. Dakar: CODESRIA.
Government of Ghana. 2003. Ghana Poverty Reduction Strategy (GPRS), 2003–2005:
An Agenda for Growth and Prosperity. Accra: Government of Ghana.
280 References
Green, E. 2006. Ethnicity and the politics of land tenure reform in Central Uganda.
Commonwealth & Comparative Politics 44(3): 370–388.
Green, E. 2008. Decentralisation and conflict in Uganda. Conflict, Security & Develop-
ment 8(4): 427–450.
Gurr, T.R. 1970. Why Men Rebel. Princeton, NJ: Princeton University Press.
Gurr, T.R. 1993. Minorities at Risk: A Global View of Ethnopolitical Conflicts. Washington,
DC: United States Institute of Peace Press.
Gurr, T.R. 1994. Peoples against states: Ethnopolitical conflict and the changing world
system. International Studies Quarterly 38(3): 347–377.
Gurr, T.R. 2002. Peoples versus States: Minorities at Risk in the New Century. Washington,
DC: United States Institute of Peace Press.
Gyimah-Boadi, E. 2003. Ghana: The political economy of ‘successful’ ethno-regional
conflict management. In Bastian, S., and Luckham, R., eds, Can Democracy be
Designed? The Politics of Institutional Choice in Conflict-torn Societies. London: Zed
Publications.
Gyimah-Boadi, E., and Asante, R. 2006. Ethnic structure, inequality and public sec-
tor governance in Ghana. In Bangura, Y., ed., Ethnic Inequalities and Public Sector
Governance. Basingstoke: Palgrave Macmillan.
Gyimah-Boadi, E., and Daddieh, C. 1999. Economic reform and political liberalization
in Ghana and Côte d’Ivoire: A preliminary assessment of implications for nation
building. In Mengisteab, K., and Daddieh, C., eds, State Building and Democratization
in Africa: Faith, Hope and Realities. Westport, CT: Preager Publishers.
Hagan, G.P. 1992. Nkrumah’s cultural policy. In Arhin, K., ed., The Life and Work of
Kwame Nkrumah. Trenton: Africa World Press.
Hartzell, C.A., and Hoddie, M. 2007. Crafting Peace: Power-sharing Institutions and the
Negotiated Settlement of Civil Wars. University Park, PA: Pennsylvania State University
Press.
Hassim, S., Kupe, T., and Worby, E., eds, 2008. Go Home or Die Here: Violence, Xenopho-
bia and the Reinvention of Difference in South Africa. Johannesburg: Wits University
Press.
Herbst, J. 1990. State Politics in Zimbabwe. Berkeley, CA: University of California
Press.
Herbst, J. 2000. States and Power in Africa: Comparative Lessons in Authority and Control.
Princeton, NJ: Princeton University Press.
Hinderink, J., and Tempelman, G.J. 1979. Development policy and development prac-
tice in Ivory Coast: A miracle or a mirage? Diskussiestukken van de Vakgroep sociale
geografie van ontwikkelingslanden 4. University of Utrecht.
Hintjens, H. 2008. Post-genocide identity politics in Rwanda. Ethnicities 8(1): 5–41.
Hirschman, A.O. 1967. Development Projects Observed. Washington, DC: Brookings
Institution.
Holborn, L. 2010. The Long Shadow of Apartheid: Race in South Africa since 1994.
Johannesburg: South African Institute of Race Relations.
Horowitz, D.L. 1985. Ethnic Groups in Conflict. Berkeley, CA: University of California
Press.
Horowitz, D.L. 1991. A Democratic South Africa? Constitutional Engineering in a Divided
Society. Berkeley, CA: University of California Press.
Horowitz, D.L. 2002. Constitutional design: Proposals versus processes. In Reynolds,
A., ed., The Architecture of Democracy: Constitutional Design, Conflict Management, and
Democracy. Oxford: Oxford University Press.
References 281
Huggins, C. 2009. Land in return, reintegration and recovery processes: Some lessons
from the Great Lakes region in Africa. In Pantuliano, S., ed., Uncharted Territory:
Land, Conflict and Humanitarian Action. London: Practical Action Publishing.
Human Rights Watch. 2008. Law and Reality: Progress in Judicial Reform in Rwanda.
New York: Human Rights Watch.
Human Rights Watch. 2010. Burundi: Closing Doors? The Narrowing of Democratic Space
in Burundi. New York: Human Rights Watch.
Hutchful, E. 1973. Military disengagement in Ghana. Unpublished Ph.D. Thesis.
University of Toronto.
ICG (International Crisis Group). 2003. Réfugiés et déplacés au Burundi: Désamorcer
la bombe foncière. Rapport Afrique 70. Bruxelles: ICG.
ICG (International Crisis Group). 2006. Burundi: Democracy and peace at risk. Africa
Report 120. Bruxelles: ICG.
ICG (International Crisis Group). 2010. Burundi: Garantir un processus électoral
crédible. Rapport Afrique 155. Bruxelles: ICG.
ICG (International Crisis Group). 2011. Burundi: Du boycott électoral à l’impasse
politique. Rapport Afrique 169. Bruxelles: ICG.
Iheduru, O.C. 2004. Black economic power and nation-building in post-Apartheid
South Africa. Journal of Modern African Studies 42(1): 1–30.
Iliffe, J. 2007. Africans: The History of a Continent. 2nd edn. Cambridge: Cambridge
University Press.
Ingelaere, B. 2009. Does the truth pass across the fire without burning? Locating
the short circuit in Rwanda’s Gacaca courts. Journal of Modern African Studies 47(4):
507–528.
Ingelaere, B. 2010. Peasants, power and ethnicity: A bottom-up perspective on
Rwanda’s political transition. African Affairs 109(435): 273–292.
ISTEEBU (Institut de Statistiques et d’Etudes Economiques du Burundi). 2009. Annuaire
statistique du Burundi 2007. Bujumbura: ISTEEBU.
Kanyenze, G. 2003. The performance of the Zimbabwean economy, 1980–2000. In
Darnolf, S., and Laakso, L., eds, Twenty Years of Independence in Zimbabwe: From
Liberation to Authoritarianism. Basingstoke: Palgrave Macmillan.
Kanyinga, K. 2006. Ethnic structure, inequality and public sector governance
in Kenya. In Bangura, Y., ed., Ethnic Inequalities and Public Sector Governance.
Basingstoke: Palgrave Macmillan.
Kanyinga, K. 2007. Governance institutions and inequality. In SID (Society for
International Development), Readings on Inequality in Kenya. Nairobi: SID.
Kasfir, N. 1998. ‘No-party democracy’ in Uganda. Journal of Democracy 13(4): 49–63.
Kibara, G. 2003. The state of constitutionalism in Kenya, 2003, http://www.
kituochakatiba.org/index2.php?option= com_docman&task= doc_view&gid= 21&
Itemid= 36 (last accessed 1 May 2013).
Kieffer, G.-A. 2000. Armée ivoirienne: le refus du déclassement. Politique Africaine 78:
26–44.
Killian, B. 2008. The state and identity politics in Zanzibar: Challenges to democratic
consolidation in Tanzania. African Identities 6(2): 99–125.
Kimenyi, M.S., and Ndung’u, N.S. 2005. Sporadic ethnic violence: Why has Kenya not
experienced a full-blown war? In Collier, P., and Sambanis, N., eds, Understanding
Civil War: Evidence and Analysis, Vol. 1: Africa. Washington, DC: World Bank.
Kimenyi, M.S., and Shughart, W.F. 1989. Political successions and the growth of
government. Public Choice 62(2): 173–179.
Klein, N. 2007. The Shock Doctrine: The Rise of Disaster Capitalism. New York: Picador.
282 References
Klugman, J. 2000. Kenya: Economic decline and ethnic politics. In Nafziger, E.W.,
Stewart, F., and Väyrynen, R., eds, War, Hunger, and Displacement: The Origins of
Humanitarian Emergencies, Vol. 2: Case Studies. New York: Oxford University Press.
Klugman, J., Neyapti, B., and Stewart, F. 1999. Conflict and Growth in Africa, Vol. 2:
Kenya, Tanzania and Uganda. Paris: OECD Development Centre.
KNBS (Kenya National Bureau of Statistics). 2010. Kenya 2009 Population and Housing
Census, http://www.knbs.or.ke/population.php (last accessed 1 May 2013).
