Party-Voter Linkage in Africa
Anja Osei
Party-Voter Linkage
in Africa
Ghana and Senegal in Comparative
Perspective
RESEARCH
ISBN 978-3-531-18612-2 ISBN 978-3-531-19140-9 (eBook)
DOI 10.1007/978-3-531-19140-9
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at .
Springer VS
© VS Verlag für Sozialwissenschaften | Springer Fachmedien Wiesbaden 2012
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole
or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical
way, and transmission or information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed.
Exempted from this legal reservation are brief excerpts in connection with reviews or schol-
arly analysis or material supplied specifically for the purpose of being entered and executed
on a computer system, for exclusive use by the purchaser of the work. Duplication of this pub-
lication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained
from Springer. Permissions for use may be obtained through RightsLink at the Copyright
Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the
date of publication, neither the authors nor the editors nor the publisher can accept any legal
responsibility for any errors or omissions that may be made. The publisher makes no warranty,

express or implied, with respect to the material contained herein.
Cover design: KünkelLopka GmbH, Heidelberg
Printed on acid-free paper
Springer VS is a brand of Springer DE. Springer DE is part of Springer Science+Business Media.
www.springer-vs.de
Anja Osei
Konstanz, Germany
 is work was submitted in ful lment of the requirements for the Degree of Doctor
of Politics at the University Leipzig, 2011.

Content
List of Abbreviations and Acronyms 9
List of Figures 12
List of Tables 13
1 Introduction: The Purpose of the Study 14
2 Conceptual Problems of Party Research in Africa 22
2.1 Introduction 22
2.2 Party Research in Africa: Findings and Problems 23
2.3 New Challenges: Neo-democracies and Party Change 28
2.4 Democratisation, Hybridisation and Political Parties 34
2.5 African Studies between Universalism and Exceptionalism 39
2.5.1 Research Paradigms in African Studies 39
2.5.2 Universalism 40
2.5.3 Exceptionalism 42
2.5.4 Spatiality, Discourse and Power Relations 44
3 Research Design and Methodology 49
3.1 Searching for a Framework of Analysis 49
3.2 Theoretical Guideline: the Concept of Linkage 54
3.3 Case selection and Comparative Design 56
3.3.1 The Comparative Method: Advantages and Limitations 56

3.3.2 The Most Different Systems Design 60
3.4. Data Collection 65
3.4.1 Collection of Qualitative Data 65
3.4.2 Collection of Quantitative Data 69
4 Background: Party Formation in Africa 72
4.1 Party Formation at Independence 72
4.2 Single-party States 77
4.3 Third Wave Parties 81
6 Content
5 Democracy, the State and Political Parties in Africa Today 86
5.1 Political Parties and Clientelism 86
5.2 Political Parties and Ethnicity 92
5.3 Political Parties and the Neoptrimonial State 97
6 Political Parties in Ghana . 100
6.1 Introduction 100
6.2 Party System, Legal and Institutional Framework 101
6.2.1 Party System 101
6.2.2 Legal and Institutional Framework 103
6.3 The Evolution of Political Parties in Ghana 104
6.3.1 Independence and First Republic 104
6.3.2 Political Parties in the Second Republic 107
6.3.3 Third Republic and PNDC Rule 109
6.3.4 Political Parties in the Fourth Republic 112
6.4 Voting Patterns in Ghana 116
6.4.1 Introduction 116
6.4.2 The Ethnic Factor 118
6.4.3 Regionalism 125
6.4.4 Party traditions 126
6.4.5 Socioeconomic Factors 130
6.4.6 Personalism 134

6.4.7 Conclusion 136
6.5 Party Organisation 138
6.5.1 Introduction 138
6.5.2 NPP 139
6.5.3 NDC 143
6.5.4 PNC 144
6.6 Ideological Orientations 146
6.6.1 Introduction 146
6.6.2 NPP 147
6.6.3 NDC 149
6.6.4 PNC 151
6.7 Party Activities 152
6.8 Voters’ Attitudes in Ghana 158
6.9 Conclusion 168
Content 7
7 Political Parties in Senegal 170
7.1 Introduction 170
7.2 Party System, Legal and Institutional Framework 171
7.2.1 Party System 171
7.2.2 Legal and Institutional Framework 174
7.3 The Evolution of Political Parties in Senegal 176
7.3.1 Independence and the Era Senghor 176
7.3.2 Political Parties under Abdou Diouf 180
7.3.3 Political Parties under Abdoulaye Wade 183
7.4 Voting Patterns 187
7.4.1 Introduction 187
7.4.2 Regional Distribution of Party Support 189
7.4.3 Ethnicity 193
7.4.4 Sociostructure 196
7.4.5 Sufi Brotherhoods and Party Support 202