KNBS (Kenya National Bureau of Statistics) and ICF Macro. 2010. Kenya Demographic
and Health Survey (DHS), 2008–09. Calverton, MD: KNBS and ICF Macro, http://
www.measuredhs.com/pubs/pdf/FR229/FR229.pdf (last accessed 1 May 2013).
Kniss, M. 2010. Walking Kenya back from the brink: A micro-level study of horizontal
inequity and civil conflict prevention. Center for International and Security Studies at
Maryland (CISSM) Working Paper June 2010. College Park, MD: CISSM, University of
Maryland.
Kriger, N. 2003. Guerrilla Veterans in Post-War Zimbabwe: Symbolic and Violent Politics,
1980–1987. Cambridge: Cambridge University Press.
Kymlicka, W. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford:
Clarendon Press.
Laakso, L. 2003. Regional voting and cabinet formation. In Darnolf, S., and Laakso,
L., eds, Twenty Years of Independence in Zimbabwe: From Liberation to Authoritarianism.
Basingstoke: Palgrave Macmillan.
Laakso, M., and Taagepera, R. 1979. Effective number of parties: A measure with
application to West Europe. Comparative Political Studies 12(1): 3–27.
Laely, T. 1992. Le destin du bushingantahe transformations d’une structure locale
d’autorité au Burundi. Genève-Afrique 30(2): 75–98.
Langer, A. 2005. Horizontal inequalities and violent group mobilisation in Côte
d’Ivoire. Oxford Development Studies 33(1): 25–45.
Langer, A. 2008. Horizontal inequalities and violent conflict: A comparative study
of Ghana and Côte d’Ivoire. In Stewart, F., ed., Horizontal Inequalities and Con-
flict: Understanding Group Violence in Multiethnic Societies. Basingstoke: Palgrave
Macmillan.
Langer, A. 2009. Living with diversity: The peaceful management of horizontal
inequalities in Ghana. Journal of International Development 21(4): 534–546.
Langer, A. 2010a. Côte d’Ivoire’s elusive quest for peace. Bath Papers in International
Development, WP11. University of Bath.
Langer, A. 2010b. The situational importance of identity in Ghana. Ethnopolitics 9(1):
9–29.
Langer, A., and Brown, G.K. 2008. Cultural status inequalities: An important dimen-
sion of group mobilization. In Stewart, F., ed., Horizontal Inequalities and Con-
flict: Understanding Group Violence in Multiethnic Societies. Basingstoke: Palgrave
Macmillan.
Langer, A., Stewart, F., and Venugopal, R., eds, 2011. Horizontal Inequalities and Post-
Conflict Development. Basingstoke: Palgrave Macmillan.
Langer, A., and Ukiwo, U. 2008. Ethnicity, religion and the state in Ghana and
Nigeria: Perceptions from the Streets. In Stewart, F., ed, Horizontal Inequalities and
Conflict: Understanding Group Violence in Multiethnic Societies. Basingstoke: Palgrave
Macmillan.
Le Pape, M. 2002. Chronologie politique de la Côte d’Ivoire, du coup d’état aux élec-
tions. In Le Pape, M., and Vidal, C., eds, Côte d’Ivoire: l’Année Terrible, 1999–2000.
Paris: Karthala.
References 283
Leibbrandt, M., Woolard, I., Finn, A., and Argent, J. 2010. Trends in South African
income distribution and poverty since the fall of Apartheid. OECD Social, Employ-
ment and Migration Working Papers 101. Paris: OECD.
Lemarchand, R. 1970. Rwanda and Burundi. London: Pall Mall Press.
Lemarchand, R. 1994. Burundi: Ethnocide as Discourse and Practice. Cambridge:
Cambridge University Press.
Lemarchand, R. 2006. Consociationalism and Power Sharing in Africa: Rwanda,
Burundi, and the Democratic Republic of the Congo. African Affairs 106(422):
1–20.
Lemarchand, R. 2009. The Dynamics of Violence in Central Africa. Philadelphia, PA:
University of Pennsylvania Press.
Lentz, C., and Nugent, P. 2000. Ethnicity in Ghana: A comparative perspectives.
In Lentz C., and Nugent, P., eds, Ethnicity in Ghana: The Limits of Intervention.
Basingstoke: Palgrave Macmillan.
Levitsky, S., and Way, L.A. 2010. Competitive Authoritarianism: Hybrid Regimes after the
Cold War. Cambridge: Cambridge University Press.
Lewis, W.A. 1965. Politics in West Africa. Toronto: Oxford University Press.
Lewis, W.A. 1985. Racial Conflict and Economic Development. Cambridge, MA: Harvard
University Press.
Lijphart, A. 1971. Cultural diversity and theories of political integration. Canadian
Journal of Political Science 4(1): 1–14.
Lijphart, A. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven,
CT: Yale University Press.
Lijphart, A. 1985. Power-Sharing in South Africa, Policy Papers in International Affairs 24.
Berkeley, CA: Institute of International Studies, University of California.
Lijphart, A. 1991. Constitutional choices for new democracies. Journal of Democracy
2(1): 72–84.
Lijphart, A. 1994. Electoral Systems and Party Systems: A Study of Twenty-Seven Democra-
cies, 1945–1990. Oxford: Oxford University Press.
Lijphart, A. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-six
Countries. New Haven, CT: Yale University Press.
Lijphart, A. 2002. The wave of power-sharing democracy. In Reynolds, A., ed., The
Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy.
Oxford: Oxford University Press.
Lijphart, A. 2008. Thinking about Democracy: Power Sharing and Majority Rule in Theory
and Practice. London: Routledge.
Lindemann, S. 2010. Exclusionary elite bargains and civil war onset: The case of
Uganda. CSRC Working Paper 76. London: CSRC (Crisis States Research Centre).
Lipton, M. 1986. Capitalism and Apartheid: South Africa, 1910–1986. Aldershot:
Wildwood House.
Luckham, R. 1971. The Nigerian Military: A Sociological Analysis of Authority and Revolt,
1960–1967. Cambridge: Cambridge University Press.
Lupogo, H. 2001. Tanzania: civil–military relations and political stability. African
Security Review 10(1): 75–86.
MacDonald, M. 2006. Why Race Matters in South Africa. Cambridge, MA: Harvard
University Press.
Makara, S., Rakner, L., and Savsand, L. 2009. Turn-round: The National Resistance
Movement and the reintroduction of multiparty system in Uganda. International
Political Science Review 30(2): 185–204.
284 References
Makina, D. 2010. Zimbabwe in Johannesburg. In Crush, J., and Tereva, D., eds,
Zimbabwe’s Exodus: Crisis, Migration, Survival. Kingston: Southern African Migration
Programme.
Mamdani, M. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late
Colonialism. Princeton, NJ: Princeton University Press.
Manby, B. 2009. Struggles for Citizenship in Africa. London: Zed Books.
Mancini, L. 2008. Horizontal inequality and communal violence: evidence from
Indonesian districts. In Stewart, F., ed., Horizontal Inequalities and Conflict: Under-
standing Group Violence in Multiethnic Societies. Basingstoke: Palgrave Macmillan.
Maquet, J. 1971. Power and Society in Africa. London: Weidenfeld & Nicolson.
Marko, J. 2005. Post-conflict reconstruction through state- and nation-building:
The case of Bosnia and Herzegovina. European Diversity and Autonomy Papers
(EDAP) 4/2005. Bolzano-Bozen: EDAP, http://www.eurac.edu/en/research/institutes/
imr/activities/Bookseries/edap/Documents/2005_edap04.pdf (last accessed 1 May
2013).
Marx, A.W. 1998. Making Race and Nation: A Comparison of the United States, South
Africa, and Brazil. Cambridge: Cambridge University Press.
Marysse, S., and André, C. 2001. Guerre et pillage économique en République
Démocratique du Congo. In Marysse, S., and Reyntjens, F., eds, L’Afrique des Grands
Lacs, Annuaire 2000–2001. Paris: L’Harmattan.
Marysse, S., Ansoms, A., and Cassimon, D. 2006. The aid ‘darling’ and ‘orphans’ of
the Great Lakes Region in Africa. IOB Discussion Paper 2006.10. Antwerp: Institute
of Development Policy and Management, University of Antwerp.
Marysse, S., and Reyntjens, F., dir., 1999. L’Afrique des Grands Lacs, Annuaire 1998–
1999. Paris: L’Harmattan.
Masunungure, E.V., ed., 2009. Defying the Winds of Change: Zimbabwe’s 2008 Elections.
Harare: Weaver Press.
Matlosa, K. 2009. The role of the Southern African Development Community in the
management of Zimbabwe’s post-election crisis. Journal of African Elections 8(2):
46–73.
Mazarire, G.C. 2009. Reflections on pre-colonial Zimbabwe, c.850–1880s. In
Raftopoulos, B., and Mlambo, A.S., eds, Becoming Zimbabwe: A History from the
Pre-Colonial Period to 2008. Harare: Weaver Press.
McDonald, D.A. 2008. World City Syndrome: Neoliberalism and Inequality in Cape Town.
London: Routledge.
McGarry, J., O’Leary, B., and Simeon, R. 2008. Integration or accommodation? The
enduring debate in conflict regulation. In Choudhry, S., ed., Constitutional Design
for Divided Societies: Integration or Accommodation? Oxford: Oxford University Press.
MERG (Macroeconomic Research Group). 1993. Making Democracy Work: A Framework
for Macroeconomic Policy in South Africa. Bellville: Centre for Development Studies,
University of the Western Cape.
Miguel, E. 2004. Tribe or nation? Nation building and public goods in Kenya versus
Tanzania. World Politics 56(3): 328–362.
Mill, J.S. 1861. Considerations on Representative Government. New York: Prometheus
(1991 edition).
Mine, Y. 2006. The political element in the works of W. Arthur Lewis: The 1954 Lewis
model and African development. Developing Economies 44(3): 329–355.
Mkandawire, T. 2001. Thinking about development states in Africa. Cambridge Journal
of Economics 25(3): 289–313.
References 285
Nyerere, J. 1973. Freedom and Development. Dar es Salaam: Oxford University Press.
Nyerere, J. 2000. Foreword. In Museveni, Y.K., What Is Africa’s Problem? Minneapolis,
MN: University of Minnesota Press.
OECD (Organisation for Economic Co-operation and Development). 2010. African
Economic Outlook. Paris: OECD.
O’Flynn, I., and Russell, D., eds, 2005. Power Sharing: New Challenges for Divided
Societies. London: Pluto Press.
Ogata, S. 2010. Development co-operation and human security. Conflict, Security &
Development 10(1): 181–188.
Okolo, A. 1999. The Nigerian census: Problem and prospects. The American Statistician
53(4): 321–325.
O’Leary, B. 2005. Debating consociational politics: Normative and explanatory
arguments. In Noel, S., ed., From Power Sharing to Democracy: Post-Conflict
Institutions in Ethnically Divided Societies. Montreal: McGill-Queen’s University
Press.
Østby, G. 2008. Horizontal inequalities, political environment and civil conflict: Evi-
dence from 55 developing countries. In Stewart, F., ed., Horizontal Inequalities and
Conflict: Understanding Group Violence in Multiethnic Societies. Basingstoke: Palgrave
Macmillan.
Othman, S. 1984. Classes, crises and coup: The demise of Shagari’s regime. African
Affairs 83(333): 441–461.
Otite, O. 1990. Ethnic Pluralism and Ethnicity in Nigeria. Ibadan: Shaneson.
Ottaway, M. 1999. Africa’s New Leaders: Democracy or State Reconstruction? Washington,
DC: Carnegie Endowment for International Peace.
Paredes, M. 2007. Fluid identities: Exploring ethnicity in Peru. CRISE Working Paper
No. 40. Oxford: CRISE (Centre for Research on Inequality, Human Security and
Ethnicity), University of Oxford.
Paris, R. 2004. At War’s End: Building Peace after Civil Conflict. Cambridge: Cambridge
University Press.
PARP (Policy Analysis and Research Project). 2007. Nigerian State Assemblies: A Study of
Legislative Activities, 1999–2005. Abuja: PARP, National Assembly of Nigeria.
PARP (Policy Analysis and Research Project). 2008. National Assembly Statistical
Information 2(1). Abuja: PARP, National Assembly of Nigeria.
PEFAR (Public Expenditure and Financial Accountability Review) Team. 2006. Local
government fiduciary assessment: Joint evaluation report. Tanzania Public Expendi-
ture Review Annual Consultation Meeting, Dar es salaam: PEFAR.
Petterson, D. 2002. Revolution in Zanzibar. Boulder, CO: Westview Press.
Platzky, L., and Walker, C. 1985. The Surplus People: Forced Removals in South Africa.
Johannesburg: Ravan Press.
Plaza, S., and Ratha, D., eds, 2011. Diaspora for Development in Africa. Washington, DC:
World Bank.
Posner, D.N. 2005. Institutions and Ethnic Politics in Africa. Cambridge: Cambridge
University Press.
Powell, G.B. Jr. 2000. Elections as Instruments of Democracy. New Haven, CT: Yale
University Press.
Prunier, G. 1995. The Rwanda Crisis: History of a Genocide, 1959–1994. London: Hurst &
Company.
Prunier, G. 1998. The Rwandan Patriotic Front. In Clapham, C., ed., African Guerrillas.
Oxford: James Currey.
288 References
Raftopoulos, B., and Sachikonye, L., eds, 2001. Striking Back: The Labour Movement and
the Post-Colonial State in Zimbabwe, 1980–2000. Harare: Weaver Press.
Regional Forecasts. 2007. Uganda Multi-Donor Group: Northern Uganda Public Expendi-
ture Review (NUPER), Final Report. Hillsborough: Regional Forecasts Limited.
Reilly, B. 2001. Democracy in Divided Societies: Electoral Engineering for Conflict Manage-
ment. Cambridge: Cambridge University Press.
Republic of Rwanda. 2003. The Constitution of the Republic of Rwanda. Kigali: Republic
of Rwanda.
Republic of Rwanda. 2007. Economic Development and Poverty Reduction Strategy, 2008–
2012. Kigali: Republic of Rwanda.
Republic of Rwanda. 2012a. Rwanda External Trade Statistics Report Fourth Quarter 2011.
Kigali: Republic of Rwanda.
Republic of Rwanda. 2012b. Summary of the Report Presented at the Closing of Gacaca
Court Activities. Kigali: Republic of Rwanda.
République de Côte d’Ivoire. 2001. Tome 1, Etat et structures de la population.
Recensement Général de la Population et de l’Habitation de 1998 (RGPH-98) IV. Abidjan:
Institut National de la Statistique.
République du Burundi. 2005. Loi portant promulgation de la constitution de la République
du Burundi. Bujumbura: République du Burundi.
République du Burundi (Ministère de la planification du développement et de la
reconstruction nationale). 2006. Enquête QUIBB 2006, Rapport final. Bujumbura:
République du Burundi.
République Rwandaise. 2002. Ministère des finances et de la planification
économique. Rapport final, Enquête intégrale sur les conditions de vie des ménages au
Rwanda, 2000–2001. Kigali: République Rwandaise.
Rettig, M. 2008. Gacaca: Truth, justice, and reconciliation in postconflict Rwanda?
African Studies Review 51(3): 25–50.
Reynolds, A. 1999. Electoral Systems and Democratization in Southern Africa. Oxford:
Oxford University Press.
Reynolds, A., ed., 2002. The Architecture of Democracy: Constitutional Design, Conflict
Management, and Democracy. Oxford: Oxford University Press.
Reyntjens, F. 1985. Pouvoir et droit au Rwanda: Droit public et évolution politique, 1916–
1973. Tervuren: Musée Royal de l’Afrique Centrale.
Reyntjens, F. 1994. L’Afrique des Grands Lacs en crise: Rwanda, Burundi, 1988–1994.
Paris: Karthala.
Reyntjens, F. 2003. Les nouveaux habits de l’empereur: Analyse juridico-politique de
la constituion rwandaise de 2003. In Marysse, S., and Reyntjens, F., eds, L’Afrique des
Grands Lacs, Annuaire 2002–2003. Paris: L’Harmattan.
Reyntjens, F. 2004. Rwanda, ten years on: From genocide to dictatorship. African
Affairs 103(411): 177–210.
Reyntjens, F. 2006. Briefing: Burundi: A peaceful transition after a decade of civil war.
African Affairs 105(418): 117–135.
Reyntjens, F. 2009. The Great African War: Congo and Regional Geopolitics, 1996–2006.
Cambridge: Cambridge University Press.