7.4.6 Personalism 206
7.4.7 Conclusion 208
7.5 Party Organisation 210
7.5.1 Introduction 210
7.5.2 PDS 210
7.5.3 PS 214
7.5.4 AFP 216
7.6 Ideological Orientations 218
7.6.1 Introduction 218
7.6.2 PDS 219
7.6.3 PS 220
7.6.4 AFP 222
7.7 Activities 223
7.8 Voters’ Attitudes in Senegal 228
7.9 Conclusion 237
8 Political Parties in Ghana and Senegal in a Comparative View 239
8.1 Overview 239
8.2 Comparing Parties 240
8.3 Comparing Social Anchorage 241
8.4 Comparing Countries 247
8 Content
9 Linkage Strategies in a Comparative View 253
9.1 Introduction 253
9.2 Direct Linkage and Indirect Linkage 253
9.3 Linkage by Reward 256
9.3.1 Clientelism 256
9.3.2 Patronage 259
9.3.3 Vote-Buying 260
9.4 Other Forms 263
9.5 Conclusion 265

10 Concluding Remarks 267
References 275
Appendix 292
Survey Ghana 2008 292
Survey Senegal 2008 299



List of Abbreviations and Acronyms
Acronym Definition
ACP Action Congress Party
AFRC Armed Forces Revolutionary Council
AFP
AJ/PADS
ANSD
Alliance des Forces de Progrès
And Jëf/ Parti Africain pour la Démocratie et le
Socialisme
Agence Nationale de la Statistique et de la
Démographie
APJ/JJ Alliance pour le Progrès et la Justice/Jëf Jël
APR/Yakar Alliance pour la République/Yaakaar
APRS Aborigines’ Rights Protection Society
BDS Bloc Démocratique Sénégalaise
BGC Bloc des Centristes Gaïnde
BMS Bloc des Masses Sénégalaises
BPS
CDR
Bloc Populaire Sénégalaise
Committee for the Defence of the Revolution

CDD Center for Democratic Development
CDP/Garab-Gi Convention des Démocrates et des Patriotes)
CNTS Confédération nationale des travailleurs du Sénégal
CODEO Coalition of Domestic Election Observers
COOC Code des Obligations Civiles et Commerciales
CPP Convention People’s Party
CRC Convergence pour Rénouveau et la Citoyenneté
DFP Democratic Freedom Party
EGLE
ENEA
Every Ghanaian Living Everywhere
École Nationale d'Économie Appliquée
FCFA
Franc de la Communauté Financière d'Afrique
FEANF Fédération des étudiants d’Afrique noire en France
FORD Forum for Democracy
FSD/BJ
GA
Front pour le Socialisme et la Démocratie/Benno Jubël
Great Alliance
GBA Ghana Bar Association
GCP Ghana Congress Party
GIGA German Institute of Global and Area Studies
IDEA International Institute for Democracy and Electoral
Assistance
IMF International Monetary Fund
JSS Junior Secondary School
10 Abbreviations and Acronyms
Acronym Definition
KANU Kenya African National Union

LD/MPT Ligue démocratique/Mouvement pour le Parti du
Travail
MAC Mouvement Autonome Casamançais
MAP Muslim Association Party
MDSD Most Different Systems Design
MFDC Mouvement des forces démocratiques de la Casamance
MFJ Movement for Freedom and Justice
MMD
MP
Movement for Multiparty Democracy
Member of Parliament
MPS Mouvement Populaire Sénégalais
MRDS Mouvement de la Réforme pour le Développement
Social
MSA Mouvement Socialiste Africain
MSSD Most Similar Systems Design
NAL National Alliance of Liberals
NCP National Convention Party
NDC National Democratic Congress
NIMD Netherland Institute for Multiparty Democracy
NIP National Independence Party
NLC National Liberation Council
NLM National Liberation Movement
NPC Northern People’s Congress
NPP New Patriotic Party
NRP National Reform Party
NUGS National Union of Ghana Students
OAU Organisation of African Unity
PAI Parti Africain de l’Indépendance
PCF Parti communiste français