Reyntjens, F. 2011. Constructing the truth, dealing with dissent, domesticat-
ing the world: Governance in post-genocide Rwanda. African Affairs 110(438):
1–34.
Rhodes, R.A.W., Wanna, J., and Weller, P. 2009. Comparing Westminster. New York:
Oxford University Press.
References 289
Rothchild, D., and Roeder, G.P. 2005. Dilemmas of state-building in divided societies.
In Roeder, G.P., and Rothchild, D., eds, Sustainable Peace: Power and Democracy after
Civil Wars. Ithaca, NY: Cornell University Press.
Rutherford, B. 2003. Belonging to the farm(er): Farm workers, farmers, and the shift-
ing politics of citizenship. In Hammer, A., Raftopoulos, B., and Jensen, S., eds,
Zimbabwe’s Unfinished Business: Rethinking Land, State and Nation in the Context of
Crisis. Harare: Weaver Press.
Rwabahungu, M., and Nintunze, A. 2009. Rapport sur l’étude d’actualisation de la
base de données sur la situation des déplacés et des rapatriés encore dans les sites.
Bujumbura: Ministère de la solidarité nationale du rapatriement des réfugiés et de la
réintégration sociale.
Sachikonye, L. 2004. Zimbabwe: Constitutionalism, the electoral system and chal-
lenges for governance and stability. Journal of African Elections 3(1): 140–159.
Sadomba, Z.W. 2011. War Veterans in Zimbabwe’s Revolution: Challenging Neo-
colonialism and Settler and International Capital. London: James Currey.
SAIRR (South African Institute of Race Relations). 1996. South Africa Survey, 1995–96.
Johannesburg: SAIRR.
Samuels, K. 2006. Post-conflict peace-building and constitution-making. Chicago
Journal of International Law 6(2): 1–20.
Samuelson, B.L., and Freedman, S.W. 2010. Language policy, multilingual education,
and power in Rwanda. Language Policy 9(3): 191–215.
Sanders, E.R. 1969. The Hamitic hypothesis: Its origin and functions in time perspec-
tive. Journal of African History 10(4): 521–532.
Sannerholm, R. 2007. Legal, judicial and administrative reforms in post-conflict soci-
eties: Beyond the rule of law template. Journal of Conflict & Security Law 12(1):
65–94.
Sartori, G. 1976. Parties and Party Systems: A Framework for Analysis. Cambridge:
Cambridge University Press.
Sartori, G. 1997. Comparative Constitutional Engineering: An Inquiry into Structures,
Incentives and Outcomes. 2nd edn. Basingstoke: Palgrave Macmillan.
Scarnecchia, T. 2008. The Urban Roots of Democracy and Political Violence in
Zimbabwe: Harare and Highfield, 1940–1964. Rochester, NY: University of Rochester
Press.
Schoenbrun, D. 1993. We are what we eat: Ancient agriculture between the Great
Lakes. Journal of African History 34(1): 1–31.
Schoenbrun, D. 1998. A Green Place, A Good Place: Agrarian Change, Gender, and Social
Identity in the Great Lakes Region to the 15th Century. Portsmouth: Heinemann.
Schulz-Herzenberg, C. 2009. Trends in party support and voter behaviour, 1994–
2009. In Southall, R., and Daniel, J. eds, ZUNAMI! The 2009 South African Elections.
Johannesburg: Jacana.
Scoones, I., Marongwe, N., Mavedzenge, B., Murimbarimba, F., Mahenehene, J., and
Sukume, C. 2010. Zimbabwe’s Land Reform: Myths and Realities. London: James
Currey.
Seekings, J. 2000. The UDF: A History of the United Democratic Front in South Africa,
1983–1991. Cape Town: David Philip.
Seekings, J., and Nattrass, N. 2005. Class, Race, and Inequality in South Africa. New
Haven, CT: Yale University Press.
Sen, A.K. 2007. Identity and Violence: The Illusion of Destiny. New York: W.W. Norton.
Shepherd, A., Gyimah-Boadi, E., Gariba, S., Plagerson, S., and Musa, W.A. 2005.
Bridging the north south divide in Ghana, Background paper for the 2006 world
290 References
development report. In World Bank. World Development Report 2006: Equity and
Development. Washington, DC: World Bank.
Silva-Leander, S. 2012. Structural violence and conflict: Vertical and horizontal
inequality in post-genocide Rwanda. In Langer, A., Stewart, F., and Venugopal,
R., eds, Horizontal Inequalities and Post-conflict Development. Basingstoke: Palgrave
Macmillan.
Sing’Oei, A.K. 2010. The ICC as arbiter in Kenya’s post-electoral violence. Minnesota
Journal of International Law Online 19: 5–20.
Sithole, M. 1999. Zimbabwe: Struggles within the Struggle. 2nd edn. Harare: Rujeko
Publishers.
Sklar, R.L. 1963. Nigerian Political Parties: Power in an Emergent African Nation.
Princeton, NJ: Princeton University Press.
Slye, R.C. 1996. The Dayton peace agreement: Constitutionalism and ethnicity. Yale
Journal of International Law 21: 459–473.
Smock, D.R., and Smock, A.C. 1975. The Politics of Pluralism: A Comparative Study of
Lebanon and Ghana. New York: Elsevier.
Southall, R. 2004. Post-election South Africa: The continuing case for electoral reform.
Journal of African Elections 3(2): 154–165.
Sparks, A. 1995. Tomorrow is Another Country: The Inside Story of South Africa’s Negotiated
Revolution. London: Heinemann.
Spears, S.I. 2002. Africa: The limits of power-sharing. Journal of Democracy 13(3):
123–136.
Statistics South Africa. 2003. Census 2001: Census in Brief. Pretoria: Statistics South
Africa.
Stewart, F. 2000. The root causes of humanitarian emergencies. In Nafziger, W.E.,
Stewart, F., and Väyrynen, R., eds, War, Hunger and Displacement: The Origins of
Humanitarian Emergencies, Vol. 1: Analysis. Oxford: Oxford University Press.
Stewart, F. 2002. Horizontal inequalities: A neglected dimension of development. QEH
Working Paper Series 81. Oxford: Queen Elizabeth House, University of Oxford.
Stewart, F. 2008a. Horizontal inequalities and conflict: An introduction and some
hypotheses. In Stewart, F., ed., Horizontal Inequalities and Conflict: Understanding
Group Violence in Multiethnic Societies. Basingstoke: Palgrave Macmillan.
Stewart, F. 2008b. Note for Discussion: Kenya, Horizontal Inequalities and the Political Dis-
turbances of 2008. Oxford: CRISE (Centre for Research on Inequality, Human Security
and Ethnicity), University of Oxford.
Stewart, F. 2010a. Horizontal inequalities in Kenya and the political disturbances
of 2008: Some implications for aid policy. Conflict, Security & Development 10(1):
133–159.
Stewart, F. 2010b. Horizontal inequalities as a cause of conflict: A review of CRISE find-
ings. CRISE Overview 1. Oxford: CRISE (Centre for Research on Inequality, Human
Security and Ethnicity), Oxford University Press.
Stewart, F., ed., 2008. Horizontal Inequalities and Conflict: Understanding Group Violence
in Multiethnic Societies. Basingstoke: Palgrave Macmillan.
Stewart, F., and Brown, G.K. 2007 Motivations for conflict: Groups and individuals.
In Crocker C.A., Hampson, F.O., and Aall, P.R., eds, Leashing the Dogs of War: Conflict
Management in A Divided World. Washington, DC: USIP (United States Institute of
Peace).
Stewart, F., Brown, G., and Langer, A. 2008. Policies towards horizontal inequalities.
In Stewart, F., ed., Horizontal Inequalities and Conflict: Understanding Group Violence in
Multiethnic Societies. Basingstoke: Palgrave Macmillan.
References 291
Stewart, F., and Langer, A. 2008. Horizontal inequalities: Explaining persistence and
change. In Stewart, F., ed., Horizontal Inequalities and Conflict: Understanding Group
Violence in Multiethnic Societies. Basingstoke: Palgrave Macmillan.
Stoneman, C. 1988. Some concluding thoughts. In Stoneman, C., ed., Zimbabwe’s
Prospects: Issues of Race, Class, State and Capital in Southern Africa. Basingstoke:
Palgrave Macmillan.
Stoneman, C. 2000. Zimbabwe land policy and the land reform. In Bowyer-Bower,
T.A.S., and Stoneman, C., eds, Land Reform in Zimbabwe: Constraints and Prospects.
Aldershot: Ashgate.
Sullivan, D.P. 2005. The missing pillars: A look at the failure of peace in Burundi
through the lens of Arend Lijphart’s theory of consociational democracy. Journal of
Modern African Studies 43(1): 75–95.
Szereszewski, R. 1966. Regional aspects of the structure of the economy. In
Birmingham, W.B., Neustadt, I., and Omaboe, E.N., eds, A Structure of Contemporary
Ghana. London: George Allen and Unwin.
Takeuchi, S. 2007. Political liberalization or armed conflicts? Political changes in post-
Cold War Africa. The Developing Economies 45(2): 172–193.
Takeuchi, S. 2011. Gacaca and DDR: The disputable record of state-building in
Rwanda. JICA-RI Working Paper No. 32. Tokyo: JICA Research Institute.
Terreblanche, S. 2002. A History of Inequality in South Africa, 1652–2002.
Pietermaritzburg: University of Natal Press.
Therkildsen O., and Tidemand, P. 2007. Staff Management and Organizational Perfor-
mance in Tanzania and Uganda: Public Servant Perspectives. Copenhagen: DIIS (Danish
Institute for International Studies).
Thompson, L. 2001. A History of South Africa. 3rd edn. New Haven, CT: Yale University
Press.
Thomson, S. 2011. The darker side of transitional justice: The power dynamics behind
Rwanda’s Gacaca courts. Africa 81(3): 373–390.
Thorp, R., Caumartin, C., and Gray-Molina, G. 2006. Inequality, ethnicity, polit-
ical mobilisation and political violence in Latin America: The cases of Bolivia,
Guatemala and Peru. Bulletin of Latin American Research 25(4): 453–480.
Tilly, C. 1992. Coercion, Capital, and European States, AD 990–1992. Oxford: Blackwell.
Transparency International. 2010. East Africa Bribery Index 2010, http://probe
international.org/library/wp-content/uploads/2010/10/EAST-AFRICA-BRIBERY-
INDEX-2010.pdf (last accessed 1 May 2013).
Tripp, A.M. 1994. Gender, political participation and the transformation of associa-
tional life in Uganda and Tanzania. African Studies Review 37(1): 107–31.
Tripp, A.M. 2010. Museveni’s Uganda: Paradoxes of Power in a Hybrid Regime. Boulder,
CO: Lynne Rienner Publishers.
Tumushabe, G.W. 2009. Trends in public administration expenditure in Uganda: The
cost of the executive and its implications on poverty eradication and governance.
Advocates Coalition for Development and Environment (ACODE) Policy Research Series
No.27. Kampala: ACODE.
Turton, D. 1997. War and ethnicity: Global connections and local violence in
Northeast Africa and Former Yugoslavia. Oxford Development Studies 25(1): 77–94.
UBOS (Uganda Bureau of Statistics). 1995. The 1991 Uganda Population and Housing
Census: Analytical Report. Kampala: UBOS.
UBOS (Uganda Bureau of Statistics). 2003. Uganda National Household Survey
2002/2003: Report on the Socio-Economic Survey. Kampala: UBOS.
292 References
UBOS (Uganda Bureau of Statistics). 2005. The 2002 Uganda Population and Housing
Census: Main Report. Kampala: UBOS.
UBOS (Uganda Bureau of Statistics). 2006. Uganda National Household Survey
2005/2006: Report on the Socio-Economic Survey. Kampala: UBOS.
UBOS (Uganda Bureau of Statistics). 2010. Uganda National Household Survey
2009/2010: Socio-Economic Module, Abridged Report. Kampala: UBOS.
Ukiwo, U. 2007. Education, horizontal inequalities and ethnic relations in Nigeria.
International Journal of Educational Development 27(2): 266–281.
UNAIDS (Joint United Nations Programme on HIV/AIDS). 2010. Report on the Global
AIDS Epidemic 2010. Geneva: UNAIDS.
UNDP (United Nations Development Programme). 1998. Zimbabwe Human Develop-
ment Report 1998. Harare: Institute of Development Studies, University of Zimbabwe.
UNDP (United Nations Development Programme). 2003a. South Africa Human Devel-
opment Report 2003. Cape Town: Oxford University Press.
UNDP (United Nations Development Programme). 2003b. Summary: Zimbabwe Human
Development Report 2003. Harare: Institute of Development Studies, University of
Zimbabwe.
UNDP (United Nations Development Programme). 2004. Human Development Report
2004: Cultural Liberty in Today’s Diverse World. Oxford: Oxford University Press.
UNDP (United Nations Development Programme). 2006. Human Development Report
2006. New York: UNDP.
UNDP (United Nations Development Programme). 2007a. Turning Vision 2020 into
Reality: From Recovery to Sustainable Human Development, National Human Develop-
ment Report, Rwanda 2007. Kigali, Rwanda: UNDP.
UNDP (United Nations Development Programme). 2007b. Human Development Report
2007/2008. New York: UNDP.
UNDP (United Nations Development Programme). 2007c. Towards a More Inclusive
Society: Ghana Human Development Report. Accra: UNDP.
UNDP (United Nations Development Programme). 2009a. Human Development Report
2009: Overcoming Barriers: Human Mobility and Development. Basingstoke: Palgrave
Macmillan.
UNDP (United Nations Development Programme). 2009b. Zanzibar Human Develop-
ment Report 2009: Towards Pro Poor Growth. Zanzibar: UNDP Zanzibar.
UNDP (United Nations Development Programme). 2009c. Nigeria National Human
Development Report. Abuja: UNDP.
UNDP (United Nations Development Programme). 2010. Human Development Report
2010, 20th Anniversary Edition: The Real Wealth of Nations: Pathways to Human
Development. Basingstoke: Palgrave Macmillan.
Union of South Africa. 1955. Summary of the Report of the Commission for the Socio-
Economic Development of the Bantu Areas within the Union of South Africa, U.G.
61/1955. Pretoria: Government Printers (Tomlinson Commission Report).
UNSC (United Nations Security Council). 2001. Report of the Panel of Experts on the
Illegal Exploitation of Natural Resources and Other Forms of Wealth of the Democratic
Republic of the Congo, S/2001/357. New York: United Nations.
UNSC (United Nations Security Council). 2002. Final Report of the Panel of Experts on
the Illegal Exploitation of Natural Resources and Other Forms of Wealth of the Democratic
Republic of the Congo, S/2002/1146. New York: United Nations.
UNSC (United Nations Security Council). 2010. Final Report of the Group of Experts on
the Democratic Republic of the Congo, S/2010/596. New York: United Nations.
References 293
USAID (United States Agency for International Development). 2010. Nigeria Conflict
Assessment. Abuja: USAID.
Uvin, P. 2010. Structural causes, development co-operation and conflict prevention in
Burundi and Rwanda. Conflict, Security & Development 10(1): 161–179.
Vambe, M., ed., 2008. The Hidden Dimensions of Operation Murambatsvina in Zimbabwe.
Harare: Weaver Press.
Van Acker, F. 2004. Uganda and the Lord’s Resistance Army: The new order no one
ordered. African Affairs 103(412): 335–357.
Van der Berg, S., Louw, M., and du Toit, L. 2009. Poverty trends since the transition:
What we know. Stellenbosch Economic Working Papers 19(9). Matieland: Department
of Economics, University of Stellenbosch.
Vandeginste, S. 2006. Théorie consociative et partage du pouvoir au Burundi, Cahier
2006.04. Anvers: Institut de politique et de gestion du développement, Université
d’Anvers.
Vandeginste, S. 2008. Burundi: Entre le modèle consociatif et sa mise en oeuvre.
In Marysse, S., Reyntjens, F., and Vandeginste, S., eds, L’Afrique des Grands Lacs,
Annuaire 2007–2008. Paris: L’Harmattan.
Vandeginste, S. 2009. Power-sharing, conflict and transition in Burundi: Twenty years
of trial and error. Africa Spectrum 44(3): 63–86.
Vandeginste, S. 2011. Power-sharing as a fragile safety valve in times of electoral tur-
moil: The costs and benefits of Burundi’s 2010 elections. Journal of Modern African
Studies 49(2): 315–335.
Vansina, J. 2001. Le Rwanda ancien: Le royaume nyiginya. Paris: Karthala.