PCP
PDC
People’s Convention Party
People’s Defence Committee
PDCI Parti Démocratique de
Côte d'Ivoire
PDS Parti Démocratique Sénégalais
PFP Popular Front Party
PHP People’s Heritage Party
PIT Parti de l’Indépendence et du Travail
PLS Parti Libéral Sénégalais
PNC People’s National Convention
PNDC Provisional National Defense Council
Abbreviations and Acronyms 11
Acronym Definition
PNDCL Provisional National Defense Council Law
PNP
PP
People’s National Party
Progress Party
PPC Parti pour le Progrès et la Citoyenneté
PRA Parti du Regroupement Africain
PS Parti Socialiste
PSA Parti Socialiste Authentique
PSAS Parti Sénégalais d'Action Socialiste
PSD/JB Parti Social-Démocrate/Jant-Bi
PSR Parti Socialiste Républicain
RDA Rassemblement Démocratique
RES Rassemblement des Écologistes du Sénégal/Les Verts
RND

RP
Rassemblement National Démocratique
Rassemblement pour le Peuple
RPS/JR Rassemblement Patriotique Sénégalaise/Jammi Rewmi
RTS Radiodiffusion Télévision Sénégalaise
SDF Social Democratic Front
SFIO Section Française de l’Internationale Ouvrière
SPD Sozialdemokratische Partei Deutschlands
TANU Tanganyika African National Union
TFP Third Force Party
TUC Trades Union Congress
UDS
UDS/R
Union Démocratique Sénégalaise
Union Démocratique Sénégalaise/Rénovation
UGCC United Gold Coast Convention
UNC United National Convention
UNIP United National Independence Party
UNP Union Nationale Patriotique
UPS Union Populaire Senegalaise
URD Union pour la Renouveau Démocratique
VAT Value Added Tax
WANC
WDC
West African National Congress of British West Africa
Workers’ Defence Committees


List of Figures
Figure 1: Voter Turnout Worldwide, 1990-2001 17

Figure 2: Party Identification in Selected African Countries 19
Figure 3: Freedom House Ratings for Africa, Selected Years. 35
Figure 4: Sub-Categories of Linkage by Reward 91
Figure 5: Party Affiliation in Ghana 117
Figure 6: Ethnic Affiliation of NPP Sympathisers 121
Figure 7: Ethnic Affiliation of NDC Sympathisers 122
Figure 8: NPP and NDC Voters Urban/Rural 131
Figure 9: Voting Intentions in Ghana 159
Figure 10: Party Membership in Ghana 161
Figure 11: Self-Perceptions and Voters’ Attitudes, NPP 163
Figure 12: Self-Perceptions and Voters’ Attitudes, NDC 164
Figure 13: Self-Perceptions and Voters’ Attitudes, PNC 165
Figure 14: Voters’ Attitudes on Selected Issues, Ghana 166
Figure 15: Factors that Influence Voting Decisions, Ghana 167
Figure 16: Party Affiliation in Senegal 188
Figure 17: Voting Intentions of Non-party Affiliates in Senegal 189
Figure 18: Ethnic Composition of Party Supporters, Senegal 195
Figure 19: Ethnic Group Never Treated Unfairly, Senegal 196
Figure 20: PDS, PS and AFP Voters Urban-rural 197
Figure 21: Age Structure of PDS, PS and AFP Voters 201
Figure 22: Voting Intentions in Senegal 229
Figure 23: Candidate Voted for in the 2007 Presidential Elections, Senegal 231
Figure 24: Self-Perception and Voters’ Attitudes, PDS 233
Figure 25: Self-Perception and Voters’ Attitudes, PS 234
Figure 26: Self-Perception and Voters’ Attitudes, AFP 235
Figure 27: Factors that Influence Voting Decisions 236
Figure 28: Voters’ Attitudes on Political Parties in Ghana 242
Figure 29: Voters’ Attitudes on Political Parties in Senegal 243
Figure 30: Satisfaction with Democracy in Ghana and Senegal 244
Figure 31: Voter Turnout in the Five Most Recent Elections 245

Figure 32: Clientelism in the Local Party 257
Figure 33: Vote-Buying Attempts in Ghana and Senegal 261


List of Tables
Table 1: Freedom House-Ratings for Ghana and Senegal (1991-2007) 62
Table 2: Dimensions of comparison 64
Table 3: Number of Seats Won by the Parties in the Parliamentary Elections, 1996-2008 102
Table 4: Regional Distribution of Seats 2004 and 2000 (NPP, NDC and PNC only) 119
Table 5: Ethnic Group Treated Unfairly (Selected Groups Only) 124
Table 6: Relative Poverty of NPP and NDC Voters 132
Table 7: Education of NPP und NDC voters 133
Table 8: Evaluation of Personal Living Conditions by NPP and NDC Voters 133
Table 9: The Nature of Party Conflict in Ghana 136
Table 10: Voting Intentions According to Constituency 160
Table 11: Parliamentary Elections Senegal, 1998 172
Table 12: Parliamentary Elections Senegal, 2001 173
Table 13: Parliamentary Elections Senegal, 2007 174
Table 14: Presidential Elections Senegal, 2007 185
Table 15: Regional Voting Patterns in Senegal 191
Table 16: Relative Poverty of PDS, PS and AFP Voters 199
Table 17: Living Conditions of PDS, PS and AFP Voters 200
Table 18: Assessment of Present Economic Conditions by PDS, PS and AFP Voters 200
Table 19: Education of PDS, PS and AFP Voters 202
Table 20: Trust in Selected Political Institutions, Senegal 207
Table 21: Voting Intentions According to Département 23 0
Table 22: Major Findings from the Case Studies in a Comparative View 248
Table 23: Direct and Indirect Linkage in Ghana and Senegal 254