Varshney, A. 2002. Ethnic Conflict and Civic Life: Hindus and Muslims in India. New
Haven, CT: Yale University Press.
Vickers, M. 2000. Ethnicity and Sub-Nationalism in Nigeria: Movement for the Mid West
State. Oxford: Worldview.
Waldmeir, P. 1997. Anatomy of a Miracle: The End of Apartheid and the Birth of the New
South Africa. New York: W.W. Norton.
Waldorf, L. 2006. Mass justice for mass atrocity: Rethinking local justice as transitional
justice. Temple Law Review 79(1): 1–87.
Watts, R.L. 2008. Comparing Federal Systems. 3rd edn. Montreal: McGill-Queen’s
University Press.
Williams, P.R., and Simpson, M.T. 2008. Rethinking the political future: An alternative
to the ethno-sectarian division of Iraq. American University International Law Review
24(2): 191–247.
Wolff, S. 2009. Complex power-sharing and the centrality of territorial self-governance
in contemporary conflict settlements. Ethnopolitics 8(1): 27–45.
Wolff, S. 2011. Post-conflict state building: The debate on institutional choice. Third
World Quarterly 32(10): 1777–1802.
Wong, R.B. 1997. China Transformed: Historical Change and the Limits of European
Experience. Ithaca, NY: Cornell University Press.
World Bank. 2006. Uganda Poverty and Vulnerability Assessment. Washington, DC:
World Bank.
World Bank. 2009. Africa Development Indicators 2008/09: Youth and Employment in
Africa: The Potential, the Problem, the Promise. Washington, DC: World Bank.
Yeager, R. 1989. Tanzania: An African experiment. 2nd edn. Boulder, CO: Westview
Press.
Yoshikuni, T. 1999. Notes on the influence of town–country relations on African urban
history, before 1957: Experiences of Salisbury and Bulawayo. In Raftopoulos, B., and
294 References
Yoshikuni, T., eds, Sites of Struggle: Essays in Zimbabwe’s Urban History. Harare: Weaver
Press.
Young, C. 1976. The Politics of Cultural Pluralism. Madison, WI: University of
Wisconsin Press.
Younis, N. 2011. Set up to fail: Consociational political structures in post-war Iraq,
2003–2010. Contemporary Arab Affairs 4(1): 1–18.
Zartman, W.I. and Delgado, C. 1984. Introduction. In: Zartman, W.I., and Delgado,
C., eds, The Political Economy of Ivory Coast. New York: Praeger Publishers.
Zimbabwe Election Support Network. 2009. Possible Electoral Systems for Zimbabwe.
Harare: Zimbabwe Election Support Network.
Zondi, S. 2011. ZANU-PF and MDC power-sharing: Zimbabwe still at crossroads?
In Besada, H., ed., Zimbabwe: Picking Up the Pieces. Basingstoke: Palgrave Macmillan.
Author Index
295
296 Author Index
Abidjan, 73, 74, 80, 81, 256 Banyankole, 127, 128, 129, 135, 136,
Accra, 83, 87, 88, 89, 209 139, 145, 146, 150, 152, 214, 216,
Acheampong, Ignatius Kutu, 84 227, 228, 229, 231, 232, 242, 243,
Acholi, 127, 128, 129, 135, 139, 144, 245
150, 214, 231, 232, 241, 242, 261, Banyoro, 127, 128, 145, 214, 225, 227,
262 228, 229, 231, 232, 240, 241, 243
Action Congress of Nigeria (ACN), 197 Baoulé, 72, 75, 76, 77, 78, 79, 92
ad hoc power-sharing, 10, 33, 95–125, Basic Human Needs (BHN), 213, 215,
254, 256–7, 258, 265, 266, 269, 272, 216, 233, 248
273 Bédié, Henri Konan, 75, 92
affirmative action, 104, 116, 193, 205 Biafran War, 259, 265
African Independent Television (AIT), Black Economic Empowerment (BEE), 99
192 Boahen, Albert Adu, 84, 85, 93
African National Congress (ANC), 98, Botswana, 14, 31, 35
105, 107, 108, 109, 110, 117, 122, budget, 17, 111, 144, 198, 271, 273
123 Buganda Kingdom, 126, 127, 128, 130,
Afrikaners, 96, 101, 110, 124 143, 144, 150, 258, 261
agriculture, 40, 43, 50, 59, 73, 87, 98, 99, Bujumbura, 62, 64, 65
123, 158, 184 Bukenya, Gilbert, 139
Akan, 70, 72, 82, 84, 85, 86, 88, 216, Bulawayo, 114, 117–21, 125
241, 242, 256 Burkina Faso, 70
Alliance for Democracy (AD), 196, 197, Burundi, 4, 7, 8, 24, 28, 31, 33, 39,
198 40–65, 108, 253, 254, 255, 261, 262,
All Nigeria Peoples Party (ANPP), 196, 264, 265, 267, 269, 272
197, 198 Bururi, 44, 45, 46, 48, 62, 64
Amin, Idi, 126, 129, 133 Busoga Kingdom, 128, 143
Angola, 8, 19, 31, 33 Buyoya, Pierre, 46, 48, 52, 61
Annan, Kofi, 167
Arab Spring, 178 cabinet
Arusha Peace and Reconciliation in Burundi, 46, 51
Agreement, 46, 50 in Côte d’Ivoire, 75, 78, 108
Ashanti, 82, 84, 85, 86, 88, 89 in Ghana, 89–91
assassination, 44, 46 in Kenya, 108, 157, 158, 176
assets, 3, 161, 213, 214, 215, 216, 239, in Nigeria, 184, 187, 196, 197–8, 202
247, 248 in Rwanda, 49
autonomy, 12, 34, 35, 43, 97, 124, in South Africa, 107, 108, 110
126–52, 191, 192, 193, 200, 257, in Tanzania, 141–3
258, 267, 269 in Uganda, 141–3
Avoidance of Discrimination Act, 85 in Zimbabwe, 106, 108, 119
Cameroon, 21
Baganda, 126, 127, 128, 129, 130, 131, Cape Town, 105, 115, 116, 117
132, 133, 134, 136, 138, 139, 142, cash crops
143, 144, 145, 146, 150, 151, 152, cocoa, 73
227, 232, 241, 245, 257, 258, 261 coffee, 73
299
300 Subject Index
Catholics, 4, 76, 83, 133, 152, 170 constitutional engineering, 8, 11, 12, 37,
Central Bank for West African States, 75, 254, 259
77 constitutions
Centre for Research on Inequality, in Burundi, 46, 50–2
Human Security and Ethnicity in Côte d’Ivoire, 75, 76
(CRISE), 3, 9 in Ghana, 84, 86, 89–90
Chama cha Mapinduzi (CCM), 133, 140, in Kenya, 154, 162–3, 164–5, 175–7
142, 143, 148 in Nigeria, 191, 193–4, 195, 196, 197,
Charte du Grand Nord, 76 198, 200, 206
child mortality, 89, 159, 160 in Rwanda, 49–50
Chimurenga, 101, 103 in South Africa, 107–8, 111
Christianity, 34, 40, 71, 72, 76, 83, 98, in Tanzania, 133, 142, 143
117, 133, 147, 156, 164, 170, 181, in Uganda, 128, 141, 144
182, 211, 212, 256, 259 in Zimbabwe, 111–12
constructivism, 3, 42, 59, 67
Civic United Front (CUF), 133, 140, 141,
Convention People’s Party (CPP), 86
143, 148, 152
corruption, 58, 63, 64, 117, 121, 140,
civil society, 50, 64, 107, 124, 164, 271
143, 167
civil war, 5, 34, 40, 45, 46, 47, 48, 49, 50,
Côte d’Ivoire, 4, 7, 8, 28, 31, 33, 35, 39,
52, 53, 56, 57, 58, 61, 63, 64, 65, 66,