1 Introduction: The Purpose of the Study

A wind of change swept the African continent in the 1990s when a large number
of countries introduced multiparty systems and numerous longstanding autocrats
lost their power.
In many cases, transitions from authoritarian rule to multiparty politics
started with popular protest by civic associations, trade unions, student and
women’s organisations, professional associations, churches and many others
(see Bratton/van de Walle 1997: 117). Multiparty democracy became the
rallying cry of these heterogeneous opposition coalitions that were kept together
by their hostility to the old regime. Yet, the more the restrictions on political
activities were lifted, the more civil society groups were replaced by political
parties. While both types of associations are channels for the aggregation and
articulation of interests, parties are able to contest elections and form
governments. As soon as founding elections are scheduled, political parties
typically emerged as the main actors in the transition.
In the early 1990s, scholarly discourse placed much hope on the democracy
promoting functions of civil society (see, among others, Diamond 1994).
Political parties, in contrast, had long been a relatively neglected topic in
African Studies. The revitalisation of African party research only began at the
end of the 20
th
century and followed the general trend of a worldwide revival of
the discipline. Two factors contributed to the rising interest in political parties:
the changing role of parties in established democracies (party change) and the
proliferation of parties in formerly authoritarian regimes in the course of the
‘Third Wave of Democratization’.
1
Both developments challenge the

conventional understanding of political parties as ‘mass parties’ that are tied to
societal interest groups. Contemporary political parties in Western democracies
are losing members, becoming less strict in their ideological orientations, and
becoming gradually more disentangled from civil society groups (see, for
example, von Beyme 2000). In a similar vein, the debate on new democracies
tends to emphasise the shortcomings or weaknesses of parties. Political parties
in Asia, Eastern Europe, Africa, Latin America and the Middle East are thought
to have weak formal structures and weak ideologies (see for example Carothers

1 This notion goes back to Samuel Huntingtons book “The Third Wave: Democratization in the Late
Twentieth Century”, published in 1991.
A. Osei, Party-Voter Linkage in Africa, DOI 10.1007/978-3-531-19140-9_1,
© VS Verlag für Sozialwissenschaften | Springer Fachmedien Wiesbaden 2012
Introduction: The Purpose of the Study 15
2006: 6-7), and little institutionalised contact with civil society groups
(Betz/Erdmann/Köllner 2004: 2). While the challenges for theory building may
differ with regard to old and new democracies, the underlying question of the
relationship between parties and societies concerns both strands of research.
Political parties are commonly seen as the “central intermediate and
intermediary structure between society and government” (Sartori 2005 [1976]:
ix). From Sartori’s statement it follows that political parties have a double
character of being active in both state and society and thereby linking the two
spheres. Especially in Western democracies, political parties have been seen as
“an outgrowth from society” (Kopecký/Mair 2003: 275). Whether or not this
understanding is applicable to non-Western societies is a controversial issue, but
what seems to be clear is that all political parties in all systems need a certain
degree of social anchorage in order to maximise their votes. Therefore, they
must be oriented towards the society in one way or another. This relation can be
discussed under the concept of linkage developed by Kay Lawson. In her book
Political parties and Linkage: A Comparative Perspective (1980) she refers to

political parties as
agencies for forging links between citizens and policy makers. Their raison d’être
is to create a substantive connection between rulers and ruled.
However, many recent publications suggest that political parties in Africa are
only weakly rooted in society (see Randall 2007; Erdmann 1999; Widner 1997).
African parties are often described as organisationally and programmatically
weak. Most of them do not have any ‘mass character’, they lack a stable and
functioning party apparatus and active membership (Erdmann 2002a: 268).
Parties do not appear as expressions of social interests, but as vehicles for the
ambitions of individual politicians (Tetzlaff 2002: 240). Even parties that
emerged from pro-democracy movements during the transitions to democracy in
the 1990s only have loose contacts to civil society groups now.
2
Widner (1997:
66) attributes this to the character of civil society itself and argues that it is not
attractive for African political parties to appeal to voluntary associations
because these groups themselves are weak and their leaders have little power to
deliver their members’ votes. Furthermore, civil society movements are mostly
urban based and do not represent the rural majority (ibid.: 78). In contrast,
Burnell (2007: 77) argues that this is only true if civil society is defined in