66–94, 108, 253, 254, 255–6, 261,
67, 69–81, 95, 107, 153, 178, 180,
262, 263, 265, 266, 267, 270
192, 193, 195, 205, 206, 208, 259,
coup d’état, 70, 77, 78, 80, 84, 92, 93,
266
129
class, 11, 12, 84, 99, 111, 112, 115, 116,
cultural status horizontal inequalities
118, 119, 138, 139, 192, 193, 257,
(HIs)
271
in Burundi, 41
Cold War, 1, 10, 106, 254, 266, 272 in Côte d’Ivoire, 76, 91
Colonialism, 23, 25, 35, 37, 42, 43, 44, in Ghana, 91
45, 47, 48, 60, 61, 69, 70, 87, 96, in Kenya, 153, 156–62
100, 118, 128, 129, 136, 142, 156, in Nigeria, 182, 191
179, 180, 215, 254, 255, 256, 258, in Rwanda, 41
261 in South Africa, 95
Coloureds, 96, 97, 98, 99, 116, 117, 123, in Tanzania, 127
124 in Uganda, 134
Comité national de salut public (CNSP), in Zimbabwe, 108
78
conflict prevention, 6, 10–39, 41, 57–9, Dahl, Robert, 12, 13, 33
253, 254, 259, 262, 264, 265, 270, Dar es Salaam, 136, 138, 144, 145, 147,
273, 274 148, 149, 152
Congress of the People (COPE), 110 decentralisation
Congress of South African Trade Unions in Burundi, 19, 22, 23, 24, 26, 43, 45,
(COSATU), 99, 110 52, 255
Conseil national pour la défense de la in Côte d’Ivoire, 19, 22, 23, 24, 26, 28,
démocratie–Forces pour la défense 31, 33
de la démocratie (CNDD-FDD), 46, in Ghana, 20, 22, 23, 24, 26
52, 54, 55, 57, 58, 63, 64 in Kenya, 20, 22, 23, 24, 26, 162–5,
consociational democracy, 11, 12 167, 176–7, 258–9
constituency, 15, 113, 119, 122, 125, in Nigeria, 20, 22, 23, 24, 26, 194,
166, 194, 195 199–203, 206
Constitutional Court, 107 in Rwanda, 20, 22, 23, 24, 26, 52
Subject Index 301
in South Africa, 21, 22, 23, 24, 26, 103, electoral systems, 14, 15, 16, 17, 18, 21,
105–11 28, 38, 112, 113, 122, 125, 138, 194,
in Tanzania, 21, 22, 23, 24, 26, 126, 195, 259, 270
131, 138, 141, 143–4, 150, 151, elite, 13, 35, 45, 47, 48, 50, 59, 60, 76,
257–8, 266 95, 103, 115, 122, 124, 128, 129,
in Uganda, 21, 22, 23, 24, 26, 126, 136, 140, 199, 260
141, 143–4, 146–7, 149, 151–2, Ethiopia, 23, 31, 34, 62
257–8 ethnic fractionalisation, 23, 24
in Zimbabwe, 21, 22, 23, 24, 26, 103, ethnicity
105–11, 113, 124 in Burundi, 41, 46, 48, 51, 53, 54, 55,
De Klerk, F.W., 106, 108 57
democracy, 10, 11, 12, 13, 14, 15, 21, 31, in Côte d’Ivoire, 69, 81
33–6, 39, 41, 49, 50, 53, 60, 82, 84, in Ghana, 82, 83, 84–5, 86
85, 107, 110, 117, 122, 126, 152, in Kenya, 153, 156, 162, 163, 168–75
167, 197, 207, 255, 256, 257, 259 in Nigeria, 179, 191
Democratic Alliance (DA), 110, 117 in Rwanda, 41, 45, 56, 57
Democratic Republic of the Congo (the in South Africa, 96, 97, 98, 101, 118
DRC), 1, 21, 23, 31, 33, 38, 53, 54, in Tanzania, 130, 140, 143, 151
62, 63, 64, 108, 111, 145, 254 in Uganda, 149
development agencies, 253, 271, 273, in Zimbabwe, 119
274 Europe, 1, 24, 66
development cooperation, 28, 64, 119, Evans-Pritchard, E.E., 107
135, 143, 146, 178, 252–74 Ewe, 82, 83, 84, 85, 90, 216, 227, 231,
disarmament, demobilisation and 232, 241, 242, 244
reintegration (DDR), 54, 111
discrimination, 48, 49, 85, 116, 189, 191,
209, 215, 227 Federal Character Commission (FCC),
194, 201, 202
distortion of perceptions, 241, 260, 262,
273 Federal Character Principle, 187, 189,
193, 194, 200, 201, 202, 203, 206,
259, 265
East Africa, 126, 131, 257 federalism, 12, 13, 18, 23, 34, 38, 108,
East Asia, 1, 123 130, 138, 142, 143, 144, 180, 184,
education, 4, 44, 47, 48, 50, 56, 87, 88, 185, 187, 189, 191, 192, 193, 194,
104, 135, 137, 145, 157, 158, 159, 195, 196, 197, 199, 200, 201, 202,
160, 162, 173, 187, 188, 202, 203, 203, 205, 206, 207, 254, 257–8, 259,
226, 227, 229, 230, 233, 261, 270 265, 267–8, 269
Egypt, 8, 178 feelings of inferiority, 233, 241, 244, 245
elections feelings of superiority, 244, 245
in Burundi, 51, 55, 57, 62, 63 FIFA World Cup, 117
in Côte d’Ivoire, 71, 75, 76, 77, 81 first-past-the-post (FPTP), 12, 13, 35, 36,
in Ghana, 85, 92 113, 114, 122, 123, 194, 254
in Kenya, 153, 163, 165, 166, 167, 172, fiscal decentralisation, 16, 266, 267–8,
173, 176, 177 269
in Nigeria, 186, 192, 194, 195 Freedom House, 31, 32
in Rwanda, 49, 53, 54, 62, 64 Frente de Libertação de Moçambique
in South Africa, 99, 107, 109, 110, 121 (FRELIMO), 138
in Tanzania, 148 Front Populaire Ivoirien (FPI), 75, 77, 80
in Uganda, 138, 141, 143 Fulani, 179, 184, 214, 216, 227, 230,
in Zimbabwe, 103, 111–15, 122, 124 232, 241, 242, 259
302 Subject Index
Japan International Cooperation Agency Kenya People’s Union (KPU), 162, 163,
(JICA), 2, 132, 149, 168, 169, 170, 164
171, 172, 173, 174, 174, 204, 209, Kenyatta, Jomo, 157, 163, 164, 258
216, 225, 230, 241, 242, 246, 247, Khama, Seretse, 95
248, 249, 253, 257, 258, 259, 260 Khoisan, 96
Jews, 98 Kibaki, Mwai, 165, 166, 167, 172, 177,
Johannesburg, 96, 105 258
Jonathan, Goodluck, 199, 202, Kigali, 62
203 Kikuyu, 154, 155, 156, 157, 163, 164,
journalism, 50, 55, 66, 211 166, 167, 169, 170, 173, 214, 216,
Judiciary 227, 231, 232, 241, 245, 258
in Burundi, 46, 51 Kikwete, Jakaya Mrisho, 140, 148
in Côte d’Ivoire, 77 King Mutebi, 139
in Ghana, 84, 85, 90 King Mutesa II, 128
in Kenya, 157–8, 162, 164, 176, 270 King Mwambutsa, 44, 45
in Nigeria, 180, 182, 184–6, 194, King Rwabugiri, 43
196–200 Kisii, 166, 227, 231, 232, 241, 245
in Rwanda, 49, 56 Krou, 70, 72, 79, 80, 92
in South Africa, 107, 110 Kufuor, John, 88, 89
in Tanzania, 142, 143–4, 146, 151
in Uganda, 130, 139, 142, 143–4, 146, Lagos, 36, 204, 228, 229, 259
151 Lake Victoria, 126, 127
in Zimbabwe, 106, 107–8, 119 Lancaster House Agreement, 103, 124
Languages, 4, 47, 82, 83, 86, 87, 97, 108,
Kabaka, 127, 128 179, 191, 192
Kagame, Paul, 53, 62 Lebanon, 34
Kalenjin, 154, 155, 156, 157, 164, 165, Legislature
166, 169, 170, 227, 258 in Burundi, 51, 54, 55
Kamba, 154, 155, 169, 170, 174, 175, in Côte d’Ivoire, 77
214, 223, 227, 228, 229, 231, 239, in Ghana, 85
243, 245, 258 in Kenya, 162, 163, 164, 176
Kampala, 127, 129, 138, 139, 145, 146, in Nigeria, 182, 184, 198, 201