2 Zambia may serve as an example. The Zambian Trade Union Congress (ZCTU) played a major
role in the struggle for democratic reform and the formation of the MMD (Movement for Multiparty
Democracy). When the MMD came to power in 1991, a gradual process of detachement began and
today the trade unions are marginalised in the political process. For more on this, see Burnell (2001).
16 Introduction: The Purpose of the Study
Western liberal democratic terms.
3
It might therefore be imaginable that
political parties in Africa have a different form of linkage to society.

A look at indicators like voter turnout, individual party identification, and
trust in political parties raises further questions about the alleged weakness of
the relationship between parties and voters in Africa. Although cross-regional
comparison should always be considered carefully, a look at the data from other
regions of the world provides an opportunity to put the findings from Africa into
perspective.
Voter mobilisation
Due to enthusiasm and the hope for change, voter turnouts were remarkably
high in founding elections, which recorded an average turnout of 64.1%
(Bratton 1998: 53). In the elections that followed the founding elections –
termed ‘second elections’ by Bratton (1998) – turnout figures declined to an
average of 55.8% (ibid.: 54-56). The reason for the decline is a ‘normalisation’
of electoral participation after the peak in the founding elections, but also a
developing disillusionment and a deterioration in the quality of multiparty
elections in some countries. In an evaluation of worldwide electoral
participation from 1990-2001 by Pintor and Gratschew (2002), Africa has the
lowest average turnout with 64%. However, the gap between Africa and the rest
of the world is relatively small (see Figure 1). Africa can keep up with the
Middle East, North America and the Caribbean, each with 65%, and Central and
South America with 69%. Only Oceania, Western Europe, Central and Eastern
Europe, and Asia scored an average turnout of more than 70% (ibid.: 77-78).

3 The usefulness of the concept ‘civil society’ for Africa is debatable. For an exemplary critique see
Kasfir (1998).

Introduction: The Purpose of the Study 17
Figure 1: Voter Turnout Worldwide, 1990-2001

Source: Pintor/Gratschew (2002: 77).
Africa’s low place in the ranking is partly caused by statistical outliers with

extremely low participation, like Mali (21.3%), Côte d’Ivoire (37%) and
Burkina Faso (41.7%), whereas in other regions even the last placed countries
had turnout rates above 45% (ibid.: 81). Secondly, the official turnout figures
are sometimes misleading because of problems with voter registration in many
African countries. For example, in the 1992 election in Ghana, the voter’s
register was bloated (Jeffries/Thomas 1993: 349), while in other cases
opposition sympathisers might be restricted from voting, as happened in the
Senegalese presidential elections 2007. Kuenzi and Lambright (2005: 6-7) find a
large gap between the official turnout and the number of people who reported
voting in recent elections in the Afrobaometer survey
4
in a number of countries.
Despite these shortcomings, electoral participation on the continent is in sum
encouraging and political parties seem to play a key role in voter mobilisation.

4 The Afrobarometer is an independent, nonpartisan research project that measures the social,
political, and economic atmosphere in Africa. Afrobarometer surveys are conducted in more than a
dozen African countries and are repeated on a regular cycle. Data are available at
www.afrobarometer.org.
79
72
72
72
69
65
65
64
0 20406080100
Oceania
Western Europe

Asia
Central/Eastern Europe
Central/South America
North America/Carrib.
Middle East
Africa
18 Introduction: The Purpose of the Study
Kuenzi and Lambright (ibid.: 17) find a strong positive relationship between
affiliating with a political party and voting. The authors conclude that political
parties are ‘mobilisation agents’ despite their organisational weakness:
those affiliating with a political party are targets of mobilisation efforts and more
connected to the political system (ibid.: 4-5).
Party identification
Levels of individual party identification are surprisingly high in Africa. In the
first round of the Afrobarometer survey, 57% of all respondents stated that they
felt close to one of the political parties in their respective country
(Bratton/Mattes/Gyimah-Boadi 2005: 257). Again, there is great variance across
the continent, with relatively low party affiliation in Zambia and notably high
levels in Namibia and Tanzania. Figure 2 shows the levels of individual party
identification in selected countries based on the Afrobarometer Survey, Round
2.
Introduction: The Purpose of the Study 19
Figure 2: Party Identification in Selected African Countries

Source: Compiled by the author.
Party identification must not be confused with party membership. The
Afrobarometer survey measures a sense of closeness to a political party with the
question: “Do you feel close to one of the political parties in the country?”,
which does not necessarily imply engaging in party activities, paying dues or
even perceiving one’s self as a proper party member in the Western sense of the

word. Notions of party membership differ significantly across countries and
regions of the world. Therefore, cross-country comparisons like that of Norris
(2002: 110-115) can lead to ambiguous conclusions. Norris found that party
membership constitutes only 7.2% of the population in Western democracies. In
contrast, in the two African cases that were included – South Africa and Nigeria
– 44.1% and 43.4% respectively perceived themselves as being party members.
The problem with this kind of comparison is that the Western understanding of a
dues-paying and card-bearing party membership is largely absent in Africa.
Therefore, the respondents in Nigeria and South Africa are better described as
sympathisers rather than as party members.
39
50,1
54,8
65,5
67,9
70,8
78,5
0 20406080
Zambia
Nigeria
Senegal
Ghana
Kenia
Tanzania
Namibia
20 Introduction: The Purpose of the Study
Trust in political parties
Across the world, political parties are among the least respected and least trusted
institutions (Carothers 2006: 4). Africa is no exception in this regard. In the
Afrobarometer Survey (Round 1), only 51% of the respondents said that they

trusted political parties ‘somewhat’ or ‘a lot’ (Bratton/Mattes/Boadi 1999:
257).
5
Again, it is telling that the trust in political parties is even much lower in
established European democracies. The Eurobarometer-Report 64 (2005)
6

shows that only 17% of the respondents in the European Union trust political
parties.
This contradiction between allegedly low roots in society, low ideological
profile and weak structures on the one hand and high voter turnouts as well as
high levels of party identification on the other provokes an interesting question:
How do political parties in Africa relate to society? How do they mobilise their
voters and sympathisers, and which strategies do they employ?
These questions have almost remained untouched in the literature so far.
Nolte et.al. (2000: 10) state that there is a lack of comparative studies on the
linkage strategies and modes of interest representation of political parties in new
democracies. While a lot of work will be needed to close this research gap, the
present study seeks to contribute to a better understanding of party-voter linkage
in Africa by comparing two country cases, namely Ghana and Senegal.
The thesis proceeds as follows. In chapter 2, the question of ‘universal‘ or
‘Africa-specific‘ research criteria will be discussed in more detail. This is
necessary because research on political parties in Africa draws on two sources:
the classical field of Western party research, which is an important and well-
developed subdiscipline of political science, and the debates on democratisation
and the state in Africa, which are more or less restricted to area specialists.
Based on this discussion, a framework for comparative research will be
developed in chapter 4. At the heart of the approach lies the assumption that a
concept for the study of political parties in African countries must be context-
sensitive and sufficiently general, and that it must involve qualitative as well as

quantitative data. Qualitative data was collected mainly in semi-structured
interviews with party representatives at different levels of the party hierarchy in
the two country cases. Quantitative data are a supplementary element of the

5 In a recent article, Logan (2008: 5) has pointed out to the fact that there is, on average, a huge gap
between the trust in ruling parties and the opposition.
6 The Eurobarometer is a yearly survey which collects data on attitudes towards political and social
developments in the European Union. Reports are available from the Eurobarometer website


Introduction: The Purpose of the Study 21
study. These include election data and data from the Afrobarometer survey,
which are available on the networks’ website (www.afrobarometer.org).
Additionally, a small survey on voters’ opinions towards political parties was
conducted in cooperation with local partner institutions in Ghana and Senegal. It
included 120 respondents in Ghana and 120 respondents in Senegal. For more
on data collection see chapters 3.3.1. and 3.3.2. in this thesis. The empirical
chapters six and seven will provide detailed information on the six selected
parties, which is put in a comparative perspective in the chapters eight and nine.
Some general conclusions that can be drawn from the research project are then
highlighted in the final section of the thesis.