147, 149, 152 in Rwanda, 50, 53
Karume, Amani Abeid, 141 in South Africa, 107, 109, 122
Karume, Sheikh Abeid Amani, in Tanzania, 133
140 in Uganda, 143, 146
Kaunda, Kenneth, 95 in Zimbabwe, 106, 111, 113, 114
Kayibanda, Grégoire, 45 Lewis, W. Arthur, 11, 12, 13, 37
Kenya, 4, 7, 8, 23, 24, 27, 28, 31, 33, 35, liberal democracy, 41, 110
36, 39, 108, 133, 153–77, 209, 214, Liberal Democratic Party (LDP), 165
216, 227, 228, 231, 232, 241, 243, Libya, 8, 178
245, 246, 253, 254, 258–9, 260, 261, Lijphart, Arend, 11, 12, 13, 14, 15, 17,
262, 264, 265, 266, 268, 269, 270, 37, 38, 39, 62, 63, 124
272 local elections, 14, 16, 18, 28, 64
Kenya African Democratic Union Local Government Authorities (LGAs),
(KADU), 162, 163 143
Kenya African National Union (KANU), Lord’s Resistance Army (LRA), 129, 135,
162, 163, 164, 165, 172, 179 139
Kenya Broadcasting Corporation (KBC), Luhya, 154, 155, 156, 157, 166, 216,
165, 166 227, 258
304 Subject Index
Luo, 154, 155, 156, 157, 166, 168–9, multi-party system, 6, 10, 13, 14, 41, 49,
170, 173, 227, 242, 245, 258 50, 53, 82, 84, 117, 126, 138, 140,
Luwero Triangle, 139 142, 147, 163, 167, 195, 257, 258
Museveni, Yoweri, 39, 62, 126, 129, 130,
138, 139, 141, 142, 145, 150, 152,
Mahama, Alhaji Aliu, 89 258
majoritarian politics, 110, 111, Musyoka, Kalonzo, 165
267
Malawi, 21, 38, 118 Nairobi, 154, 159, 160, 161, 163,
Malaysia, 4 168
Mali, 1 Nakuru, 154, 168
Mandela, Nelson, 52, 95, 99, 106, 107, National Alliance Party of Kenya (NAK),
108, 270 165
Mauritius, 14, 35 National Democratic Congress (NDC),
Mbarara, 128, 145, 146, 147, 152 85
Mbeki, Thabo, 108, 110 National Liberation Council (NLC),
Mfecane, 96, 101 84
migrant workers, 97, 99 National Liberation Movement (NLM),
military 84, 102
in Burundi, 52 National Party of Nigeria (NPN), 199
in Côte d’Ivoire, 80 National Party (NP), 97, 98, 108, 110,
in Ghana, 84 143, 199
in Kenya, 39, 133 National Patriotic Party (NPP), 85, 89
in Nigeria, 182, 192, 193, 195 National Rainbow Coalition (NRC), 84,
in Rwanda, 40, 42, 43, 45, 46 165
in South Africa, 107 National Resistance Army (NRA), 126,
129, 139
in Tanzania, 142
National Resistance Movement (NRM),
in Uganda, 133, 136
129, 138, 139, 141, 142, 143, 145,
in Zimbabwe, 107
146, 150, 258
Mill, John Stuart, 11
natural resources, 31, 73, 87, 215
minority groups, 53, 83, 118, 179, Ndadaye, Melchior, 46
258
Ndebele, 96, 101, 103, 104, 115, 117,
Minority At Risk (MAR), 208, 209 118, 119, 121, 125, 214, 216, 227,
Miracle Africain, 73 231, 232, 241, 242, 245, 257, 261,
Moi, Daniel Arap, 157, 164, 165, 172, 263, 271
259 Niger Delta, 207
Mole/Dagbani, 82, 83, 86, 90, 216, 227, Nigeria, 1, 4, 7, 8, 12, 17, 23, 28, 31, 34,
231, 232 35, 36, 39, 178–207, 209, 211, 214,
Mombasa, 154, 160, 168 216, 227, 228, 231, 232, 241, 243,
Mouvement démocratique républicain 244, 245, 246, 247, 253, 259, 260,
(MDR), 49 261, 262, 265, 266, 268, 269, 270,
Mouvement Patriotique pour la Côte 271
d’Ivoire (MPCI), 70, 81 Nigerian Civil War, 192, 259
Movement for Democratic Change See also Biafran War
(MDC), 105, 111, 112, 113, 114, Nigerian National Petroleum
119, 121 Corporation (NNPC), 202
Mugabe, Robert, 95, 103, 104, 112, Nkomo, Joshua, 103
119, 122, 124, 125, 257, Nkrumah, Kwame, 84, 85, 86, 87, 256,
270 270
Subject Index 305
quota system, 47, 122, 124 Somali, 169, 170, 214, 216, 228, 231,
232, 242
Rassemblement des républicains (RDR), South Africa, 7, 8, 12, 23, 28, 33, 34, 38,
76, 77, 80 39, 95–125, 215, 253, 254, 256–7,
Rawlings, Jerry, 84, 85, 87, 89, 93 261, 262, 263, 265, 267, 269, 270,
refugees, 44, 45, 47, 48, 61, 65 271
South African Communist Party (SACP),
religions, 40, 42, 50, 71, 72, 76, 81, 83,
98, 110
86, 87, 133, 139, 143, 147, 148, 152,
154, 170, 171, 179, 181, 191, 204, South African War, 96
259, 264 Southern African Development
Community (SADC), 112, 124
See also Christianity; Islam; traditional
Southern Mandé, 70, 78, 79
religions
state-building, 1, 40–65
Rhodes, Cecil, 13, 101
structural adjustment, 111
Rift Valley, 153, 155, 156, 158, 164, 166,
subjective inequalities, 2, 5, 6, 8, 209,
167, 258
210–15, 232, 233, 241, 245, 260, 273
Rwagasore, Prince Louis, 44, 45, 52, 61
Sudan, 1, 23, 31, 33, 38, 108, 255
Rwanda, 7, 8, 28, 35, 39, 40–65, 253,
Sukuma, 130, 131, 140
255, 261, 262, 264, 265, 266, 267,
sustainable power-sharing, 269, 270,
271
272, 274
Rwanda Demobilisation and
Swahili, 62, 131, 133, 142
Reintegration Commission (RDRC),
54, 63
Tambo, Oliver, 95
Rwandan Defence Forces (RDF), 53
Tanganyika African National Union
Rwandan Patriotic Army (RPA), 54, 63
(TANU), 140, 142
Rwandan Patriotic Front (RPF), 42, 45,
Tanzania, 7, 8, 23, 28, 34, 35, 39, 59,
46, 47, 48, 49, 50, 53, 54, 56, 58, 59,
126–52, 253, 257–8, 264, 265, 266,
61, 62, 63
268, 269, 270
Togoland Congress Party (TCP), 84
secessionism, 126 traditional religions, 83, 133, 170
senate, 14, 16, 21, 51, 53, 62, 107, 113, trait-taking and trait-making, 36
124, 176, 180, 184, 185, 186, 195, Truth and Reconciliation Commission
196 (TRC), 108
Shirazi, 131, 142, 145 Tsvangirai, Morgan, 112
Shona, 101, 103, 104, 117, 118, 119, Tutsis, 42, 43, 44, 45, 46, 47, 48, 50, 51,
121, 125, 216, 228, 231, 232, 241 52, 56, 57, 59, 60, 61, 64
Social Revolution, 44, 45, 46, 47, 50, 60, Tutu, Desmond, 95
255 Twas, 40, 51, 53, 55, 59, 63
socioeconomic horizontal inequalities
(HIs) Uganda, 7, 8, 23, 28, 31, 35, 39, 45, 47,
in Burundi, 41 59, 60, 62, 126–52, 209, 214, 216,
in Côte d’Ivoire, 69, 92, 93 227, 229, 231, 232, 241, 242, 245,
in Ghana, 68, 69 246, 253, 257–8, 259, 260, 261, 262,
in Kenya, 156, 158, 168 265, 266, 268
in Nigeria, 186 Ujamaa, 138, 140
in Rwanda, 41, 48 unemployment, 99, 123, 137, 167, 230
in South Africa, 99 Unguja island, 131, 133, 137, 140, 144,
in Tanzania, 147, 149 145, 147, 148, 149, 150, 152
in Uganda, 134 unitary state, 13, 16, 17, 126–52, 254,
in Zimbabwe, 103, 104, 111 265
Subject Index 307