2 Conceptual Problems of Party Research in Africa
2.1 Introduction
Research on political parties in Africa draws on two sources:
 the classical field of (mostly Western) party research, which is an important
and well-developed sub-discipline of political science that has produced a
large body of literature;

 debates on democratisation and the state in Africa, which are more or less
restricted to area specialists.
In these two fields of study, academic debates on theory and methodology have
largely unfolded independently from each other and may even seem
incompatible at a first glance. At least this was the outcome of some recent
studies on political parties in Africa: party research appeared to have a
methodological bias towards Western Europe, and parties on the African
continent did not easily fit into the established paradigms and assumptions.
Naturally, this raised concerns about ‘universal’ or ‘Africa-specific’ criteria for
further research – an issue which is already a perennial one in the Area Studies
(for more on this see Chabal 2005a). While area specialists are certainly more
sensitive to this problem it has only recently started to enter into the main field
of contemporary party research. The advent of the ‘Third Wave of
democratization’ in the developing countries has increasingly drawn the
attention to political parties in non-Western societies. What actually binds both
research fields together is the fact that party research cannot be disconnected
from the broader issues of statehood and democracy. In this context it is quite
obvious that the Third Wave has posed a number of challenges to conventional
notions of what democracy is and how it functions. Imperfect neo-democracies
– sometimes called hybrid regimes – in Africa, Asia, Latin America and Eastern
Europe defy some of the established assumptions. Here, Africa is no exception,
it rather seems that
A. Osei, Party-Voter Linkage in Africa, DOI 10.1007/978-3-531-19140-9_2,
© VS Verlag für Sozialwissenschaften | Springer Fachmedien Wiesbaden 2012
Party Research in Africa: Findings and Problems 23

what is often thought of as an uneasy, precarious middle ground between full-fledged
democracy and outright dictatorship is actually the most common political condition today
of countries in the developing world and the postcommunist world. (…) the disappointment
over falling short of democracy should be replaced with realistic expectations about the

likely patterns of political life in these countries (Carothers 2002: 18).
Consequently, this will have wider implications for political parties as actors in
the realm of state and society. In the following section, we will discuss some of
the findings and problems of party research in Africa and relate them to the
latest developments in the field of general party research as well as that of
democracy studies.
2.2 Party Research in Africa: Findings and Problems
Following independence most African countries had multiparty systems which
stimulated a first body of literature on political parties (see for example Hodgkin
1961; Morgenthau 1964; Coleman/Rosberg 1964 and others). From the 1960s
onwards, single-party states and authoritarian rule spread over the continent and
scholarly interest in political parties declined. In the 1990s, multipartyism
reappeared in Africa, but it took until the turn of the century before the research
on African political parties was revitalised.
As a consequence of the multidimensionality of political parties as objects
of research, the first studies took up a large variety of different problems, such
as typologies of parties and party systems (Erdmann 2002a), functions of
political parties for democratisation and consolidation (Randall/Svåsand 2002;
Emminghaus 2002), electoral systems and party systems (Boogards 2000;
Basedau 2002), clientelism and presidentialism (van de Walle 2003) or the
representation of group interests in parties (Randall 2007; Widner 1997).
Since multiparty systems on the continent are of a relatively recent origin,
only little field research has been carried out so far and there is a clear lack of
empirical evidence. Research on party systems made faster progress because it
allows working with aggregate data (election results, numbers of parties and
seat shares in parliament) while studies on individual parties usually require
extensive field research.
Nevertheless, most scholars agreed that political parties in Africa were
organisationally and programmatically weak (see for example (Tetzlaff 2002);
Randall/Svåsand 2001). Most parties were said to lack a stable and functioning

party apparatus and active membership (Erdmann 2002: 268). In a study for the
Institute for Democracy and Electoral Assistance (IDEA), Nordlund/Salih
(2007: 81) observed that 75% of the 200 West-African parties under
24 Conceptual Problems of Party Research in Africa

investigation had no offices at the polling station level, and 62% had no district
offices. Especially opposition parties are often urban-based and expose only
little activities between elections. Many parties are very poorly financed and
find it therefore difficult to maintain permanent administrative structures. State
funding is provided only in a few African countries (Saffu 2003: 22) and where
it exists, it tends to over-advantage ruling parties (Gyimah-Boadi 2007: 26). The
second commonality is the low ideological profile. For instance, political parties
in Africa cannot be placed on a traditional left-right continuum. Their
programmatic appeals – if a written programme exists at all – amount to little
more than populism: they make vague promises of fighting corruption and a
better future for all (van de Walle 2003: 304). The few parties with more policy-
based campaigns mostly have very little success (ibid.). Thus, political parties in
Africa appear less as the expression of group interests than as projects of
political entrepreneurs who can afford to pay for the campaign out of their own
pockets (Tetzlaff 2002: 243). Under these circumstances, it is hardly surprising
that party structures tend to be highly personalised and centralised around the
party leader, who often becomes synonymous with the whole party (Erdmann
1999: 380). Consequentially, internal democracy and accountability are
underdeveloped.
A careful review of the growing literature does also reveal some conceptual
problems. It is particularly noticeable that political parties in Africa tend to be
portrayed in negative terms. For example, Erdmann (2004: 65), who is one of
the pioneers in this field, characterised political parties in Africa as having the
following properties:
 hardly distinguishable programs which in most cases are meaningless to

policies
 hardly countable members
 weak bureaucratic organisation
 informal relations, partly based on clientelistic relations and patronage,
predominate the party structures
 strong personalism
 high degree of factionalism
 weak funding
 lack of internal democracy
 weak formal linkages to society
Party Research in Africa: Findings and Problems 25

 predominantly regional and/or ethnic based membership and electorate.
Erdmann’s characteristics are worth quoting at length, and not just because
they reflect the contemporary state of the debate; they also reveal the two
sources that party research in Africa draws on. This is, firstly, the debate about
democratisation and the state in Africa, which is more or less restricted to
specialised scholars of African politics. In their discourse, ethnicity and
clientelism feature prominently as elements of politics in Africa – the obvious
conclusion is that this must somehow affect political parties, too. Hypotheses
about the clientelistic nature of African parties are merely derived from state
discourses and as long as there are not enough field studies, they remain
speculative. Nevertheless, by providing a starting point for the proper questions
to be asked in a specific African context this strain of research makes a valuable
contribution. The other source that stimulated interest in African parties is party
research as a classical subfield of political science, which has produced a large
body of literature
7
and numerous different concepts that are mostly developed
with regard to political parties in established Western democracies. It soon

turned out that many of the concepts of party research were hardly applicable to
African countries. Against this background, parties in Africa were often
described by what they are not. Erdmann’s frequent usage of the expressions
weak, lack and hardly in the above quoted passage demonstrates this.
Studies that tried to test concepts of Western European party research on
African cases often ended up with falsifications:
 Contrary to their Western counterparts, political parties in Africa are not
mass parties. It was exactly for this reason that Erdmann (2002: 260-262)
criticised earlier party typologies which were, even if implicitly, based on
the conception of the mass party.
 African parties do not seem to fulfil the functions that are ascribed to them
(Emminghaus 2002; Randall/Svåsand 2001).
 Duverger (1959) had claimed that the electoral system of a country
determines its party system: a majoritarian system leads to a two-party
system, while a system of proportional representation increases the number
of parties in parliament. In Africa, dominant party systems have emerged in
many countries, irrespective of their electoral systems (Bogaards 2000:
166-167; Basedau 2003).

7 From 1945 to 2003, about 11500 books and articles that deal with parties and party systems in
Western Europe alone have been published (Gunther/Montero 2002: 2).
26 Conceptual Problems of Party Research in Africa

 Political parties have only weak organisational linkages to civil society
groups (Widner 1997; Randall 2007).
 Party formation in Africa does not represent the four classical cleavages
(urban/rural, church/state, labour/capital and centre/peripherie) which were
identified by Lipset and Rokkan (1967) as structuring the evolution of
party systems in Europe.
8


Adding to that, even some of the most famous definitions of political parties
would actually exclude African parties. For instance, Giovanni Sartori (2005
[1976]: 64) defined a political party as
any political group that presents at elections, and is capable of placing through
elections, candidate for public office.
This view is typical for a large number of definitions which emphasise the
capability of parties to contest elections. A definition by Downs even requires
free and fair elections - for him a party is
a team of men seeking to control the governing apparatus by gaining office in a
duly constituted election (1957: 25).
The focus on party competition in Sartori’s and Downs’ statements excludes
single-parties from the definition. This would mean that KANU, TANU or
UNIP were not parties during the era of the single-party state. Surely, one-party
states might be treated as a completely different phenomenon, a single-party
might even be a “contradiction in terms” (Neumann 1956: 395), but then it
would not be entirely clear, when and how a non-party becomes a proper party
in the course of democratisation and what exactly would differentiate them. The
definition is also not inclusive of opposition parties that boycott elections –
which occurs quite frequently in Africa –, of anti-system parties that do not
want to play by the rules of the game, and of parties which do not fulfil certain
requirements of the party law
9
, but which conceive of themselves as political
parties.

8 These conflicts exist in Africa, too, but they are less pronounced and did not stimulate the
formation of political parties. A partly exception is the centre/periphery-conflict, under which
ethnicity could be subsumed. For detailed discussions of the model, see Rüland (2001);
Erdmann/Weiland (2000) and Erdmann (2008).

9 In many African countries, parties must fulfil certain criteria for their registration, like building up
a nationwide structure (as it is the case in Ghana) or having a “national character”. Particularistic
parties based on ethnie, religion or race, can be deregistered in 40 out of 46 African countries with
multiparty systems. For more on ethnic party bans, see Becher (2008).